Deep learning-based scoring of tumour-infiltrating lymphocytes is prognostic in primary melanoma and predictive to PD-1 checkpoint inhibition in melanoma metastases

Research output: Contribution to journalResearch articleContributedpeer-review

Contributors

  • Eftychia Chatziioannou - , University of Tübingen (Author)
  • Jana Roßner - , Heidelberg University  (Author)
  • Thazin New Aung - , Yale University (Author)
  • David L. Rimm - , Yale University (Author)
  • Heike Niessner - , University of Tübingen (Author)
  • Ulrike Keim - , University of Tübingen (Author)
  • Lina Maria Serna-Higuita - , University of Tübingen (Author)
  • Irina Bonzheim - , University of Tübingen (Author)
  • Luis Kuhn Cuellar - , University of Tübingen (Author)
  • Dana Westphal - , Department of Dermatology, National Center for Tumor Diseases Dresden (Author)
  • Julian Steininger - , Department of Dermatology, Skin Tumor Center (Author)
  • Friedegund Meier - , Department of Dermatology, Skin Tumor Center (Author)
  • Oltin Tiberiu Pop - , Cantonal Hospital St. Gallen (Author)
  • Stephan Forchhammer - , University of Tübingen (Author)
  • Lukas Flatz - , University of Tübingen, Cantonal Hospital St. Gallen (Author)
  • Thomas Eigentler - , Charité – Universitätsmedizin Berlin (Author)
  • Claus Garbe - , University of Tübingen (Author)
  • Martin Röcken - , University of Tübingen (Author)
  • Teresa Amaral - , University of Tübingen (Author)
  • Tobias Sinnberg - , University of Tübingen, Charité – Universitätsmedizin Berlin (Author)

Abstract

Background: Recent advances in digital pathology have enabled accurate and standardised enumeration of tumour-infiltrating lymphocytes (TILs). Here, we aim to evaluate TILs as a percentage electronic TIL score (eTILs) and investigate its prognostic and predictive relevance in cutaneous melanoma. Methods: We included stage I to IV cutaneous melanoma patients and used hematoxylin-eosin-stained slides for TIL analysis. We assessed eTILs as a continuous and categorical variable using the published cut-off of 16.6% and applied Cox regression models to evaluate associations of eTILs with relapse-free, distant metastasis-free, and overall survival. We compared eTILs of the primaries with matched metastasis. Moreover, we assessed the predictive relevance of eTILs in therapy-naïve metastases according to the first-line therapy. Findings: We analysed 321 primary cutaneous melanomas and 191 metastatic samples. In simple Cox regression, tumour thickness (p < 0.0001), presence of ulceration (p = 0.0001) and eTILs ≤16.6% (p = 0.0012) were found to be significant unfavourable prognostic factors for RFS. In multiple Cox regression, eTILs ≤16.6% (p = 0.0161) remained significant and downgraded the current staging. Lower eTILs in the primary tissue was associated with unfavourable relapse-free (p = 0.0014) and distant metastasis-free survival (p = 0.0056). In multiple Cox regression adjusted for tumour thickness and ulceration, eTILs as continuous remained significant (p = 0.019). When comparing TILs in primary tissue and corresponding metastasis of the same patient, eTILs in metastases was lower than in primary melanomas (p < 0.0001). In therapy-naïve metastases, an eTILs >12.2% was associated with longer progression-free survival (p = 0.037) and melanoma-specific survival (p = 0.0038) in patients treated with anti-PD-1-based immunotherapy. In multiple Cox regression, lactate dehydrogenase (p < 0.0001) and eTILs ≤12.2% (p = 0.0130) were significantly associated with unfavourable melanoma-specific survival. Interpretation: Assessment of TILs is prognostic in primary melanoma samples, and the eTILs complements staging. In therapy-naïve metastases, eTILs ≤12.2% is predictive of unfavourable survival outcomes in patients receiving anti-PD-1-based therapy. Funding: See a detailed list of funding bodies in the Acknowledgements section at the end of the manuscript.

Details

Original languageEnglish
Article number104644
JournalEBioMedicine
Volume93
Publication statusPublished - Jul 2023
Peer-reviewedYes

External IDs

PubMed 37295047
ORCID /0000-0003-4340-0402/work/145223803
ORCID /0000-0003-4340-9706/work/145224721

Keywords

Keywords

  • Cutaneous melanoma, Digital pathology, Predictive biomarkers, Prognostic biomarkers, Tumour-infiltrating lymphocytes, Melanoma/pathology, Melanoma, Cutaneous Malignant, Prognosis, Humans, Deep Learning, Neoplasm Recurrence, Local/pathology, Lymphocytes, Tumor-Infiltrating/pathology, Skin Neoplasms/drug therapy