OBJECTIVES

Claims data are widely used to determine SCCHN treatment costs. However, identification of recurrent disease (RD) is challenging due to lack of documentation in claims. Study objectives were to understand the timing of relapse, develop proxies for its detection in claims, and validate an algorithm published by Ricketts et al in 2014 for identification of RD which used UK hospital records and claims. The algorithm established that treatment after 120 days post completion of chemoradiotherapy serves as a benchmark for RD.

METHODS

A Delphi panel of 10 US community medical oncologists currently treating ≥5 SCCHN patients discussed the patient journey and recurrence in real-world clinical practice and provided feedback on the Ricketts’ algorithm.

RESULTS

Review of a patient journey included biopsy, imaging, evaluation by surgical/radiation/medical oncologists, staging, and ancillary procedures, (e.g., central line placement, dental exam). Consensus was reached that SCCHN patients are referred to medical oncologists ≤30-45 days after diagnosis and initiate chemoradiotherapy ≤30-45 days after referral, although typically sooner. Concurrent platinum-based chemoradiotherapy is standard of care. Patients are staged by positron emission tomography (PET) and restaged 3 months after last radiation dose when a negative PET scan is interpreted as complete remission. Pre-immuno-oncology (IO), the possibility of post-radiation inflammation made interpretation of a positive PET scan equivocal, deferring the decision on salvage therapy until after a 6-month PET scan and challenging the Ricketts’ 120-day definition of relapse. With the emergence of IO, panelists stated that oncologists now commonly initiate IO if the 3-month PET is positive.

CONCLUSIONS

Based on Delphi consensus, in the IO era, delays in salvage therapy beyond 120 days represent true relapse events validating the Ricketts’ algorithm. IO initiation between day 90 and 120 may lead to difficulty in distinguishing between relapse or resistance in SCCHN.