Multi-modal risk stratification of minimum residual disease

Abstract

Methods and systems for determining a risk level of minimal residual disease (MRD) in a subject treated for cancer are described. A first type of genomic variants and a second type of genomic variants are identified from pre-treatment samples of the subject. A post treatment sample is obtained from the subject and each of the first type of genomic variants and second type of genomic variants are evaluated from the post treatment sample to make first and second determinations regarding MRD. A risk level is stratified into one of at least three risk levels of MRD in the subject based on the first determination of MRD and the second determination of MRD.

Description

MULTI-MODAL RISK STRATIFICATION OF MINIMUM RESIDUAL DISEASECROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of priority of U.S. Provisional Application Serial No. 63 / 700.040. filed September 27, 2024, the disclosure of which is incorporated herein by reference.TECHNICAL FIELD

[0002] Embodiments of the disclosure relate generally to disease detection, and more specifically to multi-modal detection and risk stratification of minimum residual disease (MRD) in cancer patients.BACKGROUND

[0003] Minimal residual disease (MRD) is a small number of cells, e.g., cancer cells, which are left in the human body after a patient receives treatment. Unfortunately, these cells may come back and cause relapse in the patient. MRD detection may be used as an approach for predicting tumor recurrence and identifying patients who may benefit from a second line of treatment, e.g. adjuvant therapy following surgery or radiotherapy. Moreover, two or more consecutive negative MRD results in patients under maintenance therapy may prompt a decision to de-escalate or stop the therapy - which reduces biological side effects of potentially unnecessary and toxic treatments for the patient and mitigates the financial burden on the patient and the healthcare system. Studies have shown the ability to detect MRD from followup cell-free DNA samples in cancer patients based on different genomic variants, but suitable methods have not been combined in a way that improves predictive performance and patient risk stratification.SUMMARY

[0004] Embodiments of the disclosure relate generally to risk stratification and detection of disease, and more specifically to methods, systems, and articles of manufacture, including computer program products, for multi-modal detection and risk stratification of minimal residual disease (MRD).

[0005] In some embodiments, a method for determining a risk level of minimal residual disease in a subject treated for cancer is provided. The method may include identifying a firsttype of reporter genomic variants indicative of minimal residual disease based on a plurality of sequencing reads obtained from a pre-treatment sample of the subject, identifying a second type of reporter genomic variants indicative of minimal residual disease based on the plurality of sequencing reads obtained from the pre-treatment sample of the subject, obtaining a plurality of sequencing reads from a post-treatment sample of the subject, evaluating the first type of reporter genomic variants in the sequencing reads from the post-treatment sample of the subject to make a first determination of MRD in the subject, evaluating the second type of reporter genomic variants from the post-treatment sample of the subject to make a second determination of MRD in the subject, and determining one of at least three risk levels of MRD in the subject based on the first determination of MRD and the second determination of MRD,

[0006] In some embodiments, the first type of reporter genomic variants includes single nucleotide variants (SNVs) indicative of MRD, and the second type of reporter genomic variants includes linked somatic variants (LSVs) indicative of MRD. For example, the first determination of MRD in the subject may include comparing a quantity of SNV reporters in the post-treatment sample to a baseline level of SNV reporters, and the second determination of MRD in the subject includes comparing a quantity of LSV reporters in the post-treatment sample to a baseline level of LSV reporters. In some instances, the at least three risk levels include a first risk level of MRD determined based on a MRD negative determination for the second MRD determination, a second risk level of MRD greater than the first risk level, the second risk level being determined based on an MRD positive determination for the first MRD determination and an MRD negative determination for the second MRD determination, and a third risk level of MRD greater than the second risk level, the third risk level being determined based on an MRD positive determination for each of the first and second MRD determinations.

[0007] In some embodiments, the method further includes determining a suggested therapy based on the determined one of at least three risk levels of MRD.

[0008] In some embodiments, a multimodal detection system for identifying the risk of MRD in a subject treated for cancer is provided. The system may include a memory and a processor coupled to the memory and configured to identify a first type of reporter genomic variants indicative of minimal residual disease based on a plurality of sequencing reads obtained from a pre-treatment sample of the subject, identify a second type of reporter genomic variants indicative of minimal residual disease based on the plurality of sequencing readsobtained from the pre-treatment sample of the subject, obtain a plurality of sequencing reads from a post-treatment sample of the subject, evaluate the first type of reporter genomic variants in the sequencing reads from the post-treatment sample of the subject to make a first determination of MRD in the subject, evaluate the second type of reporter genomic variants from the post-treatment sample of the subject to make a second determination of MRD in the subject, and determine one of at least three risk levels of MRD in the subject based on the first determination of MRD and the second determination of MRD.BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The accompanying drawings, which are incorporated in and constitute a part of this specification, show certain aspects of the subject matter disclosed herein and, together with the description, help explain some of the principles associated with the disclosed implementations. In the drawings:

[0010] FIG. 1 depicts aspects of genomic variants used for detecting MRD, consistent with implementations of the current subject mater;

[0011] FIG. 2 depicts a discrete unimodal (single modality) statistical model used to demonstrate the association of patient outcome (progression-free survival) with MRD status based on single nucleotide variants, consistent with implementations of the current subject mater;

[0012] FIG. 3 depicts a discrete unimodal (single modality) statistical model used to demonstrate the association of patient outcome (progression-free survival) with MRD status based on linked somatic variants, consistent with implementations of the current subject matter;

[0013] FIG. 4 depicts a combination of the discrete models of FIG. 2 and 3, consistent with implementations of the current subject matter;

[0014] FIG. 5 depicts a combination of the discrete models ofFIG. 2 and 3, consistent with implementations of tire current subject mater;

[0015] FIG. 6 depicts a combination of the discrete models ofFIG. 2 and 3, consistent with implementations of the current subject matter;

[0016] FIG. 7 depicts a trichotomized risk stratification model of MRD consistent with implementations of the current subject mater;

[0017] FIG. 8 depicts a flowchart illustrating an example of a method for detecting minimal residual disease, consistent with implementations of the current subject mater; and

[0018] FIG. 9 depicts a block diagram illustrating an example of a computing system consistent with implementations of the curreni subject matter.DETAILED DESCRIPTION

[0019] Aspects of the current subject matter are directed to detection of minimal residual disease (MRD) from cell free DNA (cfDNA) samples. Some aspects of the current subject matter provide for detecting MRD based on both single nucleotide variants (SNVs) indicative of MRD and linked somatic variants (LSVs) indicative of MRD.

[0020] Methods for detecting MRD by identifying SNA's in cfDNA samples have been described previously, for example, in U.S. Provisional Patent Application Nos. 63 / 560,514, filed on March 1, 2024 and U.S. Provisional Patent Application No. 63 / 645,537 filed on May 10, 2024, the contents of which are incorporated herein in their entirety for all purposes. Methods for detecting MRD by identifying USA's in cfDNA samples have also been described in various publications. While the single modality-based techniques relying solely on the identification of either SNVs or LSVs alone have advanced detection of MRD, they have not reached suitable levels of sensitivity, provide discordant (and thus confusing) results, and have not been optimized to provide increased risk stratification to identify patients with varying risks of relapse. While approaches have been published that simply evaluate multiple existing single modality techniques independently and calling a result MRD positive if any of the techniques yield a positive MRD result, such approaches do not balance sensitivity and specificity (improve one at the expense of severe deficit to the other) and therefore do not improve predictive performance with respect to patient outcomes. Moreover, by insisting on generating a binary MRD status (MRD+ vs. MRD-), current methods fail to reflect the heterogeneity of post-treatment patient physiological states. This oversimplification may lead to making suboptimal clinical decisions, and will fall short in supporting personalized treatment selection.

[0021] To overcome the limitations of current MRD detection techniques, the present disclosure is directed to a multi-modal approach to MRD detection, lire techniques described herein allow for both a highly sensitive and specific detection of MRD and a risk stratification. By using identification of both SNA’s and LSV s and combining models based thereon in a nonadditive manner in which modalities are assigned different weights, the techniques of thepresent disclosure achieve better sensitivity and specificity, and risk stratification (with potentially more than two risk strata if indicated by the data), which can better guide personalized treatment decisions.

[0022] FIG. 1 depicts aspects of genomic variants used for detecting MRD, consistent with implementations of the current subject matter. Illustrated in diagrams 100 and 105 are simplified examples of genomic variants, including single nucleotide variants (SNVs) 110 and linked somatic variants (LSVs) 130. As can be seen in diagram 100, SNVs such as SNV 110 represent variants in which a single nucleotide in a subject sequence is different from that of a reference sequence. In contrast, LSVs such as those depicted by LSVs 130 represent multiple SNVs 115, 120, and 125 present within a certain portion of a sequence, i.e., multiple SNVs within a portion of a sequence that are “linked” to each other. As used herein, LSVs are defined as more than one SNV occurring on the same read-pair in close proximity to each other, in particular, within 175 base pairs of one another. LSVs may include a set of two SNVs (doublets) or three SNVs (triplets). Tracking LSVs are useful because it facilitates tire separation of true vanants from experimental artifacts, since it would be highly unlikely for an oxidative damage during sample preparation, sequencing error or other artifactual base change to repeatedly produce a given LSV by chance (i.e., a set of matching SNVs occurring at the same distance apart). While the tracking of LSVs to identify’ MRD typically has a higher sensitivity and specificity relative to tracking only SNVs, it is worth noting that LSVs are rarer than SNVs, and in some cases, a given sample may not exhibit enough LSVs to offer adequate statistical power and enable a high confidence measurement of the potential MRD signal.

[0023] As described above, identification of the presence and quantity of both SNVs and LS Vs in follow up samples of patients who have been treated for cancer may be used to identify MRD indicative of a risk of relapse. As described m previously referenced applications U.S. Provisional Patent Application No. 63 / 560,514, filed on March 1, 2024 and U.S. Provisional Patent Application No. 63 / 645,537 filed on May 10, 2024, SNVs may be utilized to detect MRD by identifying SNVs in pretreatment samples, applying filters to remove germline mutations and / or experimental artifacts, and applying machine learning or other models to identify SNVs that are indicative of MRD. Hie detected level of signal from reporter SNVs in the follow-up sample is compared to a distribution of healthy samples. Only if the signal in the follow-up sample is significantly higher than the healthy background is it labeled ctDNA positive (MRD +). For example, a quantitative threshold of SNVs indicative of MRD may bedetermined by the models — above which a patient may be considered MRD positive, and below which the patient may be considered MRD negative. LSVs may be similarly utilized to detect MRD, using the same raw data extracted tor SNVs, except instead of analyzing simply for SNVs, identifying multiple SNVs that occur within 175 base pairs. For example, using sequencing data indicating the presence of SNVs, additional processing and filtering may be utilized to evaluate whether a given pair of SNV s are LSVs, to evaluate whether certain LSVs are indicative of MRD, and to determine a threshold quantity of LSVs above which a patient may be considered MRD positive and below which a patient may be considered MRD negative.

[0024] 264 Patients in the analyzed cohort had previously untreated CD20-positiveDLBCL and were randomized to receive Pola-R-CHP or R-CHOP. For each patient, plasma ctDNA collected at baseline and end-of-treatment, in addition to a sample of matched peripheral blood mononuclear cells (PBMC) were utilized for MRD detection. Tire primaryendpoint was investigator-assessed progression free survival (PFS) as calculated in a time-to- event analysis, in which investigator-assessed disease progression and disease relapse or death from any cause were counted as events. Data from this cohort was used to evaluate the following models described with reference to FIGS. 2-6 below.

[0025] FIGS. 2 and 3 depict survival curves associated with discrete unimodal models used to detect MRD based on single nucleotide variants and linked somatic variants, respectively, consistent with implementations of the current subject matter. Specifically, survival curve 200 of FIG. 2 depicts the survival curve for an MRD detection model that identifies tumor-specific SNVs and monitors the retention or reemergence of said tumor-specific SNVs as an indicator of MRD, and survival curve 300 of FIG. 3 depicts a survival curve for an MRD detection model that monitors multiple SNVs that co-occur on the same cfDNA molecule, i.e. LSVs. Metrics associated with each of the two models are provided in Table 1 below, where * denotes metrics calculated for the event of progression, + denotes metrics calculated for progression-free survival, CoxPH denotes the Cox proportional hazard model, HR denotes hazard ratio, p denotes the p-value, C denotes the concordance, and LRT denotes the likelihood ratio test v. null model:

[0026] As can be seen from the figures, the models provide calls of MRD positive (MRD+) or negative (MRD-) and result in only two levels of risk, where MRD- is shown by curves 210 and 310 and MRD+ is shown by curves 220 and 320. Although the SNV and LSV models have similar sensitivity, the LSV model has an increased specificity. However, neither model provides sufficient information regarding the level of risk for a given subject.

[0027] Given the above limitations, potential combinations of the two single modality techniques were evaluated, including: 1) A simple Boolean OR combination, where MRD positive results are called if either of the LSV or SNV models result in an MRD positive determination (hereinafter referred to as “SNV OR LSV”), 2) a simple Boolean AND combination, where MRD positive results are called only if both LSV and SNV models result in an MRD positive determination (hereinafter referred to as “SNV AND LSV”), 3) a modelbased combination without an interaction term (an additive model), including both SNV and LSV based MRD calls as predictors to a CoxPH model without an interaction term (hereinafter referred to as “SNV+LSV”) 4) a model-based combination with an interaction term, including both SNV and LSV based MRD calls as predictors to a CoxPH model with an interaction term (hereinafter referred to as “SNV*LSV”) and 5) a trichotomized combination of the SNV and LSV models in which the MRD status is recoded as a single three-tier categorical variable with LSV- as the baseline group (hereinafter referred to as “3-tier SNV / LSV”). Metrics associated with each of these models are provided in Table 2 below, alongside those of the individual SNV and LSV models, where * denotes metrics calculated for the event of progression, t denotes metrics calculated for progression-free survival, BA denotes balanced accuracy, CoxPH denotes the Cox proportional hazards model, HR denotes hazard ratio, p denotes the p- value, C denotes the concordance, LRT denotes the likelihood ratio test v. null model, and int denotes the interaction term.Table 2

[0028] As can be seen from the concordance scores of the combinations, the model based combinations SNV+LSV and SNV*LSV each had beter “goodness of fit” than that of the unimodal SNV or LSV model. In contrast, the simple Boolean combinations did not improve over the single modal LSV model. In order to evaluate whether to use the model with interaction (SNV*LSV) or without (SNV+LSV), further statistical analysis was performed using the Cox Proportional hazards model. Results of the analysis on the model-based combinations of SNV and LSV are excerpted in Tables 3 and 4 below. Table 3 depicts the resulting analysis of the model based combination without interaction (SNV+LSV) and Table 4 depicts the resulting analysis of the model based combination with interaction (SNV* LSV). For reference purposes, ‘‘ctdna status. x” references the MRD status based on SNV detection and “ctdna_status_y” references the MRD status based on LSV detection. When evaluating results in Table 4 of the model with the interaction term (“ctdna_status.x:ctdna_status.y”) it can be seen that the interaction term is statistically significant, which suggests that the effect of one predictor (SNV / LSV) on the outcome (MRD status) depends on the value of the other predictor. Thus, the appropriate model based combination must account for the interaction of the two parameters.Table 3

[0029] Survival curves for the combined models including the two Boolean combinations and the model based combination are provided in FIGS. 4-6. Survival curve 400 in FIG. 4 depicts the results from tire SNV OR LSV model, where curve 410 represents a call of MRD- and curve 420 represents a call of MRD+. Survival curve 500 in FIG. 5 depicts the results from the SNV AND LSV model, where curve 510 represents a call of MRD- and curve 520 represents a call of MRD+. Survival curve 600 in FIG. 6 depicts aspects of the model based combination of the LSV and SNV models where curve 610 represents the cases where both the SNV and LSV based models return an MRD- call, curve 612 represents the cases where theSNV model returns an MRD- call and the LSV model returns an MRD+ call, curve 614 represents the cases where the SNV model returns an MRD+ call and the LSV model retains an MRD- call, and curve 620 represents the cases where both the SNV and LSV models return an MRD+ call. As can be seen in survival curve 600, curves 610 and 614 greatly overlap and do not provide for significantly different levels of risk. Thus, it appears that in the case of an MRD- call based on LSV, the MRD determination based on SNV does not change the prediction of progression. In contrast, in the case of an MRD-f- call based on LSV (curves 612 and 62.0), it can be seen that tire risk is dichotomized since a corresponding MRD+ call based on SNV (curve 620) provides for a significantly poorer prognosis than a corresponding MRD- call based on SNV (curve 612). In view of these findings, it is clear that there are three risk groups associated with MRD based on the LSV and SNV status: (I) the lowest risk being associated with an MRD- status based on LSV, (2) the intermediate risk being associated with an MRD+ status based on LSV and an MRD- status based on SNV, and (3) the highest risk being associated with an MRD+ status based on both LSV and SNV.

[0030] In view of the findings described above, the combined MRD model was re-coded into a trichotomized risk model with MRD- status based on LSV as the baseline / lowest risk,MRD+ status based on LSV and an MRD- status based on SNV as the intermediate risk, and(3) an MRD+ status based on both LSV and SNV as the highest risk. The resulting recoded survival curves are depicted in FIG. 7, where 710 represents the MRD- status based on LSV, 715 represents the MRD+ LSV and MRD- SNV status, and 720 represents the MRD+ status based on both LSV and SNV. Metrics for the re-coded model are provided in Table 2 above. As can be seen in Table 2, the concordance of the re-coded multi-modal model is an improvement from existing single modality predictors of MRD as well as simple combinations. Even more importantly, the significantly different hazard ratios for LSV+ / SNV- and LSV+ / SNV+ confirms the surprising risk stratification observed and described previously and ensures that the model provides appropriate risk stratification that can be used in personalizing treatment for the subjects. Using the recoded trichotomized risk model based on two related but separate features that can be extracted from basic sequencing data (SNV and LSV), a patient’s risk of relapse can be more accurately and granularly stratified, which can lead to more informed decision-making regarding treatment plans.

[0031] FIG. 8 is a block diagram illustrating an embodiment of a method 800 for detecting minimal residual disease (MRD), according to aspects of the present disclosure. One or moreof the steps of method 800 may be performed by a processor, such as processor 910 of computing system 900. At step 810, a plurality of sequencing reads may be obtained from a pre-treatment sample of a subject diagnosed with cancer. In some instances the sequencing reads may be obtained using targeted sequencing assays. For example, sequencing reads may be obtained using sequencing assays designed to target certain genetic regions known to have single nucleotide variants or linked somatic variants indicative of minimal residual disease and / or the cancer of interest. Alternatively, sequencing reads may be obtained using whole genome sequencing techniques. In some instances, the pre-treatment sample may be a tissue sample taken from tumor tissue (fresh-frozen or have undergone FFPE treatment) of the subject. In other instances, the pre-treatment sample may be a sample taken from blood samples of the subject.

[0032] At step 820, reporter SNVs indicative of MRD may be identified based on the sequencing reads of the pre-treatment sample obtained at step 810. In order to identify SNVs indicative of MRD, the sequencing reads of tire pre-treatment sample may be evaluated for SNVs using any suitable vanant caller, and the resulting identified SNV s may be filtered (for example to remove germline and / or Clonal Hematopoiesis of Indeterminate Potential (CHIP) variants) in any suitable way to identify which of the variants are indicative of MRD. For example, germline and / or CHIP variants may be filtered by using peripheral blood mononuclear cell (PBMC) samples extracted from the blood of the subject. In some embodiments, the germline mutations and sequencing artifacts may be defined in a plurality of databases. The plurality of databases may include, but are not limited to, public databases such as Nirvana, 1000 Genomes, ExAC, COSMIC, and TCGA. It should be understood by those of ordinary skill in the art that these are merely example databases that define germline mutations and sequencing artifacts, and that other databases may be used in accordance with aspects of the present disclosure. In some embodiments, any SNVs in the publicly available databases may be annotated. For example, the annotations may be added using Nirvana or ClinVar, publicly available databases used to identify clinical -grade annotations of genomic variants. In some embodiments, filtering any germline mutations and sequencing artifacts may also be based on common biomarker SNVs stored on the nanochip workstation 120. Applying these filters improves signal -to-noise and sensitivity for detecting ctDNA in plasma. In some instances, the filtered SNVs may be fed into a machine learning model to classify the resulting identified SNVs. For example, the filtering and classifying of reporter SNV s may be carriedout as described in U.S. Provisional Patent Application No. 63 / 560,514, filed on March 1, 2024 and U.S, Provisional Patent Application No, 63 / 645,537 filed on May 10, 202.4, in which somatic variants are identified using variant callers, germline mutations and sequencing and other experimental artifacts are filtered out, and the remaining variants are fed into a random forest classifier to identify which are reporter SNVs indicative of MRD.

[0033] Once the reporter variants indicative of MRD are identified, metrics related to the quantity of reporter SN Vs may be determined to identify thresholds for use in evaluating subsequent post-treatment samples. The thresholds, quantity, and other information associated with the reporter SNVs may be stored for use in subsequent tests. For example, a threshold quantity of reporter SNVs may be determined from the pre-treatment sample above which may indicate the patient is MRD+ in subsequent post-treatment follow up sampling. Alternatively, a signal to noise ratio of the reporter SNVs relative to a background level may be used to determine if a patient is MRD+ or MRD -.

[0034] At step 830, reporter LSVs indicative of MRD may be identified based on the sequencing reads of the pre-treatment sample obtained at step 810. As described above with respect to SNVs, the sequencing reads of tire pre-treatment sample may be evaluated for SNVs using any suitable variant caller, and the resulting identified SNVs may be filtered (for example to remove germline and / or CHIP variants) in any suitable way to identify which of the variants are likely derived from the tumor and indicative of MRD. Additional processing may be carried out to identify which SNVs occur within 175 base pairs of each other in order to identify LSVs. And, once LSVs are identified, they may be classified as reporter LSVs using any suitable techniques. For example, the data from the subject’s pre-treatment sample may be fed into a classifier to be compared to healthy and cancerous sequencing data to identify reporter LSVs indicative of MRD. As with SNVs, once the reporter LSVs indicative of MRD are identified, metrics related to the quantity of reporter LSVs may be determined to identify' thresholds for use in evaluating subsequent post-treatment samples. The thresholds, quantity, and other information associated with the reporter LSVs may be stored for use in subsequent tests. For example, a threshold quantify of reporter LS Vs may be determined from the pre-treatment sample above which may indicate the patient is MRD+ in subsequent post-treatment follow up sampling. Alternatively, a signal to noise ratio of the reporter LSVs relative to a background level may be used to determine if a patient is MRD+ or MRD -.

[0035] In some instances, an early (pre -treatment) risk assessment may be made based on the quantity and or types of SNVs and LSVs identified in the pre-treatment sample. For example, the quantity and types of reporter SNVs and LSVs identified in the pre-treatment sample may be compared to reference values and based on said comparison a prognosis of post treatment risk of relapse may be made. Such prognosis may be used to advise the subject on treatment options.

[0036] At step 840, sequencing reads from a post-treatment sample of the subject may be obtained. The post-treatment sample may be obtained at any time after a subject has received treatment for cancer. For example, an initial post-treatment sample may be taken at or near the completion of the first course of treatment, m order to identify the early molecular response of the subject. Additional post-treatment samples may be taken at regular intervals during the course of treatment or after completion of treatment to evaluate a trend in any relapse and provide updated risk levels of MRD, and enable informed decision making regarding potential future treatment. Post-treatment samples may be taken from blood samples of the subject. For example, the post-treatment sample may be a plasma sample extracted from the blood of the subject. Thus, the post-treatment sample may allow for analysis of a subject’s ctDNA which may provide information regarding MRD. Any suitable sequencing techniques may be applied to the post-treatment sample in order to evaluate the sample for risk levels of MRD. For example, the sample may be sequenced using targeted enrichment to focus on areas known to have SNVs and LSVs indicative of MRD.

[0037] At step 850, the sequencing reads of the post-treatment sample may be evaluated for reporter SNVs. Using the sequencing data from the post-treatment sample, reporter SNVs may be detected in the post-treatment sample and metrics associated with the reporter SNVs may be determined. For example, a quantity and / or a signal to noise ratio of the reporter SNVs may be determined. As another example, a normalized quantity of detected variants may be determined by dividing the total number of detected variants by the total number of reads covering the variant positions. Still another example of a metric to be determined may include running a simulation to determine whether the level of signal detected in the post-treatment sample at the reporter variant positions is significantly higher than a randomly selected set of other genomic positions. The metrics associated with the reporter SNVs identified in the posttreatment sample may be compared to the metrics and / or threshold determined in step 620 to generate an SNV determination related to MRD. For example, an SNV determination relatedto MRD may be made by using a SNV model developed to identify the subject as MRD+ or MRD-. The metrics associated with the reporter SNVs and the SNV determination related to MRD may be stored in a computer for reference.

[0038] At step 860, the sequencing reads of the post-treatment sample may be evaluated for reporter LSVs. Using the sequencing data from the post-treatment sample, reporter LSVs may be detected in the post-treatment sample and metrics associated with the reporter LSVs may be determined. For example, a quantity and / or a signal to noise ratio of the reporter LSVs may be determined, a normalized quantity of detected LSVs may be determined, or the level of signal at the LSV positions may be compared to randomly selected genomic positions. The metrics associated with tire reporter LSVs identified in the post-treatment sample may be compared to the metrics and / or threshold determined in step 830 to generate an LSV determination related to MRD. For example, an LSV determination related to MRD may be made by using a LSV model developed to identify the subject as MRD+ or MRD-. The metrics associated with the reporter LSVs and the LSV determination related to MRD may be stored in a computer for reference.

[0039] At step 870, a risk level of MRD may be determined based on the evaluations of reporter SNVs and reporter LSVs. For example, based on the SNV determination and the LSV determination, a level of risk of relapse may be determined for the subject. In some instances, based on the SNV determination related to MRD and the LSV determination related to MRD, the patient may be classified as low risk of relapse, intermediate risk of relapse, or high risk of relapse. In other instances, additional categories of risk may be identified. For example, if the SNV determination and LSV determination are both MRD+, the subject may be considered high risk. If the SNV determination is MRD+ and tire LSV determination is MRD- then the subject may be considered intermediate risk. If the LS V determination is MRD- then the subject may be considered low risk. In some instances, if the post-treatment sample is not the first post-treatment sample, additional risk information may be determined based on a trend of post-treatment samples. For example, additional categories of risk may be identified if the trend of the SNV determination and LSV determination indicates additional information regarding the risk of relapse. In some instances, a report may be generated that includes the SNV determination, the LSV determination, and the risk level associated therewith.

[0040] FIG. 9 is a block diagram illustrating one embodiment of a computer system 900 configured to implement one or more aspects of the present disclosure. For example, the method 800 of FIG. 8 may be implemented using the computer system 900.

[0041] As shown in FIG. 9, computing system 900 can include a processor 910, a memory 920, a storage device 930, and input / output devices 940. Processor 910, memory 920, storage device 930, and input / output devices 940 can be interconnected via system bus 950. Processor 910 is capable of processing instructions for execution within the computing system 900. In some example embodiments, processor 910 can be a single-threaded processor. Alternatively, processor 910 can be a multi-threaded processor. Processor 910 is capable of processing instructions stored in memory 920 and / or on the storage device 930 to display graphical information for a user interface provided via the input / output device 940,

[0042] Memory' 790 is a computer readable medium such as volatile or non-volatile that stores information within computing system 900. Memory 920 can store data structures representing configuration object databases, for example. Storage device 930 is capable of providing persistent storage for computing system 900. Storage device 930 can be a floppy disk device, a hard disk device, an optical disk device, or a tape device, or other suitable persistent storage means. Input / output device 940 provides input / output operations for the computing system 700. In some example embodiments, input / output device 940 includes a keyboard and / or pointing device. In various implementations, the input / output device 940 includes a display unit for displaying graphical user interfaces.

[0043] According to some example embodiments, input / output device 940 can provide input / output operations for a network device. For example, input / output device 940 can include Ethernet ports or other networking ports to communicate with one or more wired and / or wireless networks (e.g., a local area network (LAN), a wide area network (WAN), the Internet).

[0044] In some example embodiments, computing system 900 can be used to execute various interactive computer software applications that can be used for organization, analysis and / or storage of data in various formats. Alternatively, computing system 900 can be used to execute any type of software applications. These applications can be used to perform various functionalities, e.g., planning functionalities (e.g., generating, managing, editing of spreadsheet documents, word processing documents, and / or any other objects, etc.), computing functionalities, communications functionalities, etc. The applications can include various addin functionalities or can be standalone computing products and / or functionalities. Uponactivation within the applications, the functionalities can be used to generate the user interface provided via input / output device 940. The user interface can be generated and presented to a user by computing system 900 (e.g., on a computer screen monitor, etc.).

[0045] One or more aspects or features of the subject matter described herein can be realized in digital electronic circuitry, integrated circuitry, specially designed applicationspecific integrated circuits (ASICs), field programmable gate arrays (FPGAs) computer hardware, firmware, software, and / or combinations thereof. These various aspects or features can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device. The programmable system or computing system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. Tire relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

[0046] These computer programs, which can also be referred to as programs, software, software applications, applications, components, or code, include machine instructions for a programmable processor, and can be implemented in a high-level procedural and / or object- oriented programming language, and / or in assembly / machine language. As used herein, the term "‘machine-readable medium” refers to any computer program product, apparatus and / or device, such as for example magnetic discs, optical disks, memory, and Programmable Logic Devices (PLDs), used to provide machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term “machine-readable signal” refers to any signal used to provide machine instructions and / or data to a programmable processor. The machine-readable medium can store such machine instructions n on-transitori ly, such as for example as would a non-transient solid-state memory or a magnetic hard drive or any equivalent storage medium. The machine-readable medium can alternatively or additionally store such machine instructions in a transient manner, such as for example, as would a processor cache or other random access memory associated with one or more physical processor cores.

[0047] To provide for interaction with a user, one or more aspects or features of the subject matter described herein can be implemented on a computer having a display device, such asfor example a cathode ray tube (CRT) or a liquid crystal display (LCD) or a light emitting diode (LED) monitor for displaying information to the user and a keyboard and a pointing device, such as for example a mouse or a trackball, by which the user may 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, such as for example visual feedback, auditory feedback, or tactile feedback; and input from the user maybe received in any form, including acoustic, speech, or tactile input. Other possible input devices include touch screens or other touch-sensitive devices such as single or multi-point resistive or capacitive track pads, voice recognition hardware and software, optical scanners, optical pointers, digital image capture devices and associated interpretation software, and the like.

[0048] When a feature or element is herein referred to as being “on” another feature or element, it can be directly on the other feature or element or intervening features and / or elements may also be present. In contrast, when a feature or element is referred to as being “directly on” another feature or element, there are no intervening features or elements present. It will also be understood that, when a feature or element is referred to as being “connected”, “attached” or “coupled” to another feature or element, it can be directly connected, atached or coupled to the other feature or element or intervening features or elements may be present. In contrast, when a feature or element is referred to as being “directly connected”, “directly attached” or “directly coupled” to another feature or element, there are no intervening features or elements present. Although described or shown with respect to one embodiment, the features and elements so described or shown can apply to other embodiments. It will also be appreciated by those of skill in the art that references to a structure or feature that is disposed “adjacent” another feature may have portions that overlap or underlie the adjacent feature.

[0049] Terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. For example, as used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises” and / or “comprising,” when used in this specification, specify the presence of stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof. As used herein, the term “and / or” includes any and all combinations of one or more of the associated listed items and may be abbreviated as “ / ”.

[0050] Spatially relative terms, such as “under”, “below”, “lower”, “over”, “upper” and the like, may be used herein for ease of description to describe one element or feature’s relationship to another element(s) or feature(s) as illustrated in the figures. It will be understood that the spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. For example, if a device m the figures is inverted, elements described as “under” or “beneath” other elements or features would then be oriented “over” the other elements or features. Tims, the exemplary term “under” can encompass both an orientation of over and under. Hie device may be otherwise oriented (rotated 90 degrees or at other orientations) and the spatially relative descriptors used herein interpreted accordingly. Similarly, the terms “upwardly”, “downwardly”, “vertical”, “’horizontal” and the like are used herein for the purpose of explanation only unless specifically indicated otherwise.

[0051] Although the terms “first” and “second” may be used herein to describe various features / elements (including steps), these feature s / elements should not be limited by these terms, unless the context indicates otherwise. These terms may be used to distinguish one feature / element from another feature / element. Thus, a first feature / element discussed below could be termed a second feature / element, and similarly, a second feature / element discussed below' could be termed a first feature / element without departing from the teachings of the present disclosure.

[0052] Throughout this specification and the claims which follow, unless the context requires otherwise, the word “comprise”, and variations such as “comprises” and “comprising” means various components can be co-jointly employed in the methods and articles (e.g., compositions and apparatuses including device and methods). For example, the term “comprising” will be understood to imply the inclusion of any stated elements or steps but not the exclusion of any other elements or steps.

[0053] As used herein in the specification and claims, including as used in tire examples and unless otherwise expressly specified, all numbers may be read as if prefaced by the word “about” or “approximately,” even if the term does not expressly appear. The phrase “about” or “approximately” may be used when describing magnitude and / or position to indicate that the value and / or position described is within a reasonable expected range of values and / or positions. For example, a numeric value may have a value that is + / - 0.1% of the stated value (or range of values), + / - 1 % of the stated value (or range of values), + / - 2% of the stated value (or range of values), + / - 5% of the stated value (or range of values), + / - 10% of the stated value(or range of values), etc. Any numerical values given herein should also be understood to include about or approximately that value, unless the context indicates otherwise. For example, if the value '“i O" is disclosed, then “about 10” is also disclosed. Any numerical range recited herein is intended to include all sub-ranges subsumed therein. It is also understood that when a value is disclosed that “less than or equal to” the value, “greater than or equal to the value” and possible ranges between values are also disclosed, as appropriately understood by the skilled artisan. For example, if the value “X” is disclosed the “less than or equal to X” as well as “greater than or equal to X” (e.g., where X is a numerical value) is also disclosed. It is also understood that the throughout the application, data is provided in a number of different formats, and that this data, represents endpoints and starting points, and ranges for any combination of the data points. For example, if a particular data point “10” and a particular data point “15” are disclosed, it is understood that greater than, greater than or equal to, less than, less than or equal to, and equal to 10 and 15 are considered disclosed as well as between 10 and 15. It is also understood that each unit between two particular units are also disclosed. For example, if 10 and 15 are disclosed, then 11, 12, 13, and 14 are also disclosed.

[0054] Although various illustrative embodiments are described above, any of a number of changes may be made to various embodiments without departing from the scope of the disclosure as described by the claims. For example, the order in which various described method steps are performed may often be changed in alternative embodiments, and in other alternative embodiments one or more method steps may be skipped altogether. Optional features of various device and system embodiments may be included in some embodiments and not in others. Therefore, the foregoing description is provided primarily for exemplary purposes and should not be interpreted to limit the scope of the disclosure as it is set forth m the claims.

[0055] The examples and illustrations included herein show, by way of illustration and not of limitation, specific embodiments in which the subject matter may be practiced. As mentioned, other embodiments may be utilized and derived there from, such that structural and logical substitutions and changes may be made without departing from the scope of this disclosure. Such embodiments of the inventive subject matter may be referred to herein individually or collectively by the term “invention” merely for convenience and without intending to voluntarily limit the scope of this application to any single invention or inventive concept, if more than one is, in fact, disclosed. Thus, although specific embodiments have been illustrated and described herein, any arrangement calculated to achieve the same purpose maybe substituted for the specific embodiments shown. This disclosure is intended to cover any and all adaptations or variations of various embodiments. Combinations of the above embodiments, and other embodiments not specifically described herein, will be apparent to those of skill in the art upon reviewing the above description.

Claims

CLAIMSWhat is claimed is:

1. A method for determining a risk level of minimal residual disease (MRD) in a subject treated for cancer, the method comprising: identifying a first type of reporter genomic variants indicative of minimal residual disease based on a plurality of sequencing reads obtained from a pre-treatment sample of the subject; identifying a second type of reporter genomic variants indicative of minimal residual disease based on the plurality-7of sequencing reads obtained from the pre-treatment sample of tlie subject, obtaining a plurality of sequencing reads from a post-treatment sample of the subject; evaluating the first type of reporter genomic variants in the sequencing reads from the post-treatment sample of the subject to make a first determination of MRD in the subject; evaluating the second type of reporter genomic variants from the post-treatment sample of the subject to make a second determination of MRD in the subject; determining one of at least three risk levels of MRD in the subject based on the first determination of MRD and the second determination of MRD.

2. The method of claim 1 , wherein: the first type of reporter genomic variants compri ses single nucleotide vanants (SNVs) indicative of MRD; and the second type of reporter genomic variants comprises linked somatic variants (LSVs) indicative of MRD.

3. The method of claim 1, wherein: the first determination of MRD in the subject comprises comparing a quantity of SNV reporters in the post-treatment sample to a baseline level of SNV reporters; and the second determination of MRD in the subject comprises comparing a quantity of LSV reporters in the post-treatment sample to a baseline level of LSV reporters.

4. The method of claim 2, wherein the at least three risk levels comprise: a first risk level of MRD determined based on a MRD negative determination for tire second MRD determination; a second risk level of MRD greater than the first risk level, the second risk level being determined based on an MRD positive determination for the first MRD determination and an MRD negative determination for the second MRD determination; and a third risk level of MRD greater than the second risk level, the third risk level being determined based on an MRD positive determination for each of the first and second MRD determinations.

5. The method of claim 1 , further comprising determining a suggested therapy based on the determined one of at least three risk levels of MRD.

6. A multimodal detection system for identifying the risk of MRD in a subject treated for cancer, comprising: a memory; and a processor coupled to the memory and configured to:identify a first type of reporter genomic variants indicative of minimal residual disease based on a plurality of sequencing reads obtained from a pre-treatment sample of the subject; identify a second type of reporter genomic variants indicative of minimal residual disease based on the plurality of sequencing reads obtained from the pre-treatment sample of the subject; obtain a plurality of sequencing reads from a post-treatment sample of the subject; evaluate the first type of reporter genomic variants in the sequencing reads from the post-treatment sample of the subject to make a first determination of MRD in the subject; evaluate the second type of reporter genomic variants from the post-treatment sample of the subject to make a second determination of MRD in the subject; and determine one of at least three risk levels of MRD in the subject based on the first determination of MRD and the second determination of MRD.

7. The system of claim 6, wherein: the first type of reporter genomic variants comprises single nucleotide variants (SNVs) indicative of MRD; and the second type of reporter genomic variants comprises linked somatic variants (LSVs) indicative of MRD.

8. The system of claim 7, wherein the processor is configured to: compare a quantity of SNV reporters in the post-treatment sample to a baseline level of SNV reporters to make the first determination of MRD in the subject; andcompare a quantity of LS V reporters m the post-treatment sample to a baseline level of LSV reporters to make the second determination of MRD in the subject.

9. The system of claim 7. wherein the at least three risk levels comprise a first risk level of MRD, a second risk level of MRD greater than the first risk level, and a third risk level of MRD greater than the second risk level; further wherein the processor is further configured to: determine the first risk level of MRD based on an MRD negative determination for the second MRD determination; determine the second risk level of MRD based on an MRD positive determination for the second MRD determination and an MRD negative determination for the first MRD determination; and determine the third risk level of MRD based on an MRD positive determination for the second MRD determination and the first MRD determination.

10. The system of claim 6, wherein the processor is further configured to output a suggested therapy based on the determination of risk of MRD.

Images