Reveal-MM: Retrospective evaluation of variables in early assessment and landmark trends in multiple myeloma – a US claims-based case-control study Free

Introduction Multiple myeloma (MM) presents with non-specific symptoms, often leading to diagnostic delays or mis-referrals to various specialists, including nephrologists and/or orthopedic specialists. Compared to other cancers, patients also experience the highest number of consultations before specialist referral and MM diagnosis. Such delays are associated with secondary complications and poorer clinical outcomes. While these diagnostic delays are recognized by MM specialists, real-world evidence describing medical encounters prior to MM diagnosis remain limited. This study aims to better understand healthcare patterns in the two years preceding MM diagnosis.

Methods This retrospective case-control cohort study analyzed de-identified claims data from Optum Clinformatics Data Mart (CDM), a large, closed US database containing longitudinal administrative health claims for Medicare and commercially insured individuals.

Two cohorts were defined: a pre-MM cohort and a matched control cohort. The pre-MM cohort (patients yet to be diagnosed with MM) included adults aged ≥50 years with a confirmed MM diagnosis between 1/1/2020 and 1/31/2024, defined by ≥2 MM-coded claims (ICD-10 C90.x or ICD-9 203.x) ≥30 days apart. Patients required ≥24 months of continuous data coverage before the first MM diagnosis (index date). Those with any malignant cancer diagnosis or monoclonal gammopathy of undetermined significance during this period were excluded. The control cohort had no MM diagnosis during the study period; their index date was the last claim between 1/1/2020 and 12/31/2024. The distribution of control index dates was matched to the pre-MM cohort. The same exclusion criteria applied.

Controls were matched 1:1 to the pre-MM cohort using propensity score matching based on age, gender, race, region, insurance type, Charlson Comorbidity Index (CCI), and National Cancer Institute Comorbidity Index (NCI) scores, assessed three months pre-index date. Covariate balance was achieved (standardized mean differences <0.1).

Medical history was captured from 24 months pre-index date using binary indicators (diagnosis code presence), prescription and procedure counts, and physician visits by specialty, then calculated to 12-, 9-, and 6-months pre-index date. Codes were grouped into expert-defined clinical groups for interpretation. Patient characteristics and encounter variables were summarized using descriptive statistics. Between-group comparisons used Wilcoxon rank-sum (continuous) and chi-square (categorical) tests, with Bonferroni adjustment.

Results The pre-MM cohort (n=4,733) and control cohort (n=4,733) were balanced, with a mean age of 74.1 years, 50% male, 51% White, and 45% residing in the southern US. Medicare was the primary insurer for 86% of patients. Mean NCI scores were 0.71 (pre-MM) vs. 0.67 (control), and CCI scores 2.29 vs. 2.18, respectively.

The pre-MM cohort were significantly more likely to have anemia-related diagnostic codes (including iron deficiency anemia secondary to chronic blood loss, unspecified iron deficiency anemia, nutritional anemia, and anemia in other chronic diseases) compared to the control cohort, from 12 months pre-MM diagnosis/index date (29% vs. 18%, respectively; p<0.001). Pancytopenia and neutropenia (unspecified neutropenia and decreased white blood cell count) diagnostic codes were also more frequent in the pre-MM cohort from 12 months (p<0.001).

From 12 months, musculoskeletal pain (unspecified osteoarthritis and low back pain) was more frequently coded in pre-MM patients vs. controls. Pre-MM patients had a higher prevalence of diagnostic codes unrelated to MM diagnostic criteria, including gastroesophageal reflux disease (GERD) with esophagitis and cardiovascular conditions (chronic atrial fibrillation and unspecified cardiac murmur) compared to the control cohort (5.4% vs. 2.9%, respectively; p<0.001).

ConclusionsThis study presents the largest US claims-based analysis to-date examining pre-diagnostic patterns in MM patients compared to matched controls. In the pre-MM cohort, distinct encounter patterns emerged from 12 months before MM diagnosis, including codes for anemia and musculoskeletal complaints, both common presenting symptoms of MM; and codes unrelated to MM diagnostic criteria, including other cytopenias, cardiovascular conditions, and GERD with esophagitis. Secondary analyses will explore the potential of a predictive risk model to aid earlier MM detection in clinical practice.