Real-world treatment patterns and adverse events in metastatic renal cell carcinoma from a large US claims database

Abstract

Background: Vascular endothelial growth factor (VEGF), tyrosine kinase (TK) and mechanistic target of rapamycin kinase (mTOR) inhibitors are common first-line (1 L) treatments for metastatic renal cell carcinoma (mRCC). Despite treatment availability, the 5-year survival rate in patients diagnosed at the metastatic stage is only ≈ 10%. To gain contemporary insights into RCC treatment trends that may inform clinical, scientific and payer considerations, treatment patterns and adverse events (AEs) associated with 1 L therapy were examined in a retrospective, longitudinal, population-based, observational study of patients with mRCC.

Methods: US administrative claims data (Truven Health MarketScan Commercial Databases) were used to assess trends in 1 L treatment initiation in mRCC (2006-2015) and characterize patterns of individual 1 L treatments, baseline characteristics, comorbidities and treatment-related AEs from 2011 through 2015. Outcomes were evaluated by drug class and route of administration.

Results: Ten-year trend analysis (n = 4270) showed that TK/VEGF-directed therapy rapidly became more common than mTOR-directed therapy, and oral treatments were favored over intravenous (IV) treatments. Overall, 1992 eligible patients initiated 1 L treatment for mRCC from 2011 through 2015: 1752 (88%) received TK/VEGF-directed agents and 233 (12%) received mTOR-directed agents; 1674 (84%) received oral treatments, and 318 (16%) received IV treatments. The most common 1 L treatment was sunitinib (n = 849), followed by pazopanib (n = 631), temsirolimus (n = 157) and bevacizumab (n = 154). Patient characteristics and comorbidities, including age, diabetes and congestive heart failure, were independent predictors of 1 L mRCC treatment choice. The three most common potentially 1 L treatment-related AEs were nausea/vomiting (128.2 per 100 patient-years [PY]), hypertension (69 per 100 PY) and renal insufficiency (44.6 per 100 PY). A wide variety of agents were used as second-line (2 L) therapy. Substantial latency of onset was observed for several potentially treatment-related toxicities in patients treated with TK/VEGF- or mTOR-directed agents.

Conclusions: In the US, 1 L TK/VEGF inhibitor uptake in recent years appears largely in line with national approvals and guidelines, with varied 2 L agent use. Although retrospective evaluation of claims data cannot assess underlying causality, insights from these real-world RCC treatment and AE patterns will be useful in informing medical and payer decisions.

Keywords: Administrative claims; Adverse events; Renal cell carcinoma; Targeted therapy; Treatment patterns.

Fig. 1

Study design. a Trend analysis. b Treatment patterns, patient characteristics and AE analysis. The study periods are indicated as durations between closed circles. The index date (date of 1 L treatment initiation for mRCC) could occur during the index period. Patients were followed up for a minimum of 6 months until the end of continuous enrollment. The end of 1 L treatment was defined as the last 1 L treatment claim date plus (1) the number of days of supply of the oral treatments (last claim) or (2) the labeled cycle length for the IV treatments. Time to 2 L treatment (TT2T) was defined as the initiation of a new 1 L therapy regimen > 30 days following the index date or restart of 1 L index treatment following a > 3-month gap

Fig. 2

Trends in 1 L treatment for mRCC by drug class and route of administration (2006–2015). a Trends by drug class and route of administration. TK/VEGF inhibitors included sunitinib, pazopanib, sorafenib, axitinib and bevacizumab ± IFN-α. mTOR treatments included temsirolimus and everolimus. Orally administered agents included sunitinib, pazopanib, sorafenib, axitinib and everolimus; IV treatments included bevacizumab ± IFN-α, temsirolimus and interleukin 2. For the oral (blue line) vs IV (orange line) comparison, the percentage of each group sums to 100% for each year. For the TK/VEGF (grey line) vs mTOR (yellow line) comparison, the percentage of each group may sum to < 100% because interleukin 2 was not included (accounts for < 1% of patients in each year). b Trends by individual treatment. Trends are plotted as a single value for the entire year, and approximate timings for US FDA approvals are indicated by diamond symbols in part b (pre-2006 approvals are shown adjacent to the y-axis). Approvals plotted include 1 L sunitinib (January 26, 2006), 1 L pazopanib (October 19, 2009), 2 L sorafenib (December 20, 2005), 2 L axitinib (January 27, 2012), 1 L bevacizumab ± IFN-α (July 31, 2009), 1 L temsirolimus (May 30, 2007), 2 L everolimus (March 30, 2009) and 1 L interleukin 2 (May 5, 1992). In each given year, the percentage of patients receiving each indicated agent sums to 100%

Fig. 3

Multivariate analysis. a Odds of receiving 1 L mTOR vs TK/VEGF inhibitors. b Odds of receiving IV vs orally administered treatment. Only variables found to be significant are plotted. Other variables in the multivariate analysis for the comparison of TK/VEGF inhibitor treatment vs mTOR inhibitor treatment were not significantly different (age, sex, index years 2013 and 2014 vs 2015, DCCI and nephrectomy). Other variables in the multivariate analysis for the comparison of oral vs IV administration that did not reach statistical significance were sex, employment status (active and long-term disability vs retiree), insurance type, DCCI, index years 2013 and 2014 vs 2015 and nephrectomy)

Fig. 4

1 L to 2 L treatment flow in patients with mRCC in the Truven Health MarketScan databases. Possible reasons for patients not receiving 2 L treatment include still being on 1 L treatment, no requirement for 2 L treatment, refusal of treatment and death, although these cannot be reliably obtained from the claims database