Oops, you're using an old version of your browser so some of the features on this page may not be displaying properly.

MINIMAL Requirements: Google Chrome 24+Mozilla Firefox 20+Internet Explorer 11Opera 15–18Apple Safari 7SeaMonkey 2.15-2.23

Poster display

2769 - Slow natural history predicts higher response rate to nivolumab and pembrolizumab in advanced melanoma patients


09 Oct 2016


Poster display


Nora Kramkimel


Annals of Oncology (2016) 27 (6): 379-400. 10.1093/annonc/mdw379


N. Kramkimel1, V. Heidelberger2, O. Huillard2, P. Boudou Rouquette3, J. Chanal1, N. Franck1, J. Arrondeau2, J. Alexandre2, B. Blanchet4, J. Mullaert5, S. Aractingi1, N. Dupin1, F. Goldwasser2

Author affiliations

  • 1 Department Of Dermatology, Hôpital Cochin, 75014 - Paris/FR
  • 2 Medical Oncology, Hôpital Cochin, 75014 - Paris/FR
  • 3 Medical Oncology, Hôpital Cochin, 75679 - Paris/FR
  • 4 Functional Unit Of Pharmacokinetics And Pharmacochemistry, Hôpital Cochin, 75014 - Paris/FR
  • 5 Biostatistics, Hopital Bichat Claude Bernard, Paris/FR


Abstract 2769


Anti PD-1 antibodies nivolumab and pembrolizumab are checkpoint inhibitors widely used in metastatic melanoma with a 40% response rate. Little is known on predictive factors of response. Given the mechanism of action, we investigated whether features of natural history of melanoma correlated with response.


All melanoma patients treated with anti PD-1 between August 2014 and January 2016 in our center were retrospectively reviewed. Objective response to treatment was defined as complete response or partial response according to 1.1 RECIST criteria. Patients who received only 1 infusion were excluded. No clear definition of lymphatic or hematogenous dissemination in melanoma could be found in the literature. We defined lymphatic dissemination as exclusive lymphatic metastases or occurrence of a lymph node metastasis prior to a visceral metastasis. The rest was considered hematogenous dissemination. We excluded patients with only in-transit metastases (stage N2c). Time-in-node was defined as the delay between first lymphatic metastasis and first visceral metastasis. Time-to-treatment was defined as the delay between melanoma diagnosis and anti PD-1 therapy initiation.


65 patients were included (31 females; median age 65 years [21 to 90]). Treatment was initiated in patients having disease progression. 73% received pembrolizumab and 27% nivolumab. 28% of tumors harbored BRAF V600 mutations. Anti PD-1 was the first line therapy for 36% of the patients. Dissemination was lymphatic in 23 patients (37%) and hematogenous in 39 (63%). Objective response rate in the total population was 40%. Mean time-in-node was 26 months [2 to 132 months]. Mean time-to-treatment was 71 months [2 to 409 months]. There was no statistical correlation between response and either lymphatic vs hematogenous dissemination or time-in-node. Time-to-treatment was statistically associated with response (mean time-to-treatment 99 months in responders and 53 in non-responders, p = 0, 01). The same analysis with time from diagnosis to first line therapy was also positive (85 months in responders and 45 in non-responders, p = 0, 02).


Melanomas with slow natural history exhibit a higher sensitivity to nivolumab and pembrolizumab.

Clinical trial identification

Legal entity responsible for the study

Hôpital Cochin APHP


Hôpital Cochin APHP


All authors have declared no conflicts of interest.

This site uses cookies. Some of these cookies are essential, while others help us improve your experience by providing insights into how the site is being used.

For more detailed information on the cookies we use, please check our Privacy Policy.

Customise settings