Gynecologic Oncology Reports (Aug 2021)

Pattern of disease and response to pembrolizumab in recurrent cervical cancer

  • Kathryn M. Miller,
  • Olga T. Filippova,
  • Sara A. Hayes,
  • Nadeem R. Abu-Rustum,
  • Carol Aghajanian,
  • Vance Broach,
  • Lora H. Ellenson,
  • Pier Selenica,
  • Elizabeth L. Jewell,
  • Chrisann Kyi,
  • Yuliya Lakhman,
  • Jennifer J. Mueller,
  • Roisin E. O'Cearbhaill,
  • Kay J. Park,
  • Yukio Sonoda,
  • Dmitriy Zamarin,
  • Britta Weigelt,
  • Mario M. Leitao, Jr,
  • Claire F. Friedman

Journal volume & issue
Vol. 37
p. 100831

Abstract

Read online

Objective: Since the approval of pembrolizumab for advanced or recurrent PD-L1 positive (CPS > 1%) cervical cancer, the clinical characteristics associated with response have remained undefined. We sought to characterize the clinicopathologic features of patients with advanced cervical cancer at our institution who derived durable clinical benefit from treatment with pembrolizumab. Methods: We conducted a retrospective cohort study of 14 patients with recurrent or metastatic cervical cancer who received pembrolizumab monotherapy from August 2017 to November 2019 and were followed until November 1, 2020. Reviewed clinical data included age, histology, tumor molecular profiling results, stage at diagnosis, treatment history, baseline pattern of metastatic disease at initiation of anti-PD-1 therapy, and outcomes. Treatment response was evaluated by computed tomography using RECIST v1.1 criteria. Results: The objective response rate was 21% (n = 3), including two partial responses and one complete response. Two patients (14%) had stable disease of six months or greater, for an observed durable clinical benefit rate of 36%. When stratified by those who derived clinical benefit, metastatic spread to lung and/or lymph node only at baseline was associated with improved response to pembrolizumab (n = 7, p = 0.02) and associated with significantly improved PFS and OS. Tumor mutational burden was higher in those with durable clinical benefit compared to non-responders (median 12.7 vs. 3.5 mutations/megabase, p = 0.03). Conclusions: Our findings highlight clinical features that may select for a population most likely to benefit from pembrolizumab monotherapy and underscores the need for identification of additional biomarkers of response.

Keywords