General Information
Conflict of Interest
Request Clinical Trials
Associated Trial(s):- NCT02136134 - Phase 3 Study Comparing Daratumumab, Bortezomib and Dexamethasone (DVd) vs Bortezomib and Dexamethasone (Vd) in Subjects With Relapsed or Refractory Multiple Myeloma
- NCT02076009 - Phase 3 Study Comparing Daratumumab, Lenalidomide, and Dexamethasone (DRd) vs Lenalidomide and Dexamethasone (Rd) in Subjects With Relapsed or Refractory Multiple Myeloma
- NCT03180736 - A Phase 3 Study Comparing Pomalidomide and Dexamethasone With or Without Daratumumab in Subjects With Relapsed or Refractory Multiple Myeloma Who Have Received at Least One Prior Line of Therapy With Both Lenalidomide and a Proteasome Inhibitor
Request Clinical Trials
Data Request Status
Status: OngoingResearch Proposal
Project Title: Dynamic immune-control index and pre-progression threshold in multiple myeloma using longitudinal laboratory data
Scientific Abstract:
Background: Multiple myeloma can progress after treatment, and routine laboratory tests may show changes in immune status, inflammation, anemia, nutrition, and disease burden before progression.
Objective: To derive a clinically interpretable dynamic immune-control index from longitudinal laboratory measurements and evaluate whether worsening of this index precedes progression in relapsed or refractory multiple myeloma treated in daratumumab trials.
Study Design: Retrospective secondary analysis of de-identified participant-level data from randomized daratumumab trials. Laboratory values will be aligned by study day, treatment cycle, and time before progression.
Participants: Participants with relapsed or refractory multiple myeloma and available longitudinal laboratory, treatment, response, and follow-up data.
Primary and Secondary Outcome Measure(s): The primary outcome is progression-free survival and documented progressive disease. Secondary outcomes include overall survival and treatment response.
Statistical Analysis: Candidate components will include lymphocyte-related, myeloid/inflammatory, anemia/nutritional, and disease-burden markers. We will describe trajectories before progression, compare participants with and without progression, and evaluate associations using Cox models, landmark analyses, and joint models. Exploratory thresholds will be assessed by threshold crossing, future progression risk within 3, 6, and 12 months, false-positive burden, and lead time before progression. Results will be hypothesis-generating and require prospective validation.
Brief Project Background and Statement of Project Significance:
Multiple myeloma remains an incurable plasma-cell malignancy for many patients, despite major therapeutic advances including daratumumab-based regimens. Patients with relapsed or refractory multiple myeloma often experience repeated cycles of response, remission, and progression. In clinical practice and clinical trials, disease status is monitored using response assessments and laboratory markers such as M-protein, serum free light chains, hemoglobin, albumin, creatinine, calcium, LDH, beta-2 microglobulin, and complete blood count components. These routinely measured laboratory values may contain dynamic information about immune status, systemic inflammation, anemia, nutrition, organ function, and disease burden before progression is formally documented.
Most prognostic analyses in multiple myeloma focus on baseline risk factors, treatment response categories, cytogenetic risk, or static prediction of progression-free survival and overall survival. However, a clinically important question is whether longitudinal laboratory changes can identify a pre-progression period during which the patient is moving from a controlled disease state toward a higher-risk state. Such a dynamic approach may be useful for designing future monitoring strategies and prospective studies of immune-based or immunomodulatory intervention timing.
The significance of this project is to develop a clinically interpretable dynamic immune-control index using laboratory measurements that are already routinely collected in multiple myeloma trials. Rather than attempting to directly test the efficacy of a new intervention, this study aims to determine whether deterioration of a laboratory-based immune-control index occurs before documented progression and whether exploratory thresholds can identify a clinically meaningful lead time before progression. This would create generalizable medical knowledge by clarifying how longitudinal routine laboratory data can be used to characterize disease control, immune-inflammatory deterioration, and potential pre-progression windows.
The requested daratumumab trial data are well suited for this purpose because they include randomized clinical trial populations with relapsed or refractory multiple myeloma, standardized treatment information, serial laboratory measurements, response assessments, progression-free survival, overall survival, and follow-up data. The analysis may provide a reproducible framework for transforming routine laboratory measurements into a dynamic clinical index. If validated in future independent and prospective studies, such an index could help inform risk-adapted monitoring and the design of future studies evaluating when immune-based interventions, including tumor vaccine strategies, may be most appropriately timed.
This study is exploratory and hypothesis-generating. It will not attempt to identify individual participants, and it will not claim that laboratory thresholds alone can determine treatment decisions without prospective validation.
Specific Aims of the Project:
Aim 1: Construct a clinically interpretable dynamic immune-control index from routinely collected longitudinal laboratory variables in relapsed/refractory multiple myeloma, including markers of immune status, inflammation, anemia/nutrition, organ function, and myeloma disease burden when available.
Aim 2: Describe how this index and its individual laboratory components change over time and before documented progressive disease, and compare trajectories between participants who experience progression and those who remain progression-free.
Aim 3: Evaluate associations between baseline values, time-updated values, changes from baseline, and slopes of the index with progression-free survival, documented progressive disease, overall survival, and treatment response.
Aim 4: Explore hypothesis-generating threshold-crossing patterns that may identify increased future progression risk within 3, 6, and 12 months and estimate lead time before documented progression. Any threshold will be exploratory and will require independent prospective validation before clinical use.
Study Design: Meta-analysis (analysis of multiple trials together)
What is the purpose of the analysis being proposed? Please select all that apply.: Participant-level data meta-analysis Meta-analysis using only data from the YODA Project Develop or refine statistical methods Research on clinical prediction or risk prediction
Software Used: Python, R, RStudio
Data Source and Inclusion/Exclusion Criteria to be used to define the patient sample for your study:
Data source: De-identified participant-level data from the requested YODA daratumumab multiple myeloma trials: NCT02136134 (CASTOR), NCT02076009 (POLLUX), and NCT03180736 (APOLLO), including available demographic, clinical, treatment, response, progression/PFS, OS, follow-up, and longitudinal laboratory data.
Inclusion criteria: enrolled/randomized participants with relapsed or refractory multiple myeloma in these trials who have usable follow-up and outcome data. For longitudinal laboratory analyses, participants must have at least one relevant baseline or post-baseline laboratory assessment.
Exclusion criteria: participants without usable follow-up/outcome data, invalid or non-positive time-to-event values, or no relevant laboratory data for a given laboratory-based analysis. Missing individual laboratory variables will not exclude a participant from all analyses; analyses will use available data by variable.
No external participant-level dataset will be pooled with YODA data in the primary analysis. Literature, protocols, SAPs, data dictionaries, and aggregate results may be used only for background, variable definition, and interpretation.
Primary and Secondary Outcome Measure(s) and how they will be categorized/defined for your study:
Primary outcome: progression-free survival (PFS), defined according to the original trial definitions as time from randomization or treatment initiation to documented progressive disease or death, with censoring based on trial-specific rules.
Key event outcome: documented progressive disease/progression status and time to progression when available.
Secondary outcomes: overall survival (OS), treatment response categories, best overall response, time to response, and duration of response when available.
Exploratory outcomes: future progression within 3, 6, and 12 months after a landmark laboratory assessment; first threshold crossing of the dynamic immune-control index; and lead time from first threshold crossing to documented progression. These exploratory outcomes will be used only for hypothesis generation and will not be interpreted as clinically validated decision thresholds.
Main Predictor/Independent Variable and how it will be categorized/defined for your study:
The main independent variable will be a time-updated dynamic clinical immune-control index derived from longitudinal routine laboratory measurements. The index will be designed so that higher values indicate better immune-control status and lower values indicate worse immune-inflammatory, anemia/nutritional, organ-function, or disease-burden status.
Candidate components will be selected from laboratory markers available in the requested trials, including lymphocytes, neutrophils or granulocytes when available, monocytes, platelets, hemoglobin, albumin, LDH, CRP, creatinine, calcium, beta-2 microglobulin, M-protein, serum free light chains, and derived immune-inflammatory ratios such as neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, systemic immune-inflammation index, and systemic inflammation response index when the required components are available.
Laboratory values will be aligned by study day and treatment cycle. Continuous laboratory variables will be harmonized across trials, checked for units and distributions, and standardized within trial when appropriate. The immune-control index will be evaluated as a continuous time-updated variable, as change from baseline, as longitudinal slope over defined time windows, and as threshold-crossing status.
For threshold analyses, candidate thresholds will be used to classify each laboratory assessment as above or below a potential immune-control risk boundary. Threshold crossing will be defined as the first time a participant's dynamic index enters the higher-risk range before documented progression. These thresholds will be exploratory and will be evaluated for their association with progression within 3, 6, and 12 months and for lead time before documented progressive disease.
Other Variables of Interest that will be used in your analysis and how they will be categorized/defined for your study:
Other variables of interest will include demographic, disease-related, treatment-related, and trial-related factors used to describe the study population, evaluate subgroup patterns, and adjust for potential confounding.
Demographic variables may include age, sex, race or ethnicity when available, body mass index, and geographic region if available. Disease-related variables may include baseline ECOG performance status, ISS or R-ISS stage, cytogenetic or FISH risk markers, baseline disease burden, baseline M-protein or serum free light chain levels, beta-2 microglobulin, albumin, LDH, calcium, creatinine, renal function markers, refractory status, number of prior lines of therapy, prior exposure to proteasome inhibitors or immunomodulatory drugs, and other baseline characteristics available in the trial datasets.
Treatment-related variables will include trial, treatment arm, daratumumab exposure, background regimen, treatment start date or study day, treatment cycle, dose modifications, treatment discontinuation, reason for discontinuation, and available concomitant medications when relevant. Trial indicator will be included to account for differences across NCT02136134, NCT02076009, and NCT03180736.
In multivariable analyses, core adjustment variables will include age, sex, trial, treatment arm, baseline ECOG performance status, ISS or R-ISS stage when available, cytogenetic risk when available, number of prior treatment lines, refractory status, and baseline disease-burden markers. Additional variables may be included in sensitivity analyses if they are clinically relevant and sufficiently complete. Post-baseline response variables will be summarized descriptively and used in exploratory analyses, but they will not be treated as baseline confounders in the primary models.
Statistical Analysis Plan:
All analyses will be conducted using de-identified participant-level data within the approved YODA data environment. We will first review the protocols, statistical analysis plans, case report forms, data dictionaries, and clinical study reports to understand trial design, visit schedules, laboratory variables, outcome definitions, censoring rules, and missing data patterns.
Data preparation: We will harmonize participant identifiers, trial indicators, treatment arms, study days, visits/cycles, laboratory names, values, units, response assessments, progression status, PFS, OS, and follow-up information. Laboratory values will be evaluated for unit consistency, implausible values, extreme observations, and missingness patterns. Variables with sufficient longitudinal coverage and clinical relevance will be retained for subsequent analyses.
Dynamic index construction: We aim to develop a longitudinal immune-control index that captures temporal changes in patients' biological status before disease progression. The index will integrate multiple biologically relevant domains derived from routinely collected clinical laboratory measurements, including immune reserve, nutritional and hematologic status, systemic inflammation and myeloid activation, and tumor burden/coagulation-related characteristics when available.
Potential input variables may include immune-related markers (e.g., lymphocyte, neutrophil, monocyte counts, and derived inflammatory ratios such as NLR, PLR, LMR, SII, and SIRI), nutritional and hematologic markers (e.g., albumin, hemoglobin, platelet-related parameters), and myeloma-related disease burden markers (e.g., M-protein, free light chains, beta-2 microglobulin, LDH, calcium, creatinine) when available.
Continuous biomarkers will be appropriately normalized before integration, and transformations will be considered for highly skewed variables. Domain-level scores will be constructed by aggregating related biomarkers within each biological category. A composite dynamic immune-control index will then be developed by integrating protective biological features (such as immune reserve and nutritional status) and adverse biological features (such as systemic inflammation and disease burden).
The weighting strategy for combining individual biomarkers and biological domains will not be prespecified as fixed values. Instead, weights will be determined using predefined statistical modeling procedures based on the available YODA dataset, considering factors such as longitudinal association with disease progression, survival outcomes, and model performance. Clinically meaningful thresholds for identifying potential periods of immune deterioration will also be explored within the dataset.
If a simple weighted composite index does not provide sufficient predictive performance or robustness, alternative approaches, including more flexible statistical or machine-learning-based longitudinal modeling strategies, may be considered to define the optimal dynamic index while maintaining interpretability.
The final index will be evaluated by examining its longitudinal trajectory and its association with clinical outcomes using appropriate survival analysis methods, including time-dependent Cox proportional hazards models and additional predictive performance assessments where appropriate.
Narrative Summary: Multiple myeloma can return or progress after treatment. Routine blood tests are collected repeatedly during clinical trials and may show changes in immune status, inflammation, anemia, nutrition, and disease burden before progression is formally diagnosed. This study will use de-identified participant-level data from daratumumab multiple myeloma trials to develop a dynamic clinical immune-control index from longitudinal laboratory measurements. We will examine whether worsening of this index occurs before documented progression and estimate exploratory thresholds that may identify a pre-progression window for future immune-based interventions. The study is a secondary analysis only and will not attempt to identify participants.
Project Timeline:
If the data request is approved, we anticipate starting the project within 1 month after completion of the Data Use Agreement and data access setup.
Months 1-2: Review trial protocols, statistical analysis plans, clinical study reports, data dictionaries, and laboratory data structures. Harmonize participant-level datasets across the requested trials and define analysis-ready variables.
Months 3-4: Conduct descriptive analyses, assess missingness, summarize longitudinal laboratory availability, and construct candidate laboratory components and the dynamic immune-control index.
Months 5-7: Perform longitudinal trajectory analyses, Cox models, landmark analyses, and joint modeling of longitudinal laboratory markers and progression-free survival.
Months 8-9: Evaluate exploratory threshold-crossing definitions, progression risk within 3, 6, and 12 months, false-positive burden, and lead time before documented progression. Conduct sensitivity and trial-specific analyses.
Months 10-11: Prepare tables, figures, interpretation, and manuscript draft.
Month 12: Submit the manuscript for publication or finalize the report if manuscript submission is not yet complete. Results will be reported back to the YODA Project according to YODA requirements. If additional time is needed, we will request an extension in accordance with the Data Use Agreement.
Dissemination Plan:
The anticipated products include an academic manuscript, statistical analysis code generated within the approved secure environment when permissible, summary tables, non-identifiable figures, and an aggregate report of study findings. The target audience includes hematologists, oncologists, clinical trial researchers, biomarker researchers, and investigators interested in dynamic risk monitoring and immune-based intervention timing in multiple myeloma.
We plan to submit the completed manuscript to a peer-reviewed journal in hematology, oncology, cancer biomarkers, or medical informatics. Potential target journals may include Blood Advances, British Journal of Haematology, European Journal of Haematology, Clinical Lymphoma Myeloma and Leukemia, BMC Medicine, BMC Cancer, or other appropriate peer-reviewed journals depending on the final scope and strength of the results.
Results may also be presented at academic conferences, institutional seminars, or research meetings. All reported results will be aggregate and non-identifiable. No individual participant-level data will be disclosed. Any dissemination will acknowledge the YODA Project, Johnson & Johnson/Janssen, the original clinical trial investigators, and data contributors according to applicable requirements. The proposed thresholds will be clearly described as exploratory and hypothesis-generating, requiring independent and prospective validation before clinical use.
Bibliography:
Palumbo A, Chanan-Khan A, Weisel K, et al.; CASTOR Investigators. Daratumumab, Bortezomib, and Dexamethasone for Multiple Myeloma. New England Journal of Medicine. 2016;375(8):754-766. doi:10.1056/NEJMoa1606038.
Dimopoulos MA, Oriol A, Nahi H, et al.; POLLUX Investigators. Daratumumab, Lenalidomide, and Dexamethasone for Multiple Myeloma. New England Journal of Medicine. 2016;375(14):1319-1331. doi:10.1056/NEJMoa1607751.
Dimopoulos MA, Terpos E, Boccadoro M, et al. Daratumumab plus pomalidomide and dexamethasone versus pomalidomide and dexamethasone alone in previously treated multiple myeloma (APOLLO): an open-label, randomised, phase 3 trial. Lancet Oncology. 2021;22(6):801-812. doi:10.1016/S1470-2045(21)00128-5.
Rajkumar SV. Multiple myeloma: 2024 update on diagnosis, risk-stratification, and management. American Journal of Hematology. 2024.
Ferle M, Grieb N, Kreuz M, Platzbecker U, Neumuth T, Reiche K, Oeser A, Merz M. Predicting progression events in multiple myeloma from routine blood work. npj Digital Medicine. 2025.
Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. New York: Springer; 2000.
Rizopoulos D. Joint Models for Longitudinal and Time-to-Event Data: With Applications in R. Boca Raton: Chapman and Hall/CRC; 2012.
Harrell FE Jr. Regression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis. 2nd ed. Cham: Springer; 2015.
Heagerty PJ, Lumley T, Pepe MS. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics. 2000;56(2):337-344.
Uno H, Cai T, Pencina MJ, D’Agostino RB, Wei LJ. On the C-statistics for evaluating overall adequacy of risk prediction procedures with censored survival data. Statistics in Medicine. 2011;30(10):1105-1117.