Validation of algorithms to select patients with multiple myeloma and patients initiating myeloma treatment in the national Veterans Affairs Healthcare System

Cumulative Excess Body Mass Index and MGUS Progression to Myeloma

Importance

Obesity is a risk factor associated with multiple myeloma (MM) and its precursor, monoclonal gammopathy of unknown significance (MGUS). However, it is unclear how cumulative exposure to obesity affects the risk of MGUS progression to MM.

Objective

To determine the association of cumulative exposure to excess body mass index (EBMI), defined as BMI (calculated as weight in kilograms divided by height in meters squared) greater than 25, with risk of MGUS progression to MM.

Design, Setting, and Participants

This cohort study included patients with MGUS, including immunoglobin G, immunoglobin A, or light chain MGUS, from the nationwide US Veterans Health Administration database from October 1, 1999, to December 31, 2021. A published natural language processing-assisted model was used to confirm diagnoses of MGUS and progression to MM. Data were analyzed from February 12 to November 4, 2024.

Exposures

Cumulative EBMI was calculated by area under the curve of measured BMI subtracting the reference BMI at 25 during the first 3 years after MGUS diagnosis.

Main Outcomes and Measures

The main outcome was progression from MGUS to MM. Multivariable Fine-Gray time-to-competing-event analyses, with death as the competing event, were used to determine associations.

Results

The cohort included 22 429 patients with MGUS (median [IQR] age, 70.5 [63.5-77.9] years; 21 613 [96.4%] male), with 8329 Black patients (37.1%) and 14 100 White patients (62.9%). There were 4862 patients (21.7%) with reference range BMI (18.5 to <25), 7619 patients (34.0%) with BMI 25 to less than 30, and 8513 patients (38.0%) with BMI 30 or greater at the time of MGUS diagnosis. Compared with reference range BMI at MGUS diagnosis, patients with BMI 25 to less than 30 (adjusted hazard ratio [aHR], 1.17; 95% CI, 1.03-1.34) or 30 or greater (aHR, 1.27; 95% CI, 1.09-1.47) at MGUS diagnosis had higher risk of progression to MM. In patients with reference range BMI at MGUS diagnosis, each 1-unit increase of EBMI per year was associated with a 21% increase in progression risk (aHR, 1.21; 95% CI, 1.04-1.40). However, for patients with BMI 25 or greater at MGUS diagnosis, the incremental risk associated with cumulative EBMI exposure was not statistically significant.

Conclusions and Relevance

This cohort study found that, for patients with BMI 18.5 to less than 25 at the time of MGUS diagnosis, cumulative exposure to BMI 25 or greater was associated with an increased risk of progression. These findings suggest that for these patients, maintaining a healthy and stable weight following MGUS diagnosis may prevent progression to MM.

Use of disease modifying anti-rheumatic drugs and risk of multiple myeloma in US Veterans with rheumatoid arthritis

BACKGROUND

Biologic (b) and targeted synthetic (ts) disease-modifying anti-rheumatic drugs (DMARDs) used in the management of rheumatoid arthritis (RA) target inflammatory pathways implicated in the pathogenesis of multiple myeloma (MM). It is unknown whether use of b/tsDMARDs affects the incidence of MM.

METHODS

In this cohort study using Veterans Health Administration (VHA) data, we identified Veterans newly diagnosed with RA from 1/1/2002 to 12/31/2018 using diagnostic codes and medication fills. DMARD exposure was categorized as follows: conventional synthetic (cs)DMARDs; bDMARDs, which included tumor necrosis factor inhibitors (TNFi), non-TNFi; and a tsDMARD, tofacitinib. A Cox proportional hazards model with time-varying exposure was used to estimate the hazard ratio for developing MM among those who received b/tsDMARD medications relative to b/tsDMARD-naïve persons.

RESULTS

27,540 veterans with RA met eligibility criteria of whom 8322 (30%) took a b/tsDMARD during follow-up. There were 77 incident cases of MM over 192,000 person-years of follow-up. The age-adjusted incidence rate (IR) of MM among b/tsDMARD-naïve patients was 0.37 (95% CI 0.28-0.49) per 1000 person-years and 0.42 among current or former b/tsDMARD users (95% CI 0.25-0.65). Adjusting for age and other demographic characteristics, the hazard ratio for MM associated with use of b/tsDMARDs was 1.32 (95% CI 0.78, 2.26).

CONCLUSION

In this study of Veterans with RA, the rate of MM did not differ between b/tsDMARD and csDMARD users. The relatively short duration of follow-up and few events limited our power to detect treatment-related differences in MM risk.

Association between glucagon-like peptide-1 receptor agonist use and progression of monoclonal gammopathy of uncertain significance to multiple myeloma among patients with diabetes

BACKGROUND

In patients with diabetes and monoclonal gammopathy of uncertain significance (MGUS), the impact of glucagon-like peptide-1 (GLP-1) receptor agonists on the natural history of MGUS is unknown. We aimed to assess the association of GLP-1 receptor agonist use in the progression of MGUS to multiple myeloma in patients with diabetes.

METHODS

This is a population-based cohort study of veterans diagnosed with MGUS from 2006 to 2021 with a prior diagnosis of diabetes. A validated natural language processing algorithm was used to confirm MGUS and progression to multiple myeloma. We performed 1:2 matching for individuals with and without GLP-1 receptor agonist exposure. The Gray test was performed to detect the difference in cumulative incidence functions for progression by GLP-1 receptor agonist use status. The association between time-varying GLP-1 receptor agonist use and progression was estimated through multivariable-adjusted hazard ratio using a stratified Fine-Gray distribution hazard model, with death as a competing event and stratum for the matched patient triad.

RESULTS

Our matched cohort included 1097 individuals with MGUS who had ever used GLP-1 receptor agonists and the matched 2194 patients who had never used GLP-1 receptor agonists. Overall, 2.6% of individuals progressed in the GLP-1 receptor agonist ever use group compared with 5.0% in the GLP-1 receptor agonist never use group. Cumulative incidence functions were statistically significantly different between the exposed and unexposed groups (P = .02). GLP-1 receptor agonist use vs no use was associated with decreased progression to multiple myeloma (hazard ratio = 0.45, 95% confidence interval = 0.22 to 0.93, P = .03).

CONCLUSIONS

For patients with diabetes and MGUS, GLP-1 receptor agonist use is associated with a 55% reduction in risk of progression from MGUS to multiple myeloma compared with no use.

Natural Language Processing Algorithm to Extract Multiple Myeloma Stage From Oncology Notes in the Veterans Affairs Healthcare System

PURPOSE

Stage in multiple myeloma (MM) is an essential measure of disease risk, but its measurement in large databases is often lacking. We aimed to develop and validate a natural language processing (NLP) algorithm to extract oncologists' documentation of stage in the national Veterans Affairs (VA) Healthcare System.

METHODS

Using nationwide electronic health record (EHR) and cancer registry data from the VA Corporate Data Warehouse, we developed and validated a rule-based NLP algorithm to extract oncologist-determined MM stage. To that end, a clinician annotated MM stage within over 5,000 short snippets of clinical notes, and annotated MM stage at MM treatment initiation for 200 patients. These were allocated into snippet- and patient-level development and validation sets. We developed MM stage extraction and roll-up algorithms within the development sets. After the algorithms were finalized, we validated them using standard measures in held-out validation sets.

RESULTS

We developed algorithms for three different MM staging systems that have been in widespread use (Revised International Staging System [R-ISS], International Staging System [ISS], and Durie-Salmon [DS]) and for stage reported without a clearly defined system. Precision and recall were uniformly high for MM stage at the snippet level, ranging from 0.92 to 0.99 for the different MM staging systems. Performance in identifying for MM stage at treatment initiation at the patient level was also excellent, with precision of 0.92, 0.96, 0.90, and 0.86 and recall of 0.99, 0.98, 0.94, and 0.92 for R-ISS, ISS, DS, and unclear stage, respectively.

CONCLUSION

Our MM stage extraction algorithm uses rule-based NLP and data aggregation to accurately measure MM stage documented in oncology notes and pathology reports in VA's national EHR system. It may be adapted to other systems where MM stage is recorded in clinical notes.

Development and validation of algorithms for identifying lines of therapy in multiple myeloma using real-world data

Aim: To validate algorithms based on electronic health data to identify composition of lines of therapy (LOT) in multiple myeloma (MM). Materials & methods: This study used available electronic health data for selected adults within Henry Ford Health (Michigan, USA) newly diagnosed with MM in 2006-2017. Algorithm performance in this population was verified via chart review. As with prior oncology studies, good performance was defined as positive predictive value (PPV) ≥75%. Results: Accuracy for identifying LOT1 (N = 133) was 85.0%. For the most frequent regimens, accuracy was 92.5-97.7%, PPV 80.6-93.8%, sensitivity 88.2-89.3% and specificity 94.3-99.1%. Algorithm performance decreased in subsequent LOTs, with decreasing sample sizes. Only 19.5% of patients received maintenance therapy during LOT1. Accuracy for identifying maintenance therapy was 85.7%; PPV for the most common maintenance therapy was 73.3%. Conclusion: Algorithms performed well in identifying LOT1 - especially more commonly used regimens - and slightly less well in identifying maintenance therapy therein.

More intensive therapy as more effective treatment for frail patients with multiple myeloma [corrected]

Although randomized controlled trial data suggest that the more intensive triplet bortezomib-lenalidomide-dexamethasone (VRd) is superior to the less intensive doublet lenalidomide-dexamethasone (Rd) in patients newly diagnosed with multiple myeloma (MM), guidelines have historically recommended Rd over VRd for patients who are frail and may not tolerate a triplet. We identified 2573 patients (median age, 69.7 years) newly diagnosed with MM who were initiated on VRd (990) or Rd (1583) in the national US Veterans Affairs health care System from 2004 to 2020. We measured frailty using the Veterans Affairs Frailty Index. To reduce imbalance in confounding, we matched patients for MM stage and 1:1 based on a propensity score. Patients who were moderate-severely frail had a higher prevalence of stage III MM and myeloma-related frailty deficits than patients who were not frail. VRd vs Rd was associated with lower mortality (hazard ratio [HR], 0.81; 95% confidence interval [CI], 0.70-0.94) in the overall matched population. Patients who were moderate-severely frail demonstrated the strongest association (HR 0.74; 95% CI, 0.56-0.97), whereas the association weakened in those who were mildly frail (HR, 0.80; 95% CI, 0.61-1.05) and nonfrail (HR, 0.86; 95% CI, 0.67-1.10). VRd vs Rd was associated with a modestly higher incidence of hospitalizations in the overall population, but this association weakened in patients who were moderate-severely frail. Our findings confirm the benefit of VRd over Rd in US veterans and further suggest that this benefit is strongest in patients with the highest levels of frailty, arguing that more intensive treatment of myeloma may be more effective treatment of frailty itself.

The Association of Agent Orange (AO) Exposure with Monoclonal Gammopathy of Undetermined Significance (MGUS) to Multiple Myeloma (MM) Progression: A Population-based Study of Vietnam War Era Veterans

Background

Herbicide and pesticide exposure (e.g., agent orange [AO]) is associated with an increased risk of multiple myeloma (MM) due to the contaminant, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Monoclonal gammopathy of undetermined significance (MGUS) is the precursor state to MM; however, not all patients with MGUS progress to MM. It is unclear whether AO exposure increases the risk of progression of MGUS to MM.

Purpose

We aimed to determine the association between AO exposure and progression to MM in a nation-wide study of U.S. Veterans with MGUS.

Patients and Methods

This is a population-based cohort study of Vietnam Era Veterans diagnosed with MGUS. A natural language processing (NLP) algorithm was used to confirm MGUS and progression to MM. The association between AO and progression was analyzed using multivariable Fine-Gray subdistribution hazard model with death as a competing event. Veterans who served during the Vietnam War Era from 1/9/1962-5/7/1975 and were diagnosed with MGUS between 10/1/1999-12/31/2021 were included. We excluded patients with missing BMI values, progression within 1 year after MGUS diagnosis date, non-IgG or IgA MGUS, or birth years outside of the range of the AO exposed group, and race other than Black and White. AO exposure and service during 1/9/1962-;5/7/1975 and stratified according to TCDD exposure levels by three time periods: 1/9/1962-11/30/1965 (high), 12/1/1965-12/31/1970 (medium), or 1/1/1971-5/7/1975 (low). The association between AO and progression was analyzed using multivariable Fine-Gray subdistribution hazard model with death as a competing event.

Results

We identified 10,847 Veterans with MGUS, of whom 7,996 had AO exposure. Overall, 7.4% of MGUS patients progressed to MM over a median follow-up of 5.2 years. In multivariable analysis, AO exposure from 1/9/1962-11/30/1965, high TCDD exposure, was associated with an increased risk of progression (adjusted hazard ratio 1.48; 95% confidence interval 1.02-2.16), compared to Veterans with no exposure.

Conclusions

In patients with MGUS, the high Agent Orange exposure time period is associated with a 48% increased risk of progression to multiple myeloma. This suggests that patients with MGUS and prior Agent Orange exposure or occupational exposure to TCDD (eg. Agricultural workers) may require thorough screening for plasma cell dyscrasias.

Natural Language Processing-Assisted Classification Models to Confirm Monoclonal Gammopathy of Undetermined Significance and Progression in Veterans' Electronic Health Records

PURPOSE

To develop and validate natural language processing (NLP)-assisted machine learning (ML)-based classification models to confirm diagnoses of monoclonal gammopathy of undetermined significance (MGUS) and multiple myeloma (MM) from electronic health records (EHRs) in the Veterans Health Administration (VHA).

MATERIALS AND METHODS

We developed precompiled lexicons and classification rules as features for the following ML classifiers: logistic regression, random forest, and support vector machines (SVMs). These features were trained on 36,044 EHR documents from a random sample of 400 patients with at least one International Classification of Disease code for MGUS diagnosis from 1999 to 2021. The best-performing feature combination was calibrated in the validation set (17,826 documents/200 patients) and evaluated in the testing set (9,250 documents/100 patients). Model performance in diagnosis confirmation was compared with manual chart review results (gold standard) using recall, precision, accuracy, and F1 score. For patients correctly labeled as disease-positive, the difference between model-identified diagnosis dates and the gold standard was also computed.

RESULTS

In the testing set, the NLP-assisted classification model using SVMs achieved best performance in both MGUS and MM confirmation with recall/precision/accuracy/F1 of 98.8%/93.3%/93.0%/96.0% for MGUS and 100.0%/92.3%/99.0%/96.0% for MM. Dates of diagnoses matched (±45 days) with those of gold standard in 73.0% of model-confirmed MGUS and 84.6% of model-confirmed MM.

CONCLUSION

An NLP-assisted classification model can reliably confirm MGUS and MM diagnoses and dates and extract laboratory results using automated interpretation of EHR data. This algorithm has the potential to be adapted to other disease areas in VHA EHR system.