Paper Details

ABSTRACT

Introduction: Baseline elevated body mass index (BMI) is associated with worse progression free (PFS) and overall survival (OS) in multiple myeloma (MM). However, little is known about how changes in weight and metabolic health during induction therapy impact outcomes. To address this gap, we longitudinally profiled weight, body composition, markers of insulin resistance and metabolomics during MM induction.

Methods: In this single center observational study of 355 newly diagnosed MM (NDMM) patients who completed induction with lenalidomide, dexamethasone and carfilzomib (KRd, n=185) or bortezomib (VRd, n=170), we measured BMI at baseline (BL), end of induction (EOI), and its longitudinal change. BMI was defined as elevated if ≥25kg/m². A significant loss or gain of BMI was defined as >5% change.

Body composition metrics were calculated at BL (n=92) and EOI (n=43) from computed tomography scans using open source deep learning anatomic segmentation models. Metrics included muscle (MT) and adipose tissue – visceral (VAT), subcutaneous (SAT), and intramuscular (IMAT). 

Adiponectin and leptin levels to calculate an AL ratio (low ratio implies greater insulin resistance) were measured from paired plasma samples collected at BL (n=128) and EOI (n=57). Of these, 118 underwent metabolomics via LC-MS as an exploratory analysis.

These metrics of metabolic health were correlated to PFS and OS. Multivariable analyses (MVA) studying AL ratio included adjustments for age, sex, cytogenetics. BMI and metabolomic analyses were additionally adjusted for stage (RISS), cardiac history, and autologous stem cell transplant. Wilcoxon rank-sum tests were performed to compare quantitative measures between groups.

Results: In our cohort, 76% of patients had an elevated BMI. While 63% of patients maintained stable weight during induction, 20% had >5% BMI increase and 17% had >5% BMI decrease. Older patients had greater BMI reduction (−0.03 kg/m²/year age increase, p<0.01).

Elevated BMI was associated with reduced PFS (BL p=0.01; EOI p=0.01). These associations remained significant after MVA (BL HR 1.71, p=0.02; EOI HR 1.73, p=0.01). Elevated BMI did not associate with reduced OS (BL HR 2.34, p=0.08; EOI HR 1.70 p=0.2). BMI increase did not have increased hazard of disease progression or death compared to BMI reduction (PFS HR 1.34, p=0.5; OS HR 1.47, p=0.5)

Median SAT (BL p=0.03; EOI p=0.7) and VAT volumes (BL p=0.01; EOI p=0.02) were higher and MT volumes (BL p<0.01; EOI p<0.01) were lower in patients with an elevated BMI compared to normal BMI. There was an increase in median VAT (p<0.01) and IMAT (p=0.02), with a decrease in MT (p=0.04) and no change in SAT (p=0.12) by EOI compared to BL.

At BL and EOI, patients with an elevated BMI had lower median AL ratio, compared to those with a normal BMI (p≤0.01). In patients with paired samples, median insulin sensitivity improved by EOI (AL ratio change +0.26, p=0.01). On MVA, AL ratio had a non-significant trend for better PFS at EOI (HR 0.23; p=0.07). While AL ratios increased by 31% from BL to EOI in the stable weight group (n=44; p=0.01), patients with ≥5% BMI decrease showed a 2.8-fold increase (n=7; p=0.02). There was no change in patients with a ≥5% BMI increase (n=6; p=0.8). This supports weight loss during induction is associated with improved insulin sensitivity and, by extension, overall metabolic health.

Metabolomics revealed metabolic pathway changes and only metabolites with p<0.05 in both univariate and MVA are reported. Elevated fumarate (worse PFS and OS), glutamine (worse PFS), 4-methylaminobutyrate (better PFS) as well as leucine, isoleucine, dimethylglycine, and N-methylalanine (better OS) correlated with clinical outcomes. These results suggest that metabolic activity may have clinical relevance beyond reporting on adiposity and insulin sensitivity.

Conclusions: This is the first study to comprehensively profile changes in metabolic markers during MM induction. Elevated baseline BMI correlated with higher abdominal fat, insulin resistance and decreased PFS. Induction therapy generally led to unfavorable metabolic changes despite stable BMI, including increased VAT and reduced muscle volume. These data support evaluating established biomarkers of metabolic health beyond BMI and provide evidence to prospectively study dietary interventions to improve metabolic health and clinical outcomes in NDMM (NUTRIVENTION5; NCT07226609).