Abstract 1902
Background
VEGF- and mTOR-directed therapies are the mainstay of mRCC treatment. Guidelines recommend risk stratification for treatment selection. Here, we examined how demographics and comorbidities may impact treatment selection.
Methods
Patients (pts) initiating 1L mRCC therapy from 2011 to 2013 were identified in MarketScan, a US claims database with ≈40 million pts with employer-sponsored insurance. Pts with non-RCC histologies and previous primary cancers were excluded. Pts were stratified by 1L drug category (VEGF- or mTOR-directed) and dosage form (intravenous [IV] or oral [PO]). Clinical and demographic characteristics (eg, age, sex, employment, region, insurance type), baseline comorbidities, and Charlson Comorbidity Index (CCI) were assessed. Multivariate logistic regression models examined associations between these characteristics and receipt of (1) VEGF- vs mTOR-directed and (2) IV vs PO therapy.
Results
1262 mRCC pts initiated 1L therapy between 2011 and 2013 and met selection criteria. Median age was 62 y; 70% were male; 84% had a CCI of 0-1, with diabetes (26%) and chronic kidney disease (19%) being the most frequent. VEGF-directed therapy was more common than mTOR-directed therapy (87% vs 13%). PO therapy was more common than IV therapy (83% vs 17%). Diabetes was more common in pts taking VEGF-directed therapy, while the opposite was seen for CHF; these factors remained independently associated with 1L therapy in multivariate analyses (Table). Pts receiving IV therapy were older than those receiving PO therapy (mean, 65 vs 62 y), and age remained significant in the multivariate-adjusted model (P = .04).
Association between baseline diabetes and CHF and receipt of mTOR-directed therapy in 1L mRCC
| Adjusted odds of receiving mTORa | ||||
|---|---|---|---|---|
| Characteristic | mTOR N = 165 | VEGF N = 1094 | OR (95% CI) | P value |
| Diabetes | ||||
| No | 81% | 72% | Ref | |
| Yes | 19% | 28% | 0.57 (0.37-0.87) | .01 |
| CHF | ||||
| No | 87% | 94% | Ref | |
| Yes | 13% | 6% | 2.39 (1.4-41) | .001 |
a Adjusted for age, sex, employment status, geographic region, insurance type, CCI score, and other baseline comorbidities that were statistically significant (P
Conclusions
Baseline diabetes and CHF as well as age are independent predictors of 1L mRCC treatment selection. Differences in treatment safety profiles and pt health may drive this difference. These results will inform ongoing studies that compare treatment selection in 1L mRCC.
Clinical trial identification
NA
Legal entity responsible for the study
F. Hoffmann La-Roche
Funding
F. Hoffmann La-Roche
Disclosure
S. Mhatre, S-W. Lin, H. Wallen, J. Simpson, S. Ogale: Employee of Genentech, Inc. A. Surinach: Employee of Genesis Research. R. Vohra: Former employee of Genentech, Inc. All other authors have declared no conflicts of interest.