Article Text

Download PDFPDF

606 Biomarker analyses of baseline tumor specimens and on-treatment changes in sera samples of patients enrolled in the RELATIVITY-047 trial to characterize LAG-3 biology
  1. Sonia Dolfi1,
  2. Tracy Tang1,
  3. Georgina Long AO2,
  4. Paolo Ascierto3,
  5. F Stephen Hodi4,
  6. Evan Lipson5,
  7. Dirk Schadendorf6,
  8. John Wojcik1,
  9. Jennifer Postelnek1,
  10. Yu Wang1,
  11. Anne Marie Sobiesk1,
  12. Kaushal Mishra1,
  13. Sarah Keidel1,
  14. Karen Miller-Moslin1,
  15. Hussein Tawbi7 and
  16. Charlie Garnett-Benson1
  1. 1Bristol Myers Squibb, Princeton, NJ, United States
  2. 2Melanoma Institute Australia, The University of Sydney; Royal North Shore Hospital; Mater Hospital, Wollstonecraft, Australia
  3. 3Istituto Nazionale Tumori IRCCS Fondazione G. Pascale, Milan, Italy
  4. 4Dana-Farber Cancer Institute, Boston, MA, United States
  5. 5Bloomberg-Kimmel Institute for Cancer Immunotherapy, Johns Hopkins Sidney Kimmel Comprehensive Cancer Center, Baltimore, MD, United States
  6. 6University Hospital Essen, and German Cancer Consortium, Essen, Germany
  7. 7University of Texas MD Anderson Cancer Center, Houston, TX, United States


Background The phase 2/3 RELATIVITY-047 clinical trial (NCT03470922) in patients with previously untreated metastatic or unresectable melanoma met its primary endpoint of improved progression-free survival with a combination of nivolumab plus relatlimab (NIVO+RELA) compared with NIVO monotherapy.1 To better understand the biology underlying combined programmed death-1 (PD-1) and lymphocyte-activation gene 3 (LAG-3) inhibition, a series of exploratory biomarker analyses was performed. Analyses included longitudinal changes in immunomodulatory cytokines, free soluble LAG-3 (sLAG-3), and selected immune-cell markers within tumor biopsies collected at baseline.

Methods Peripheral blood serum samples collected prior to treatment, and prior to study drug during week 4 and week 8 of treatment, were evaluated for changes in inflammatory cytokines including interferon gamma (IFNγ) and sLAG-3. Baseline tumor samples were collected prior to treatment and were analyzed by immunohistochemistry for LAG-3 (17B4), CD8 (C8/144B), and tumor cell programmed death-ligand 1 (PD-L1) expression (Agilent/Dako PD-L1 28-8 IHC pharmDx).1–3 Correlative analyses were performed.

Results Four weeks after treatment initiation, IFNγ levels increased over baseline in both treatment arms (table 1), but increases were significantly larger with NIVO+RELA than with NIVO (fold change 2.17 and 1.54, respectively; P <0.0001). Eight weeks after treatment initiation, sLAG-3 levels were modestly increased with NIVO (fold change 1.12; P <0.001), and significantly decreased with NIVO+RELA (fold change 0.60; P <0.001) (table 2), suggesting RELA-specific target engagement in the combination arm. No significant correlation was observed between baseline sLAG-3 and tissue-based LAG-3 expression in pretreatment tumors (Spearman’s r=0.21). In baseline tumor biopsies, LAG-3 showed moderate correlation with PD-L1 expression (r=0.53) (figure 1). The majority of LAG-3–negative tumors (<1%) were also PD-L1 negative (<1%). PD-L1 expression varied among LAG-3–expressing tumors, with differences observed in the range of PD-L1 expression (high vs low) across LAG-3 subgroups. CD8 levels correlated more strongly with LAG-3 than with PD-L1. Furthermore, patients whose tumors had higher LAG-3 expression (≥1% and ≥5%) had numerically higher median progression-free survival and objective response rates in both the NIVO+RELA and NIVO arms.

Conclusions These exploratory analyses support the hypothesis that NIVO+RELA enhances immune activation compared with NIVO alone, and that LAG-3 may be a promising biomarker of tumor inflammation. Although LAG-3 alone may not be a useful predictive biomarker for selection of patients with melanoma for NIVO+RELA vs NIVO regimens, further work is required to assess the clinical utility of LAG-3, likely as part of a composite biomarker of immunotherapy susceptibility.

Acknowledgements Editorial support was provided by Sandra J Page, PhD, of Spark Medica Inc.

Trial Registration Clinicaltrials. gov. NCT03470922.


  1. Tawbi HA et al. Relatlimab and nivolumab versus nivolumab in untreated advanced melanoma. N Engl J Med. 2022;386:24–34.

  2. Johnson L et al. Development of a LAG-3 immunohistochemistry assay for melanoma. J Clin Pathol. 2022;jclinpath-2022-208254. doi:10.1136/jclinpath-2022-208254.

  3. Szabo PM et al. Development and performance of a CD8 gene signature for characterizing inflammation in the tumor microenvironment across multiple tumor types. J Mol Diagn. 2021;23:1159–1173.

Ethics Approval All trial investigators received approval from their respective institutional review boards. The trial was conducted in accordance with the International Council for Harmonisation Good Clinical Practice guidelines, and all patients provided written informed consent before participation. An independent data monitoring committee provided oversight to assess the efficacy and safety profile of relatlimab and nivolumab.

Abstract 606 Table 1

Serum IFNγ increases after treatment with NIVO+RELA and NIVO.

Abstract 606 Table 2

Soluble LAG-3 changes after treatment with NIVO+RELA and NIVO.

Abstract 606 Figure 1

Correlation between LAG-3 and PD-L1 expression by immunohistochemistry.

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.