Radiosensitivity, Immune Activation, and Mutational Frequency of Melanoma Metastases Across Tissue Types

Melanoma is a more radioresistant and immunogenic histology, associated with a higher mutational burden. However, the influence of metastasis tissue location is unknown. We hypothesize exploiting these differences may allow therapy personalization. We identified metastatic melanoma samples across 6...

Full description

Saved in:
Bibliographic Details
Published inInternational journal of radiation oncology, biology, physics Vol. 111; no. 3; p. S42
Main Authors Sim, A.J., Shetty, P.A., Welsh, E., Teer, J.K., Eschrich, S.A., Mule, J.J., Torres-Roca, J.F., Ahmed, K.A., Grass, G.D.
Format Journal Article
LanguageEnglish
Published Elsevier Inc 01.11.2021
Online AccessGet full text

Cover

Loading…
More Information
Summary:Melanoma is a more radioresistant and immunogenic histology, associated with a higher mutational burden. However, the influence of metastasis tissue location is unknown. We hypothesize exploiting these differences may allow therapy personalization. We identified metastatic melanoma samples across 6 tissues (brain, liver, lung, lymph node [LN], skin, soft tissue [ST]), which were profiled for gene expression by microarray chips. Intrinsic radiosensitivity was estimated by the radiosensitivity index (RSI), a rank-based gene expression signature, where a higher RSI (scale 0-1) indicates greater radioresistance, with a cut point of 0.375 in other histologies. Degree of inflammation in the microenvironment was calculated by an existing 12-gene chemokine signature (12CK) where a higher 12CK score indicates greater inflammation. A subset (n = 94) underwent targeted exome sequencing (1,327 cancer-associated genes) and somatic mutations (MF) were enriched after filtering known germline/silent mutations and artifacts. Differences by tissue were analyzed with the Kruskal-Wallis H test. Correlations between RSI, 12CK, and MF were tested using Spearman's rho. From 1995 to 2011, 337 metastatic samples were identified from unique patients. Median age at diagnosis was 61 (range 20-97). With a median follow-up of 162 months (95% CI 142.6-181.4), median survival was 59.8 months (95% CI 49.2-70.4). Median values for RSI, 12CK, and MB were 0.451 (interquartile range [IQR] 0.331-0.552), 8.700 (IQR 7.218-9.746), and 101.5 (IQR 73.75-137), respectively. Significant differences existed among tissues for both RSI (P = 0.003) and 12CK (P < 0.001), but not MF (P = 0.498) (Table 1). No pair-wise comparisons were significant after Bonferroni correction for RSI, but for 12CK, Brain-ST (P = 0.044), Brain-Lung (P = 0.013), Brain-LN (P = 0.000), and Liver-LN (P = 0.023) were significant. Correlation between RSI and 12CK was significant (rho = -0.661, P < 0.001), but correlations between MB and RSI/12CK were not significant. In this novel composite genomic analysis of melanoma metastases, all were relatively radioresistant, but higher RSI values for liver and skin metastases may warrant radiotherapy dose escalation. Similarly, the lower inflammatory scores of liver and brain metastases may have implications for selection of patients for immunotherapy. More inflamed tumors may be more responsive to radiation. In contrast to current homogenous approaches to metastases, these results indicate personalized approaches to types/sequencing of therapy for melanoma metastases by tissue type may be warranted.
ISSN:0360-3016
1879-355X
DOI:10.1016/j.ijrobp.2021.07.118