Metabolic expression patterns from peripheral blood distinguish between low and high risk of metastasizing in uveal melanoma

Abstract

Purpose : Uveal melanoma (UM) is the most common intraocular primary malignancy. Patients with UM are predominantly treated by irradiation. This eye-sparing treatment results in absence of tumor tissue for mutation-analyses. Mutations in BAP1-mutated tumors metastasize regularly within 5 years, SF3B1-mutated tumors pose an intermediate risk and metastasize in 5-15 years and EIF1AX mutations rarely metastasize. Without tumor tissue, the gene mutation related prognosis is unobtainable.
We aimed to elucidate prognostic information non-invasively by deriving serum metabolic expression patterns from patients harboring either a BAP1, SF3B1 or EIF1AX mutated tumor.

Methods : We have performed untargeted liquid chromatography tandem mass spectrometry (LC-MS/MS) using an Orbitrap system on plasma isolated from peripheral blood from 58 UM-patients in our discovery cohort. Additionally, plasma samples from a replication cohort and non-UM controls were used. Blood was drawn prior to treatment between 1998 and 2021 and gene-mutation was used for group stratification. Afterwards, the obtained metabolite expression was analyzed using MetaboAnalyst 5.0 and supervised clustering for metabolic patterns was performed using R.

Results : We have obtained patterns in differentially expressed metabolites between patients harboring an EIF1AX (low metastatic risk) and BAP1 (high metastatic risk) mutated tumors (Figure 1). In the metabolite expression we have identified two clusters, each with two subclusters. These subclusters are correlated with a gradual worsening of survival from patients harboring an EIF1AX to SF3B1 and BAP1-mutated tumor (p < 0.0001) (Figure 2).

Conclusions : Our preliminary data show that metabolomics performed on plasma of peripheral blood can distinguish between low and high risk of metastasizing and these promising results can be combined with other modalities for an integrative noninvasive prognosticator.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

 

Patients harboring an EIF1AX-mutated tumor are depicted in blue, SF3B1 in gray and BAP1 in red. Two clusters were identified, separating high risk (BAP1-mutated) and low risk (EIF1AX-mutated) tumors.

 

Survival plot based on (sub)clusters derived from differential expression of metabolites in plasma of patients harboring an EIF1AX, SF3B1 or BAP1-mutated tumor, showing a worse survival in cluster 2 compared to cluster 1 based on metabolite expression.