System for and method of personalized screening and diagnosis with hematological setpoints

Abstract

Systems and methods for receiving data representing one or more measurements of a hematological parameter of a subject, each of the one or more measurements being collected from a corresponding complete blood count (CBC) test of a plurality of CBC tests performed on blood samples of the subject taken from the subject at one or more times and calculating, using the data representing the one or more measurements of the hematological parameter of the subject, a hematological setpoint of the hematological parameter of the subject. The method further includes determining, from the hematological setpoint for the hematological parameter of the subject, a pathophysiological state of the subject and / or a risk for the subject to develop the pathophysiological state.

Description

[0001] Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0002] System for and Method of Personalized Screening and Diagnosis with Hematological Setpoints

[0003] CLAIM OF PRIORITY

[0004] This application claims the benefit of U.S. Provisional Application Serial No. 63 / 695,679, filed on September 17, 2024. The entire contents of the foregoing are incorporated herein by reference.

[0005] FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0006] This invention was made with government support under grant numbers 5R01DK123330-04 and 5R01HD104756-03 (agreements 2019A015196 and 2021A003180) from the NIH. The government has certain rights in the invention.

[0007] BACKGROUND

[0008] The complete blood count (CBC) is an important and versatile clinical test that is ordered more frequently than any other clinical test across nearly all medical contexts. It provides a valuable non- specific assessment of the hematologic and immunologic state of patients by measuring numbers of red blood cells (RBC), white blood cells (WBC), and platelets (PLT) per unit volume of blood, along with some cell population statistics. Because blood cells are continuously exposed to almost all tissues and organs, the CBC provides timely information on a very wide range of disease processes.

[0009] CBC indices vary dramatically among healthy adults, with the upper limit of the reference interval more than twice as large as the lower in some cases (e.g., 4.5-11 x 10 / il for WBC). These one-size-fits-all reference intervals currently used for interpretation are very wide and can therefore be insensitive to signs of disease and recovery. The use of screening laboratory studies for healthy adults has not changed significantly in many years and still relies primarily on the generic application of a few population-wide approaches to screening.

[0010] SUMMARY

[0011] Disclosed herein is a family of diagnostic and screening technologies that identify a hematologic setpoint of a patient and uses that hematological setpoint to identify the patient’s risk for developing cardiovascular, metabolic, renal disease. Atorney Docket No. 29539-0847WO1 / MGH 2024-526 cancer, common morbidities like osteoporosis where early intervention alters disease course, and other morbidities. These technologies, for example, calculate a hematologic setpoint by using a Gaussian mixture model and apply the hematological setpoint as part of screening the patient.

[0012] These technologies also improve the sensitivity and specificity of existing screening programs relying on HbAlc to screen for prediabetes and diabetes and eGFR / creatinine to screen for kidney disease.

[0013] The complete blood count (CBC) is an important screening tool for healthy adults and is the most commonly ordered test at periodic physical exams. However, results are usually interpreted relative to one-size- fits-all reference intervals, undermining the goal of precision medicine to tailor medical care for individual patients based on their unique characteristics. This disclosure describes the results of tens of thousands of diverse patients at an academic medical center and shows that routine complete blood count indices tend to fluctuate around stable values or setpoints. These setpoints are patient-specific, with the typical healthy adult’s ten blood count setpoints distinguishable as a group from those of 98% of other healthy adults, and with these differences persisting for at least 20 years. Hematologic setpoints reflect a deep physiologic phenotype enabling investigation of both acquired and genetic determinants of hematologic regulation and its variation in healthy adults, including discovery of novel loci by GWAS. Setpoints in apparently healthy adults w ere associated with significant variation in clinical risk: absolute risk of some common diseases and morbidities varied by >2% (heart attack and stroke, diabetes, kidney disease, osteoporosis), and absolute risk of all-cause 10-year mortality' varied by >5%. Setpoints also define patient specific reference intervals and personalize the interpretation of subsequent test results. In retrospective analysis, setpoints improved sensitivity and specificity for evaluation of some common conditions including diabetes, kidney disease, thyroid dysfunction, iron deficiency, and myeloproliferative neoplasms. This disclosure shows complete blood count setpoints are sufficiently stable and patient-specific to help realize the promise of precision medicine for healthy adults.

[0014] In general, in a first aspect, a method includes receiving data representing one or more measurements of a hematological parameter of a patient. Each of the one or more measurements being collected from a corresponding complete blood count Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0015] (CBC) test of one or more CBC tests performed on blood samples of the patient taken from the patient at one or more times. The method includes calculating, using the data representing the one or more measurements of the hematological parameter of the patient, a hematological setpoint of the hematological parameter of the patient. The method includes determining, from the hematological setpoint for the hematological parameter of the patient, a pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state.

[0016] In general, in a second aspect, combinable with the first aspect, determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, from the hematological setpoint of the hematological parameter of the patient and a reference interval for the hematological parameter of a population, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

[0017] In general, in a third aspect, combinable with the first aspect, determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, based on the one or more measurements of the hematological parameter of the patient being within a reference interval, that the one or more measurements of the hematological parameter do not indicate the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

[0018] In general, in a fourth aspect, combinable with the first aspect, determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, from the hematological setpoint of the hematological parameter of the patient and one or more hematological setpoints of a population, that the one or more measurements of the hematological parameter do not indicate the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

[0019] In general, in a fifth aspect, combinable with the first aspect, the pathophysiological state includes diabetes, kidney disease, thyroid dysfunction, iron deficiency, myeloproliferative neoplasm and other malignancy, atrial fibrillation, myelodysplastic syndrome, and osteoporosis, cardiovascular events, and bone fractures. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0020] In general, in a sixth aspect, combinable with the first aspect ,a quantity of the one or more CBC tests is no less than five.

[0021] In general, in a seventh aspect, combinable with the first aspect, the one or more times of the one or more CBC tests are at least 90 days apart from each other.

[0022] In general, in an eighth aspect, combinable with the first aspect, calculating, using the data representing the one or more measurements of the hematological parameter of the patient, the hematological setpoint of the hematological parameter of the patient includes calculating, using the data representing the one or more measurements of the hematological parameter of the patient, a variability7of the one or more measurements for calculating the hematological setpoint of the hematological parameter of the patient and determining, from the variability of the one or more measurements for the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

[0023] In general, in a ninth aspect, combinable with the first aspect, calculating the variability of the one or more measurements includes calculating the variability based on a difference between the hematological setpoint of the hematological parameter of the patient and a most recent measurement of the one or more measurements.

[0024] In general, in a tenth aspect, combinable with the first aspect, calculating, using the data representing the one or more measurements of the hematological parameter of the patient, the hematological setpoint of the hematological parameter of the patient includes calculating, using the data representing the one or more measurements of the hematological parameter of the patient, one or more hematological setpoints of the hematological parameter of the patient, the one or more hematological setpoints including the hematological setpoint. Additionally, in the tenth aspect, combinable with the first aspect, determining, from the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, from a variability of the one or more hematological setpoints of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

[0025] In general, in an eleventh aspect, combinable with the first aspect, determining, from the hematological setpoint of the hematological parameter of the Atorney Docket No. 29539-0847WO1 / MGH 2024-526 patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, from an interval at least partially defined by the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

[0026] In general, in a twelfth aspect, combinable with the eleventh aspect, computing the interval based on a variability of the one or more measurements of the hematological parameter with the hematological setpoint for the hematological parameter of the patient.

[0027] In general, in a thirteenth aspect, combinable with the first aspect, determining, from the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, from a measurement of the hematological parameter of the patient and the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

[0028] In general, in a fourteenth aspect, combinable with the thirteenth aspect ,the pathophysiological state is a ferritin-deficient or iron-deficient state, and the hematological parameter is a red blood cell (RBC) hemoglobin content.

[0029] In general, in a fifteenth aspect, combinable with the fourteenth aspect, the measurement of the hemoglobin of the patient and a reference interval for the hemoglobin of a population do not indicate the ferritin-deficient or iron-deficient state.

[0030] In general, in a sixteenth aspect, combinable with the first thirteenth, the pathophysiological state is a state of elevated thyroid stimulating hormone (TSH) and the hematological parameter is a mean RBC volume.

[0031] In general, in a seventeenth aspect, combinable with the sixteenth aspect, the measurement of the mean RBC volume of the patient and a reference interval for the mean RBC volume of a population do not indicate the state of elevated TSH.

[0032] In general, in an eighteenth aspect, combinable with the first thirteenth, the pathophysiological state is a state of JAK2 mutation disorder, and the hematological parameter is a platelet count. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0033] In general, in a nineteenth aspect, combinable with the eighteenth aspect ,the measurement of the platelet count of the patient and a reference interval for the platelet count of a population do not indicate the state of JAK2 mutation disorder.

[0034] In general, in a twentieth aspect, combinable with the first aspect, determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes detennining information indicative of a degree of morbidity’ of the patient, the information indicative of the degree of morbidity’ including at least one of: information indicative of a presence of an infection, information indicative of a presence of malignancy, information indicative of a presence of anemia, information indicative of a presence of diabetes, or information indicative of a dysfunction.

[0035] In general, in a twenty first aspect, combinable with the first aspect, determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining the risk for the patient to develop the pathophysiological state, the risk for the patient to develop the pathophysiological state being a risk to develop at least one of: an infection, a malignancy, a disease, anemia, or diabetes.

[0036] In general, in a twenty’ second aspect, combinable with the first aspect, the method includes providing information indicative of a recommended treatment regimen for the patient, wherein the recommended treatment regimen includes diagnostic screening on the patient specific to the pathophysiological state.

[0037] In general, in a twenty third aspect, combinable with the first aspect, the method includes identifying, from the hematological setpoint for the hematological parameter of the patient, a stratification group for the patient for a clinical trial for a treatment for the pathophysiological state.

[0038] In general, in a twenty fourth aspect, combinable with the first aspect, the information indicative of the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes information indicative of a degree of morbidity of the patient, the information indicative of the degree of morbidity' including at least one of: information indicative of a presence of an infection, information indicative of a presence of malignancy, information indicative of a presence of anemia, infomiation indicative of a presence of diabetes, or information indicative of a dysfunction. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0039] In general, in a twenty fifth aspect, combinable with the first aspect, the information indicative of the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes information indicative of an efficacy of treatment of the patient for the pathophysiological state.

[0040] In general, in a twenty sixth aspect, combinable with the first aspect, the hematological parameter includes a measure of red blood cell (RBC) volume, a measure of RBC hemoglobin mass, a measure of RBC size, a measure of RBC hemoglobin concentration, a measure of platelet size, a measure of hemoglobin content, a measure of platelet content, a measure of RBC count, a measure of RBC size variation, or white blood cell (WBC) count.

[0041] In general, in a twenty-seventh aspect, combinable with the first aspect, calculating, using the data representing the one or more measurements of the hematological parameter of the patient, the hematological setpoint includes computing a mean of a Gaussian component of the one or more measurements identified from applying a Gaussian mixture model to the one or more measurements, the mean corresponding to the hematological setpoint.

[0042] In general, in a twenty-eighth aspect, combinable with the first aspect, calculating, using the data representing the one or more measurements of the hematological parameter of the patient, the hematological setpoint includes computing a mean of a distribution of one or more measurements identified from applying a Bayesian method to the one or more measurements, the mean corresponding to the hematological setpoint.

[0043] In general, in a twenty -ninth aspect, combinable with the first aspect, the method further includes receiving data representing a measurement of a hematological parameter of a patient, the measurements being collected from a corresponding complete blood count (CBC) test performed on a blood sample of the patient taken at a time. The method further includes calculating, using the hematological setpoint of the hematological parameter and the data representing the measurement of the hematological parameter of the patient, an updated hematological setpoint of the hematological parameter of the patient and determining, from the updated hematological setpoint for the hematological parameter of the patient, a pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0044] In general, in a thirtieth aspect, combinable with any of the first through twenty-nine aspects, a system includes one or more processors and memory storing instructions which, when executed by the one or more processors, cause the one or more processors to perform operations including: receiving data representing one or more measurements of a hematological parameter of a patient, each of the one or more measurements being collected from a corresponding complete blood count (CBC) test of one or more CBC tests performed on blood samples of the patient taken from the patient at one or more times; calculating, using the data representing the one or more measurements of the hematological parameter of the patient, a hematological setpoint of the hematological parameter of the patient; and determining, from the hematological setpoint for the hematological parameter of the patient, a pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Methods and materials are descnbed herein for use in the present invention; other, suitable methods and materials known in the art can also be used. The materials, methods, and examples are illustrative only and not intended to be limiting. All publications, patent applications, patents, sequences, database entries, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control.

[0046] Other features and advantages of the invention will be apparent from the following detailed description and figures, and from the claims.

[0047] DESCRIPTION OF DRAWINGS

[0048] FIG. 1 is a diagram of an example system for determining a recommended treatment for a patient based on one or more hematological setpoints.

[0049] FIG. 2A shows a graph indicative of a computation of a hematological setpoint using a Gaussian mixture model.

[0050] FIG. 2B show s a graph indicative of setpoints as personal reference points in a population wide reference interval.

[0051] FIG. 2C shows a graph indicative of enhanced comparisons between new results and hematological setpoints within a population wide reference interval. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0052] FIG. 2D shows a graph indicative of a personalized reference interval for a first patient based on a hematological setpoint.

[0053] FIG. 2E shows a graph indicative of a personalized reference interval for a second patient based on a hematological setpoint.

[0054] FIG. 3 is a flowchart of an example process for determining a pathophysiological state of a patient based on one or more hematological setpoints.

[0055] FIG. 4A shows a graph indicative of intra-patient and inter-patient variation in complete blood counts (CBCs).

[0056] FIG. 4B shows a graph indicative of ratios of intra- and inter-patient coefficient of variations (CVs).

[0057] FIG. 4C shows a graph indicative of long term inter- and intra-patient CVs for CBC indices based on sex or ethnicity.

[0058] FIG. 4D shows a graph indicative of one or more levels of hematological setpoints within a population reference interval.

[0059] FIG. 5A shows a graph indicative of setpoint and single CBC correlation between partners.

[0060] FIG. 5B shows a graph indicative of setpoint and single CBC correlation between first-degree relatives.

[0061] FIG. 5C shows a graph indicative of heritability estimates based on setpoints and literature values.

[0062] FIG. 5D shows a graph indicative of single nucleotide polymorphism (SNP)- heritability estimates using setpoints and single CBCs.

[0063] FIG. 5E shows a graph indicative of a Manhattan plot for a genome-wide association study (GWAS) of a hemoglobin setpoint.

[0064] FIG. 5F shows a graph indicative of a comparison of p-values from coefficients for each SNP in a GWAS using hemoglobin setpoints vs. using single CBCs.

[0065] FIG. 5G shows a graph indicative of a comparison of effect sizes from coefficients for each SNP in a GWAS using hemoglobin setpoints vs. using single CBCs.

[0066] FIG. 5H shows a graph indicative of the number of significant genome-wide hits based on a setpoint or a single CBC measurement of the hematological parameter. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0067] FIG. 51 shows a set of graphs indicative of correlation of polygenic score quintiles with mean setpoint values.

[0068] FIG. 6A shows a set of graphs indicative of mortality rates associated with various setpoint quintiles.

[0069] FIG. 6B show s a graph indicative of mortality7hazard ratios adjusted for age and sex of patients for various setpoint values.

[0070] FIG. 6C shows a graph indicative of mortality hazard ratios for various setpoint variations at two different medical centers.

[0071] FIG. 7A show s a set of graphs indicative of the association of the one or more hematological setpoints with a risk of one or more respective major diseases.

[0072] FIG. 7B shows a graph indicative of disease hazard ratios adjusted for age and sex for various hematological setpoint variations.

[0073] FIG. 7C shows a graph indicative of a mortality risk associated w ith variations of a measurement of a hematological parameter with the setpoint of the hematological parameter for the patient.

[0074] FIG. 7D shows a graph indicative of a mortality hazard ratio for a patient in w hich the current measurement of a hematological parameter is outside the patient’s hematological setpoint based interval for a hematological parameter compared to the mortality hazard ratio for a patient in which the current measurement is outside the population-wide reference interval.

[0075] FIG. 7E show s a set of graphs indicative of a likelihood of progressing to latestage kidney disease based on variations from an RBC volume fraction setpoint.

[0076] FIG. 7F shows a graph indicative of likelihood to have future blood tests in diabetic range for prediabetic patients based on variations from one of the hematological setpoints.

[0077] FIG. 7G shows a set of graphs indicative of a likelihood of low test results based on variations from one of the hematological setpoints for the patient.

[0078] FIG. 7H shows a graph indicating a likelihood for high test results based on variations from one of the hematological setpoints for the patient.

[0079] FIG. 71 shows a graph indicating a likelihood for test positivity rates based on various values for one of the hematological setpoints.

[0080] FIG. 8 shows a table indicative of CBC reference intervals for the one or more hematological parameters. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0081] FIG. 9 shows a table indicative of age and gender effects on the values of the one or more hematological parameters.

[0082] FIG. 10 shows a graph indicative of heritabihty estimates for each hematological parameter using one or more studies.

[0083] FIG. 11 A shows a set of graphs indicative of a stratification by age of mean inter- and intra-patient CVs for each hematological setpoint of the one or more hematological parameters.

[0084] FIG. 11 B shows a graph indicative of a mean intra-patient CV over various time periods.

[0085] FIG. 11C shows a set of graphs indicative of associations between hematological setpoints and hematological setpoint CV.

[0086] FIG. HD shows a set of graphs indicative of mean absolute difference between hematological setpoints with variable numbers of datapoints for each of the one or more hematological setpoints.

[0087] FIG. 12A shows a graph indicative of correlation of hematological setpoints with setpoints of other hematological markers.

[0088] FIG. 12B shows a graph indicative of correlation of hematological setpoints with mean marker values for a range of common laboratory7tests.

[0089] FIG. 12C shows a set of graphs indicative of correlation of hematological setpoints with common laboratory test differences.

[0090] FIG. 13 A shows a graph indicative of a hematological setpoint shift in a patient pre- and post-menopause.

[0091] FIG. 13B shows a graph indicative of hematological setpoint shifts after pathophysiological events such as menopause.

[0092] FIG. 13C shows a graph indicative of hematological setpoint shifts after diagnosis of hypothyroidism.

[0093] FIG. 13D shows a graph indicative of hematological setpoint shifts after a splenectomy operation.

[0094] FIG. 13E shows a graph indicative of hematological setpoint shifts after a diagnosis of liver disease.

[0095] FIG. 13F shows a graph indicative of hematological setpoint shifts after pregnancy. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0096] FIG. 13G shows a graph indicative of similar effect sizes when using hematological setpoints or a randomly chosen single isolated CBC measurement of the hematological parameter.

[0097] FIG. 13H shows a graph indicative of higher precision of effect size estimates using setpoints based on a ratio of p-values using hematological setpoints or isolated CBC measurements.

[0098] FIG. 14A shows a set of graphs indicative of associations of each hematological setpoint between first degree relatives.

[0099] FIG. 14B shows a set of graphs indicative of associations of each hematological setpoint between partners.

[0100] FIG. 15 A shows a set of graphs indicative of Manhattan plots for a genomewise association study (GWAS) of one or more hematological setpoints for one or more hematological parameters.

[0101] FIG. 15B shows a set of graphs indicative of quintile-quintile plots for each hematological setpoint GWAS.

[0102] FIG. 16A shows a set of graphs indicative of comparison of beta coefficients for SNPs in a GWAS.

[0103] FIG. 16B shows a set of graphs indicative of comparison of p-values for SNPs in a GWAS.

[0104] FIG. 17 shows a set of graphs indicative of a number of significant loci identified when using average CBC measurements compared to the hematological setpoint of the hematological parameter.

[0105] FIG. 18A shows a set of graphs indicative of hematological setpoint mortality associations over 2 years.

[0106] FIG. 18B shows a set of graphs indicative of hematological setpoint mortality associations over 5 years.

[0107] FIG. 18C shows a set of graphs indicative of hematological setpoint mortality' associations over 10 years in a less-restricted cohort of patients.

[0108] FIG. 18D shows a set of graphs indicative of hematological setpoint mortalityassociations over 5 years in a distinct cohort from University of Washington Medical Center (UWMC).

[0109] FIG. 19A shows a set of graphs indicative of associations between hematological setpoints and future diagnosis. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0110] FIG. 19B shows a graph indicative of hazard ratios for future disease diagnosis based on a standard increase in each hematological setpoint adjusted for age and sex.

[0111] FIG. 19C shows a graph indicative of time from presentation with a prediabetic Ale until a diabetic Ale stratified by variation from a hemoglobin concentration setpoint.

[0112] FIG. 19D shows a graph indicating likelihood of an elevated thyroid stimulating hormone (TSH) result stratified by a current value of red blood cell (RBC) size and the RBC size setpoint.

[0113] FIG. 19E shows a set of graphs indicating likelihood of low ferritin stratified by the current hemoglobin (HGB) count and the HGB setpoint.

[0114] FIG. 20 shows a set of graphs indicative associations with hematological setpoints and current measurements of the corresponding hematological parameters with 1 -year mortality rates.

[0115] DETAILED DESCRIPTION

[0116] As described in this disclosure, complete blood count (CBC) parameters in healthy individuals are tightly regulated over long periods of time around a patientspecific hematological setpoint. The patient-specific hematological setpoint can represent a personalized baseline for a patient within a population wide reference interval for each of one or more hematological parameters measured in a CBC test. Described herein are one or more hematological setpoints corresponding to the one or more hematological parameters measured in a CBC test. The one or more hematological setpoints are computed based on data from a multitude of CBC tests for a patient taken over an extended period of time and can be used for enhanced diagnostic and screening purposes for common medical conditions. Upon computation of the one or more hematological setpoints for the one or more hematological parameters, systems and methods described herein determine a pathophysiological state of the patient or a risk for the patient to develop a pathophysiological state with greater sensitivity and specificity. More specifically, the one or more hematological setpoints can provide a personalized reference point within a population wide reference interval to enable improvements in the sensitivity' and specificity for clinical investigations of common medical conditions. Additionally, the one or more hematological setpoints can be used to provide complementary risk Atorney Docket No. 29539-0847WO1 / MGH 2024-526 stratification information for common morbidities and all cause-mortality, and setpoint derived personalized reference intervals can further enable enhanced identification and detection of medical conditions in a patient. Furthermore, the one or more hematological setpoints can be used to identify a stratification group for a patient in a clinical trial and to evaluate the efficacy of a treatment of the patient for a medical condition.

[0117] Examples of hematological parameters include red blood cell (RBC) volume 122 (also referred to as HCT), RBC hemoglobin mass (also referred to as MCH), RBC size (also referred to as MCV), RBC hemoglobin concentration (also referred to as MCHC), hemoglobin (also referred to as HGB), platelet size (also referred to as MPV), platelet count (also referred to as PLT), RBC count (also referred to as RBC). RBC size variation (also referred to as RDW), and white blood cell (WBC) count (also referred to as WBC).

[0118] The systems and methods of the present disclosure can have one or more of the following advantages.

[0119] First, the hematological setpoint of a hematological parameter provides a patient-specific reference point for determining deviations of measurements of the hematological parameter that indicate a pathophysiological state and / or the risk of a pathophysiological state of the patient. That is, the existence of the hematological setpoint enables enhanced specificity’ and sensitivity in screening, diagnosis, and treatment management for a patient. Particularly, the hematological setpoint enables enhanced specificity and sensitivity in large patient wide reference intervals in which the degree to which hematological setpoints can vary significantly among healthy people.

[0120] For example, a first hematological setpoint for a first patient can be a hematological setpoint at the top of the reference interval and a second hematological setpoint for a second patient can be a hematological setpoint at the bottom of the reference interval. On the most recent CBC test, the measurement of the hematological parameter of the first patient can be a value on the low end of the reference interval. In a traditional testing method, the first patient’s value is still within the reference interval and may not draw attention to a potential pathophysiological state. A significant deviation of the measurement of the hematological parameter from the hematological setpoint (e.g., the homeostatic Atorney Docket No. 29539-0847WO1 / MGH 2024-526 baseline for the patient), however, can indicate a pathophysiological state and / or the risk to develop a pathophysiological state for the patient. In this manner, the hematological setpoint can enhance diagnostic and screening processes.

[0121] Furthermore, the enhanced sensitivity for screening processes can enable cheaper and more efficient preliminary' screening for medical conditions in which traditional screening processes are more invasive and / or expensive.

[0122] For example, for hypothyroidism, traditional thyroid stimulating hormone (TSH) tests (TSH level, T3 and T4 levels) are >2x more likely to be elevated depending on how a patient’s current red blood cell (RBC) size compares to the patient's hematological setpoint for RBC size. In this example, the comparison of a recent measurement of the RBC size and the hematological setpoint of the RBC size of a patient can be used as a preliminary screening measure to determine whether further TSH tests are needed, which are significantly more expensive (lOx) than a CBC test. Similarly, comparison of current measurements of RBC to the RBC setpoint, WBC to the WBC setpoint, or PLT to the PLT setpoint, or a multivariate combination of current measurements can help identify patients at increased and decreased risk of hematologic malignancies or bone marrow suppression or failure, helping determine whether subsequent studies like a bone marrow biopsy are necessary for optimal clinical care or are unnecessary and therefore entail unnecessary expense and unnecessary patient discomfort and risk. Additionally, the existence of the hematological setpoint as a personal reference point can enhance analysis of the efficacy of treatment of the patient for a pathophysiological state. By collecting measurements of the hematological parameter of the patient during the treatment of the patient for the pathophysiological state, the new measurement can be compared with the hematological setpoint of the patient to determine whether the treatment is w orking for the patient, or if a new' treatment plan needs to be explored.

[0123] FIG. 1 is a block diagram of an example system 100 that obtains patient data from one or more complete blood count (CBC) tests 110 including measurements for each of one or more hematological parameters 120 and computes, for each of the one or more hematological parameters 120, a hematological setpoint 145 of the hematological parameter. Using the computed hematological setpoint 145, a medical practitioner can determine a recommended treatment plan for the patient 155.

[0124] At a high level, the hematological setpoint 145 for the hematological Atorney Docket No. 29539-0847WO1 / MGH 2024-526 parameter represents a homeostatic baseline for the patient for the hematological parameter. In this manner, for a healthy individual, the hematological setpoint can act as a patient-specific setpoint e.g., within a population wide reference interval or outside a population wide reference interval, where setpoints may vary significantly among healthy patients. That is, the hematological setpoint can act as a more-specific reference point for the patient against which new measurements of a hematological parameter can be compared. The population wide reference interval corresponds to a range of values typically observed in a healthy, demographically similar group of patients. The population wide reference interval is used by medical providers to help interpret a patient’s results, indicating whether they fall within the expected range for their population. For example, a patient can go into a clinic, get a CBC test and the patient and doctor can be provided with the static measurement value of the patient’s CBC test and the population wide reference interval as the CBC test results.

[0125] Instead of a diagnostic process that exclusively uses a population wide reference numeral, a hematological setpoint used in a diagnostic process can enable increased sensitivity and specificity in the analysis of measurements of hematological parameters of a patient. For example, a first patient can have a hematological setpoint on the top end of the reference interval while a second patient can have a hematological setpoint on the lower end of the reference interval. For the first patient, a measure of the hematological parameter that falls in the lower end of the reference interval may be a sign of a pathophysiological state and / or a risk to develop a pathophysiological state where it would not be for the second patient, simply based on the hematological setpoints of the patients. In this manner, the hematological setpoint for the hematological parameter enables a more specific and sensitive identification of particular pathophysiological states for a patient.

[0126] To receive the data representing multiple measurements from multiple CBC tests 110 and compute one or more corresponding hematological setpoints, the system 100 can include one or more computing devices. For example, the system 100 can include an input device, a network, and one or more computers (e.g., one or more local or cloud-based processors). The one or more computers can include an input processing engine that receives the data representing the multiple measurements from the multiple CBC tests 110 and a setpoint calculation engine to calculate the hematological setpoint 145 of the hematological parameter from the data representing Atorney Docket No. 29539-0847WO1 / MGH 2024-526 the multiple measurements from the multiple CBC tests 110. In some implementations, the computer can be a server.

[0127] The system 100 can receive data, from multiple CBC tests 110 taken over a period of time, representing multiple measurements of a hematological parameter of a patient, where each of the multiple measurements is collected from a corresponding CBC test. The system 100 can receive data from multiple CBC tests 110, e.g., 2 or more tests, 3 or more tests. 4 or more tests, 5 or more tests, 10 or more tests, 20 or more tests, 100 or more tests, between 2 and 5 tests, between 2 and 10 tests, between 2 and 20 tests, between 2 and 50 tests, between 2 and 100 tests, etc. In some implementations, the system 100 can receive data from at least 5 CBC tests. The multiple CBC tests 110 can be taken with any interval of time in between the tests such as, for example, 7 or more days, 14 or more days, 30 or more days, 60 or more days, 90 or more days, 180 or more days, between 7 and 14 days, between 14 and 30 days, between 7 and 60 days, between 7 and 90 days, between 7 and 180 days, etc. In some implementations, the interval of time between each test is at least 90 days apart. As a specific example, in some implementations, the multiple CBC tests 110 can be at least 5 tests, each taken at least 90 days apart (e.g., the multiple CBC tests 110 can span multiple years). For example, the multiple CBC tests 110 can be 10 tests, each taken a year apart, spanning 10 years. The subsequent calculated hematological setpoints (e.g., the hematological setpoint 145) can represent homeostatic measurements of a patient over multiple years.

[0128] Each CBC test of the multiple tests 110 can include data representing a measurement of each of one or more hematological parameters 120. The one or more hematological parameters 120 can be any hematological parameter that can be measured using a CBC test. Examples of hematological parameters 120 include HCT 122, MCH 124, MCV 126, MCHC 128, HGB 130, MPV 132, PLT 134, RBC 136, RDW 138, and WBC 140. The one or more hematological parameters 120 can be any combination of the one or more above-described hematological parameters as well as any hematological parameter measured in a CBC test. For example, the one or more hematological parameters 120 can be a single hematological parameter (e.g., RBC volume 122). As another example, the one or more hematological parameters can be three hematological parameters (e.g., RBC volume 122, RBC hemoglobin mass 124, and platelet count 134). As depicted in FIG. 1. for example, the system 100 can Atorney Docket No. 29539-0847WO1 / MGH 2024-526 receive data from three CBC tests: patient CBC A 112, patient CBC B 114, and patient CBC C 116. From each CBC test of the CBC tests 110, the system 100 can receive data representing a measurement for each of one or more hematological parameters 120. More specifically, for patient CBC B 114, the system 100 can receive data representing a measurement for RBC volume 122, RBC hemoglobin mass 124, RBC size 126. RBC hemoglobin concentration 128, hemoglobin 130, platelet size 132, platelet count 134, RBC count 136. RBC size variation 138, and WBC count 140.

[0129] The system 100 can obtain the data from the multiple CBC tests 110 in any appropriate manner. For example, the system 100 can obtain the data from the multiple CBC tests 110 from an electronic data warehouse that can obtain the data, e.g., by accessing medical records of a patient, and transmit the patient data to another device such as a computer across a network. In some implementations, the electronic data warehouse can obtain the data that can be accessed by one or more other input devices such as a computer (e.g.. desktop, laptop, tablet, etc.), a smartphone, or a server. In such instances, the one or more other input devices can access the patient data obtained by the electronic data warehouse and transmit the obtained patient data to a computer via a network. The network can include one or more of a wired Ethernet network, a wired optical network, a wireless WiFi network, a LAN, a WAN, a Bluetooth network, a cellular network, the Internet, or other suitable network, or any combination thereof. In some implementations, the electronic data warehouse and the computer are the same.

[0130] For each of the one or more hematological parameters, the system 100 can receive the data representing multiple measurements of the particular hematological parameter and calculate, using the data, a hematological setpoint 145 of the hematological parameter of the patient. For example, as depicted in FIG. 1, the system 100 can calculate a hematologic setpoint 145 of RBC volume 122 of the patient from multiple measurements of RBC volume 122 collected from the patient from the multiple CBC tests 110 (e.g., patient CBC A 112, patient CBC B 114, and patient CBC C 116). The system 100 can calculate the hematological setpoint 145 of a hematological parameter (e.g., RBC volume 122) using any appropriate techniques. For example, the system 100 can calculate the hematological setpoint 145 using a Gaussian mixture model. In particular, the system 100 can calculate the hematological Atorney Docket No. 29539-0847WO1 / MGH 2024-526 setpoint 145 by computing a mean of a Gaussian component of the multiple measurements identified from applying a Gaussian mixture model to the multiple measurements, the mean corresponding to the hematological setpoint. The use of a Gaussian mixture model to determine the hematological setpoint 145 is described in further detail below with reference to FIG. 2. As another example, the system 100 can calculate the hematological setpoint 145 using a Bayesian method. In particular, the system 100 can compute a mean of a distribution of the multiple measurements identified from applying a Bayesian method to the multiple measurements, the mean corresponding to the hematological setpoint. More specifically, the system 100 can calculate a posterior distribution of the multiple measurement, where the posterior distribution is calculated by blending a prior distribution (the initial belief about the mean) with a likelihood function (which shows how probable the data is for different possible mean values). The mean of the posterior distribution can represent a weighted average determined by the likelihood function. In some implementations, more recent measurements can be weighed more heavily to determine the mean (e.g., the hematological setpoint) as they may be more representative of a homeostatic baseline of the hematological parameter of the patient.

[0131] After the system 100 computes the hematological setpoint 145 for the hematological parameter, a pathophysiological state of the patient and / or risk for the patient to develop the pathophysiological state can be determined. Examples of pathophysiological states include diabetes, kidney disease, thyroid dysfunction, iron deficiency, myeloproliferative neoplasm, atrial fibrillation, myelodysplastic syndrome, osteoporosis, cardiovascular events, and bone fractures. The risk for the patient to develop the pathophysiological state can include determining the risk to develop at least one of: an infection, a malignancy, a disease, anemia, or diabetes. In some implementations, the pathophysiological state of the patient and / or risk for the patient to develop the pathophysiological state can be determined from the combination of two or more hematological setpoints. Combining multiple setpoints will enhance accuracy for most diagnostic inferences because, as FIG 12B shows, many of the setpoints are only modestly correlated, meaning they are not redundant, and adding information on the deviation of one measurement from its setpoint provides information complementary to that derived from the deviation of another measurement from its setpoint. Conditions associated with multilineage perturbations Atorney Docket No. 29539-0847WO1 / MGH 2024-526 will be more accurately detected by combining multiple measurements and setpoints. For instance, hematologic malignancies are often associated with pancytopenia, and considering RBC 136, WBC 140, and PLT 134 measurements relative to their setpoints can enhance the accuracy of this diagnosis. Iron deficiency is associated with perturbations to MCV 126, MCH 124, and RDW 138, and comparison of all three to their setpoints enhances both specificity and sensitivity. Pregnancy involves physiologic changes in RBC 136, WBC 140, and PLT 134 populations, and consideration of all three in relation to their setpoints can provide more precise information available on risk of obstetric complications like preterm delivery, preeclampsia, and preterm premature rupture of membranes.

[0132] The pathophysiological state of the patient and / or risk for the patient to develop the pathophysiological state can be determined in one or more ways.

[0133] As a first example, the pathophysiological state and / or risk to develop the pathophysiological state for the patient can be determined by detennining the hematological setpoint for the patient. More specifically, determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state can include determining, from the hematological setpoint of the hematological parameter of the patient and one or more hematological setpoints of the hematological parameter of one or more other patients in the general population, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. That is, the pathophysiological state and / or risk to develop the pathophysiological state for the patient can be determined by comparing the hematological setpoint of the patient to the hematological setpoints of the population.

[0134] For example the one or more hematological setpoints can be used to identify an absolute increase in the risk of disease of >2.5% for significant fractions of healthy adults (>20%, representing 10s of millions of adults in the US alone) depending on the associated setpoint quintile of the hematological setpoint of the patient compared to the setpoints of the general population. Many current screening programs, for example for cardiovascular disease and cancers including colon and breast, have different recommendations for groups whose absolute risk of 10-year disease-related mortality differs by 2.5%-5%'3. because these risk differences have previously been shown to be large enough to warrant enhanced screening or early lifesty le modifications for disease prevention in the contexts of cancer1’2and cardiovascular Atorney Docket No. 29539-0847WO1 / MGH 2024-526 disease3. As depicted in FIG. 6A, the 10-year survival rate is dependent on the particular setpoint quintile of the hematological setpoint for the hematological parameter. By calculating the one or more hematological setpoints for a patient and comparing to the general population, the diagnostic process can enhance risk screening for patients and lead to further screening or lifesty le modifications to reduce risk.

[0135] For example, determining, using the hematological setpoints, a high risk for heart attack and stroke, as identified by an MCHC 128 setpoint in the bottom quartile, can lead to early lifestyle modifications as well as anti-hypertensives, statins, and anticoagulants delay to reduce future morbidity and risk of death. As another example, determining, using the hematological setpoints, a high risk for diabetes, as identified by a WBC 140 setpoint in the top quartile, can lead to early lifestyle modifications and glucose-lowering medications delay to reduce future morbidity' and risk of death. As another example, determining, using the hematological setpoints, a high risk for kidney failure, as identified by a HCT 122 setpoint in the bottom quartile, can lead to dietary’ modification as well as steroids and other pharmacologic therapy' delay to reduce future morbidity' and risk of death. As another example, determining, using the hematological setpoints, a high risk for atrial fibrillation, as identified by an RDW 138 setpoint in the top quartile, can lead to cardioversion and anticoagulation therapy to reduce risk of future thromboembolism and related morbidity' and mortality risk. As another example, determining, using the hematological setpoints, a high risk for osteoporosis, as identified by an MCV 126 setpoint in the bottom quartile, can lead to lifesty le modification and bisphosphonate therapy delay onset to reduce intensity of future morbidity’.

[0136] As another example, a diagnostic process, using the one or more hematological setpoints, can determine the pathophysiological state of the patient is a state of JAK2 mutation disorder based on the hematological setpoint of the platelet count 134 of the patient compared to the hematological setpoints of the general population. For patients with JAK2 mutation testing, which is commonly ordered to evaluate thrombocytosis, those whose platelet count 134 setpoints were in the highest quintile at least 1 year before testing were 9x more likely to have a mutation identified (4% vs. 35%) as seen in FIG. 71. That is, the hematological setpoint of the platelet count 134 can be used as a pre-screening mechanism to determine whether further Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0137] JAK2 mutation testing should be ordered for the patient. In some implementations, the measurement of the platelet count 134 of the patient and a reference interval for the platelet count 134 of a population do not indicate the state of JAK2 mutation disorder. In this implementation, further JAK2 mutation testing can still be ordered and &JAK2 mutation disorder can be diagnosed.

[0138] As another example, a diagnostic process can, using the one or more hematological setpoints, determine personalized RBC 136 or HGB 130 targets based on a patient’s RBC 136 or HGB 130 setpoint to improve utilization of blood transfusions.

[0139] As another example, a diagnostic process can determine the pathophysiological state and / or risk to develop the pathophysiological state for the patient by determining that the hematological setpoint for the patient is outside a population wide reference interval for the hematological parameter. More specifically, determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state can include determining, from the hematological setpoint of the hematological parameter of the patient and a reference interval for the hematological parameter of a population, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. That is, a diagnostic process can determine the pathophysiological state of the patient and / or the risk of the patient to develop the pathophysiological state by comparing the hematological setpoint of the patient with a population wide reference interval. The diagnostic process can determine the pathophysiological state and / or the risk to develop a pathophysiological state of the patient by determining that the hematological setpoint is outside a population wide reference interval.

[0140] As another example, a diagnostic process can determine the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state by determining that the hematological setpoint is within a population wide reference interval for the hematological parameter and thus, that the patient does have a pathophysiological state. More specifically, determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, based on the multiple measurements of the hematological parameter of the patient being within the reference interval, that the multiple measurements of the hematological parameter do not indicate the Atorney Docket No. 29539-0847WO1 / MGH 2024-526 pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. That is. the diagnostic process can determine that the patient does not have the pathophysiological state and / or the risk to develop the pathophysiological state of the patient by determining that the hematological setpoint is within a population wide reference interval.

[0141] As another example, a diagnostic process can determine the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state by determining that a variability of the measurements of the hematological parameters indicate the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. In particular, the diagnostic process can calculate, using the data representing the multiple measurements of the hematological parameter of the patient, a variability of the multiple measurements for calculating the hematological setpoint of the hematological parameter of the patient. The diagnostic process can calculate the variability of the multiple measurements for calculating the hematological setpoint in one or more ways. For example, the diagnostic process can calculate the variability of the multiple measurements by determining the coefficient of variation (CV) during calculation of the hematological setpoint. As another example, the diagnostic process can calculate the variability' based on a difference between the hematological setpoint of the hematological parameter of the patient and a most recent measurement of the multiple measurements.

[0142] As another example, a diagnostic process can determine one or more hematological setpoints of the hematological parameter and determine from a variability the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. For example, the diagnostic process can determine one or more hematological setpoints of the hematological parameter due to significant variation betw een multiple measurements of the hematological parameter. Based on the variation between the one or more hematological setpoints, the diagnostic process can determine the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. In particular, calculating, using the data representing the multiple measurements of the hematological parameter of the patient, the hematological setpoint of the hematological parameter of the patient can include calculating, using the data representing the multiple measurements of the Atorney Docket No. 29539-0847WO1 / MGH 2024-526 hematological parameter of the patient, one or more hematological setpoints of the hematological parameter of the patient, the one or more hematological setpoints including the hematological setpoint. In this example, determining, from the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes detennining, from a variability of the one or more hematological setpoints of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. A diagnostic process can determine the variability' of the one or more hematological setpoints using any appropriate method. For example, a diagnostic process can determine the difference between the one or more hematological setpoints.

[0143] As another example, determining, from the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, from an interval at least partially defined by the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. The inter al can be a patient-specific interval detennined by the hematological setpoint. In particular, the interval can be computed based on a variability of multiple measurements of the hematological parameter with the hematological setpoint for the hematological parameter of the patient. For example, in an example in which the hematological setpoint is calculated using a Gaussian mixture model, the interval can be calculated using a coefficient of variation (CV) calculated by the Gaussian mixture model. More specifically, the interval can be a CV above and below the hematological setpoint of the patient. The interval can act as a patient specific interval for the hematological setpoint in which the system 100 can determine a pathophysiological state of the patient if a measurement of the patient is not within the patient specific reference interval.

[0144] As another example, determining, from the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state includes determining, from a measurement of the hematological parameter of the patient and Atorney Docket No. 29539-0847WO1 / MGH 2024-526 the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state. That is, a diagnostic process can determine the pathophysiological state of the patient and / or risk of the patient to develop a pathophysiological state based on a measurement of the hematological parameter (e.g., the most recent measurement) and the hematological setpoint of the hematological parameter. In other words, a diagnostic process can determine that the measurement of the hematological parameter is a deviation from the hematological setpoint of the patient for the hematological parameter. From that deviation, a diagnostic process can determine the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

[0145] As a first example, a diagnostic process can determine the pathophysiological state of the patient is a ferritin-deficient or iron-deficient state based on a measurement of the hemoglobin 130 of the patient and the hematological setpoint of the hemoglobin 130 of the patient. Ferritin is the most commonly ordered test to assess iron status to determine whether a patient is in a ferritin-deficient state. As seen in FIG. 7G, among patients with ferritin tests, the likelihood of a low ferritin result was significantly associated with the relationship between the patient’s current measurement of HGB 130 and the hemoglobin 130 setpoint. For instance, among female patients with current anemia (HGB <10g / dL) and a ferritin test, those for whose HGB was no more than 0.5 g / dL below' their setpoint were 7x less likely to have a low ferritin than those HGB was > 2 g / dL below' their setpoint (6% vs. 44%, P <0.001). That is, the deviation of a current measurement of a hematological parameter (e.g., hemoglobin 130) from the hematological setpoint of the hematological parameter can be used as a preliminary' screening and diagnostic tool for common conditions (such as anemia, or a ferritin deficient state). Iron studies (e.g., ferritin, total iron binding capacity, transferrin saturation) often cost 5x as much as another blood count, and this cost can be avoided in some cases using the hematological parameters of the patient as a preliminary screening tool. Based on the results of this pre-screening mechanism, those more expensive, more specific tests can be ordered to further investigate the possibility' of a current ferritin deficient state or assess the response to iron supplementation (e.g., assess the efficacy of a treatment). In this case, usage of a hematological setpoint can enable a more accurate estimation of pre-test Atorney Docket No. 29539-0847WO1 / MGH 2024-526 probabilities of positive results, enhancing test utilization and clinical decision accuracy. Furthermore, treatment plans for a ferritin deficient state can be personalized, adjusting dose by comparing a patient’s current RBC 136 or HGB 130 measurement to the patient’s RBC 136 or HGB 130 setpoint. In some implementations, the measurement of the RBC 136 or HGB 130 of the patient and a reference interval for the RBC 136 or HGB 130 of a population do not indicate the ferritin-deficient state. In this implementation, further tests, diagnosis, and treatments can still be ordered / implemented.

[0146] As another example, a diagnostic process can determine the pathophysiological state of the patient is a state of elevated thyroid stimulating hormone (TSH) based on a measurement of the RBC volume 122 of the patient and the hematological setpoint of the RBC volume 122 of the patient. Patients with a recent measurement indicating an elevated RBC volume 122 relative to their hematological setpoints for RBC volume 122 were more >2x more likely to have high TSH, consistent with the hypothyroidism implied by high TSH and its known association with macrocytosis as seen in FIG. 7H. That is. the deviation of a current measurement of a hematological parameter (e.g., RBC volume 122) from the hematological setpoint of the hematological parameter can be used as a preliminary screening and diagnostic tool for common conditions (such as hypothyroidism or an overactive thyroid gland). TSH tests (e.g., TSH level, T3 and T4 levels) can cost lOx more than a blood count test, and this cost can be avoided by using the patient’s hematological setpoint of RBC volume 122 as a pre-screening tool. Based on this prescreening mechanism, those more expensive, more specific TSH tests can be performed to further investigate the possibility of thyroid dysfunction. In some implementations, the measurement of the mean RBC volume 122 of the patient and a reference interval for the mean RBC volume 122 of a population do not indicate the state of elevated TSH. In this implementation, further tests, diagnosis, and treatments can still be ordered / implemented for thyroid dysfunction.

[0147] As another example, for kidney disease, patients with a first-time estimated glomerular filtration rate (eGFR) in the early-stage kidney disease range had a 3x increased risk of progression to late-stage kidney disease depending on how the current RBC volume 122 compares to the patient’s RBC volume 122 setpoint. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0148] As another example, the system 100 can, using the one or more hematological setpoints, determine a risk for progression of diabetes. More specifically, patients with a first-time HbAlc in the prediabetic range had a 50% increased risk of progression to diabetes depending on how the current RBC hemoglobin concentration 128 compares to the patient’s RBC hemoglobin concentration 128 setpoint. A related calculation identifies patients whose HbAlc falls just below the prediabetic range but for whom a comparison of CBC measurements to CBC setpoints determines that the HbAlc is artifactually decreased and that the patient’s true level of glycemia is high enough to constitute prediabetes, and there patients are at increased risk of progression to diabetes, with a risk that is equivalent to or greater than that for patients with HbAl c in the prediabetic range.

[0149] As another example, relationships between current CBC indices and a patient’s CBC setpoints are correlated with prostate cancer risk in patients with borderline positive PSA test results. Using this relationship, the system 100 can optimize surveillance frequency and biopsy decisions.

[0150] As another example, relationships between a pregnant patient’s current CBC indices to the corresponding CBC setpoint can identify pregnant patients at > 2x risk of pre-eclampsia, preterm delivery', and other adverse outcomes. The system 100 can utilize this early detection of elevated risk to guide preventive treatments like low- dose aspirin, calcium supplementation, and increased surveillance and frequency of antenatal visits.

[0151] As another example, comparison of a current WBC count 140 of a patient with the WBC count 140 setpoint of the patient can enhance the sensitivity and specificity of the use of the WBC count 140 to detect chronic latent infections or acute infections. This comparison can guide enhanced surveillance, follow-up microbiological testing, or utilization of empiric antibiotic treatment. As another example, WBC counts 140 can be compared to the patient’s WBC 140 setpoint to provide early indication to halt therapy and avoid agranulocytosis or other blood cell dyscrasias during screening for drug toxicity in patients treated with clozapine (or chlorpromazine, thioridazine, olanzapine, quetiapine, carbamazepine, and valproate), which can be a side effect.

[0152] As another example, comparison of recent CBC parameters to their hematological setpoints can be used to decide whether to proceed with invasive Atorney Docket No. 29539-0847WO1 / MGH 2024-526 diagnostic testing (e.g., bone marrow biopsy) for patients suspected of hematologic malignancy, enhancing diagnosis and treatment response.

[0153] As another example, a diagnostic process can determine the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state by determining information indicative of a degree of morbidity of the patient. That is, a diagnostic process can determine, using the hematological setpoint of the hematological parameter, information indicative of a degree of morbidity of the patient. The information indicative of the degree of morbidity of the patient can be any information indicative of the degree of morbidity7such as, for example, information indicative of the presence of an infection, information indicative of the presence of malignancy, information indicative of the presence of anemia, information indicative of the presence of diabetes, or information indicative of a dysfunction.

[0154] After determining the pathophysiological state of the patient, the system 100 can utilize the hematological setpoint and the determined pathophysiological state for one or more clinical uses.

[0155] For example, a medical provider can determine a treatment plan for the patient 155 using the hematological setpoint and the determined pathophysiological state and / or determined risk for the pathophysiological state. As seen in the abovedescribed examples, the system 100. in conjunction with a medical provider, can provide information indicative of a recommended treatment regimen for the patient, where the recommended treatment regimen includes diagnostic screening on the patient specific to the pathophysiological state. For example, for a ferritin-deficient state, an elevated TSH level, etc., more expensive, and specific tests can be ordered to further investigate pathophysiological states based on the pre-screening results using the hematological setpoints.

[0156] As another example, the system 100 can identify a stratification group for the patient for a clinical trial for a treatment for the pathophysiological state.

[0157] As another example, the system 100 can determine information indicative of an efficacy of treatment of the patient for the pathophysiological state. As seen in the above described examples, the system 100 can determine the efficacy of treatments of the patient based on deviations between a hematological setpoint for the hematological parameter and a measurement of the hematological parameter taken Atorney Docket No. 29539-0847WO1 / MGH 2024-526 during treatment. Furthermore, the hematological setpoint of the hematological parameter can be used to personalize doses of treatment for the patient.

[0158] In some implementations, the hematological setpoint of the patient for the hematological parameter can be updated with a new measurement of the hematological parameter from a new CBC test. In particular, the system 100 can receive data representing a measurement of a hematological parameter of a patient, the measurement being collected from a corresponding complete blood count (CBC) test performed on a blood sample of the patient taken at a time. The system 100 can calculate, using the hematological setpoint of the hematological parameter and the data representing the measurement of the hematological parameter of the patient, an updated hematological setpoint of the hematological parameter of the patient and determine, from the updated hematological setpoint for the hematological parameter of the patient, a pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state. That is, the system 100 can calculate an updated hematological setpoint of the patient for the hematological parameter by including a new measurement of the hematological parameter. The system 100 can then utilize the updated hematological setpoint to determine a pathophysiological state of the patient and / or risk for the patient to develop a pathophysiological state using any of the above-described methods.

[0159] FIG. 2A shows a graph indicative of a computation of a hematological setpoint 243 using a Gaussian mixture model.

[0160] As described above with reference to FIG. 1, in some implementations, the hematological setpoint 243 of a hematological parameter 223 can be computed using a Gaussian mixture model on multiple measurements of the hematological parameter 223 in multiple CBC tests.

[0161] As depicted in FIG. 2 A, the system 100 can compute the hematological setpoint 243 of the hematological parameter 223 of a WBC count (e.g., the WBC hematological parameter 140 of FIG. 1) using 19 CBC tests taken over the course of 15+ years, including CBC C 216 and CBC I 218.

[0162] To calculate the hematological setpoint 243, the system 100 can apply a Gaussian mixture model with one or more components to the set of multiple measurements of the hematological parameter 223 and take the mean of the largest component, e.g., the Gaussian component 247. More specifically, the system 100 can Atorney Docket No. 29539-0847WO1 / MGH 2024-526 apply a Gaussian mixture model to the set of multiple measurements of the hematological parameter 223 using an iterative process called Expectation- Maximization (EM). The EM algorithm starts with an initial guess for the parameters of each Gaussian component — its mean, variance, and mixing weight. In the Expectation (E-step), the system 100 can calculate the probability' that each data point belongs to each of the Gaussian components. In the Maximization (M-step), the system 100 can, using the probabilities, re-estimate the parameters of each Gaussian component, e.g., Gaussian component 247. For instance, the system 100 can calculate the new mean of a component as a weighted average of all data points, with the weights being the probabilities computed in the E-step. The system 100 can repeat the two-step process until the model parameters no longer change significantly, meaning the Gaussian mixture model has converged to the best fit for the data. To get the hematological setpoint 243, the system 100 can take the overall weighted average of the means of all components, or in the example depicted in FIG. 2, the system 100 can determine the hematological parameter by taking the mean of the component that has the largest mixing weight, e.g.. the largest component, which represents the most dominant cluster in the set of multiple measurements of the hematological parameter 223.

[0163] As depicted in FIG. 2A, the multiple measurements of the hematological parameter 223 can include one or more temporary disruptions or multiple measurements that deviate from the hematological setpoint. The one or more temporary disruptions can be outliers from a temporary illness, or temporary condition (e.g., pregnancy) or any other disruption that may temporarily cause marker alterations over months.

[0164] FIG. 2B shows a graph indicative of a hematological setpoint 261 as a personal reference point in a population wide reference interval 252. As depicted in FIG. 2B and described above with reference to FIG. 1, a hematological setpoint (e.g., hematological setpoint for patient A 261) for a hematological parameter (e.g., platelet count) can act as a patient-specific reference point. In this example, the hematological setpoint for patient A 261 can act as a more specific reference point to which to compare new results (e.g., new result 263) within a population wide reference interval 252. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0165] FIG. 2C further depicts the usage of hematological setpoints as a patientspecific reference point for enhanced comparisons between new measurements (or results) and patient-specific hematological setpoints. The graph of FIG. 2C includes three patient-specific hematological setpoints: hematological setpoint for patient A 261, hematological setpoint for patient B 271, and hematological setpoint for patient C 281. As depicted in FIG. 2C, the hematological setpoints (e.g., hematological setpoints 261, 271. and 281) can vary significantly within a population wide reference interval. For example, hematological setpoint for patient A 261 is 350 k / uL for platelet count, while the hematological setpoint for patient C is 181k / UL for platelet count. This significant variability can lead to very different interpretations of new measurement (or results) of platelet count from a new CBC test. While each of the new results of the three patients (e.g., new result of patient A 261, new result patient B 273, and new result patient C 283) are within the population wide reference interval 252, the patient-specific reference intervals allow enhanced specificity7and sensitivity to deviations of new measurements from the hematological setpoints.

[0166] For example, the new result of patient A 263 deviates significantly from the hematological setpoint of patient A 261 and may indicate a pathophysiological state of the patient A. However, the new result of patient A 263 does not deviate that far from the hematological setpoint of patient C 281 and if the result was a measurement of the platelet count of patient C, may not indicate a pathophysiological state of the patient C. Similarly, the new result of patient C 283 deviates significantly from the hematological setpoint of patient C 281 and may indicate a pathophysiological state and / or risk to develop a pathophysiological state of patient C. However, the new result of patient C 283 does not deviate that far from the hematological setpoint of patient A 261 and if the result was a measurement of the platelet count of patient A, may not indicate a pathophysiological state of the patient C. That is, the healthy baselines (e.g., hematological setpoints) of patients differ greatly within a population wide reference interval and the patient-specific hematological setpoints described herein enable enhanced screening and diagnostic processes with more specific and sensitive reference points to which to compare new measurements for a patient. So, while all of the new results for the patient fall within the reference interval, the patient-specific hematological setpoints offer greater insight as to whether a new Atorney Docket No. 29539-0847WO1 / MGH 2024-526 measurement of a hematological parameter is a true deviation from a patient’s homeostatic baseline.

[0167] To give more context to the deviations, in some implementations, the hematological setpoints of the patients can not only act as a reference point for new measurements of the hematological parameter, but the hematological setpoint can influence the determination of a patient specific reference interval for the hematological parameter.

[0168] FIGS. 2D and 2E depict graphs that include a personalized reference interval for a patient based on the hematological setpoint for the hematological parameter. As described above with reference to FIG. 1, a patient-specific reference interval can be determined based on at least a hematological setpoint of the patient. In some implementations, the patient specific reference interval can be determined by the hematological setpoint and a coefficient of variation (CV) of the hematological setpoint. For example, as depicted in FIGS. 2D and 2E, the patient specific reference intervals can be determined based on the hematological setpoint and the CV of the setpoint (e.g., 10.2% for the hematological setpoint of patient A 261, and 8.0% for the hematological setpoint of patient B 271). As a specific example, the patient specific reference interv als can be determined as any value within 2 CV in either direction of the hematological setpoint.

[0169] First, FIG. 2D shows a graph indicative of a new result 263 outside of a patient-A specific reference interval 265 based on a hematological setpoint of patient A 261 for platelet count. As seen in FIG. 2D, the new result of patient A 263 falls outside of the patient A specific reference interval 265, while falling within the population wide reference interval 252. So, while in a traditional method where the value may not be noticed as a deviation from a patient’s “normal” baseline, it is very clear that there is a deviation of the new result of patient A 263 from the hematological setpoint of patient A 261 and that the value falls outside a patient A specific reference interval 265. In this manner, this deviation can be caught, and further testing, diagnosis and treatment can be ordered to further investigate the deviation. As opposed to traditional methods in which only the population wide reference interv al 252 is consulted, the usage of the patient-specific reference interval (e.g., patient A specific reference interval 265) enables more sensitive screening and Atorney Docket No. 29539-0847WO1 / MGH 2024-526 diagnosis and can beter inform a medical provider as to the current state of the patient.

[0170] FIG. 2E shows a graph indicative of a new result 273 within a patient-B specific reference interval 275 on a hematological setpoint of patient B 271 for platelet count. As seen in FIG. 2E, the new result of patient B 273 falls within the patient B specific reference interval 275. while falling within the population wide reference interval 252 as well. In this manner, while a medical provider can observe that the new result 273 is within a population wide reference interval, the medical provider can further observe that the new result of patient B is within a patient-B specific reference interval 275 of the patient and thus, within a nonnal homeostatic baseline of the patient. That is. just as a value outside of a patient specific reference interval can indicate a pathophysiological state and / or risk of a pathophysiological state of the patient, a value inside a patient-specific reference interval (as depicted in FIG. 2E) does not indicate a pathophysiological state and / or risk of the pathophysiological state for the patient.

[0171] At a high level, the hematological setpoint and / or a patient-specific reference interval based on the hematological setpoint can enable more specific and sensitive diagnostic and screening processes as well as provide more significant insight as to the current state of a patient in comparison to a traditional, population wide reference interval.

[0172] FIG. 3 is a flowchart of an example process for determining a pathophysiological state of a patient based on one or more hematological setpoints.

[0173] The process 300 will be described as being performed by a system of one or more computers programmed appropriately in accordance with this specification. For example, the system 100 of FIG. 1 can perform at least a portion of the example process. In some implementations, various steps of the process 300 can be run in parallel, in combination, in loops, or in any order.

[0174] The below steps describe the process of determining a hematological setpoint of a hematological parameter, where a measurement of the hematological parameter is collected in a corresponding CBC test. The below process can be repeated for each hematological parameter of a set of hematological parameters.

[0175] The system receives data representing multiple measurements of a hematological parameter of a patient, each of the multiple measurements being Atorney Docket No. 29539-0847WO1 / MGH 2024-526 collected from a corresponding complete blood count (CBC) test of multiple CBC tests performed on blood samples of the patient taken from the patient at one or more times (302). As described above with reference to FIG. 1, the system can receive data from multiple CBC tests, where each CBC test can provide a measurement of a hematological parameter of the patient. The multiple CBC tests can be performed on blood samples of the patient taken at one or more different times. As a specific example, the multiple CBC tests can be at least five CBC tests, with each CBC test taken at least 90 days from the previous test.

[0176] The hematological parameter can be any hematological parameter whose measurements can be determined from a CBC test. Examples of hematological parameters can include a measure of red blood cell (RBC) volume, a measure of RBC hemoglobin mass, a measure of RBC size, a measure of RBC hemoglobin concentration, a measure of hemoglobin, a measure of platelet content, a measure of RBC count, a measure of RBC size variation, or white blood cell (WBC) count. The set of hematological parameters can include any of the above-mentioned hematological parameters as well as any other measurements determined from a CBC test on a blood sample. In some implementations, the hematological parameters can further include a measure of platelet size.

[0177] The system can calculate, using the data representing the multiple measurements of the hematological parameter of the patient, a hematological setpoint of the hematological parameter of the patient (304). As described above with reference to FIGS. 1 and 2, the system can calculate the hematological setpoint using one or more methods such as, for example, a Gaussian mixture model or a Bayesian method. As a specific example, the system can calculate the hematological setpoint of the hematological parameter of the patient by applying a Gaussian mixture model to the multiple measurements and computing the mean of the largest component of the Gaussian mixture model, where the mean corresponds to the hematological setpoint. As another example, the system can calculate the hematological setpoint of the hematological parameter of the patient by applying a Bayesian method to the multiple measurements and computing a mean of the distribution of the measurements, where the mean corresponds to the hematological setpoint.

[0178] The system can determine, from the hematological setpoint for the hematological parameter of the patient, a pathophysiological state of the patient Atorney Docket No. 29539-0847WO1 / MGH 2024-526 and / or a risk for the patient to develop the pathophysiological state (306). As described above with reference to FIG. 1, the system can determine the pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state in one or more ways using the hematological setpoint for the hematological parameter. At a high level, the system can determine the pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state by determining the hematological setpoint is outside a population wide reference interval, determining the hematological setpoint is inside a population wide reference interval, determining a variability' of the multiple measurements of the hematological parameter, determining a variability' of one or more hematological setpoints for the hematological parameter, determining a variation from a patient specific reference interval, etc..

[0179] This specification uses the term ‘’configured” in connection with systems and computer program components. For a system of one or more computers to be configured to perform particular operations or actions means that the system has installed on it software, firmware, hardware, or a combination of them that in operation cause the system to perform the operations or actions. For one or more computer programs to be configured to perform particular operations or actions means that the one or more programs include instructions that, when executed by data processing apparatus, cause the apparatus to perform the operations or actions.

[0180] Embodiments of the patient matter and the functional operations described in this specification can be implemented in digital electronic circuitry, in tangibly- embodied computer software or firmware, in computer hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them. Embodiments of the patient matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible non transitory' storage medium for execution by, or to control the operation of, data processing apparatus. The computer storage medium can be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or a combination of one or more of them. Alternatively or in addition, the program instructions can be encoded on an artificially generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal that is generated to Atorney Docket No. 29539-0847WO1 / MGH 2024-526 encode information for transmission to suitable receiver apparatus for execution by a data processing apparatus.

[0181] The term '‘data processing apparatus" refers to data processing hardware and encompasses all kinds of apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or computers. The apparatus can also be, or further include, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit). The apparatus can optionally include, in addition to hardware, code that creates an execution environment for computer programs, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.

[0182] A computer program, which may also be referred to or described as a program, software, a software application, an app, a module, a software module, a script, or code, can be written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages; and it can be deployed in any form, including as a standalone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data, e.g., one or more scripts stored in a markup language document, in a single file dedicated to the program in question, or in multiple coordinated files, e.g., files that store one or more modules, sub programs, or portions of code. A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a data communication network.

[0183] In this specification, the term “database” is used broadly to refer to any collection of data: the data does not need to be structured in any particular way, or structured at all, and it can be stored on storage devices in one or more locations. Thus, for example, the index database can include multiple collections of data, each of which may be organized and accessed differently.

[0184] Similarly, in this specification the term “engine” is used broadly to refer to a software-based system, subsystem, or process that is programmed to perform one or more specific functions. Generally, an engine will be implemented as one or more software modules or components, installed on one or more computers in one or more Atorney Docket No. 29539-0847WO1 / MGH 2024-526 locations. In some cases, one or more computers will be dedicated to a particular engine; in other cases, multiple engines can be installed and running on the same computer or computers.

[0185] The processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows can also be performed by special purpose logic circuitry, e.g., an FPGA or an ASIC, or by a combination of special purpose logic circuitry and one or more programmed computers.

[0186] Computers suitable for the execution of a computer program can be based on general or special purpose microprocessors or both, or any other kind of central processing unit. Generally, a central processing unit will receive instructions and data from a read only memory or a random access memon or both. The essential elements of a computer are a central processing unit for performing or executing instructions and one or more memory devices for storing instructions and data. The central processing unit and the memory can be supplemented by, or incorporated in, special purpose logic circuitry. Generally, a computer wi 11 also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto optical disks, or optical disks. However, a computer need not have such devices. Moreover, a computer can be embedded in another device, e.g., a mobile telephone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a Global Positioning System (GPS) receiver, or a portable storage device, e.g., a universal serial bus (USB) flash drive, to name just a few.

[0187] Computer readable media suitable for storing computer program instructions and data include all forms of non-volatile memory', media, and memoir devices, including by way of example semiconductor memory' devices, e.g., EPROM, EEPROM, and flash memory' devices; magnetic disks, e.g., internal hard disks or removable disks; magneto optical disks; and CD ROM and DVD-ROM disks.

[0188] To provide for interaction with a user, embodiments of the patient matter described in this specification can be implemented on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid cry stal display) monitor, for displaying information to the user and a keyboard and a pointing device, e.g.. a mouse Atorney Docket No. 29539-0847WO1 / MGH 2024-526 or a trackball, by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback, e.g., visual feedback, auditory feedback, or tactile feedback; and input from the user can be received in any form, including acoustic, speech, or tactile input. In addition, a computer can interact with a user by sending documents to and receiving documents from a device that is used by the user; for example, by sending web pages to a web browser on a user’s device in response to requests received from the web browser. Also, a computer can interact with a user by sending text messages or other forms of message to a personal device, e.g., a smartphone that is running a messaging application, and receiving responsive messages from the user in return.

[0189] Data processing apparatus for implementing machine learning models can also include, for example, special-purpose hardware accelerator units for processing common and compute-intensive parts of machine learning training or production, i.e., inference, workloads.

[0190] Machine learning models can be implemented and deployed using a machine learning framework, e.g., a TensorFlow framework, a Microsoft Cognitive Toolkit framework, an Apache Singa framework, or an Apache MXNet framework.

[0191] Embodiments of the patient matter described in this specification can be implemented in a computing system that includes a back end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front end component, e.g., a client computer having a graphical user interface, a web browser, or an app through which a user can interact with an implementation of the patient matter described in this specification, or any combination of one or more such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network (LAN) and a wide area network (WAN), e.g., the Internet.

[0192] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each Atorney Docket No. 29539-0847WO1 / MGH 2024-526 other. In some embodiments, a server transmits data, e.g., an HTML page, to a user device, e.g., for purposes of displaying data to and receiving user input from a user interacting with the device, which acts as a client. Data generated at the user device, e.g., a result of the user interaction, can be received at the server from the device.

[0193] While this specification contains many specific implementation details, these should not be construed as limitations on the scope of any invention or on the scope of what may be claimed, but rather as descriptions of features that may be specific to particular embodiments of particular inventions. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features may be described above as acting in certain combinations and even initially be claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination.

[0194] Similarly, while operations are depicted in the drawings and recited in the claims in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Moreover, the separation of various system modules and components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0195] EXAMPLES

[0196] The following examples further describe examples related to using hematological setpoints for determining a pathophysiological state of a patient and other aspects of the diagnostic processes and systems described in this disclosure and do not limit the scope of the invention described in the claims.

[0197] Methods Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0198] The following methods were used in the Examples below.

[0199] Patient data collection

[0200] Three cohorts were defined (A, B, and C) by analyzing complete blood count (CBC) data for adult Mass General Brigham (MGB) outpatients during three overlapping periods: 2002-2021. 2002-2006, 2017-2021. Patients were included if they met the following criteria: 5+ isolated CBCs (outpatient and >90d apart from other blood tests) during the period, no >48h inpatient stay during the period, and alive at the endpoint. Patients in cohort A were excluded from cohort B-C, and patients in cohort B were excluded from cohort C, such that the cohorts contained no overlap. After exclusions, the sizes of the cohorts A-C were 12.407, 14.371, and 20,062, respectively. Patient demographics, blood counts, procedures, medications, and diagnoses were collected using the MGB Research Patient Data Registry7(RPDR) and Electronic Data Warehouse (EDW). Patient deaths were collected from RPDR and EDW, which are updated frequently using the United States National Death Index and US Social Security Death Master File to capture deaths external to the hospital. Diagnosis data was converted from ICD9 and ICD10 codes to disease phenotypes using PheCodes. The CBCs included ten parameters: hematocrit (HCT), hemoglobin (HGB), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), mean platelet volume (MPV), platelet count (PLT), red cell count (RBC), red cell distribution width (RDW), and white cell count (WBC). MPV values were not routinely reported in the medical record before 2015 and are not reported for cohort B. CBCs were run on a variety of hematology analyzers, reflecting changes in MGB laboratory equipment over time. Most instruments have very low analytic variation, though even this little variation may introduce a small, fluctuating bias to marker results over time. The MGB system comprises multiple medical centers across Massachusetts, with each having slight differences in CBC reference intervals. For consistency, the reference intervals currently in use at Massachusetts General Hospital (MGH) were used. MGH reference intervals for each marker are given in FIG. 8.

[0201] Each cohort was evaluated for health, with all three showing similar or better 1-year mortality' rates (0.3%, 0.9%, 0.8% for A-C) than a similarly aged general US population (1.1% for 55-60-year olds in 2019). Analysis of disease phenotypes in cohort A showed no evidence for significant numbers of diagnoses that would be Atorney Docket No. 29539-0847WO1 / MGH 2024-526 unexpected in a general healthy adult population of similar age - with diagnoses pain, hypertension, and hyperlipidemia being among the most prevalent.

[0202] Key results were validated using a cohort derived from the University of Washington Medical Center (UWMC) in Seattle, USA. Laboratory test, demographic, and diagnosis data for this cohort were derived from a similar electronic data warehouse at UWMC. Mortality data was similarly linked to the US Social Security Death Master File. Unless otherwise specified, UWMC was processed in the same way as MGB data.

[0203] For various specific analyses, additional patient cohorts were identified, with equivalent data pulled from the electronic health records. Details of these individual cohorts are defined in the relevant sections below.

[0204] Setpoint calculation

[0205] The setpoint is defined as the mean of a patient’s regulated healthy biological marker distribution and was estimated by fitting a Gaussian mixture model with up to 3 components to each patient’s set of ‘isolated” CBCs (defined above) and taking the mean of the largest component. Multi-component models were fit to help isolate the dominant (presumed physiologic) marker value distribution from F, as in FIG. 2). Optimal component number was chosen based on the Akaike information criteria score. However, to ensure the dominant distribution was captured, a multi-component model was used only if one of its components was significantly larger than the others, comprising either 70% (2 components) or 45% (3 components). The coefficient of variation (CV - the standard deviation expressed as a percentage of the mean) was calculated from the component variance. From inclusion criteria each patient had at least 5 CBCs available for setpoint calculation - however, if a patient had more measurements, the entire set was used. In cohorts A-C respectively, 81.7%, 22.0%, and 30.5% of patients had at least 10 isolated CBCs available for setpoint estimation.

[0206] Inter- and intra-patient variation in blood count markers was compared to short-term intra-patient marker variation rates reported in the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Biological Variation database8. Estimates of inter- and intra-patient marker variation in FIG. 4A-4C were calculated based on each patient’s set of isolated CBCs, without use of models, in order to reflect overall variation including outlier values. These results were adjusted for age- associated drift using linear regression over the entire study cohort. A mixed model Atorney Docket No. 29539-0847WO1 / MGH 2024-526 regression was not used, as coefficient estimates were similar when using all patient CBCs compared to using a single CBC from each patient. More precise estimation of age and gender effects may be valuable for future research. Rates of age-associated drift were also very low for all markers (FIG. 9). Unless otherwise specified, other results did not use age-adjustment, to enable unbiased comparison to reference intervals. Age and sex effects were estimated using all markers from the cohort and were not corrected for repeat observations from patients using a mixed effects model. Effects were also estimated using single markers and were of similar size.

[0207] Heritability analysis

[0208] Setpoint heritability was estimated using patient relationship data in the electronic health record (EHR), similar to prior reports. All patients in cohorts A-C with first-degree familial (parent-child or sibling) or partner (spouse or life partner) relationships also in one of the cohorts were retained (N: 439, 440 pairs respectively). Familial relationships were assumed biological unless otherwise noted (e.g., stepfather, etc.). Heritability was estimated as (pfamiiiai - ppartner) / / ?, where pfamiiiai and Ppaitner are the correlations between familial and partner setpoints and h is the genetic strength (0.5 for first-degree relatives), with ppartner acting as a control for environmental effects. All setpoints were age- and sex-corrected via linear regression prior to correlation estimation. Heritability estimates were compared to literature estimates from five studies: 2 twin studies, 1 pedigree study, 1 multi-generation study, and 1 large-scale EHR study. Estimates were also calculated using a randomly chosen isolated CBC from each patient. Corresponding heritability estimates are given in FIG. 10

[0209] Genome-wide association studies

[0210] To assess setpoint-genome associations, data from the MGB Biobank - a biorepository of genotyped samples from consented MGB patients - was used. All Biobank patients who had genotype data, at least 5 isolated CBCs, and who were alive as of Apr-01-2023 were included (N: 32,093). Patient genotyping was performed by the MGB Biobank, using the Illumina Multi-Ethnic Genoty ping (MEGA), and Expanded MEGA (MEGAEX) array covering 1,416,020 and 1,741,376 SNPs. respectively. Results were imputed by the MGB Biobank team using the Minimac3, on the Michigan Imputation Server, with the Haplotype Reference Consortium (rl . 1 Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0211] 2016) panel. Samples with high SNP missingness (first pass filter 0.2, second pass filter 0.02). high heterozygosity, sex discrepancy, or high relatedness (kinship coefficient >0.2) were excluded. Analysis of genetic principal components in the subset of MGB Biobank patients for whom CBC setpoints could be estimated was consistent with a large group (> 85%) with inferred ancestry' consistent with predominantly European based on comparison of multidimensional scaling plots to the 1000 Genomes reference panel. Sub-cohorts with other predominant ancestry based on the same reference panel were much smaller (< 1500) and underpowered for most single-cohort GWAS. Analysis of correlations betw een genetic principal components and setpoints was consistent with the presence of significant non-genetic effects associated with inferred ancestry, increasing the risk of confounding in a combined mixed model analysis. Future analysis in larger and more diverse cohorts with extensive CBC histories will be an important extension of this study.

[0212] SNPs were excluded if they had low minor allele frequency (<0.05) or poor imputation (R2< 0.3) or were outside Hardy-Weinberg equilibrium (p<le-10). GWAS analysis was performed using a linear model with age, sex, and 10 genetic principal components as covariates. Setpoints w ere calculated using the same procedure as described previously. Analysis was also performed using a randomly chosen isolated CBC from each patient. Loci were identified using the plink clump feature with significance thresholds of 5e-8 and 1 for primary and secondary SNPs, a linkage disequilibrium threshold of R2> 0.2, and a clumping region of 250kbp. A locus w as deemed novel if there w ere no reported significant associations to any of the 9 CBC markers for any SNP within 250kbp of the locus’s primary' SNP, as determined by queries to the GWAS catalog. All novel loci were subsequently manually reviewed to confirm novelty. SNP heritability estimates were derived from the GWAS summary^ statistics using the sum-hers function in LDAK version 5.2, with the LDAK-Thin tagging file (as supplied on the LDAK website) and excluding any predictors which explained more than 1% of phenotype variation. All other analysis was implemented using bcftools. plink, and plink2, with a significance threshold of 5e-8 unless otherwise specified. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0213] Polygenic score analysis

[0214] A polygenic score (PGS) was calculated for each setpoint using results of the GWAS analysis. After quality control, the total cohort (n=25,254) was split randomly into 80% for score development, and 20% for validation. A new GWAS was performed for each setpoint using the first 80% of the cohort. The SNPs from this GWAS were clumped (with an R2threshold of 0. 1, and region size of 250kb), and any clumps with a lead SNP p-value < le-5 were retained. Each lead SNP was then used to calculate the PGS via the score function in plink. PGS associations with the phenotype were evaluated in the held-out 20% test cohort. PGS quintiles were calculated separately for males and females for HGB, HCT, and RBC given their known strong sex differences. PGS associations with mortality were not evaluated due to insufficient power from both the smaller cohort size and due to short follow-up time leading to low event rates.

[0215] Setpoint outcome associations

[0216] Setpoint associations with mortality were estimated in cohort B, using Kaplan- Meier curve analysis and Cox proportional -hazards models. Patients were censored 10 years after the study close date (01-01-2017). Setpoint hazard ratios were normalized by the mean intra-patient CV, while CV hazard ratios were normalized to a 1 % change. Joint associations of setpoint and current marker values with outcomes were investigated by comparing setpoints in cohort B to the lowest (HCT, HGB, MCH, MCV, MCHC, PLT, RBC) or highest (RDW, WBC) outpatient test value in the year after setpoint estimation (01-01-2007 to 01-01-2008). Mortality and likelihood of a future abnormal test result were measured over the year following this (01-01-2008 to 01-01-2009). Setpoint associations with future disease development were measured in the same cohort, using ICD codes, after exclusion of all patients with a diagnosis prior to the study end date (01-01-2007). Diseases chosen for analysis were atrial fibrillation, chronic kidney disease (at any stage), type 2 diabetes, myelodysplastic syndrome (MDS), osteoporosis, and major adverse cardiovascular events (MACE; defined as the composite of stroke, myocardial infarction, or heart failure). MDS diagnoses were validated by comparison with a manually-curated MGB MDS patient database. Diagnosis hazard ratios were computed using this data, with patients censored at time of death or last date of data collection, and using age (on 01-01- 2007) and sex as covariates. The following ICD codes were used for each diagnosis: Atorney Docket No. 29539-0847WO1 / MGH 2024-526 atrial fibrillation (ICD9: 427.3, ICD10: 148); chronic kidney disease (ICD9: 585, ICD10: N18); type 2 diabetes (ICD9: 250.x0, 25O.x2, ICD10: El l). MACE (ICD9: 410, 428, 431, 432, 434, ICD10: 121, 150, 161, 162, 163); MDS (ICD9: 238.72-238.75, ICD10: D46), osteoporosis (ICD9: 733.0, 733.1, ICD10: M80, M81). Each ICD code includes all of its sub-codes, and 250. xO refers to codes 250.00, 250.10,... 250.90, and equivalently for 250. 2.

[0217] Key results were validated using a cohort from UWMC. Setpoints were estimated in all UWMC patients who had 5+ isolated CBCs between 2014-2018 (N: 13,864. Mortality and diagnosis rates were measured in the five years following (2019-2023), with patients censored at time of death or 01-01-2024. Similarly to MGB, patients with the relevant diagnosis prior to 01-01-2019 were excluded.

[0218] Test positivity’ analysis

[0219] Setpoint associations with future positivity of ferritin (FER), thyroid stimulating honnone (TSH), estimated glomerular filtration rate (eGFR), hemoglobin Ale, and JAK2 tests were estimated using available retrospective data for each test. Results were included if a patient had a CBC measured in the 48 hours prior to the test and had 5+ isolated CBCs during the study period prior to the test. If multiple tests were available, the first test for each patient was considered. For eGFR and Ale, to ensure patients were undergoing regular screening (for chronic kidney disease and diabetes respectively) the patients were limited to those presenting with their first mildly abnormal value (eGFR: 60-90 mL / min / 1.73m2; Ale: 5.7-6.4%) who had at least 2 prior normal values (eGFR > 90 mL / min / 1.73m2; Ale < 5.7%). Likelihood of future disease progression (eGFR < 60 mL / min / 1.73m2; Ale > 6.5%) stratified by setpoint and presenting CBC (at time of the mild abnormality) was measured. For JAK2, to decrease risk of bias, setpoints were calculated using only CBCs from at least 1 year prior to the JAK2 test date.

[0220] Results for FER, TSH, and Ale were validated using cohorts from UWMC. For FER and TSH, all patients with a test between 01-01-2023 and 07-01-2024 were selected, and the first test with an estimable pre-test setpoint (5+ isolated CBCs) for each patient was retained (n: 6285, 7510 for FER and TSH respectively). Rates of low ferritin (<10ng / mL) and high TSH (>5 mIU / L) were estimated after stratification bypresenting marker and pre-test setpoint. For Ale analyses, patients were limited to those presenting with their first pre-diabetic Ale, with estimable CBC setpoints prior Atorney Docket No. 29539-0847WO1 / MGH 2024-526 to Ale, and a CBC at time of this Ale (n: 2173). eGFR results could not be validated, as UWMC coded eGFR results between 60-90 asL‘>60”. JAK2 results could not be validated due to limited genetic test data availability at UWMC.

[0221] Statistical analysis

[0222] Statistical significance of data was calculated using t-tests for continuous variables and chi-square for categorical variables unless otherwise specified. Survival analysis was performed using Kaplan-Meier curves, with statistical significance calculated using log-rank tests. For time-to-event analysis, patients were censored at the date of last EHR data collection (Jan-01-2023 unless otherwise noted). Thresholds for significance for all analyses were set at p = 0.05 unless otherwise specified. All non-genomic data analysis was performed using MATLAB 2023b and Python 2.7. Genomic data was analyzed using bcftools 1.21 via the Windows Linux Subsystem, and using plink and plink2.

[0223] Example 1. CBC indices are tightly regulated

[0224] A cohort of 12,407 healthy patients who had at least 5 CBCs measured under apparently stable conditions over a 20-year period was studied. Intra-patient variation in 10 CBC indices were used: RBC, WBC, PLT, hematocrit (HCT), hemoglobin (HGB), mean red cell volume (MCV), mean red cell hemoglobin content (MCH), mean red cell hemoglobin concentration (MCHC), mean platelet volume (MPV), and red cell distribution width (RDW) (FIG. 2). Standard CBC reference intervals are based on inter-patient variation in CBC indices, which can be quantified by the ratio of the standard deviation to the mean (coefficient of variation or CV).The inter-patient CV for CBC indices ranged from 5-30% (FIG. 4A). Intra-patient CVs were much lower over the 20-year period, ranging from 2-15%, representing only 30-70% of the variation reflected by the equivalent inter-patient CV (see FIG. 4B and Setpoint calculation in Methods). Twenty -year intra-patient CVs remained closer to previously reported intra-patient CVs determined over much shorter periods of a few weeks or months8. The magnitude of the CV for each setpoint did not vary systematically with sex, self-reported race or ethnicity7, or age (FIG. 4C, FIG. 11A- 11B), suggesting that this tight regulation is a feature of normal physiology. The CVs were also stable across a wide range of values for each setpoint (FIG. 11C). consistent with the hypothesis that regulatory processes are equally efficient for a range of setpoints. An Atorney Docket No. 29539-0847WO1 / MGH 2024-526 individual’s setpoints could typically be accurately inferred from as few as 4 CBCs that met inclusion criteria (FIG. 11D). Overall, each patient's CBC indices appear to be regulated to stay within a sub-interval of the larger inter-patient reference interval for decades, suggesting that these hematologic setpoints represent a well-defined state of health (FIG. 4D).

[0225] Example 2. Patient-specific healthy state

[0226] Healthy individuals were distinguishable by comparing their setpoints. For instance, the three healthy individuals in FIG. 4D can be distinguished for most of the 20-year period based on just their PLT count. The CBC setpoints as a group defined a hematologic state for each patient in the cohort, with MPV excluded because it was not consistently available throughout the 20-year period. The state defined by the remaining 9 setpoints for a typical patient was distinguishable from that of 98% of others in the cohort. In other words, for the typical patient, only 2% of the remaining cohort had setpoints that all fell within 2*CV of that patient's setpoints. This high degree of patient-specificity when considering all 9 setpoints as a group implies only modest intra-patient correlations between at least some pairs of setpoints (FIG. 12A). Some indices are correlated by definition (e.g., HCT, HGB, and RBC; MCV and RDW), and their setpoints were correlated as expected, as were setpoints for indices that are known to be co-regulated (e.g., MCH and MCV). The WBC setpoint was more strongly associated with the relative lymphocyte count setpoint than the relative neutrophil count setpoint, implying that healthy adults with higher WBC setpoints generally have higher circulating lymphocyte counts. The MPV setpoint was positively associated with the immature platelet fraction and platelet distribution width setpoints, consistent with the hypothesized negative association between PLT size and PLT age. The PLT setpoint w as modestly negatively correlated with RBC- related setpoints except RDW and modestly positively associated with the WBC setpoint, suggesting some degree of co-regulation between the RBC, WBC. and PLT populations in this cohort of healthy adults. The RDW setpoint was also associated with the setpoints for immature reticulocyte fraction and fragmented and nucleated red cell counts in patients for w hom several of those measurements were available, suggesting chronic RDW elevation may in some cases reflect sustained dysregulated RBC production or the presence of schistocytes19. These complex correlations between CBC setpoints suggest the presence of persistent differences among healthy Atorney Docket No. 29539-0847WO1 / MGH 2024-526 adults in the underlying processes of cell production, trafficking, or clearance, whether those differences be acquired or genetic.

[0227] Example 3. Acquired origins of setpoint variation

[0228] The heterogeneity' of hematologic setpoints among healthy adults raises the question of origin, and the extent to which acquired or genetic factors may contribute to variation. Differences in disease history or exposures could potentially cause hematologic setpoints to diverge. For instance, chronic disease is often associated with low-level inflammation which might raise the WBC setpoint, along with other inflammatory markers. In a small prospective study and retrospective analysis, there was no evidence for large associations between CBC setpoints and setpoints for non- hematologic markers, other than iron status marker setpoints (FIG. 12B) which were correlated with red cell setpoints as expected and likely reflect the high prevalence of chronic and sub-clinical iron deficiency anemia. CBC setpoints in this cohort showed very small age-related changes, less than the equivalent of 1% of the CV over 10 years (FIG.9), consistent with prior studies. For instance, the median WBC setpoint was 6.32 (x!03 / pL), and the average age-related increase would raise it only to 6.35 over 10 years. Setpoint CVs were not significantly age-dependent (FIG. 11A-11B). Average setpoints shifted following normal physiologic changes of pregnancy and menopause, after development of the chronic diseases of hypothyroidism or hepatitis, and after splenectomy (FIG. 13A-13H). These patterns are consistent with prior single-CBC studies, and in each case, setpoints provided improved precision for estimation of effect sizes compared to single CBC measurements (FIG. 13A-13H). These results demonstrate that setpoints are modifiable, but the magnitudes of demonstrated change are modest, leaving the majority of setpoint variation in the healthy adult population unexplained.

[0229] Example 4. Genetic origins of setpoint variation

[0230] Given the limited available evidence that acquired factors explain a significant fraction of setpoint variation, the genetic basis of setpoint differences by analyzing hcritability and performing GWAS w as investigated. From electronic health record (EHR)-derived familial relationships in all three study cohorts, setpoints were found to be highly correlated between first-degree relatives but not between partners, who are likely to share an environment (FIG. 5A-5B, FIG. 14A-14B). Except for MCHC, Atorney Docket No. 29539-0847WO1 / MGH 2024-526 all CBC setpoints showed strong heritability (FIG. 5C; h2range: 0.37-0.52). Estimating heritability from EHR-derived relationship data can be imprecise, but estimates using setpoints in these cohorts were consistent with previous reports using single-CBC results (FIG. 14C).

[0231] A GWAS was performed using both setpoints and single CBC measurements. SNP -based heritability estimates were consistently higher when using setpoints (FIG. 5D). GWAS of the 9 CBC setpoints identified 397 associated loci. Eight loci (lead SNPs: rs60528951 , rs2047265, rsl 2522573, rs6997857, rsl 0021975, rsl0043270, rsl 17912622, rs869243453) appeared to be novel after comparison to the GWAS catalog28and to results derived from much larger studies that relied on single measurements (FIG. 5E, FIG. 15A). Setpoint GWAS produced comparable effect size estimates to single-CBC GWAS but with increased precision and thus significance (FIG. 5F-5G, FIG. 16A-16B), leading to an average 3.6-fold increase in SNPs with statistically significant (p < 5e-8) associations (FIG. 5H). Setpoints also provided improved discover)’ over 2, 4, and 8-CBC averages, demonstrating the benefits of identifying the underlying physiologic distnbution and systematically excluding outlier values (FIG. 17). A polygenic score (PGS) reflected variation in each CBC setpoint (FIG. 51) but explained only a small fraction of the inter-patient variance (median R2= 0.10). These results suggest that setpoints are partially genetically determined but likely also reflect substantial acquired effects, most of which have not yet been defined.

[0232] Example 5. Setpoint associations with mortality

[0233] In view of the showing that hematologic setpoints are stable for decades and vary among healthy adults in ways that are not currently utilized in clinical settings, it was investigated whether setpoint differences are associated with clinical outcomes. Associations betw een setpoints and all-cause mortality' in a cohort of healthy patients who had 15 years of follow-up after setpoint estimation were studied. This cohort was restricted to patients whose setpoints all fell within the population-wide reference intervals, and most of this setpoint variation might therefore have been expected to represent largely equivalent patient-specific states of health with little association with clinical outcomes. In contrast to this expectation, FIG. 6A shows that 10-year mortality risk varied significantly as a function of setpoint. Most setpoints showed roughly monotonic relationships with mortality, while HCT and HGB were associated Atorney Docket No. 29539-0847WO1 / MGH 2024-526 with minimal mortality risk in the middle and increases toward both extremes. Associations remained significant after adjusting for age and sex (FIG. 6B) over different periods of time (FIG. 18A-18B) and under replication in both an independent cohort at main study hospitals and a distinct cohort from an independent academic medical center (FIG. 18C-18D). Increased setpoint variability as quantified by CV was also associated with higher mortality risk, and this pattern persisted after adjusting for age, sex, and setpoint value (FIG. 6C) for all markers except MCHC. The association of some of the extreme setpoint values (e.g., high WBC, high RDW, low HGB, and low7PLT) with elevated risk is consistent with existing epidemiologic studies of CBCs, but the continuous variation of risk for many setpoints throughout the reference intervals was unexpected. These results overall imply that significant fractions (>20%) of the healthy adult population with CBC results within populationwide reference intervals have absolute mortality risk increases that exceed 3% and up to 5%.

[0234] Example 6. Setpoint associations with disease risk

[0235] Given setpoint associations with all-cause mortality, associations between setpoints and onset of major illness and morbidity in the same patient cohort (Cohort B, FIG. 7A) were analyzed. An MCHC setpoint in the lowest quartile was associated with an increased rate of major adverse cardiovascular events (heart attack, stroke, heart failure), higher WBC setpoint with increased risk of ty pe 2 diabetes, higher MCV setpoint with osteoporosis, lower HCT setpoint with chronic kidney disease, higher RDW setpoint with atrial fibrillation, and low er RBC setpoint with myelodysplastic syndrome (MDS). Associations remained significant after adjustment for age and sex (FIG. 7B) and were validated at an independent medical center (FIG. 19A-19E). The directions of some of these associations are consistent with prior observations for individual CBC measurements. For instance, lower MCHC has been reported to be associated with heart attack outcomes, RDW with atrial fibrillation and many other conditions, MCV with hip fracture, WBC with diabetes development, and RBC with MDS risk. These prior studies largely relied on single CBC measurements and typically found smaller effect sizes even though they were often not restricted to healthy patients. The larger effect sizes identified for setpoints suggest that they provide a stronger and more precise pathophysiologic signal. These results overall imply that significant fractions (-25%) of the healthy adult population may have Atorney Docket No. 29539-0847WO1 / MGH 2024-526 absolute increases in risk of major diseases that exceed 2% or even 5%, levels that warrant enhanced screening in some contexts including cancer and cardiovascular disease.

[0236] Example 7. Setpoints for more accurate prognosis

[0237] Setpoints thus provided prognostic information on their own. They also served as patient-specific reference points that could be used in combination with subsequent CBC measurements to provide additional prognostic information. The WBC setpoint was used to calculate between 2002-2006 for individuals in Cohort B as a benchmark for subsequent outpatient WBC results in 2007. All patients had setpoints that fell within the population-wide reference interval. Patients with WBC setpoints in the higher part of the reference interval who then had anew WBC in 2007 in the lower part of the reference interval had a significantly elevated one-year mortality rate (6.9%) compared to the rest of those with CBCs in 2007, and the same was true for patients with a WBC setpoint in the lower part of the reference interval and a subsequent WBC in the higher part (4.5%) (FIG. 7C). Significant stratification was found for other setpoints used as reference for new CBC results (FIG. 20). Setpoints can also define personalized reference intervals, extending 2*CV in both directions from the setpoint. The width of these setpoint-derived reference intervals is analogous to the reference change value (RCV) which integrates expected instrument error and intra-patient biological variation into a confidence interval2,9, but by adding knowledge of the setpoint to the RCV, the personalized reference interval can be anchored within the population-wide range and used to help interpret all test results for that patient not just the next one over a short period of time as is typically discussed in the context of RCVs. Ten-year mortality rates were calculated for patients in Cohort B as a function of whether their current CBC result fell outside their setpoint-derived reference interval or not, and the mortality rate stratification was compared to that estimated using the population-wide reference intervals. Hazard ratios were higher for the setpoint-based intervals for all CBC indices except MCHC (FIG. 7D)

[0238] Example 8. Setpoints for more accurate diagnosis

[0239] Whether interpretation of new CBCs in the context of a patient’s established setpoints might enhance diagnostic accuracy for other tests was investigated as well. Atorney Docket No. 29539-0847WO1 / MGH 2024-526

[0240] It was hypothesized that comparing a new CBC to the patient’s setpoints provides insight into subtle changes in the patient's hematologic state. Screening and diagnostic scenarios where the presence or absence of subtle hematologic perturbations might alter positive and negative predictive values was analyzed.

[0241] Kidney disease is often associated with erythropoietic dysfunction, and whether setpoint-based interpretation of a CBC might help detect the presence or absence of mild erythropoietic dysfunction and enable more accurate interpretation of the most common screening test for kidney disease, serum creatinine and the derived estimated glomerular filtration rate (eGFR) was investigated. Data from patients at study hospitals who had a first-time eGFR in the early-stage kidney disease range (60 - 90mL / min / 1.73m2) and a concurrent CBC was analyzed, as well as sufficient CBC history for setpoint estimation regardless of whether their setpoints fell within the population-wide reference intervals (see Test positivity analysis in Methods). It was found that patients with current HCTs below their HCT setpoints at the time of a first eGFR in the early-stage kidney disease range were significantly more likely to progress to late-stage kidney disease (FIG. 7F). The HCT below its setpoint may reflect mild erythropoietic dysfunction that would often be expected in the setting of underlying kidney disease.

[0242] HbAlc is the most commonly-used screening testing for prediabetes and diabetes, but false positives are common, in some cases due to transient alterations in RBC kinetics2that may be difficult to identify. Whether combining setpoints with a new CBC might improve the accuracy of HbAlc interpretation was investigated by studying patients at study hospitals who had a first HbAlc screening test in the prediabetic range (5.7%-6.4%) and a concurrent CBC, as well as sufficient CBC history for setpoint estimation regardless of whether their setpoints fell within the population-wide reference intervals. It was found that those patients whose MCHCs were below their MCHC setpoints were less likely to be diagnosed with diabetes in the future (FIG. 7E), suggesting that the prediabetic HbAlc may have been a false positive. The reduced MCHC compared to its setpoint may reflect subtle alterations in RBC kinetics that would increase the likelihood of a glycemia-independent increase in HbAlc, as has been shown to occur in the setting of iron deficiency.

[0243] Setpoints alone and in combination with current CBCs also provided useful context when interpreting diagnostic tests commonly ordered to investigate conditions Atorney Docket No. 29539-0847WO1 / MGH 2024-526 associated with hematologic abnormalities. Setpoints and CBCs in patients at study hospitals who had one of the following tests along with a concurrent CBC and enough CBC history for setpoint estimation were studied, regardless of whether their setpoints fell within the population-wide reference intervals: ferritin, thyroid stimulating hormone (TSH), and JAK2 genetic analysis. Ferritin is the most commonly ordered test to assess iron status, and Fig. 4g shows that among patients with ferritin tests, the likelihood of a low ferritin result was significantly associated with the relationship between the patient’s current HGB and HGB setpoint, independent of the actual HGB level. For instance, among female patients with current anemia (HGB < lOg / dL) and a ferritin test, those whose HGB was no more than 0.5 g / dL below' their setpoint were 7x less likely to have a low ferritin than those HGB was > 2 g / dL below their setpoint (6% vs. 44%, P < 0.001). Similar patterns were seen for MCV in patients with TSH testing. Patients with MCV elevated relative to their setpoints were more likely to have high TSH, consistent with the hy pothyroidism implied by high TSH and its known association with macrocytosis (FIG. 7H). For patients with JAK2 mutation testing, which is commonly ordered to evaluate thrombocytosis, those whose PLT setpoints were in the highest quintile at least 1 year before testing were 9x more likely to have a mutation identified (4% vs. 35%) (FIG. 71). Understanding the clinical context of these test orders is important because, for instance, ferritin may have been ordered to investigate the possibility of a current iron deficiency or alternatively to assess the response to iron supplementation. In either case, consideration of setpoints may enable more accurate estimates of pre-test probabilities of positive results, which could enhance general test utilization and clinical decision accuracy.

[0244] OTHER EMBODIMENTS

[0245] It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

Atorney Docket No. 29539-0847WO1 / MGH 2024-526WHAT IS CLAIMED IS:

1. A method, comprising: receiving data representing a plurality of measurements of a hematological parameter of a patient, each of the plurality' of measurements being collected from a corresponding complete blood count (CBC) test of a plurality of CBC tests performed on blood samples of the patient taken from the patient at a plurality of times; calculating, using the data representing the plurality of measurements of the hematological parameter of the patient, a hematological setpoint of the hematological parameter of the patient; and determining, from the hematological setpoint for the hematological parameter of the patient, a pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state.

2. The method of claim 1. wherein: determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: determining, from the hematological setpoint of the hematological parameter of the patient and a reference interval for the hematological parameter of a population, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

3. The method of claim 1, wherein: determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: determining, based on the plurality of measurements of the hematological parameter of the patient being within a reference interval, that the plurality of measurements of the hematological parameter do not indicate the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

4. The method of claim 1, wherein: determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises:Atorney Docket No. 29539-0847WO1 / MGH 2024-526 determining, from the hematological setpoint of the hematological parameter of the patient and one or more hematological setpoints of a population, that the plurality of measurements of the hematological parameter do not indicate the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

5. The method of claim 1. wherein: the pathophysiological state comprises: diabetes, kidney disease, thyroid dysfunction, iron deficiency, myeloproliferative neoplasm and other malignancy, atrial fibrillation, myelodysplastic syndrome, and osteoporosis, cardiovascular events, and bone fractures.

6. The method of claim 1, wherein: a quantity of the plurality of CBC tests is no less than five.

7. The method of claim 1, wherein: the plurality of times of the plurality of CBC tests are at least 90 days apart from each other.

8. The method of claim 1. wherein: calculating, using the data representing the plurality of measurements of the hematological parameter of the patient, the hematological setpoint of the hematological parameter of the patient comprises: calculating, using the data representing the plurality of measurements of the hematological parameter of the patient, a variability of the plurality of measurements for calculating the hematological setpoint of the hematological parameter of the patient; and determining, from the variability of the plurality of measurements for the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.Atorney Docket No. 29539-0847WO1 / MGH 2024-5269. The method of claim 8, wherein calculating the variability of the plurality of measurements comprises: calculating the variability based on a difference between the hematological setpoint of the hematological parameter of the patient and a most recent measurement of the plurality7of measurements.

10. The method of claim 1. wherein: calculating, using the data representing the plurality of measurements of the hematological parameter of the patient, the hematological setpoint of the hematological parameter of the patient comprises: calculating, using the data representing the plurality of measurements of the hematological parameter of the patient, a plurality7of hematological setpoints of the hematological parameter of the patient, the plurality of hematological setpoints comprising the hematological setpoint; and determining, from the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: determining, from a variability7of the plurality7of hematological setpoints of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

11. The method of claim 1, wherein: determining, from the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: determining, from an interval at least partially defined by the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.Atorney Docket No. 29539-0847WO1 / MGH 2024-52612. The method of claim 11, comprising: computing the interval based on a variability of the plurality of measurements of the hematological parameter with the hematological setpoint for the hematological parameter of the patient.

13. The method of claim 1, wherein: determining, from the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: determining, from a measurement of the hematological parameter of the patient and the hematological setpoint of the hematological parameter of the patient, the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state.

14. The method of claim 13, wherein: the pathophysiological state is a ferritin-deficient or iron-deficient state: and the hematological parameter is a red blood cell (RBC) hemoglobin content.

15. The method of claim 14, wherein: the measurement of the hemoglobin of the patient and a reference interval for the hemoglobin of a population do not indicate the ferritin-deficient or iron-deficient state.

16. The method of claim 13, wherein: the pathophysiological state is a state of elevated thyroid stimulating hormone (TSH); and the hematological parameter is a mean RBC volume.

17. The method of claim 16, wherein: the measurement of the mean RBC volume of the patient and a reference interval for the mean RBC volume of a population do not indicate the state of elevated TSH.Atorney Docket No. 29539-0847WO1 / MGH 2024-52618. The method of claim 13, wherein: the pathophysiological state is a state of JAK2 mutation disorder; and the hematological parameter is a platelet count.

19. The method of claim 18, wherein: the measurement of the platelet count of the patient and a reference interval for the platelet count of a population do not indicate the state of JAK2 mutation disorder.

20. The method of claim 1, wherein: determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: determining information indicative of a degree of morbidity of the patient, the information indicative of the degree of morbidity comprising at least one of: information indicative of a presence of an infection, information indicative of a presence of malignancy, information indicative of a presence of anemia, information indicative of a presence of diabetes, or information indicative of a dysfunction.

21. The method of claim 1. wherein: determining the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: determining the risk for the patient to develop the pathophysiological state, the risk for the patient to develop the pathophysiological state being a risk to develop at least one of: an infection, a malignancy, a disease, anemia, or diabetes.

22. The method of claim 1, comprising: providing infonnation indicative of a recommended treatment regimen for the patient, wherein the recommended treatment regimen comprises diagnostic screening on the patient specific to the pathophysiological state.Atorney Docket No. 29539-0847WO1 / MGH 2024-52623. The method of claim 1, comprising: identifying, from the hematological setpoint for the hematological parameter of the patient, a stratification group for the patient for a clinical trial for a treatment for the pathophysiological state.

24. The method of claim 20, wherein: the information indicative of the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: information indicative of a degree of morbidity of the patient, the information indicative of the degree of morbidity comprising at least one of: information indicative of a presence of an infection, information indicative of a presence of malignancy, information indicative of a presence of anemia, information indicative of a presence of diabetes, or information indicative of a dysfunction.

25. The method of claim 20, wherein: the information indicative of the pathophysiological state of the patient and / or the risk for the patient to develop the pathophysiological state comprises: information indicative of an efficacy of treatment of the patient for the pathophysiological state.

26. The method of claim 1, wherein: the hematological parameter comprises: a measure of red blood cell (RBC) volume, a measure of RBC hemoglobin mass, a measure of RBC size, a measure of RBC hemoglobin concentration, a measure of platelet size, a measure of hemoglobin, a measure of platelet content, a measure of RBC count, a measure of RBC size variation, or white blood cell (WBC) count.

27. The method of claim 1. wherein calculating, using the data representing the plurality of measurements of the hematological parameter of the patient, the hematological setpoint comprises:Atorney Docket No. 29539-0847WO1 / MGH 2024-526 computing a mean of a Gaussian component of the plurality of measurements identified from applying a Gaussian mixture model to the plurality of measurements, the mean corresponding to the hematological setpoint.

28. The method of claim 1, wherein calculating, using the data representing the plurality of measurements of the hematological parameter of the patient, the hematological setpoint: computing a mean of a distribution of plurality of measurements identified from applying a Bayesian method to the plurality of measurements, the mean corresponding to the hematological setpoint.

29. The method of claim 1, further comprising: receiving data representing a measurement of a hematological parameter of a patient, the measurements being collected from a corresponding complete blood count (CBC) test performed on a blood sample of the patient taken at a time; calculating, using the hematological setpoint of the hematological parameter and the data representing the measurement of the hematological parameter of the patient, an updated hematological setpoint of the hematological parameter of the patient; and determining, from the updated hematological setpoint for the hematological parameter of the patient, a pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state.

30. A system, comprising: one or more processors; and memory storing instructions which, when executed by the one or more processors, cause the one or more processors to perform operations comprising: receiving data representing a plurality of measurements of a hematological parameter of a patient, each of the plurality of measurements being collected from a corresponding complete blood count (CBC) test of a plurality of CBC tests performed on blood samples of the patient taken from the patient at a plurality of times;Atorney Docket No. 29539-0847WO1 / MGH 2024-526 calculating, using the data representing the plurality of measurements of the hematological parameter of the patient, a hematological setpoint of the hematological parameter of the patient; and determining, from the hematological setpoint for the hematological parameter of the patient, a pathophysiological state of the patient and / or a risk for the patient to develop the pathophysiological state.

Images