Extending Lives ... at the Expense of the Job?

Oncologists are increasingly being forced into a bind, as costly novel therapeutic strategies with revolutionary clinical gains must be weighed against external pressures to reduce utilization.

By Steve Y. Lee, MD

In April, results of KEYNOTE-189, a phase III trial of combination chemotherapy and immunotherapy in first-line advanced non-squamous NSCLC, landed with a bang.  Simultaneously published in NEJM, the trial reported practice-changing improvements in all outcome measures including overall survival which promised greater uptake of the combination strategy previously approved in May 2017. However, exuberance was immediately tempered by concerns of soaring costs, with some estimates of $150,000 in additional per-patient costs.

Cost concerns further resonate for oncologists practicing under third party bundled payment systems or the CMS Quality Payment Program’s Merit-Based Incentive Payment System (MIPS) or Oncology Care Model (OCM), in which increased drug costs weigh negatively against reimbursement.  For MIPS, the composite score will adjust physician reimbursement +/- 4% by 2019, increasing to +/- 9% by 2022.  Federal advocacy from specialty groups including ASCO for more equitable risk adjustment or alternative scoring mechanisms such as clinical pathways has not yet resulted in change.

Studies such as KEYNOTE-189 in which massive clinical gains are coupled with potential increased toxicity raise further concern for demerits in both cost and non-cost measures of quality.  OCM Core Measures for quality include incidence of all-cause emergency department visits or inpatient admissions; similarly, CMS and private insurers aim to include the reduction of preventable ED visits and hospitalization as part of hospital outpatient prospective payment systems.

In this context, review of KEYNOTE-189 adverse events shows an increased level of toxicity with combination therapy.  Despite similar overall toxicity counts, further analysis reveals a near doubling of grade 3-to-5 adverse events that led to discontinuation of all treatment (11.9% vs 6.9%), any treatment component (20% vs 10.9%), immunotherapy (15.8% vs 8.4%), or pemetrexed chemotherapy (17% vs 8.4%).  As increased incidence of grade 3-to-5 toxicities has been associated with a five-fold increase in emergency department services and hospitalizations, early adopters of novel therapies run the risk of losses in measures of both cost and quality.

Unfortunately, CMS risk stratification mechanisms remain primitive, and while including certain clinical and demographic factors, they currently lack critical oncology-specific criteria such as cancer staging or therapy-defining biomarkers such as EGFR status or PD-L1 tumor proportion score.  Further, regimen-specific toxicity such as in KEYNOTE-189 in NSCLC, or the difference between one versus two immunotherapeutic agents in melanoma have yet to be addressed.

These competing demands can ultimately leave oncologists seeking the best for their patients in financial jeopardy both within their practices as well as from external payers.  Physicians under MIPS will have scores reported publically via a five-star system on the CMS Physician Compare website with score designations which can affect reputation and referrals or possibly employment or compensation status by the physician’s employer.

In the end, cancer patients may become unwitting victims of incompletely risk-adjusted quality programs, as financial hazards for oncologists pose strong disincentives against prescribing what may be the best contemporary therapies.