Metabolomic signatures in liquid biopsy are associated with overall survival in metastatic melanoma patients treated with immune checkpoint inhibitor therapy

Abstract

Background: Immune checkpoint inhibitors (ICIs), such as anti-Cytotoxic T-Lymphocyte Antigen 4(CTLA-4) and anti-Programmed cell death protein 1 (PD-1) agents, have improved the prognosis of patients with metastatic melanoma. However, a proportion of patients develop resistance to these treatments, leading to a poor prognosis. Therefore, identifying potential non invasive and easy to measure biomarkers is crucial for guiding treatment strategies in patients with metastatic melanoma.

Methods: A retrospective single-center study was conducted involving patients with metastatic stage IV melanoma who received first-line treatment with anti-CTL4 and/or anti-PD-1 agents. The patients were categorized into two groups on the basis of their 1-year overall survival (OS): those with good outcomes (long-term OS ≥ 1 year) and those with poor outcomes (short-term OS < 1 year). Peripheral metabolomics was performed using baseline sera from 132 patients via 600 MHz Nuclear Magnetic Resonance (NMR) spectroscopy. Enriched functional analysis was conducted to identify the metabolic pathways in which significant metabolites were involved.

Results: Sparse partial least squares discriminant analysis (sPLS-DA) and loading plots obtained by analyzing the metabolomics profiles of samples collected before ICI treatment revealed significantly different levels of metabolites between the two groups (long-term OS vs. short-term OS). Specifically, lactate, tryptophan and valine significantly predicted the OS of the whole study population subjected to ICI immunotherapy; alanine, asparagine, glutathione, histidine, isoleucine and phenylalanine significantly predicted the OS of patients treated with ipilimumab; glucose, glutamine, histidine and proline significantly predicted the OS of patients treated with nivolumab; and lactate, lysine and proline significantly predicted the OS of patients treated with ipilimumab plus nivolumab. Notably, tryptophan levels were correlated with treatment response in the overall patient group, whereas histidine and lactate levels were associated with response in patients treated with ipilimumab and with ipilimumab plus nivolumab, respectively. Interestingly, higher pretreatment levels of histidine were commonly found in long-term OS subgroups of patients treated with ipilimumab, nivolumab or ipilimumab plus nivolumab. Interestingly, considering only those metabolites that predict OS after univariate analysis, higher histidine, and lower lactate and proline levels resulted as associated with favorable OS in at least two patient cohorts.

Conclusions: Overall, this exploratory liquid biopsy study revealed a strong correlation between the pretreatment levels of some metabolites and the OS of patients with metastatic stage IV melanoma treated with anti-CTL4 and/or anti-PD-1 antibodies in the first-line setting and revealed the potential of these molecules to predict outcomes and define personalized management and treatment strategies.

Keywords: Immune checkpoint inhibitors; Melanoma; Metabolomics; Serum biomarkers.

Fig. 1

Score plot (A) and loading plot (B) related to metabolomic profiling of the sera of all metastatic melanoma patients treated with immunotherapy at first-line enrolled in training set, and subdivided accordingly to OS in good (long-term; OS ≥ 1 year) and poor (short-term; OS < 1 year) outcome

Fig. 2

Kaplan–Meier curves of OS accordingly to lactate, tryptophan and valine in the sera from metastatic melanoma patients treated with immunotherapy at first-line enrolled in training set (log-rank p values are reported). * symbols indicate p values < 0.05

Fig. 3

Score plot (A) and loading plot (B) related to metabolomic profiling of the sera of metastatic melanoma patients treated with ipilimumab at first-line enrolled in training set, and subdivided accordingly to OS in good (long-term; OS ≥ 1 year) and poor (short-term; OS < 1 year) outcome

Fig. 4

Kaplan–Meier curves of OS accordingly to asparagine, phenylalanine, glutathione, alanine, histidine and isoleucine in the sera of metastatic melanoma patients treated with ipilimumab at first-line enrolled in training set (log-rank p values are reported). * and *** symbols indicate p values < 0.05 and < 0.0001, respectively

Fig. 5

Score plot (A) and loading plot (B) related to metabolomic profiling of the sera of metastatic melanoma patients treated with nivolumab at first-line enrolled in training set, and subdivided accordingly to OS in good (long-term; OS ≥ 1 year) and poor (short-term; OS < 1 year) outcome

Fig. 6

Kaplan–Meier curves of OS accordingly to LDH, glutamine, glucose, histidine and proline in the sera of metastatic melanoma patients treated with nivolumab at first-line enrolled in training set (log-rank p values are reported). *, ** and *** symbols indicate p values < 0.05, < 0.01 and < 0.0001, respectively

Fig. 7

Score plot (A) and loading plot (B) related to metabolomic profiling of the sera of metastatic melanoma patients treated with ipilimumab plus nivolumab at first-line enrolled in training set, and subdivided accordingly to OS in good (long-term; OS ≥ 1 year) and poor (short-term; OS < 1 year) outcome

Fig. 8

Kaplan–Meier curves of OS accordingly to lactate, lysine and proline in the sera of metastatic melanoma patients treated with ipilimumab plus nivolumab at first-line enrolled in training set (log-rank p values are reported). * and ** symbols indicate p values < 0.05 and < 0.01, respectively