Article Text

Download PDFPDF

Patterns of response in patients with advanced melanoma treated with Pembrolizumab (MK-3475) and evaluation of immune-related response criteria (irRC)
  1. F Stephen Hodi1,
  2. Antoni Ribas2,
  3. Adil Daud3,
  4. Omid Hamid4,
  5. Caroline Robert5,
  6. Richard Kefford6,
  7. Wen-Jen Hwu7,
  8. Tara C Gangadhar8,
  9. Anthony M Joshua9,
  10. Peter Hersey10,
  11. Jeffrey Weber11,
  12. Richard W Joseph12,
  13. Hassane Zarour13,
  14. Roxana Dronca12,
  15. Linda Gammage14,
  16. Darcy Hille14,
  17. Dahai Xue14,
  18. S Peter Kang14,
  19. Patrick Chun14,
  20. Scot W Ebbinghaus14,
  21. Andrea Perrone14 and
  22. Jedd D Wolchok15
  1. Aff1 grid.65499.370000000121069910Dana-Farber Cancer Institute Boston MA USA
  2. Aff2 grid.19006.3e0000000096326718University of California Los Angeles CA USA
  3. Aff3 grid.266102.10000000122976811University of California San Francisco CA USA
  4. Aff4 grid.488730.0The Angeles Clinic and Research Institute Los Angeles CA USA
  5. Aff5 grid.14925.3b0000000122849388Gustave Roussy Villejuif France
  6. Aff6 grid.419690.30000000404916278Westmead Hospital and Melanoma Institute of Australia NSW Australia
  7. Aff7 grid.267308.80000000092062401The University of Texas, MD Anderson Cancer Center Houston TX USA
  8. Aff8 grid.25879.310000000419368972Abramson Cancer Center of the University of Pennsylvania Philadelphia PA USA
  9. Aff9 grid.231844.80000000404740428Princess Margaret Cancer Centre Toronto ON Canada
  10. Aff10 grid.1013.3000000041936834XUniversity of Sydney Sydney NSW Australia
  11. Aff11 grid.170693.a000000012353285XH. Lee Moffitt Cancer Center Tampa FL USA
  12. Aff12 grid.66875.3a000000040459167XMayo Clinic Rochester MN USA
  13. Aff13 grid.21925.3d0000000419369000University of Pittsburgh Pittsburgh PA USA
  14. Aff14 grid.417993.10000000122600793Merck & Co USA
  15. Aff15 grid.51462.340000000121719952Memorial Sloan Kettering Cancer Center New York NY USA

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Meeting abstracts


Unique patterns of response have been observed with immunotherapies. Notably, objective response and prolonged disease stabilization can occur after an initial increase in tumor burden. irRC were developed to better characterize response to immunotherapy based on data for Ipilimumab. We previously showed that patients with melanoma treated with the anti-PD-1 monoclonal antibody Pembrolizumab may also experience unique patterns of response and that conventional response criteria may underestimate the therapeutic benefit of Pembrolizumab [1]. We updated our initial analysis to include an additional 6 months of follow-up.


Patients from 3 melanoma cohorts treated with Pembrolizumab 2 mg/kg Q3W, 10 mg/kg Q3W, or 10 mg/kg Q2W in the Phase I KEYNOTE-001 trial served as the source population. Tumor imaging was performed every 12 weeks. Response was assessed by irRC and RECIST v1.1 by independent central review. Patients were managed by irRC by investigator. Early and delayed pseudo-progression were identified using centrally assessed irRC data among patients treated with Pembrolizumab who were followed by imaging for ≥28 weeks as of May 2014. Early pseudo-progression was defined as unconfirmed PD at assessment 1 (ie, week 12) and non-PD at assessment 2. Delayed pseudo-progression was defined as PD at any time point followed by non-PD at the next assessment. Survival data as of May 2014 were analyzed in all 411 patients enrolled in KEYNOTE-001.


Following an incremental assessment of 6 months, there were an additional 16 patients with ≥28 weeks of imaging follow-up (n = 208 total), 1 additional patient who experienced early pseudo-progression (n = 8 total; 3.8%), and 2 additional patients who experienced delayed pseudo-progression (n = 9 total; 4.3%). This corresponded to a 0.9% increase in atypical responses compared with the previous assessment. Based on analysis of best overall response in the total population (n = 411), 2 additional patients had PD by RECIST but CR/PR/SD by irRC (n = 53 total). These 53 patients had favorable OS compared with the 145 patients who had PD by both criteria (Figure 1).

Figure 1

Kaplan-Meier Estimates of OS.


Pembrolizumab-treated patients with melanoma may experience unique patterns of response and should be managed accordingly. Analysis of OS suggests that conventional RECIST may underestimate the benefit of Pembrolizumab in approximately 10% of patients. These and other data suggest that new standards for response criteria should be considered for PD-1 inhibitors and other immunotherapies.

Registration, unique identifier NCT01295827


  1. 1.