System and method for estimating tumor growth

Abstract

A method for estimating tumor growth kinetics includes obtaining, by one or more processors, progression-free lifetime (PFS) data for a plurality of patients, the PFS data indicating (i) a plurality of observation times, and (ii) how many of the plurality of patients have occurred a PFS event within a recent window of time at each of the plurality of observation times; determining, by the one or more processors, a population distribution of one or more patient-specific parameters based on the PFS data; obtaining, by the one or more processors, measured tumor growth data of the particular patient undergoing the drug treatment; estimating, by the one or more processors, tumor growth of the particular patient based on (i) the measured tumor growth data and (ii) a population distribution of the one or more patient-specific parameters; and causing a display to present, by the one or more processors, a visual indication of the estimated tumor growth of the particular patient.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Patent Application Serial No. 63 / 408,885, filed on September 22, 2022, entitled “SYSTEMS AND METHODS FORESTIMATING TUMOR GROWTH,” the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present application relates generally to tumor growth analysis in oncology patients, and more particularly to modeling and estimating contributions to tumor growth rate. Background Art

[0004] Tumor growth analysis is essential for experimental oncology research. The principle of anticancer treatment is to reduce the rate of tumor growth over time by shrinking the tumor and / or slowing tumor growth, thereby improving patients' symptoms and prolonging overall survival. Progression-free survival (PFS) can be the length of time a patient lives with the disease without it getting worse during and after treatment. PFS can predict overall survival. Summary of the invention

[0005] In one aspect, a computer-implemented method for estimating tumor growth includes: obtaining, by one or more processors, progression-free survival (PFS) data for a plurality of patients, the PFS data indicating (i) a plurality of observation times, and (ii) how many of the plurality of patients had a PFS event within a most recent time window at each of the plurality of observation times; determining, by the one or more processors, a population distribution of one or more patient-specific parameters based on the PFS data; obtaining, by the one or more processors, measured tumor growth data for a particular patient undergoing drug treatment; estimating, by the one or more processors, the tumor growth of the particular patient based on (i) the measured tumor growth data and (ii) the population distribution of the one or more patient-specific parameters; and causing a display, by the one or more processors, to present a visual indication of the estimated tumor growth of the particular patient.

[0006] In some embodiments, determining the population distribution of the one or more patient-specific parameters includes determining a growth curve function that includes the one or more patient-specific parameters. In some embodiments, the growth curve function includes at least one of an exponential growth function, a logistic growth function, or an ordinary differential function. In some embodiments, the one or more patient-specific parameters include a baseline normalized sum of longest diameters (SLD) measurement parameter. In some embodiments, the one or more patient-specific parameters include a growth rate parameter.

[0007] In some embodiments, the growth rate parameter comprises a baseline growth rate parameter without treatment. In some embodiments, the one or more patient-specific parameters comprise a parameter of a proportion of drug-sensitive tumor cells in a patient of the plurality of patients. In some embodiments, the parameter value of the proportion of drug-sensitive tumor cells in a patient of the plurality of patients ranges from 0 to 1. In some embodiments, the growth curve function is time-dependent. In some embodiments, determining the population distribution of the one or more patient-specific parameters comprises fitting the growth curve function to the observed values ​​at the plurality of observation times.

[0008] In some embodiments, the growth curve function is a logistic growth function; and the one or more patient-specific parameters include multiple parameters of the logistic growth function. In some embodiments, estimating the tumor growth of the specific patient includes: modeling the tumor growth rate of the specific patient using a pharmacokinetic-pharmacodynamic (PKPD) model, the model having a first term representing the change in tumor size of the specific patient without the drug treatment and a second term representing the contribution to the change in tumor size of the specific patient due to the drug treatment; and jointly estimating the one or more parameters of the first term and the one or more parameters of the second term by fitting the PKPD model to the measured tumor growth data, in part by setting constraints on the one or more parameters of the first term using the population distribution of the one or more patient-specific parameters.

[0009] In some embodiments, the one or more parameters of the second item include one or more of the following: drug concentration in the plasma of the particular patient; maximum killing rate of the particular patient; and half-maximal effective concentration (EC50). In some embodiments, estimating the tumor growth further includes obtaining an overall remission rate for the particular patient undergoing the drug treatment. In some embodiments, estimating the tumor growth further includes obtaining one or more non-target events for the particular patient undergoing the drug treatment.

[0010] In some embodiments, the observations include, for each of the plurality of patients, a first observation at a first time indicating that the patient had not had a PFS event prior to the first time, and a second observation at a second time indicating that the patient had had a PFS event prior to the second time. In some embodiments, the PFS data corresponds to a specific cancer type; and estimating the tumor growth of the specific patient includes estimating the tumor growth of patients diagnosed with the specific cancer type. In some embodiments, causing the display to present a visual indication of the estimated tumor growth of the specific patient includes causing the display to display a tumor growth trajectory for the specific patient when the specific patient has not yet been treated with the drug.

[0011] In some embodiments, the population distribution of the one or more patient-specific parameters includes a lognormal distribution. In some embodiments, the PFS data includes at least one of a digitized PFS graph or a PFS risk table. In some embodiments, the PFS data indicates how many of the multiple patients have a baseline normalized sum of longest diameters (SLD) measurement of at least 1.2 or new lesions at each observation time in the multiple observation times and within the most recent time window. In some embodiments, the multiple patients represented by the PFS data are patients known to be associated with ineffective drug treatment. In some embodiments, the method further includes adjusting the dose of the drug treatment based on the estimated tumor growth of the particular patient.

[0012] In another aspect, a computer system for estimating tumor growth comprises: a data storage device storing processor-readable instructions; and a processor configured to execute the instructions to perform a method comprising: obtaining, by one or more processors, progression-free survival (PFS) data for a plurality of patients, the PFS data indicating (i) a plurality of observation times, and (ii) how many of the plurality of patients had a PFS event within a most recent time window at each of the plurality of observation times; determining, by the one or more processors, a population distribution of one or more patient-specific parameters based on the PFS data; obtaining, by the one or more processors, measured tumor growth data for a particular patient undergoing drug treatment; estimating, by the one or more processors, the tumor growth of the particular patient based on (i) the measured tumor growth data and (ii) the population distribution of the one or more patient-specific parameters; and causing a display to present, by the one or more processors, a visual indication of the estimated tumor growth of the particular patient.

[0013] In yet another aspect, a non-transitory computer-readable medium contains instructions for estimating tumor growth, which when executed by a processor causes the processor to perform a method comprising: obtaining, by one or more processors, progression-free survival (PFS) data for a plurality of patients, the PFS data indicating (i) a plurality of observation times, and (ii) how many of the plurality of patients had a PFS event within a most recent time window at each of the plurality of observation times; determining, by the one or more processors, a population distribution of one or more patient-specific parameters based on the PFS data; obtaining, by the one or more processors, measured tumor growth data for a particular patient undergoing drug treatment; estimating, by the one or more processors, the tumor growth of the particular patient based on (i) the measured tumor growth data and (ii) the population distribution of the one or more patient-specific parameters; and causing a display, by the one or more processors, to present a visual indication of the estimated tumor growth of the particular patient. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Those skilled in the art will appreciate that the drawings described herein are included for illustrative purposes and are not intended to limit the present disclosure. The drawings are not necessarily drawn to scale, but rather focus on illustrating the principles of the present disclosure. It should be understood that in some cases, various aspects of the described embodiments may be shown exaggerated or enlarged to aid in understanding the described embodiments. In the drawings, similar reference numerals generally refer to components that are similar in function and / or structure throughout the drawings.

[0015] Figure 1 is a simplified block diagram of an example system that may implement the tumor growth modeling and estimation techniques disclosed herein.

[0016] Figure 2 Depicted are example progression-free survival (PFS) graphs and associated risk tables in accordance with some embodiments of the technology described herein.

[0017] Figure 3 is a graph depicting an example selection of a single fixed PFS event time for a particular patient within a particular PFS window and an associated exponential growth curve according to some embodiments of the techniques described herein.

[0018] Figure 4 is a graph depicting three exemplary possible PFS event times and associated exponential growth curves for a particular patient within a particular PFS window according to some embodiments of the technology described herein.

[0019] Figure 5 is a graph depicting noise in SLD measurements of an example patient over time according to some embodiments of the techniques described herein.

[0020] Figure 6 Depicted are tumor doubling times measured in different patient populations for each of a variety of cancer types according to some embodiments of the technology described herein.

[0021] Figure 7 Depicted are estimated drug concentrations in plasma (top row) and estimated tumor growth (bottom row) for three example combinations of drug doses and cancer types according to some embodiments of the technology described herein, where the estimated tumor growth shows tumor growth with and without the drug.

[0022] Figure 8 is a flow chart of an example method for estimating tumor growth in accordance with some embodiments of the techniques described herein.

[0023] Fig. 9An exemplary generation process of data according to some embodiments of the techniques described herein is depicted.

[0024] Fig.10 Depicted are one or more exemplary posterior samplings of the SLD of a single patient from a control group who progressed at a second post-baseline scan with ground truth superimposed in white according to some embodiments of the technology described herein.

[0025] Fig.11 Depicted are exemplary marginal posterior values ​​for each patient estimated from the model and ground truth values ​​according to some embodiments of the techniques described herein.

[0026] Fig.12 An exemplary a posteriori sampling of trial parameters and group-specific parameters is shown with their ground truth values ​​superimposed in accordance with some embodiments of the techniques described herein.

[0027] Fig.13 Depicted are exemplary graphs associated with PFS curves and the total number of remissions per sampling overlaid with observed values ​​in a posterior predictive check (PPC) to assess the adequacy of the fit in accordance with some embodiments of the techniques described herein.

[0028] Fig.14 Shown are simulated numbers of remissions and PFS curves for four groups of a simulation trial according to some embodiments of the technology described herein.

[0029] Fig.15 is a schematic diagram of an illustrative computing device that can be used to implement various aspects described herein. DETAILED DESCRIPTION

[0030] To assess the efficacy of a drug in an oncology patient (e.g., during an early drug trial), observations of tumor / lesion size may be collected over a period of time during which the drug is administered. While tumor growth rate measurements are helpful, they may not be sufficient because tumor growth rate may depend on two competing factors. One factor may be the rate at which a tumor grows in the absence of drug treatment, and another factor may be the rate at which the administered drug shrinks the tumor. Unless the individual contributions of each of these factors can be accurately estimated, it may be difficult to assess whether a drug is fully effective, let alone determine the best dosing regimen, or determine which patient population may benefit most from the drug.

[0031] A technique (referred to herein as the "Kay technique") has been proposed to separate and evaluate tumor growth rates in the absence of drug treatment, in which data from a progression-free survival (PFS) curve plot is used to estimate tumor size doubling time based on the "sum of the longest diameter" (SLD) indicator of the target lesion. See Kay et al., The AAPS Journal [American Association of Pharmaceutical Scientists Journal], 21:27, Estimation of Solid Tumor Doubling Times from Progression-Free Survival Plots Using a Novel Statistical Approach [Estimation of Solid Tumor Doubling Times from Progression-Free Survival Plots Using a Novel Statistical Approach] (2019). However, the Kay technique cannot accurately explain specific uncertainties in the PFS curve plot, such as uncertainty about the location of the PFS event within a given time window and uncertainty caused by SLD measurement error / noise. In addition, the Kay technique strictly applies certain assumptions (e.g., exponential tumor growth).

[0032] Some techniques and modeling methods can describe the sum of longest diameter (SLD) measurement and its continuous response over time to a certain drug treatment, however, for these techniques and modeling methods, the data is generally not publicly available. In addition, some methods can extract certain publicly available data and assume exponential growth due to their model assumptions, but the extracted publicly available data may not provide information on the underlying tumor dynamics. Therefore, there is still a need for tumor growth modeling that can be both more accurate and more flexible.

[0033] In order to solve the above-mentioned problems related to tumor growth modeling, the systems and methods disclosed herein can provide a more flexible modeling method to more accurately learn one or more parameters (e.g., patient-specific parameters associated with tumor growth). The systems and methods disclosed herein can describe the longitudinal sum of the longest diameters (SLD) measurement and its continuous response to a certain drug treatment over time in a specific longitudinal scheme. This can produce several advantages over traditional methods that usually consider time snapshots. First, the systems and methods disclosed herein can capture the effects on responses that may occur when exposure changes over time, for example, when patients reduce doses, change dosing regimens, or are exposed to lower anti-drug antibodies. This happens. Secondly, because the entire longitudinal history of SLD is modeled, multiple quantities of interest can be captured by a unified model, such as maximum reduction, progression time, time to maximum reduction, or doubling time. The systems and methods disclosed herein can incorporate non-target progression events as time-event data modeled by a hazard function, which can go further and can capture common endpoints such as PFS and overall response rate (or ORR). Again, due to its semi-mechanistic nature, one or more parameters can have a mechanistic explanation, which provides rich possibilities for simulation and is conducive to promotion to new schemes. For example, estimates of certain parameters, such as kinetic constants (eg, killing rate), can be borrowed or extrapolated from preclinical data or published studies to compare treatments under similar conditions or to estimate the efficacy of a novel treatment.

[0034] The systems and methods disclosed herein may also address: how early Phase 1 results can be generalized to earlier lines of treatment by using one or more parameters (e.g., patient-specific parameters or non-patient-specific parameters) to capture the kinetics of patient groups with different lines of treatment; how a drug performs when directly compared to an existing standard of care by using one or more parameters to capture drug-specific effects; how a drug performs when combined with an existing standard of care by combining one or more parameters to simulate combinations; and how many patients should be enrolled in a trial to demonstrate these results in a statistically significant manner by simulating an entire trial using the systems and methods disclosed herein to account for all levels of variation.

[0035] Although individual SLD data are not usually publicly available, endpoints such as PFS and ORR derived directly from the individual data are usually published. Because the relationship between such published data and the underlying individual SLD data is causal and unidirectional, publicly available endpoints such as PFS and ORR can be used to infer parameters in the systems and methods disclosed herein according to a specific set of rules. In some cases, certain parameters can be estimated based on individual SLD data from external data sets. One or more parameters that capture the dynamics of a cohort with an earlier treatment line can be learned from data from those cohorts that are not available in Phase 1. Tumor killing rates for other drugs can be learned from data on patients taking the drug.

[0036] The systems and methods disclosed herein can use Bayesian generative models or probabilistic graphical models to provide a very flexible and general framework for modeling the way a set of observed data may be generated from a set of root causes. Such models can be represented by probability graphs that encode the joint probability distribution between unobserved variables and observed data and can be commonly used in Bayesian statistics and machine learning methods. Subsequently, by adding conditions to the observed data and applying Bayes' rule, this joint distribution can be used to infer causal relationships and parameters of interest. Recently, with the emergence of several probabilistic programming languages ​​such as BUGS, JAGS, Stan, PyMC3, Turing and Pyro, this inference process of Bayesian inversion can be made particularly simple and flexible. The advantages of using Bayesian generative models can include: specifying the data generation process (conditional distribution) is natural and can provide a joint distribution; the model can be specified in a probabilistic programming language (PPL) by only encoding the generation process, because Bayesian inference can include a joint distribution; and implementing any complex distribution, numerical method or constraint.

[0037] The system and method using Bayesian generative model disclosed herein can estimate general tumor dynamic information from published studies containing PFS and ORR information. By directly modeling tumor growth from information in PFS curve graphs and / or risk tables, rather than randomly selecting a fixed event time for each patient (e.g., as in Kay technology), the uncertainty of PFS data can be better considered. In particular, the system and method disclosed herein can use the starting point and end point of the time window as censored observations to determine that a specific patient has not yet occurred or has each occurred an event where the baseline normalized SLD reaches 1.2. By not artificially constraining the event time of each patient to a specific fixed time, the overall distribution of the patient population reflected by the PFS data can more accurately reflect uncertainty, including uncertainty in the PFS event time within a given window and uncertainty caused by SLD measurement error or noise. In addition, the system and method disclosed herein may not assume a specific type of growth (e.g., exponential growth). The system and method disclosed herein can also be used to jointly estimate tumor growth parameters and PKPD parameters of a pharmacokinetic-pharmacodynamic (PKPD) model. This can include parameters indicating the contribution of both baseline (untreated) tumor growth and drug-associated tumor shrinkage to the overall tumor growth rate, which would otherwise be difficult to estimate without the use of supplemental and / or internal PFS data. The systems and methods disclosed herein can also be used to supplement existing non-public studies.

[0038] The systems and methods disclosed herein can provide a deeper and more accurate understanding of the contribution of different indications (e.g., bile duct cancer, pancreatic cancer, testicular cancer, etc.) to patient tumor growth. In addition, the systems and methods disclosed herein can allow modeling of the effects of a drug without running a control group that has not been treated with the drug. This in turn can greatly reduce the amount of time, money, and / or other resources that can be spent before making important decisions such as whether to advance drug research to the next stage (e.g., whether to expand drug trials to a larger patient population).

[0039] The various concepts introduced above and discussed in greater detail below may be implemented in any of a variety of ways, and the concepts described are not limited to any particular implementation. For purposes of illustration, examples of implementation are provided.

[0040] Figure 1 1 is a simplified block diagram of an example system 100 in which the systems and methods disclosed herein (e.g., tumor growth modeling and estimation techniques) may be implemented. System 100 includes a computing system 110 coupled to a patient database 112. Computing system 110 may be a single computing device, or include multiple co-located and / or distributed computing devices communicatively coupled via one or more networks. Figure 1In the illustrated example embodiment, the computing system 110 includes a processing unit 120, a network interface 122, a display 124, a user input device 126, and a memory 128. The processing unit 120 includes one or more processors, each of which may be a programmable microprocessor that executes software instructions stored in the memory 128 to perform some or all of the functions of the computing system 110 as described herein. Alternatively, one, some, or all of the processors in the processing unit 120 may be other types of processors (e.g., an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), etc.), and the functions of the computing system 110 as described herein may be implemented in part or in whole in hardware instead. The memory 128 may include one or more physical memory devices or units including volatile and / or non-volatile memory. Any suitable one or more memory types may be used, such as a read-only memory (ROM), a solid-state drive (SSD), a hard disk drive (HDD), etc.

[0041] The network interface 122 may include any suitable hardware (e.g., front-end transmitter and receiver hardware), firmware, and / or software configured to communicate with external devices and / or systems (e.g., client devices, or one or more servers maintaining the patient database 112) via one or more networks using one or more communication protocols. For example, the network interface 122 may be or include an Ethernet interface, and / or include a wireless local area network (LAN) interface, etc.

[0042] Display 124 may use any suitable display technology (e.g., LED, OLED, LCD, etc.) to present information to the user, and user input device 126 may be a keyboard or other suitable input device. In some embodiments, display 124 and user input device 126 are integrated into a single device (e.g., a touch screen display). Typically, display 124 and user input device 126 may be combined to enable a user to interact with a user interface (e.g., a graphical user interface (GUI)) provided by computing system 110 (such as those discussed in further detail below). However, in some embodiments, computing system 110 does not include display 124 and / or user input device 126, or one or both of display 124 and user input device 126 are included in another computer or system (e.g., a touch screen display) that is communicatively coupled to computing system 110. Figure 1 In a client device not shown in FIG.

[0043] The memory 128 stores instructions for one or more software applications, including a tumor growth estimation application 130 (also referred to herein as a "TGE application 130"). The TGE application 130, when executed by the processing unit 120, is typically configured to determine / learn the distribution of one or more parameters (e.g., tumor growth model parameters) based on progression-free survival (PFS) data stored in the patient database 112, and to determine / learn (jointly estimate) pharmacokinetic-pharmacodynamic (PKPD) model parameters for each of one or more specific patients using the determined (multiple) distributions. The TGE application 130 can display the estimated tumor growth or any information associated with the estimated tumor growth. The TGE application 130 includes a user interface unit 140, a data extraction unit 142, a population modeling unit 144, and a patient modeling unit 146. In general, the user interface unit 140 manages interaction with a user (e.g., a user operating the user input device 126 and viewing the display 124), the data extraction unit 142 manages retrieval of PFS data from the patient database 112 (and / or any pre-processing of the PFS data), the population modeling unit 144 learns distributions of baseline (i.e., no drug treatment or ineffective drug treatment) tumor growth for different patient populations represented by the PFS data, and the patient modeling unit 146 uses the population distribution learned for a particular patient and the tumor size observations over time to model the contribution of both baseline tumor growth and drug treatment to the overall growth rate of the patient's tumor(s). The operation of the TGE application 130 and its various units 140-146 will be discussed in further detail below.

[0044] Patient database 112 may include one or more databases, which may be stored in one or more memories at one or more co-located or remote locations. In some embodiments, patient database 112 is local to computing system 110 (e.g., stored in memory 128). Patient database 112 includes any information related to multiple patients, such as PFS data for each of one or more cancer indications / types for multiple patients. For example, patient database 112 may include a digitized PFS graph and / or a PFS risk table for a particular cancer type (e.g., Figure 2 ), may also include multiple PFS curves and / or PFS risk tables that each correspond to a different cancer type (e.g., pancreatic cancer, bile duct cancer, breast cancer, etc.). In another example, the patient database 112 may include ORR.

[0045] The data extraction unit 142 is generally responsible for retrieving / obtaining the desired PFS data from the patient database 112. In some embodiments, the data extraction unit 142 retrieves PFS data for a particular type of cancer based on user input detected by the user interface unit 140. For example, the user interface unit 140 may generate and / or populate a GUI and cause the display 124 to present the GUI to the user. The user may then operate the user input device 126 to input an indication of the type of cancer of interest via the GUI, and the data extraction unit 142 may retrieve PFS data (e.g., a graph and / or a risk table) corresponding to the indicated type of cancer. In some embodiments, the patient database 112 includes raw data (e.g., anonymous encounter data indicating dates and diagnoses / measurements / etc. from a health care provider), and the data extraction unit 142 constructs a PFS graph and / or a PFS risk table, or some other similar data structure(s). For example, the data extraction unit 142 can construct a digitized PFS curve graph from the raw data, or it can generate the data in a more user-friendly form (e.g., an indexed list of patient-specific entries, each patient-specific entry indicating the beginning and end of the time window in which the corresponding patient experienced a PFS event).

[0046] The PFS data obtained by the data extraction unit 142 may be data corresponding to patients who have not yet received effective drug treatment (e.g., patients who received experimental drug treatment but later proved to be ineffective, and / or patients who chose not to receive drug treatment at all). In this way, the population modeling unit 144 may use the PFS data to learn a “baseline” growth rate pattern for tumors of a particular cancer type.

[0047] The population modeling unit 144 uses the obtained PFS data (possibly after the data extraction unit 142 formats, cleans, and / or otherwise preprocesses the PFS data) to determine a population distribution of one or more parameters (e.g., patient-specific parameters) of the growth curve function, i.e., to determine the distribution across the patient population represented by the obtained PFS data. To this end, the population modeling unit 144 first uses the PFS data to learn one or more parameters (e.g., one or more growth curve parameters) for each patient in the population. In some embodiments, the growth curve of a particular patient (patient i) over time may generally be referred to herein as f(t, θ i ), where θ irepresents one or more growth curve parameters specific to the patient (also referred to herein as "patient-specific parameters" of the growth curve function). In some other embodiments, the growth curve over time for a particular patient (patient i) may be generally referred to herein as R(t), where t represents time. In some embodiments, the growth curve function R(t) includes one or more parameters (e.g., patient-specific parameters) including and f (i) , where t represents time and i is an index representing the i-th patient among a plurality of patients. In some embodiments, the population modeling unit 144 fits a particular type of distribution (e.g., a lognormal distribution) to the censored observations of the patient. The population modeling unit 144 learns the growth curve parameters (multiple) of a given patient as a distribution that provides a range of possible growth curves. It is worth noting that the information in the PFS curve graph or risk table may not exclude the possibility that a PFS event occurs anywhere within a given time window. Reference Figure 2 For example, from the risk table below the curve, it can be seen that two patients in the population had a PFS event somewhere between 4 and 6 months. That is, the PFS event for each of these two patients may have occurred at any time during the 2-month time window.

[0048] Some PFS graphs can provide more detailed information (e.g. Figure 2 , where the trace indicates that the timing of PFS events is much more refined than the 2 month window of the PFS risk table). In this case, the population modeling unit 144 may be able to utilize the more specific timing information to learn a more accurate distribution of the growth curve parameters. However, in some cases, the finer timing is the result of factors that are not fully captured in the modeling (e.g., a patient becomes sicker and is tested before his / her next medication treatment), in which case the population modeling unit 144 may instead ignore the finer timing information of the PFS curve graph and instead use the information of the PFS risk table. This can result in significant uncertainty about the trajectory of a patient's growth curve. As Figure 4As seen in Figure 4, for example, within a time window marked by start time 402 and end time 404, a PFS event for a particular patient may occur at a time corresponding to PFS event 410, PFS event 412, or PFS event 414, or at any other time within the window. This uncertainty in the estimate for each patient affects the uncertainty in the estimate for the entire population. To avoid artificially removing this uncertainty (and thereby losing useful information), the population modeling unit 144 does not assume a fixed time for each patient's PFS event, but instead directly uses the start and end times of the time windows in which it is known that PFS events have occurred (e.g., Figure 4 The time 402 and 404 in the PFS data are used as censored observations to learn the distribution of the growth curve parameters. For example, according to the PFS data (affected by measurement error), the patient's baseline normalized SLD is less than 1.2 at time 402 and at least 1.2 at time 404.

[0049] The population modeling unit 144 fits a particular type of distribution (e.g., a lognormal distribution) to the censored observations of the patient. In one embodiment, for example, the population modeling unit 144 assumes an exponential growth curve such that:

[0050]

[0051] Among them, k g，i is a patient-specific parameter that controls the rate at which the growth curve accelerates, and t represents time. The population modeling unit 144 can then use the two missing observations to determine k g，i In another example embodiment, the population modeling unit 144 assumes a logistic growth function with multiple parameters, for example, such that:

[0052]

[0053] Among them, L i is the patient-specific maximum value of the growth curve, k L，i is a patient-specific parameter that controls the steepness of the growth curve, t is time, and t 0,i is the patient-specific time value at the midpoint of the growth curve. In this example, the population modeling unit 144 may determine the patient-specific parameter k L，i ,t 0，i and L i (or possibly only k L，i and t 0，i , where L i is constant among all patients, etc. In yet other embodiments, population modeling unit 144 may use any other suitable growth curve function and / or growth curve parameter(s).

[0054] Using y(t) to represent the baseline normalized SLD, the population modeling unit 144 can determine the population distribution of y(t) according to the following:

[0055] y(t)~LogNormal(f(t,θ i ),σ),

[0056] P(y(t)<1.2|f(t,θ),σ),

[0057] P(y(t)>1.2|f(t,θ),σ),

[0058] θ i ~LogNormal(μ,ω).

[0059] In the above expression, σ represents a noise parameter, which the population modeling unit 144 can jointly estimate with the parameter(s) θ i The parameters μ and ω are the learned θ i The mean and variance of the lognormal distribution (for example, if θ i Including more than one parameter, there are multiple means and variances). The probability (P) is the probability / likelihood that a given patient has had a PFS event (baseline normalized SLD exceeds 1.2) or that a given patient has not had the event at any time t. In particular, they are the integrals of the probability density functions of the lognormal distribution from 0 to 1.2 and from 1.2 to infinity, respectively.

[0060] In some embodiments, the growth curve of a particular patient (patient i) over time may generally be referred to herein as R(t), where t represents time. In some embodiments, the growth curve function R(t) includes one or more patient-specific parameters, including and f (i) , where t represents time, and i is an index representing the i-th patient in a plurality of patients. In some embodiments, the population modeling unit 144 fits a particular type of distribution (e.g., a lognormal distribution) to the censored observations of the patient. In one embodiment, for example, the population modeling unit 144 assumes that the growth curve is as follows:

[0061]

[0062] Where R0 is the baseline SLD, k g is the baseline growth rate without treatment, k s is the shrinkage / killing rate of drug-sensitive tumor cells. In some embodiments, the killing rate of drug-sensitive tumors k s can be a fixed constant that is greater than any credible baseline growth rate kg This is consistent with empirical observations, but can be further relaxed in the future. f is a ratio between zero and one, indicating the proportion of tumor cells that are sensitive to the drug.

[0063] To estimate the likelihood for a single subject, X it The discrete SLD observations for the i-th patient at observation time t can be included. The observations can be modeled as log-normally distributed around the curve with some observation / model misspecification noise σ:

[0064]

[0065] in, and f (i) is a patient-specific parameter (e.g., an individual-specific tumor growth / inhibitory effect). In some embodiments, since X i0 is the baseline, so it can be modeled as being noise-free and simply

[0066] The progression criterion can be the first time that both of the following are true: 1. The observed SLD value is 1.2 times the minimum observed SLD; 2. The observed SLD value is at least 5 mm greater than the minimum observed SLD.

[0067] At each observation point, the progress criteria may include a threshold above which an observation is considered to be progressing. Formally, the threshold may be

[0068] max{1.2min{X i0 ,…,X 0,t-1},min{X i0 ,…,X 0,t-1}+5}

[0069] If it is assumed that the patient is at time T i , then given their patient-specific parameters, the likelihood of the observed value for this patient can be

[0070]

[0071] …

[0072]

[0073] If the patient is right-censored after time Ti-1, the last term can be removed because the patient's progression status is unknown at this time.

[0074] The above likelihoods may involve multidimensional integrals for which there is no known closed-form solution. When enforcing constraints in Stan, we can itThe likelihood can therefore be a probability density function (PDF for each observation:

[0075]

[0076] With the following constraints:

[0077] X it ≤max{1.2min{X i0 ,…,X i，Ti-1},min{X i0 ,…,X i，Ti-1}+5}

[0078] For observations before the progression time, where , the inequality is greater than for observations at the progression time.

[0079] To summarize the growth dynamics to the population level, the individual / patient specific likelihood can be multiplied by the population distribution, and the product for all patients can be:

[0080]

[0081] In addition to patient-specific observations and population observations, patient-specific parameters can be sampled in Stan instead of marginalizing the patient-specific parameters.

[0082] ORR information from historical studies can be included in the likelihood. ORR or equivalently the number of patients experiencing a response can be given. The overall response rate can be the minimum observed value for a patient's tumor to be less than 0.7 times the baseline. If the minimum observed value for each patient is Mi: = min{X i1 ,···,X iT}, then ORR can be

[0083]

[0084] In some embodiments, a method using obeyed soft constraints may be used for sampling in Stan, and the approximate Bayesian inference may be:

[0085]

[0086] where κ controls the smoothness of the approximation and can be manually tuned from 10 to 50 (e.g., 20). In some embodiments, the likelihood of the density can be

[0087]

[0088] Among them, τ is tuned to force the ORR of each posterior sample of the entire data set to be close to the truly observed ORR. Since ORR is a discrete number, most of the posterior samples generated by this satisfy the ORR constraint, but a few may not. The method disclosed herein can use posterior sampling and use soft constraints to enforce the ORR condition. In some embodiments, all posterior samples can be rejected when the ORR constraint does not satisfy the approximate Bayesian calculation (ABC).

[0089] In some embodiments, progression may include when the SLD reaches a certain threshold. In some other embodiments, the progression criteria may include not only when the SLD reaches a certain threshold, but also non-target progression events, such as the appearance of new lesions, progression of non-target lesions, worsening symptoms, or death. To this end, these non-target progression events can be modeled using a hazard function (which itself is a function of tumor growth inhibition dynamics) in the following form:

[0090]

[0091] The hazard function can include a survival function that can be incorporated into the likelihood of an individual patient Specifically, for patients who progress at time T, their observed SLD value X at that time is iT may not be greater than the SLD progression threshold and therefore no longer has a lower limit. However, if X iT If t is below the threshold, and the patient progresses at time T, then the patient is likely to have had a non-target progression event during the interval between T-1 and T. Therefore, the likelihood can be increased by:

[0092]

[0093] Otherwise, if X iT If the SLD progression threshold is greater than the SLD progression threshold, the patient may not have had non-target progression before time T, but the possibility is not excluded and therefore its likelihood is increased by:

[0094]

[0095] In some embodiments, the GUI provided by the user interface unit 140 enables the user to select or input a desired type of growth curve function and / or (multiple) growth curve parameters (e.g., from a drop-down list of options), and the selected function / (multiple) parameters are used by the population modeling unit 144. In yet other embodiments, the GUI provided by the user interface unit 140 enables the user to select a specific cancer type, after which the data extraction unit 142 obtains (from the patient database 112) PFS data corresponding to a patient population with that cancer type, and the population modeling unit 144 automatically selects an a priori growth function suitable for the selected cancer type.

[0096] As with the growth function, the population modeling unit 144 may assume any suitable type of distribution for the distribution of the entire patient population represented by PFS. i Lognormal distribution, indicating that a particular parameter μ itself can follow the distribution μ j ~LogNormal(α, τ), or where x i As an indicator or to study the distribution of a specific covariate j ~LogNormal(α+βx i , τ), etc. In some embodiments, the GUI provided by the user interface unit 140 enables the user to select or input (multiple) parameters θ i The desired type of distribution is selected (e.g., from a drop-down list of options), and population modeling unit 144 uses the selected distribution.

[0097] As described above, modeling directly from censored observations (i.e., two censored observations per patient in the population) can provide several advantages. Through a direct modeling approach, the inherent uncertainty in the PFS curve graph and risk table (discussed above) can be more accurately quantified, thereby more accurately depicting when a PFS event may occur. The system and method can better account for other sources of uncertainty, such as SLD measurement / observation noise. Noisy SLD measurements may be very common and may result in, for example, detecting a PFS event in the wrong time window (e.g., if the PFS event actually occurs just before the start of the time window, but the measurement for (multiple) tumor sizes is slightly biased). Figure 5 An example of SLD noise is shown in , where each point represents a different observation / measurement value.

[0098] In addition, the direct modeling method is highly flexible, allowing different growth curve functions and / or distribution types, as discussed above. For example, by directly modeling from PFS data with a probabilistic programming language, the growth curve can take any form, as long as the trajectory inserted into the likelihood / probability can be forward solved. Different studies can also be modeled as having non-zero effects, as long as certain assumptions are made about specific PKPD characteristics. The population distribution of the growth curve can also be summarized in a probabilistic programming language. For example, a more flexible distribution with more parameters or specific characteristics (e.g., biological modalities, sparse or heavier tails) can be used. Indication level parameters can come from a hierarchical distribution across indications (cancer types), thereby allowing the collection of information (especially for studies with less information). The population modeling unit 144 can regress the indication level parameters, for example, to model the progression rate according to the previous treatment line. Figure 6 Some example population distributions for different cancers / disease indications / types are shown in .

[0099] As yet another advantage, the parameters (e.g., one or more parameters θ i Joint estimation of σ and noise parameters σ can better inform the learning of these parameters.

[0100] After the population modeling unit 144 determines / learns the distribution of parameters of baseline tumor growth in the population represented by the PFS data, the patient modeling unit 146 uses the distribution(s) and the observed / measured SLD values ​​of a particular patient over time to model the contribution of both baseline tumor growth and drug treatment to the overall tumor growth rate of the patient. More specifically, the patient modeling unit 146 uses the PKPD model to jointly estimate expression parameters of the characteristics that include both contributions. The patients to whom the PKPD model is applied can be patients with the same cancer type as the patients reflected in the PFS data, or patients with cancer types with known tumor growth characteristics that are very similar.

[0101] As described above, the techniques described herein may assume or apply any suitable type of growth curve (exponential, logistic, etc.). In some embodiments assuming exponential growth, the patient modeling unit 146 may jointly estimate tumor growth parameters for a particular patient using the following PKPD model:

[0102]

[0103] Where T' is the patient's tumor growth rate, T is the patient's tumor size, and k g is the exponential growth parameter of the patient (e.g., according to Equation 1), k maxis the maximum killing rate of the patient, C is the drug concentration in the patient's plasma, and EC50 is the half-maximal effective concentration of the patient. In some embodiments, the patient modeling unit 146 models the tumor growth of the patient using equations with more or fewer parameters and / or in different formats. For example, each instance of C in Equation 4 can be multiplied by another tumor perfusion parameter p, which represents the ratio of the drug concentration at the patient's (multiple) tumor sites to the drug concentration in the patient's plasma. The perfusion parameters can also be jointly estimated, or fixed values ​​can be assumed (e.g., depending on the cancer type), etc.

[0104] The patient modeling unit 146 can learn the patient-specific parameters of Equation 4 (or a patient-specific distribution of the parameters of Equation 4) by attempting to fit Equation 4 to the measured / observed patient tumor size / growth subject to certain constraints (i.e., over time and during a series of two or more patient encounters). For example, if exponential growth is assumed, the patient modeling unit 146 can use the determined k g The population distribution is used to set the k in Equation 4 g As a more specific example, the patient modeling unit 146 may determine the k corresponding to the 90% certainty interval in the population distribution. g upper and lower limits, and then in the case of k for a particular patient g Use these values ​​as upper and lower bounds when making estimates.

[0105] The patient modeling unit 146 can also use the information prior to set constraints on one or more other parameters in Equation 4. For example, k max The informative prior for can be 0.025 (eg, for the mean of the lognormal distribution), as provided by the fastest responding patients in the population, but variations from this value are allowed in the joint estimation process.

[0106] Once the patient modeling unit 146 has estimated the parameters and / or parameter distributions, the user interface unit 140 may cause a display (e.g., display 124) to present a visual indication of the estimated tumor growth for the particular patient in any suitable format. For example, for a particular patient, the user interface unit 140 may output Figure 7 The graphs / information shown in any of the columns shown in the Figure 7 In the graph, for a particular patient, each column represents a specific combination of drug dose (in this example, 20 mg or 60 mg) and cancer indication (in this example, testicular cancer or colorectal cancer).

[0107] In the top row of each column, each dot represents the observed / measured drug concentration in the plasma of the corresponding patient, while the solid line / trace represents the average of the concentrations in plasma estimated by the patient modeling unit 146 using a PKPD model similar to Equation 4. Figure 7 4, but the shaded area around the solid line / trace can be displayed (e.g., to represent a 90% confidence interval for the concentration in plasma). In the bottom row of each column, each point represents an observed / measured baseline normalized SLD, and the top horizontal line indicates that the value of the baseline normalized SLD is 1.2. The solid line / trace in the bottom row represents the average of the estimated baseline normalized SLD for the corresponding patient, and the dashed line / trace in the bottom row represents the estimated baseline normalized SLD for the corresponding patient under the counterfactual scenario where the drug is not administered to the patient (i.e., the estimated tumor growth of the patient in the absence of drug treatment). In the bottom row, the solid line / trace corresponds to T' of Equation 4, and the dashed line / trace corresponds to the first term on the right side of Equation 4. The shaded area around the dashed line represents a 90% confidence interval. Although in Figure 7 Not shown, but the shaded area around the solid line may also represent a 90% confidence interval.

[0108] By showing the tumor growth trajectory of a particular patient relative to the actual trajectory (i.e., relative to the real-world observations and / or the estimated overall growth rate) under the counterfactual scenario that the particular patient did not receive drug treatment, such as Figure 7 Graphs such as those shown in can enable a user to accurately assess the efficacy of a drug at the indicated dose level and for the indicated cancer type. For example, for 60 mg / colorectal cancer, a user may be able to determine that the low overall growth rate is due to the low baseline tumor growth for that patient, rather than the efficacy of the drug / dose for that patient. On the other hand, for 20 mg / testicular cancer, a user may determine that the negative growth rate is largely due to the efficacy of the drug, given that the baseline growth would have been positive. And for 60 mg / testicular cancer, although the growth rate is quite high, a user may determine that the drug is effective because the baseline growth rate is significantly higher than that growth rate.

[0109] Figure 8 is a flow chart of an example method 800 for estimating tumor growth. For example, method 800 may be performed by computing system 100 (eg, processing unit 120) when executing instructions of TGE application 130. In some embodiments, method 800 may include a Bayesian generative model.

[0110] Method 800 includes step 802, in which one or more processors (e.g., one or more processors of computing system 110) can obtain (e.g., from patient database 112 or local memory) progression-free survival (PFS) data of multiple patients. Patients (or subjects) can include any organism or non-organism, including but not limited to humans (e.g., male humans, female humans, fetuses, pregnant women, children, etc.), non-human animals, plants, bacteria, fungi, or protists. Any human or non-human animal can serve as a patient, including but not limited to mammals, reptiles, birds, amphibians, fish, ungulates, ruminants, bovines (e.g., cattle), horses (e.g., horses), caprines and ovines (e.g., sheep, goats), swine (e.g., pigs), camelids (e.g., camels, llamas, alpacas), monkeys, apes (e.g., gorillas, chimpanzees), bears (e.g., bears), poultry, dogs, cats, mice, rats, fish, dolphins, whales, and sharks. In certain embodiments, the patient (or subject) is a male or female (e.g., male, female, or child) at any stage. The multiple patients can be patients associated with one or more drug treatments. The one or more drug treatments can include effective drug treatments (e.g., successful trials), neutral drug treatments, or ineffective drug treatments (e.g., failed trials). The details of one or more drug treatments are described elsewhere herein.

[0111] Progression-free survival (PFS) may include the time from the start of treatment to the occurrence of disease progression or death. In some embodiments, disease progression may be performed according to the solid tumor remission evaluation criteria (RECST), that is, the sum of the maximum tumor diameter increases by at least 20%, any new lesions appear, or unmeasurable malignant diseases clearly increase. In some embodiments, PFS may be used as an endpoint. The endpoint may include the targeted results of a clinical trial to determine the efficacy and safety of the therapy being studied. The endpoint of a clinical trial may include one or more clinical outcome assessments and / or surrogate endpoints. In some embodiments, the clinical outcome assessment may include cure, clinical deterioration, and death. Surrogate endpoints may be clinical trial endpoints, used as a direct measure of a patient's sense, function, or survival mode. PFS data may include any information associated with PFS, including but not limited to a PFS graph, a digitized PFS graph, or a PFS risk table, as described elsewhere herein. PFS data may include published PFS data or unpublished PFS data. PFS data may indicate at each observation time in a plurality of observation times and / or within a recent time window, how many patients in a plurality of patients have a baseline normalized sum of the longest diameters (SLD) measurement of at least 1.2 or new lesions appear. The multiple patients represented by the PFS data may include patients known to be associated with one or more drug treatments, as described elsewhere herein.

[0112] PFS data can indicate (i) multiple observation times, and / or (ii) how many patients have PFS events in each observation time of the multiple patients in the multiple observation times. The multiple observation times can include any time range, for example, at least one month, two months, three months, four months, five months, six months, one year, five years or longer. In certain embodiments, the multiple observation times can include any time range, for example, at most five years, one year, six months, five months, four months, three months, two months, one month or shorter time. The multiple observation times can be at least 1 hour, 1 day, 1 month, 1 quarter, 1 year or longer time. In some other embodiments, the multiple observation times can be at most 1 year, 1 quarter, 1 month, 1 day or shorter time. PFS events can include any event associated with PFS, including but not limited to the progression of non-target lesions, the SLD observation value of the baseline normalized SLD is 1.2 times the minimum SLD observed, the SLD observation value of the baseline normalized SLD is at least 5mm larger than the minimum SLD observed, or the situation of new lesions appearing in the most recent time window. The most recent time window may be the same or different from a plurality of observation times. The most recent time window may be at least 1 hour, 1 day, 1 month, 1 quarter, 1 year or longer. In some other embodiments, the most recent time window may be at most 1 year, 1 quarter, 1 month, 1 day or shorter. The number of multiple patients at each observation time in the plurality of observation times where a PFS event has occurred may be at least 10, 100, 1000, 10000 or more. The number of multiple patients at each observation time in the plurality of observation times where a PFS event has occurred may be at most 10000, 1000, 100, 10 or less. In some embodiments, PFS data may include how many patients have had a PFS event in a plurality of patients at a subset of a plurality of observation times. The subset of the plurality of observation times may be at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more of the plurality of observation times. In some embodiments, the subset of the plurality of observation times may be at most 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10% or less of the plurality of observation times. The PFS data may correspond to one or more specific cancer types. The one or more specific cancer types may include, but are not limited to, breast cancer, colorectal cancer, esophageal cancer, head / neck cancer, lung cancer, lymphoma, ovarian cancer, pancreatic cancer, prostate cancer, kidney cancer, or uterine cancer.

[0113] In step 804, one or more processors may determine a population distribution of one or more patient-specific parameters of a growth curve function. A population distribution may include the distribution of one or more parameters or characteristics (e.g., patient-specific parameters) between multiple individuals in a population. The one or more parameters may include patient-specific parameters and non-patient-specific parameters (e.g., population parameters). The one or more patient-specific parameters may include tumor growth parameters, PKPD parameters, or growth curve parameters. The one or more patient-specific parameters may include individual parameters, SLD observations, or non-target progression events. Population parameters may be determined by a population, for example, by prostate cancer patients who did not receive chemotherapy in a Phase 3 study of anti-androgen therapy. For example, in Fig. 9 In , the population parameters may include μ~HalfNormal(0,0.1) and T~HalfNormal(0,1). Individual / patient specific parameters may be plotted conditional on the population parameters. For example, in Fig. 9 Individual / patient specific parameters may include and f~LogitNormal(μ f , τ f ). SLD observations and non-target progression event times can be plotted conditioned on the kinetics of individual / patient specific parameters. Fig. 9 In the example, the SLD observation can be X t ~LogNormal(R(t),σ), and the non-target progression event can be N~h(t)=exp{λ0+λ1R(t)+λ2R′(t)}.

[0114] In some embodiments, determining the population distribution of the one or more patient-specific parameters may include determining a growth curve function that includes the one or more patient-specific parameters. The growth curve function may include an exponential growth function, a logistic growth function, or an ordinary differential function. The growth curve function may include Equation 1, Equation 2, or Equation 3, as disclosed elsewhere herein. The details of the growth curve function are described elsewhere herein. The population distribution of the growth curve can be summarized in a probabilistic programming language. The growth curve function may be time-dependent. In this case, the growth curve function may vary based on multiple observation times. In some embodiments, the method may further include determining the population distribution of one or more parameters via one or more processors.

[0115] In some embodiments, the one or more patient-specific parameters may include θ i , and with θ i The associated growth curve function can be f(t, θ i ), as described elsewhere herein. In some embodiments, a growth curve function f(t, θ i )i , where t represents time and i is an index representing the i-th patient among a plurality of patients. If exponential growth is assumed, then for example the patient-specific parameter(s) θ i You can include the parameter k g , and the growth curve function may be as shown in Equation 1. In some embodiments, the one or more patient-specific parameters may include and f (i) , and with The growth curve function associated with f(i) may be R(t), as described elsewhere herein. In some embodiments, the growth function R(t) may be determined and f (i) , where t represents time, and i is an index representing the i-th patient in a plurality of patients. In some embodiments, A baseline normalized sum of longest diameters (SLD) measurement may be included. In some embodiments, f (i) The ratio of drug-sensitive tumor cells in the i-th patient among the plurality of patients may be included. In some embodiments, f (i) can range from 0 to 1. In some embodiments, the growth curve function is a logistic growth function; and the one or more patient-specific parameters include a plurality of parameters of the logistic growth function.

[0116] In various embodiments, determining the population distribution of one or more patient-specific parameters may include determining a single population distribution of a single patient-specific parameter, determining the corresponding population distribution of two or more patient-specific parameters, or determining a single joint population distribution of two or more patient-specific parameters. The one or more patient-specific parameters may include a baseline normalized sum of longest diameters (SLD) measurement parameter. The baseline normalized SLD measurement may be determined based on PSF data. The details of the baseline normalized SLD measurement are described elsewhere herein. The one or more patient-specific parameters may include a growth rate parameter (e.g., a tumor growth parameter). The growth rate parameter may include a baseline growth rate parameter under treatment or without treatment. The one or more patient-specific parameters may include a parameter for the proportion of drug-sensitive tumor cells in the patient's body of the multiple patients. The parameter value range of the proportion of drug-sensitive tumor cells in the patient's body of the multiple patients is 0 to 1. In some embodiments, the parameter value of the proportion of drug-sensitive tumor cells in the patient's body of the multiple patients may be at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more. In some embodiments, the parameter value of the proportion of drug-sensitive tumor cells in patients among the plurality of patients may be at most 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10% or less.

[0117] Determining the population distribution of the one or more patient-specific parameters may include fitting the growth curve function to the observations at the multiple observation times. Determining the one or more population distributions includes fitting the growth curve function to the observations at the subset of the multiple observation times. The subset of the multiple observation times may be automatically selected via a mathematical model or manually selected by one or more users. For example, these observations may include indicating at a first time for each patient in the multiple patients that the patient did not have a first observation of a PFS event before the first time, and indicating at a second time that the patient had a second observation of a PFS event before the second time. (Multiple) population distributions may include any type of distribution, including but not limited to lognormal distribution, Bernoulli distribution, uniform distribution, binomial distribution, normal or Gaussian distribution, exponential distribution or Poisson distribution.

[0118] In step 806, one or more processors may obtain measured tumor growth data for a particular patient undergoing drug therapy (e.g., from the patient database 112 or a local memory such as the memory 128). The tumor growth data may include any information related to tumor growth, including but not limited to the diameter of the tumor over time, the volume of the tumor over time, or the tumor growth rate. The tumor growth data may be obtained by any type of technology, such as imaging studies, including mammography, ultrasound (US), and magnetic resonance imaging (MRI). The drug therapy may include one or more cancer therapeutic agents, including but not limited to chemotherapeutic agents, targeted cancer therapeutic agents, differentiation therapeutic agents, hormone therapeutic agents, and immunotherapeutic agents. For example, the treatment may be one or more chemotherapeutic agents selected from the group consisting of: alkylating agents, antimetabolites, anthracyclines, antitumor antibiotics, cytoskeletal disruptors (taxanes), topoisomerase inhibitors, mitotic inhibitors, corticosteroids, kinase inhibitors, nucleotide analogs, platinum-based agents, and any combination thereof. In certain embodiments, treatment is one or more targeted cancer therapeutic agents selected from the group consisting of the following: signal transduction inhibitors (e.g., tyrosine kinase and growth factor receptor inhibitors), histone deacetylase (HDAC) inhibitors, retinol receptor agonists, proteosome inhibitors, angiogenesis inhibitors and monoclonal antibody conjugates. In certain embodiments, treatment is one or more differentiation therapy agents, including retinoids, such as retinoic acid, alitretinoin and bexarotene. In certain embodiments, treatment is one or more hormone therapy agents selected from the group consisting of the following: antiestrogens, aromatase inhibitors, progestins, estrogens, antiandrogens and GnRH agonists or analogs. In one embodiment, treatment is one or more immunotherapeutics including monoclonal antibody therapy.

[0119] In step 808, one or more processors may estimate tumor growth for a particular patient (e.g., a patient having the same cancer type as the plurality of patients) based on (i) the measured tumor growth data and (ii) the determined population distribution. Step 808 may include modeling the tumor growth rate of the particular patient using a PKPD model (e.g., Equation 4 or the like), the model having a first term representing a change in tumor size for the particular patient without drug treatment and a second term representing a contribution to the change in tumor size for the particular patient due to drug treatment, and jointly estimating one or more parameters of the first term and one or more parameters of the second term by fitting the PKPD model to the measured tumor growth data. The joint estimation may include setting the first term (e.g., k) using the population distribution(s) determined at block 804. g ). For example, the second parameter(s) may include the drug concentration in the patient's plasma, the patient's k max (maximum killing rate) and / or half maximal effective concentration (EC50). Constraints can include any population priors.

[0120] Estimating the tumor growth may further include obtaining an overall response rate (ORR) for a particular patient undergoing drug therapy. ORR may include multiple patients experiencing remission. Remission may include observing that the patient's tumor growth satisfies one or more remission conditions. Remission conditions may include that the patient's tumor size is less than 0.7 times the baseline SLD (or 30% less than their baseline SLD). In some embodiments, ORR may be calculated for each sample and fixed to a value. For example, accurate inference may sum all combined probabilities of remitters. In contrast, this may be approximated by setting the simulated proportion of remitters close to what is observed: The simulated responder status of each patient can be represented by the inverse logarithm of the approximate response condition: whether the patient had an SLD measurement or not, X t All are 30% smaller than their baseline SLD (remission condition). In some embodiments, estimating tumor growth can further include obtaining one or more non-patient-specific parameters. In this case, these non-patient-specific parameters can include fixed values. Such non-patient-specific parameters can include a kill rate.

[0121] Estimating tumor growth can further include obtaining one or more non-target progression events for a specific patient undergoing drug treatment. In the case where the progression is not due to SLD reaching a certain threshold, non-target progression events (such as the appearance of new lesions, the progression of non-target lesions, symptom worsening or death) can be included in a population distribution or growth curve to determine tumor growth. The one or more non-target progression events can be included by modeling the time of these non-target events using a hazard function. The hazard function may include a tumor growth inhibition kinetic function. In some embodiments, the PFS data corresponds to a specific cancer type; and estimating the tumor growth of the specific patient may include estimating the tumor growth of a patient diagnosed with the specific cancer type.

[0122] In step 812, one or more processors may cause a display (e.g., display 124) to present a visual indication of the estimated tumor growth of a particular patient. For example, step 812 may include generating and / or populating a user interface. In some embodiments, step 812 includes causing a display to display a tumor growth trajectory of a particular patient when the particular patient has not yet undergone drug therapy. In this case, the visual indication of the estimated tumor may be presented to a health / research professional on a user interface, and the health / research professional may prescribe one or more drug therapies. In some other embodiments, step 812 includes causing a display to display a tumor growth trajectory of a particular patient when the particular patient has undergone one or more drug therapies. In this case, the visual indication of the estimated tumor may be presented to a health / research professional on a user interface, and the health / research professional may adjust the dosage, regimen, or anti-drug antibodies associated with the one or more drug therapies. In this case, such changes in dosage, regimen, or anti-drug antibodies associated with one or more drug therapies may be displayed to the patient via one or more processors.

[0123] The methods and systems disclosed herein may further include adjusting the dosage of the drug therapy based on the estimated tumor growth of the particular patient. For example, if the tumor growth of the particular patient is below a predetermined threshold (e.g., a predetermined tumor growth rate), the dosage of the drug therapy can be adjusted to a lower dose. In another example, if the tumor growth of the particular patient is above a predetermined threshold, the dosage of the drug therapy can be adjusted to a higher dose. In addition, in some embodiments, the estimated tumor growth of the particular patient can be used to evaluate the efficacy of the drug therapy, evaluate the efficacy of a combination of one or more drug therapies, or compare the direct comparison results of each drug therapy as a monotherapy and as a combination of one or more drug therapies.

[0124] Additional considerations related to the present disclosure will now be addressed.

[0125] Some figures described herein show example block diagrams with one or more functional components. It will be understood that such block diagrams are for illustrative purposes, and the described and illustrated devices may have additional, fewer, or alternative components than those shown. Additionally, in various embodiments, components (and the functions provided by the corresponding components) may be associated with or otherwise integrated as a part of any suitable component.

[0126] Embodiments of the present disclosure relate to non-transitory computer-readable storage media having computer code thereon for performing various computer-implemented operations. The term "computer-readable storage medium" is used herein to include any medium capable of storing or encoding a series of instructions or computer code for performing the operations, methods, and techniques described herein. The medium and computer code may be a medium and computer code specifically designed and constructed for the purposes of embodiments of the present disclosure, or the medium and computer code may be of a type known and accessible to those skilled in the art of computer software. Examples of computer-readable storage media include, but are not limited to: magnetic media, such as hard disks, floppy disks, and tapes; optical media, such as CD-ROMs and holographic devices; magneto-optical media, such as optical disks; and hardware devices specifically configured to store and execute program code, such as ASICs, programmable logic devices ("PLDs"), and ROM and RAM devices.

[0127] The example of computer code includes the machine code such as produced by the compiler and the file containing the higher level code executed by the computer using the interpreter or the compiler.For example, Java, C++ or other object-oriented programming languages ​​and development tools can be used to implement the embodiment of the present disclosure.Additional examples of computer code include encrypted code and compressed code.In addition, the embodiment of the present disclosure can be downloaded as a computer program product, which can be transmitted to the requesting computer (for example, client computer or different server computers) from a remote computer (for example, a server computer) via a transmission channel.Another embodiment of the present disclosure can replace machine executable software instructions with hard-wired circuitry or be implemented in combination with machine executable software instructions.

[0128] Examples

[0129] Bayesian generative model for published PFS and ORR data

[0130] Method and system disclosed herein are developed for estimating tumor growth.In certain embodiments, a Bayesian generative model is developed, which specifies the joint distribution of one or more parameters, and the one or more parameters include test-specific parameters, group-specific parameters, individual patient parameters, individual SLD observations, non-target progression events, and PFS and ORR reported in published studies.The model includes semi-mechanism population components, to capture the longitudinal tumor dynamics at individual and population levels.In addition, the model includes a novel component based on RECIST standards, which specifies individual progression time and relief distribution conditions with individual tumor dynamics as conditions.RECIST standards are standards for determining whether a tumor disappears, shrinks, remains intact, or becomes larger, including complete remission (CR), partial remission (PR), stable disease (SD), and disease progression (PD).Joint distribution allows one or more parameters to be jointly estimated using standard Bayesian estimation with the observed data from published studies as conditions.

[0131] The dynamics of individual tumors were described using a standard two-state model representing two tumor subpopulations, each with linear growth and lethality:

[0132] R1=k g R1-k kkill R1

[0133] R′2=k g R2-k kkillr R2.

[0134] Here, R1 is a parameter representing the drug-sensitive cell subpopulation, and R2 represents the killing rate k kkillr <k kkill (k kkillr is the killing rate of more resistant cells, and k kkill The baseline growth rate is k. g The subpopulation is associated with the proportion of drug-sensitive tumor cells in the patients of the plurality of patients. The overall tumor burden / function R(t) is equal to the weighted sum of R1(t) and R2(t), where the parameter f determines the proportion of drug-sensitive tumor cells in the patients of the plurality of patients and the proportion of drug-insensitive tumor cells in the patients of the plurality of patients. Therefore, the linear model has the following closed form solution through the familiar double exponential form:

[0135]

[0136] Where R0 is the initial SLD, k s :=k g -k kkill , and kr :=k g -k kkillr .

[0137] Initial SLD, growth rate, and proportion of drug-sensitive cells are individual-specific parameters (or patient-specific parameters) and were assigned according to the following distribution among patients in the same study regardless of the group:

[0138]

[0139] f~LogitNormal(μ f , τ f )

[0140] where σ can be, and RegLogNormal(μ, σ, c) represents the regularized log-normal distribution, such that if Y ≈ N(μ, τ), then This distribution is log-normal on (0, c). For numerical stability and to avoid k g The value of is too large, c is fixed at 0.02 / day, which is equivalent to the progression time of log(1.2) / 0.02≈9 days. kkill and k kkillr It is not modeled patient-specifically but is determined by the group of the study that the patient belongs to. In practice, fixed values ​​for the kill rate parameter within a group are sufficient to describe the data and are justified in later trials where doses within groups are standardized.

[0141] The SLD observation value X at time t t is modeled as a lognormal distribution with respect to R(t) conditioned on patient-specific parameters:

[0142] X t ~LogNormal(R(t),σ).

[0143] The time to a non-target progression event according to RECIST criteria - event N, i.e., the time to progression of non-target lesions or appearance of new lesions, was modeled conditional on patient-specific parameters using a hazard model based on:

[0144] h(t)=exp{λ0+λ1R(t)+λ2R′(t)}.

[0145] The hazard is modeled as proportional to the absolute tumor burden or growth rate and its rate of change. This captures the following situation: the dynamics of non-target lesions are related to target lesions, and when the existing tumor burden is large and growing, it is intuitively more likely that new lesions will appear. Although the parameters σ and λ cannot be identified solely by PFS and ORR data, their values, or at least the prior distribution of their values, are obtained by fitting the model on actual SLD data that are generally not publicly available. The parameters are set to the following fixed values ​​obtained from point estimates of the model fitted to private SLD data from an internal program: σ = 0.1, λ1 = -6.4, λ2 = 0.014, λ2 = 2.6. For the values ​​of λ, all parameters were found to be significant in the sense that at least 90% of the posterior masses were far from zero.

[0146] The last observation time of a patient and whether they progressed at that time is captured by the tuple (T, E), as Fig. 9 This quantity is modeled conditional on a vector of SLD measurements for one or more patients (denoted as X) and the time of a non-target progression event N. T is set to the last observed time of the patient. E is determined conditional on X and N according to RECIST criteria. For example, when E=1, if a patient has target or non-target progression at time T, then progression occurs at T=1. By definition, when N=T, a non-target progression event occurs at time T, while the relationship between target progression and X is slightly more complex. According to RECIST, target progression occurs at time T when X T 1.2 times greater than the minimum value of X observed up to time T, and / or X T is also at least 5 mm greater than the observed minimum. In addition, for all t < T, X t Less than or equal to this threshold, target progression will not occur earlier. This is expressed as the following inequality, which satisfies:

[0147]

[0148] Similarly, when a patient drops out at time T without progression, or more specifically when N>T and all X (including up to time T) are less than the aforementioned threshold, E=0. Bayesian inference conditions on or holds fixed the data (T, E), and the values ​​of the unknown quantities X and N that explain the fixed (T, E) are estimated via Bayes' rule and in this case Markov Chain Monte Carlo (MCMC) sampling. To capture right censoring in published studies, missingness is treated as a time-event occurring at time T when E=0. It is modeled using a constant group-specific hazard rate d.

[0149] Whether the patient experienced any remission and whether they experienced a complete remission is captured by the tuple (O, C). Like time to progression and status, this quantity is also modeled conditional on X and N. If X t If the minimum value of X is less than 0.7R0 and no non-target progression event has occurred up to (and including) that time, then O is set to 1, otherwise O is set to 0. t If the minimum value of X is less than 2 mm, C, which indicates complete remission, is set to 1. Although complete remission is described as complete disappearance of all lesions in the RECIST criteria, this boundary was chosen for two reasons. First, numerically, X t The minimum value of is a positive number and thus the actual threshold is selected. Secondly, this is a sufficiently small value in practice so that the disappearing lesion is no longer found in the imaging.

[0150] The entire generation process of the data (e.g., the joint distribution structure) is Fig. 9 is specified in and summarized by the following joint distribution of data and parameters:

[0151]

[0152] exist Fig. 9 The bottom left graph shows that individual specific parameters have a large impact on latent tumor dynamics. The bottom middle graph shows the potential X, N observations associated with the actual potential SLD observations. The bottom right graph shows that the set of random variables T, E, O, C is associated with the actual published PFS and ORR.

[0153] Bayesian inference of model parameters given observed data

[0154] Inference of the model parameters is performed using the standard Bayesian inversion of the data generating process facilitated by MCMC posterior sampling. Specifically, prior values ​​for the model parameters are specified and multiplied by the above conditional distribution to obtain the joint distribution of the model parameters and the observed data. The observed data are then fixed to the observed values, e.g., conditioned on the observed values, and the resulting unnormalized posterior density of the model parameters is sampled using MCMC.

[0155] For most parameters, priors were set to allow for a wide range of plausible values ​​that allow the model to explain the data from the various examples in published studies shown in the Results section, while also reducing the probability that the values ​​are unrealistic and numerically unstable. Specifically, some parameters are and k kkill ~HalfNormal(0,0.02). The latter places the prior quality at k kkill , while avoiding the rapid kill rate that would lead to numerical instability in the danger score. kkillr~Exponential(1000) achieves two goals. First, it places a significant amount of prior mass at zero, which facilitates the g to determine the growth kinetics of the more resistant cells in the control / placebo group. This helps with identifiability issues without having to consider the k kkillr Second, because the exponential distribution has a heavier tail, it allows enough flexibility in practice to better capture the possible long-term benefits of treatment relative to control or placebo, as the faster killing rate k s , mainly capturing early remission rather than longer term progression behavior. f , we tried a fixed value of 0, corresponding to f = 0.5. However, we found that a more relaxed μ f ~N(0, 0.5) provides a better fit in practice, while still promoting the SLD trajectory to maintain the characteristic double exponential U-shape often seen in practice. The prior values ​​of the variance parameters are set to and τ f ~HalfNormal(0,2). This avoids large inter-individual variability and numerical instability while allowing sufficient flexibility to fit observational data and, in particular, the observed number of complete responses seen in some published studies. and is set to a fixed value that provides a reasonable R0 value based on the distribution R0 seen in internal data and the number of measurable patients reported in published studies, for example, an initial R0 greater than 10 mm. For the study in the example section, and Because this allows for measurements in approximately half of the patients, which is roughly the number typically seen in published studies used in the Results section.

[0156] Observational data (e.g., PFS data) were extracted from published studies, including PFS curves and risk tables, ORR, and the number of complete responders to incorporate as much information as possible into the model. The risk table and PFS curve were combined using the standard formula of Kaplan-Meier estimation to obtain the number of patients M who progressed and withdrew at each time τ in the risk table, respectively. τ1 and M τ0 . The likelihood of the progress data is then set according to the following equation:

[0157]

[0158] Where the triple subscripts denote the i-th patient in the group whose last observation time is τ and who has progressed or not progressed, respectively. Because T and E are determined from X and N via the RECIST rules, when the former is set to a fixed value, this provides information about the possible values ​​that the latter may take, which in turn provides information about the individual kinetic parameters to form the data generation chain. In embodiments, the individual likelihood term p(T τ·i =τ,E τ·i =·|X τ·i , N τ·i ) is not a literal expression for the joint probability density included in Stan. Instead, it is an expression for X τEi and N τEi A set of constraints for R0 are implemented as inequality constraints on those specific parameters in Stan. As a simple example, if T equals the first post-baseline scan time, and E=1, then X1 is constrained to be greater than both 1.2R0 and 5+R0 or N=1.

[0159] Information about the number of overall responders and complete responders (e.g., patients who responded to treatment) is incorporated into the likelihood in a slightly different way. Published studies typically specify the number of overall responders and complete responders, but do not specify which patients are responders, thus making the likelihood τ·i and C τ·i The value of is unknown. Nevertheless, the flexibility of probabilistic programming languages ​​such as Stan allows conditioning on this information in the joint probability density. The number of observed total responders and complete responders is expressed as and The information is incorporated into the model using the following probability statements:

[0160]

[0161] where the sum is taken over the remission status of all patients and the σ term is an auxiliary hyperparameter that is set to a small value to keep the sum values ​​of these quantities in the model close to their observed values. Typically, these are set to 1.

[0162] Simulating data

[0163] To demonstrate the model's ability to recover tumor kinetics using PFS and ORR information, simulated data sets with known ground truth parameter values ​​were simulated and extrapolated back. The population-level growth kinetic parameters shared between groups were set to μ f =0.5, μ f = 2. Two groups, control group and treatment group, were simulated with 100 patients, each with a group-specific killing parameter k kill-ctrl =0.01, k killr-ctrl = 0 and kkill-trt =0.02, k killr = 0.003. Scans were performed every two months. As in the case of the real data, the model uses information from the resulting risk table and the number of mitigaters.

[0164] Fig.10 Several posterior samples of the SLD for a single patient (e.g., X) from the control group who progressed at the second post-baseline scan are shown, with the ground truth superimposed in red. Conditioning on this information is equivalent to placing constraints on X and generating posterior samples from the space of possible SLD values ​​that are likely to cause the patient to progress at the second post-baseline scan. X has a joint posterior distribution that also depends on other parameters not depicted here, such as R0 and N. For example, several samples show that the SLD decreases over time, but this does not contradict the fact that the patient has progressed, because in these cases, the patient's progression is explained by non-target progression.

[0165] Fig.11 The marginal posterior values ​​estimated from the model and the ground truth are summarized for each patient. The top and middle subplots show the k g The marginal posterior medians and 90% credible intervals for the f and f parameters are shown. The ground truth values ​​are superimposed, showing good posterior coverage of the true parameter values. The estimates for each patient are shown by the patient’s time to progression or withdrawal, T, revealing two noteworthy patterns. First, the later the patient’s known time to progression occurred, the greater the probability that k g Second, because the model estimates a patient’s time to progression or exit, T, and whether they actually progressed at that time, E, rather than their actual SLD measurements, patients with the same values ​​of T and E have the same posterior distributions for their individual parameters. Fig.11 The two middle bottom subplots show the marginal posterior probability of each individual patient being one of the responders, where the actual ground truth responders are depicted with dashed lines. As mentioned before, the number of responders is known, not which patients are responders, so in each posterior sample, a different combination of patients forms the responder group. Intuitively, patients with later progression have a higher chance of belonging to this group, and this is captured by the model in terms of their higher posterior probabilities.

[0166] Fig.12 The posterior sampling of the trial and group-specific parameters is shown, with their ground truth values ​​superimposed (note that k is killrParameters omitted). The posterior distributions show good coverage of the ground truth, demonstrating the model's ability to capture population-level dynamics using summary PFS and ORR information. Each pair of plots also reflects the complex nonlinear relationships between the dynamic parameters captured by the posterior. For example, (exist Fig.12 where mu_kg) and μ f (exist Fig.12 mu_f in the figure has a correlated posterior distribution because higher values ​​of f are dominated by higher k g The values ​​are offset to fully describe the data.

[0167] Published research

[0168] To illustrate the ability of the model to capture information about tumor dynamics from published studies, the method was applied to three published studies in metastatic castration-resistant prostate cancer (mCRPC). These studies are summarized in Table 1. All models were fitted according to standard guidelines until the R-hat values ​​for all parameters were below 1.05. To assess the fit of the model to the published data, a posterior predictive check (PPC) was performed in which each group of the experiment was repeated with the individual-level parameters R0,k g All new values ​​of ,f are drawn conditional on each posterior sampling of the population-level parameter. The entire generation of the data is then simulated once for each posterior sampling to obtain the PFS curve and the total number of responses for each sampling. These simulated values ​​are then overlaid with the observed values ​​in the PPC to assess the adequacy of the fit. Fig.13 The resulting plots are shown in . Although ground truth values ​​were not available for either population-level, individual-level, or SLD parameters, the model showed good ability to capture the PFS curves and response data observed in various published studies.

[0169] Table 1

[0170]

[0171]

[0172] Use published research to simulate novel test conditions

[0173] Posterior estimates from two different published studies in the previous subsection were combined in silico to 1) evaluate the efficacy of a drug treatment tested in chemotherapy-naive patients; 2) evaluate the efficacy of a combination of one or more drug treatments; and 3) compare the direct comparative results of each drug treatment as a monotherapy and as a combination of one or more drug treatments.

[0174] In the chemotherapy-free setting, the PREVAIL study in the chemotherapy-free setting described in the previous subsection μ f ,τ f To estimate the parameters, four groups of 500 patients were simulated (e.g., patients who had not received chemotherapy). The estimated k for the corresponding group from the same model fit was used. kkill The parameters were used to simulate the placebo and Enza groups. The k kkill Parameters to simulate Finally, Enza and A fourth group of combination therapy was simulated by assigning a kill rate to this group that was the Enza and The sum of the fitted killing rates. Given that these treatments act via different mechanisms, additive killing is a reasonable first-pass assumption. To account for the estimated uncertainty in these parameters and other sources of variation (e.g., trial-to-trial variation), simulations were performed for each posterior sampling of the trial parameters and group-specific parameters. Fig.14 The simulated remission numbers and PFS curves for the four groups of the simulation trial are shown. In terms of PFS, PR+CR and CR, the Enza group seems to perform better than the Pluvicto group, while in terms of CR, the combination seems to be superior to the Enza group. These simulations are used to provide more formal probability estimates. Specifically, the Enza group outperformed the Pluvicto group in PR+CR, CR and median PFS, with probabilities of 0.997, 0.999 and 0.898, respectively, while the combination group outperformed the pure Enza group in the same three categories, with probabilities of 0.730, 0.995 and 0.437, respectively.

[0175] A Bayesian generative model that combines tumor dynamics with published PFS and remission data is introduced. The model allows important tumor dynamic parameters and information to be estimated using published data. Several results are shown. First, an example using simulated data is shown, where the ground truth is known, to demonstrate the ability of the model to recover tumor dynamic parameters using published data. Second, the model is applied to three published studies from real mCRPC trials and has a good fit to them. Finally, these estimated parameter values ​​are combined to compare these therapies in computer simulation trials in new trial settings and in combination with each other.

[0176] Fig.15An illustrative implementation of a computer system 1500 that can be used in conjunction with any of the embodiments of the technology described herein is shown in . The computer system 1500 includes one or more processors 1510 and one or more articles of manufacture including non-transitory computer-readable storage media (e.g., memory 1520 and one or more non-volatile storage media 1530). The processor 1510 can control writing data to and reading data from the memory 1520 and the non-volatile storage device media 1530 in any suitable manner, as the various aspects of the technology described herein are not limited to any particular technology for writing or reading data. In order to perform any of the functions described herein, the processor 1510 can execute one or more processor-executable instructions stored in one or more non-transitory computer-readable storage media (e.g., memory 1520), which can be used as a non-transitory computer-readable storage medium for storing processor-executable instructions for execution by the processor 1510.

[0177] The computer system 1500 may also include a network input / output (I / O) interface 1540 via which the computing device can communicate with other computing devices (e.g., over a network), and one or more user I / O interfaces 1550 via which the computing device can provide output to a user and receive input from a user. The user I / O interfaces may include devices such as a keyboard, a mouse, a microphone, a display device (e.g., a monitor or a touch screen), a speaker, a camera, and / or various other types of I / O devices.

[0178] The above-described embodiments may be implemented in any of a variety of ways. For example, embodiments may be implemented using hardware, software, or a combination thereof. When implemented in software, the software code may be executed on any suitable processor (e.g., a microprocessor) or processor set, whether provided in a single computing device or distributed among multiple computing devices. It should be understood that any component or component set that performs the above-described functions may be generally considered to be one or more controllers that control the above-described functions. The one or more controllers may be implemented in a variety of ways, such as using dedicated hardware or using general-purpose hardware (e.g., one or more processors) that is programmed using microcode or software to perform the above-described functions.

[0179] In this regard, it should be understood that one implementation of the embodiments described herein includes at least one computer-readable storage medium (e.g., RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, cassette, tape, magnetic disk storage or other magnetic storage device, or other tangible non-transitory computer-readable storage medium), the at least one computer-readable storage medium is encoded with a computer program (i.e., a plurality of executable instructions) that performs the above-mentioned functions of one or more embodiments when executed on one or more processors. The computer-readable medium may be transportable so that the program stored thereon can be loaded onto any computing device to implement various aspects of the technology described herein. In addition, it should be understood that the reference to a computer program that performs any of the above-mentioned functions when executed is not limited to an application program running on a host. Rather, the terms computer program and software are used herein in a general sense to refer to any type of computer code (e.g., application software, firmware, microcode, or any other form of computer instructions) that can be employed to program one or more processors to implement various aspects of the technology described herein.

[0180] The above description of the embodiments provides illustration and description, but is not intended to be exhaustive or limit the embodiments to the precise form disclosed. Modifications and variations are possible in accordance with the above teachings, or may be obtained from the practice of the embodiments. In other embodiments, the methods depicted in these figures may include fewer operations, different operations, operations in different orders, and / or additional operations. Further, non-dependent blocks may be executed in parallel.

[0181] It is apparent that in the embodiments shown in the figures, the example aspects described above can be implemented in many different forms of software, firmware, and hardware. Further, some parts of these embodiments can be implemented as "modules" that perform one or more functions. The module can include hardware (such as a processor, an application specific integrated circuit (ASIC), or a field programmable gate array (FPGA)) or a combination of hardware and software.

[0182] Having described several aspects and embodiments of the technology set forth in this disclosure, it should be understood that various changes, modifications and improvements will be readily conceived by those skilled in the art. Such changes, modifications and improvements are intended to be within the spirit and scope of the technology described herein. For example, a person of ordinary skill in the art will easily envision various other means and / or structures for performing the functions described herein and / or obtaining results and / or one or more advantages, and each of such changes and / or modifications is considered to be within the scope of the embodiments described herein. Those skilled in the art will recognize or be able to determine many equivalents of the specific embodiments described herein using only routine experiments. Therefore, it should be understood that the aforementioned embodiments are presented only by way of example, and within the scope of the appended claims and their equivalents, the inventive embodiments may be practiced in a manner different from that specifically described. In addition, any combination of two or more features, systems, articles, materials, kits and / or methods described herein, if such features, systems, articles, materials, kits and / or methods are not mutually inconsistent, is included within the scope of this disclosure.

[0183] The above embodiments can be implemented in any of a variety of ways. One or more aspects and embodiments of the execution of the process or method disclosed herein can utilize program instructions that can be executed by a device (e.g., a computer, a processor, or other device) to execute these processes or methods or control their execution. In this regard, different inventive concepts can be embodied as a computer-readable storage medium (or multiple computer-readable storage media) (e.g., a computer memory, one or more floppy disks, compact disks, optical disks, tapes, flash memories, circuit configurations in field programmable gate arrays or other semiconductor devices, or other tangible computer storage media), which is encoded with one or more programs, which execute one or more methods of implementing the various embodiments described above when executed on one or more computers or other processors. The one or more computer-readable media can be transportable, so that one or more programs stored thereon can be loaded onto one or more different computers or other processors to implement various aspects of the above aspects. In some embodiments, the computer-readable medium can be a non-transient medium.

[0184] The term "program" or "software" is used herein in a general sense to refer to any type of computer code or computer executable instruction set that can be employed to program a computer or other processor to implement the various aspects described above. In addition, it should be understood that, according to one aspect, one or more computer programs that perform the methods of the present disclosure when executed need not reside on a single computer or processor, but can be distributed in a modular manner among multiple different computers or processors to implement various aspects of the present disclosure.

[0185] Computer executable instructions can be in many forms, such as program modules, that are executed by one or more computers or other devices. Typically, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. Typically, in various embodiments, the functionality of the program modules can be combined or distributed as desired.

[0186] Moreover, the data structure can be stored in a computer-readable medium in any suitable form. For simplicity of presentation, the data structure can be shown as having fields that are related by location in the data structure. Similarly, such relationships can be achieved by assigning locations in a computer-readable medium that convey the relationship between the fields for storage of the fields. However, any suitable mechanism can be used to establish the relationship between the information in the fields of the data structure, including by using pointers, tags, or other mechanisms that establish relationships between data elements.

[0187] When implemented in software, the software code can be executed on any suitable processor or collection of processors, whether provided in a single computer or distributed among multiple computers.

[0188] Moreover, the computer may have one or more input and output devices. Among other things, these devices may be used to present a user interface. Examples of output devices that may be used to provide a user interface include a printer or display screen for visually presenting the output and a speaker or other sound generating device for auditorily presenting the output. Examples of input devices that may be used for a user interface include a keyboard and a pointing device, such as a mouse, a touch pad, and a digitized tablet computer. As another example, the computer may receive input information by voice recognition or in other audible formats.

[0189] Such computers can be interconnected by one or more networks of any suitable form, including local area networks or wide area networks (such as enterprise networks) and intelligent networks (IN) or the Internet. Such networks can be based on any suitable technology and can operate according to any suitable protocol, and can include wireless networks, wired networks or fiber optic networks.

[0190] Moreover, as described, some aspects may be embodied as one or more methods. The actions performed as part of a method may be ordered in any suitable manner. Thus, embodiments may be constructed in which the actions are performed in an order different from that shown, which may include performing some actions simultaneously (even though the actions are shown as being sequential in the illustrative embodiments).

[0191] As used herein, the singular terms "a," "an," and "the" may include plural referents unless the context clearly indicates otherwise.

[0192] As used herein, the terms "connect," "connected," and "connector" refer to an operational coupling or link. Connected components may be coupled to each other directly or indirectly, such as through another set of components.

[0193] The phrase "and / or" as used herein in the specification and claims should be understood to mean "either or both" of the elements so combined, i.e., elements that are present in combination in some cases and separately in other cases. Multiple elements listed with "and / or" should be interpreted in the same manner, i.e., "one or more" of the elements so combined. Other elements other than the elements specifically identified by the "and / or" clause may optionally be present, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, when used in conjunction with open language such as "comprising", a reference to "A and / or B" may refer to only A (optionally including elements other than B) in one embodiment; to only B (optionally including elements other than A) in another embodiment; to both A and B (optionally including other elements) in yet another embodiment; etc.

[0194] As used herein in the specification and claims, when referring to a list of one or more elements, the phrase "at least one" should be understood to mean at least one element selected from any one or more of the elements in the list of elements, but not necessarily including at least one of each element specifically listed within the list of elements, and not excluding any combination of elements in the list of elements. This definition also allows that elements other than the elements specifically identified within the list of elements to which the phrase "at least one" refers may optionally be present, whether related to or unrelated to those elements specifically identified. Thus, as a non-limiting example, "at least one of A and B" (or equivalently "at least one of A or B", or equivalently "at least one of A and / or B") may refer to at least one (optionally including more than one) A in one embodiment, where B is absent (and optionally including elements other than B); to at least one (optionally including more than one) B in another embodiment, where A is absent (and optionally including elements other than A); to at least one (optionally including more than one) A and at least one (optionally including more than one) B (and optionally including other elements) in yet another embodiment; etc.

[0195] In the claims and the above specification, all transitional phrases, such as "comprising", "including", "carrying", "having", "containing", "involving", "holding", "composed of", etc., should be understood as open-ended, that is, meaning including but not limited to. Only the transitional phrases "consisting of" and "consisting essentially of" should be closed or semi-closed transitional phrases, respectively.

[0196] As used herein, the terms "approximately", "substantially", "substantially" and "about" are used to describe and explain small changes. When used in conjunction with an event or situation, these terms can refer to situations where the event or situation just occurs and situations where the event or situation approximately occurs. For example, when used in conjunction with a numerical value, these terms can refer to a range of variation where the numerical value is less than or equal to ±10%, such as less than or equal to ±5%, less than or equal to ±4%, less than or equal to ±3%, less than or equal to ±2%, less than or equal to ±1%, less than or equal to ±0.5%, less than or equal to ±0.1%, or less than or equal to ±0.05%. For example, if the difference between the values ​​is less than or equal to ±10% of the average value of the values, such as less than or equal to ±5%, less than or equal to ±4%, less than or equal to ±3%, less than or equal to ±2%, less than or equal to ±1%, less than or equal to ±0.5%, less than or equal to ±0.1%, or less than or equal to ±0.05%, then two numerical values ​​can be considered to be "substantially" the same.

[0197] In addition, quantities, ratios, and other numerical values ​​are sometimes presented in a range format herein. It should be understood that this range format is used for convenience and brevity, and should be flexibly interpreted to include the values ​​explicitly specified as the limits of the range, but also include all individual values ​​and subranges contained within the range, just as if each value or subrange was explicitly specified.

[0198] Although the disclosure has been described and shown with reference to the specific embodiments of the disclosure, these descriptions and illustrations do not limit the disclosure. It should be understood by those skilled in the art that various changes can be made and equivalents can be substituted without departing from the true spirit and scope of the disclosure as defined by the appended claims. These illustrations may not necessarily be drawn to scale. Due to manufacturing processes, tolerances and / or other reasons, there may be differences between the artistic reproduction in the disclosure and the actual device. There may be other embodiments of the disclosure that are not specifically shown. The specification (except for the claims) and the drawings should be considered illustrative rather than restrictive. Modifications may be made to adapt specific situations, materials, material compositions, technologies, or processes to the purpose, spirit and scope of the disclosure. All these modifications are intended to fall within the scope of the appended claims. Although the technology disclosed herein has been described with reference to specific operations performed in a specific order, it should be understood that these operations can be combined, subdivided, or reordered to form equivalent technologies without departing from the teachings of the disclosure. Therefore, unless specifically indicated herein, the order and grouping of operations are not limitations on the disclosure.

Claims

1. A computer-implemented method for estimating tumor growth, the method comprising: Obtaining, by one or more processors, progression-free survival (PFS) data for a plurality of patients, the PFS data indicating (i) a plurality of observation times, and (ii) how many of the plurality of patients had a PFS event within a most recent time window at each of the plurality of observation times; determining, by the one or more processors, a population distribution of one or more patient-specific parameters based on the PFS data; obtaining, by the one or more processors, measured tumor growth data for a specific patient undergoing drug treatment; estimating, by the one or more processors, tumor growth for the specific patient based on (i) the measured tumor growth data and (ii) the population distribution of the one or more patient-specific parameters; and A display is caused, by the one or more processors, to present a visual indication of the estimated tumor growth for the particular patient.

2. The computer-implemented method of claim 1, wherein: Determining a population distribution of the one or more patient-specific parameters includes determining a growth curve function that includes the one or more patient-specific parameters.

3. The computer-implemented method of claim 2, wherein: The growth curve function includes at least one of an exponential growth function, a logistic growth function, or an ordinary differential function.

4. The computer-implemented method of any one of claims 1 to 3, wherein: The one or more patient-specific parameters include a baseline normalized sum of longest diameters (SLD) measurement parameter.

5. The computer-implemented method of any one of claims 1 to 4, wherein: The one or more patient-specific parameters include a growth rate parameter.

6. The computer-implemented method of any one of claims 5, wherein: The growth rate parameter comprises a baseline growth rate parameter in the absence of treatment.

7. The computer-implemented method of any one of claims 1 to 6, wherein: The one or more patient-specific parameters include a parameter of a proportion of drug-sensitive tumor cells in a patient among the plurality of patients.

8. The computer-implemented method of claim 7, wherein: The parameter value of the proportion of drug-sensitive tumor cells in the patients among the plurality of patients ranges from 0 to 1.

9. The computer-implemented method of claim 2, wherein: The growth curve function is time dependent.

10. The computer-implemented method of claim 2, wherein: Determining the population distribution of the one or more patient-specific parameters includes fitting the growth curve function to the observed values ​​at the plurality of observation times.

11. The computer-implemented method of claim 3, wherein: The growth curve function is a logistic growth function; and the one or more patient-specific parameters include a plurality of parameters of the logistic growth function.

12. The computer-implemented method of any one of claims 1 to 11, wherein: Estimating tumor growth for this particular patient includes: modeling the tumor growth rate for the particular patient using a pharmacokinetic-pharmacodynamic (PKPD) model having a first term representing the change in tumor size for the particular patient without treatment with the drug and a second term representing the contribution to the change in tumor size for the particular patient due to treatment with the drug; and The one or more parameters of the first term and the one or more parameters of the second term are jointly estimated by fitting the PKPD model to the measured tumor growth data, in part by setting constraints on the one or more parameters of the first term using the population distribution of the one or more patient-specific parameters.

13. The computer-implemented method of claim 12, wherein: The one or more parameters of the second item include one or more of the following: the drug concentration in the plasma of that particular patient; The maximum kill rate for that particular patient; and Half maximal effective concentration (EC50).

14. The computer-implemented method of any one of claims 1 to 13, wherein: Estimating the tumor growth further includes obtaining an overall response rate for the particular patient undergoing treatment with the drug.

15. The computer-implemented method of any one of claims 1 to 14, wherein: Estimating the tumor growth further includes obtaining one or more non-target events for the particular patient undergoing treatment with the drug.

16. The computer-implemented method of claim 10, wherein: The observations include, for each patient of the plurality of patients, a first observation at a first time indicating that the patient had not had a PFS event before the first time, and a second observation at a second time indicating that the patient had had a PFS event before the second time.

17. The computer-implemented method of any one of claims 1 to 16, wherein: The PFS data corresponds to a specific cancer type; and Estimating tumor growth for the particular patient includes estimating tumor growth for patients diagnosed with the particular cancer type.

18. The computer-implemented method of any one of claims 1 to 17, wherein: Causing the display to present a visual indication of the estimated tumor growth for the particular patient includes causing the display to display a trajectory of the tumor growth for the particular patient if the particular patient has not yet been treated with the drug.

19. The computer-implemented method of any one of claims 1 to 18, wherein: The population distribution of the one or more patient-specific parameters includes a log-normal distribution.

20. The computer-implemented method of any one of claims 1 to 19, wherein: The PFS data includes at least one of a digitized PFS graph or a PFS risk table.

21. The computer-implemented method of any one of claims 1 to 20, wherein: The PFS data indicates how many of the plurality of patients had a baseline normalized sum of longest diameters (SLD) measurement of at least 1.2 or the development of new lesions at each observation time in the plurality of observation times and within the most recent time window.

22. The computer-implemented method of any one of claims 1 to 21, wherein: The plurality of patients represented by the PFS data are patients associated with ineffective drug treatment.

23. The computer-implemented method of any one of claims 1 to 22, further comprising adjusting the dosage of the drug therapy based on the estimated tumor growth of the particular patient.

24. A computer system for estimating tumor growth, the computer system comprising: a data storage device storing processor-readable instructions; as well as A processor configured to execute the instructions to perform a method comprising: Obtaining progression-free survival (PFS) data of a plurality of patients, the PFS data indicating (i) a plurality of observation times, and (ii) how many of the plurality of patients had a PFS event within a most recent time window at each of the plurality of observation times; determining a population distribution of one or more patient-specific parameters based on the PFS data; obtaining measured tumor growth data for a specific patient undergoing drug treatment; estimating tumor growth for the specific patient based on (i) the measured tumor growth data and (ii) the population distribution of the one or more patient-specific parameters; and A display is caused to present a visual indication of the estimated tumor growth for that particular patient.

25. The computer system of claim 24, wherein: Determining a population distribution of the one or more patient-specific parameters includes determining a growth curve function that includes the one or more patient-specific parameters.

26. The computer system of claim 25, wherein: The growth curve function includes at least one of an exponential growth function, a logistic growth function, or an ordinary differential function.

27. A computer system as claimed in any one of claims 24 to 26, wherein: The one or more patient-specific parameters include a baseline normalized sum of longest diameters (SLD) measurement parameter.

28. A computer system as claimed in any one of claims 24 to 27, wherein: The one or more patient-specific parameters include a growth rate parameter.

29. The computer system of any one of claims 28, wherein: The growth rate parameter comprises a baseline growth rate parameter in the absence of treatment.

30. The computer system of any one of claims 24 to 29, wherein: The one or more patient-specific parameters include a parameter of a proportion of drug-sensitive tumor cells in a patient among the plurality of patients.

31. The computer system of claim 30, wherein: The parameter value of the proportion of drug-sensitive tumor cells in the patients among the plurality of patients ranges from 0 to 1.

32. The computer system of claim 25, wherein: The growth curve function is time dependent.

33. The computer system of claim 25, wherein: Determining the population distribution of the one or more patient-specific parameters includes fitting the growth curve function to the observed values ​​at the plurality of observation times.

34. The computer system of claim 26, wherein: The growth curve function is a logistic growth function; and the one or more patient-specific parameters include a plurality of parameters of the logistic growth function.

35. The computer system of any one of claims 24 to 34, wherein: Estimating tumor growth for this particular patient includes: modeling the tumor growth rate for the particular patient using a pharmacokinetic-pharmacodynamic (PKPD) model having a first term representing the change in tumor size for the particular patient without treatment with the drug and a second term representing the contribution to the change in tumor size for the particular patient due to treatment with the drug; and The one or more parameters of the first term and the one or more parameters of the second term are jointly estimated by fitting the PKPD model to the measured tumor growth data, in part by setting constraints on the one or more parameters of the first term using the population distribution of the one or more patient-specific parameters.

36. The computer system of claim 35, wherein: The one or more parameters of the second item include one or more of the following: the drug concentration in the plasma of that particular patient; The maximum kill rate for that particular patient; and Half maximal effective concentration (EC50).

37. A computer system as claimed in any one of claims 24 to 36, wherein: Estimating the tumor growth further includes obtaining an overall response rate for the particular patient undergoing treatment with the drug.

38. A computer system as claimed in any one of claims 24 to 37, wherein: Estimating the tumor growth further includes obtaining one or more non-target events for the particular patient undergoing treatment with the drug.

39. The computer system of claim 33, wherein: The observations include, for each patient of the plurality of patients, a first observation at a first time indicating that the patient had not had a PFS event before the first time, and a second observation at a second time indicating that the patient had had a PFS event before the second time.

40. A computer system as claimed in any one of claims 24 to 39, wherein: The PFS data corresponds to a specific cancer type; and Estimating tumor growth for the particular patient includes estimating tumor growth for patients diagnosed with the particular cancer type.

41. A computer system as claimed in any one of claims 24 to 40, wherein: Causing the display to present a visual indication of the estimated tumor growth for the particular patient includes causing the display to display a trajectory of the tumor growth for the particular patient if the particular patient has not yet been treated with the drug.

42. A computer system as claimed in any one of claims 24 to 41, wherein: The population distribution of the one or more patient-specific parameters includes a log-normal distribution.

43. A computer system as claimed in any one of claims 24 to 42, wherein: The PFS data includes at least one of a digitized PFS graph or a PFS risk table.

44. A computer system as claimed in any one of claims 24 to 43, wherein: The PFS data indicates how many of the plurality of patients had a baseline normalized sum of longest diameters (SLD) measurement of at least 1.2 or the development of new lesions at each observation time in the plurality of observation times and within the most recent time window.

45. A computer system as claimed in any one of claims 24 to 44, wherein: The plurality of patients represented by the PFS data are patients associated with ineffective drug treatment.

46. ​​The computer system of any one of claims 24-45, further comprising adjusting the dosage of the drug therapy based on the estimated tumor growth of the particular patient.

47. A non-transitory computer readable medium comprising instructions for estimating tumor growth, which instructions, when executed by a processor, cause the processor to perform a method comprising: Obtaining progression-free survival (PFS) data of a plurality of patients, the PFS data indicating (i) a plurality of observation times, and (ii) how many of the plurality of patients had a PFS event within a most recent time window at each of the plurality of observation times; determining a population distribution of one or more patient-specific parameters based on the PFS data; obtaining measured tumor growth data for a specific patient undergoing drug treatment; estimating tumor growth for the specific patient based on (i) the measured tumor growth data and (ii) the population distribution of the one or more patient-specific parameters; and A display is caused to present a visual indication of the estimated tumor growth for that particular patient.

48. The non-transitory computer readable medium of claim 47, wherein: Determining a population distribution of the one or more patient-specific parameters includes determining a growth curve function that includes the one or more patient-specific parameters.

49. The non-transitory computer readable medium of claim 48, wherein: The growth curve function includes at least one of an exponential growth function, a logistic growth function, or an ordinary differential function.

50. The non-transitory computer readable medium of any one of claims 47 to 49, wherein: The one or more patient-specific parameters include a baseline normalized sum of longest diameters (SLD) measurement parameter.

51. The non-transitory computer readable medium of any one of claims 47 to 50, wherein: The one or more patient-specific parameters include a growth rate parameter.

52. The non-transitory computer readable medium of any one of claims 51, wherein: The growth rate parameter comprises a baseline growth rate parameter in the absence of treatment.

53. The non-transitory computer readable medium of any one of claims 47 to 52, wherein: The one or more patient-specific parameters include a parameter of a proportion of drug-sensitive tumor cells in a patient among the plurality of patients.

54. The non-transitory computer readable medium of claim 53, wherein: The parameter value of the proportion of drug-sensitive tumor cells in the patients among the plurality of patients ranges from 0 to 1.

55. The non-transitory computer readable medium of claim 48, wherein: The growth curve function is time dependent.

56. The non-transitory computer readable medium of claim 48, wherein: Determining the population distribution of the one or more patient-specific parameters includes fitting the growth curve function to the observed values ​​at the plurality of observation times.

57. The non-transitory computer readable medium of claim 49, wherein: The growth curve function is a logistic growth function; and the one or more patient-specific parameters include a plurality of parameters of the logistic growth function.

58. The non-transitory computer readable medium of any one of claims 47 to 57, wherein: Estimating tumor growth for this particular patient includes: modeling the tumor growth rate for the particular patient using a pharmacokinetic-pharmacodynamic (PKPD) model having a first term representing the change in tumor size for the particular patient without treatment with the drug and a second term representing the contribution to the change in tumor size for the particular patient due to treatment with the drug; and The one or more parameters of the first term and the one or more parameters of the second term are jointly estimated by fitting the PKPD model to the measured tumor growth data, in part by setting constraints on the one or more parameters of the first term using the population distribution of the one or more patient-specific parameters.

59. The non-transitory computer readable medium of claim 58, wherein: The one or more parameters of the second item include one or more of the following: the drug concentration in the plasma of that particular patient; The maximum kill rate for that particular patient; and Half maximal effective concentration (EC50).

60. The non-transitory computer readable medium of any one of claims 47 to 59, wherein: Estimating the tumor growth further includes obtaining an overall response rate for the particular patient undergoing treatment with the drug.

61. The non-transitory computer readable medium of any one of claims 47 to 60, wherein: Estimating the tumor growth further includes obtaining one or more non-target events for the particular patient undergoing treatment with the drug.

62. The non-transitory computer readable medium of claim 56, wherein: The observations include, for each patient of the plurality of patients, a first observation at a first time indicating that the patient had not had a PFS event before the first time, and a second observation at a second time indicating that the patient had had a PFS event before the second time.

63. The non-transitory computer readable medium of any one of claims 47 to 62, wherein: The PFS data corresponds to a specific cancer type; and Estimating tumor growth for the particular patient includes estimating tumor growth for patients diagnosed with the particular cancer type.

64. The non-transitory computer readable medium of any one of claims 47 to 63, wherein: Causing the display to present a visual indication of the estimated tumor growth for the particular patient includes causing the display to display a trajectory of the tumor growth for the particular patient if the particular patient has not yet been treated with the drug.

65. The non-transitory computer readable medium method of any one of claims 47 to 64, wherein: The population distribution of the one or more patient-specific parameters includes a log-normal distribution.

66. The non-transitory computer readable medium of any one of claims 47 to 65, wherein: The PFS data includes at least one of a digitized PFS graph or a PFS risk table.

67. The non-transitory computer readable medium of any one of claims 47 to 66, wherein: The PFS data indicates how many of the plurality of patients had a baseline normalized sum of longest diameters (SLD) measurement of at least 1.2 or the development of new lesions at each observation time in the plurality of observation times and within the most recent time window.

68. The non-transitory computer readable medium of any one of claims 47 to 67, wherein: The plurality of patients represented by the PFS data are patients associated with ineffective drug treatment.

69. The non-transitory computer readable medium of any one of claims 47 to 68, further comprising adjusting the dosage of the drug therapy based on the estimated tumor growth of the particular patient.

Images