ASCO 2015: Pembrolizumab (Keytruda) - characteristics predictive of response, atypical responses, and effect on brain mets so far....study still enrolling!

Clinical characteristics predictive of response to pembrolizumab in advanced melanoma.  J Clin Oncol 33, 2015.  Tsai, Loo, Khurar, Daud, et al.

Advanced melanoma patients (110) from 12/2011 - 10/2013, were given Pembro in 1 of 3 dosing patterns:  2mg/kgq3wk, 10q3wk, or 10q2wks.  In this set, overall response rate was 40%.  Factors that correlated with significantly higher overall response were:  LDH at or below normal (ORR = 52%), no previous ipi (48%), lung mets (52.8%).  Patient with liver mets had a worse response (ORR = 18%) as did those with liver and lung mets (31%).  Conclusion:  Normal LDH, no prior ipi, and lung mets correlated with better response to pembro.  Correlations were observed regardless of BRAF status, presence of brain mets, or site of primary (cutaneous vs uveal).

Atypical patterns of response in patients with metastatic melanoma treated with pembrolizumab.  J Clin Oncol 33, 2015.  Wolchok, Hamid, Ribas, Robert, Weber, Hodi, et al.

Immune-related response criteria were developed to better characterize the atypical response patterns observed with ipi.  Previously we showed that 7.2% of melanoma patients treated with the anti-PD1 monoclonal antibody pembro also demonstrate atypical response patterns and that immune-related response criteria (irRC) may better represent the clinical benefit of pembro than conventional RECIST.  Patients were given pembro at 2mg/kg q 3 wks, 10mg/kg q 3 wks, or 10mg/kg q 2 wks with imaging done q 12wks.  Early pseudoprogression was defined as more than 25% increase in tumor burden at first assessment that was not confirmed as progressive disease at the next assessment performed about 4 wks later.  Delayed pseudoprogression was defined as more than 25% increase in tumor burden at any point after the first assessment, followed by non-progressive disease at the next assessment.  Results:  655 patients.  327 had more than 28 wk f/u by imaging at the time of analysis and were assessed for atypical responses.  Overall, 29 (8.9%) patients experienced atypical response.  Early pseudoprogression was observed in 15 (4%) Late pseudoprogression was observed in 14 patients.  In the 592 patients who survived more than 12 wks, 331 (56%) had best overall response of non-progressive disease per RECIST and irRC.  177 (30%) had progressive disease per both criteria.  84 (14%) was progressive disease per RECIST by non-progressive disease per irRC. CONCLUSION:  Results of this expanded analysis are consistent with previous reports suggesting that pembro may result in atypical response patterns and that conventional response criteria may underestimate the therapeutic benefit. 

Safety and activity of pembrolizumab in melanoma patients with untreated brain metastases.
J Clin Oncol 33, 2015.  Kluger, Goldberg, Sznol, Chiang, et al.

Brain mets develop in 40% of metastatic melanoma patients.  Untreated brain mets are excluded from most clinical trials. Prior trials of metastatic melanoma show treatment with pembro produced a response rate of more than 30%.  A phase II study was started in patients with previously untreated or progressing melanoma brain mets (recruitment is ongoing).  Patients with brain mets from melanoma or non-small cell lung cancer are accepted if they have at least 1 asymptomatic 5-20mm brain met not requiring immediate local therapy or systemic steriods, and at least 1 brain met amenable to biopsy or resection.  Prior PD-1/PD-L1 are excluded.  Pembro is given 10mg/kg q 2 wks. Brain MRI is done q 4 wks and restaging is done q 8 wks.  RESULTS thus far:  Between April and December of 2014, 17 patients were accrued.  6 had BRAF mutations.  10 had previously received ipi.  5 were unevaluable for brain met response due to: rapid progression of disease in the body (3) and intralesional hemorrhage (1) and 1 patient was too early in treatment.  Of 12 evaluable patients:  brain met partial responses were observed in 3 patients (1 with prior ipi), stable disease was found in 2, progressive disease in 7 (2 with a mixed response and 1 with progressive disease by imaging but with pseudoprogression on histology). Brain met responses are ongoing at 7+, 6+, and 3+ months.  One complete response and 3 partial responses were observed in extra-cerebral metastatic disease, 3 of these 4 with concordant brain met response.  The only grade 3 adverse event from pembro was liver function abnormalities.  2 patients had seizures, 1 from perilesional edema, 1 from tumor growth.  They were treated with anti-convulsants and transient use of steroids.  

Here is the link to the study...as noted accrual is ongoing:  Pembro for brain mets

No corticosteroid treatment for symptomatic disease allowed, though low dose replacement therapy is.  No leptomeningeal or autoimmune disease allowed.  No other concurrent treatments.  No radiotherapy for 14 days prior.  Yale is only recruiting site listed.

For what it's worth! Wishing you all my best. - c