Expressed mutated genes in Sezary syndrome and their potential prognostic value in patients treated with extracorporeal photopheresis

Abstract

Background: Sézary syndrome (SS) is an aggressive and leukemic variant of Cutaneous T-cell Lymphoma (CTCL) with an incidence of 1 case per million people per year. It is characterized by a complex and heterogeneous profile of genetic alteration ns that has so far precluded the development of a specific and definitive therapeutic intervention.

Methods: Deep-RNA-sequencing (RNA-seq) data were used to analyze the single nucleotide variants (SNVs) carried by 128 putative CTCL-driver genes, previously identified as mutated in genomic studies, in longitudinal SS samples collected from 17 patients subjected to extracorporeal photopheresis (ECP) with Interferon-α. Results obtained were integrated with Whole Exome Sequencing (WES) data. SNVs were validated using the Sanger method. Pathway analysis was performed with g:Profiler web server (https://biit.cs.ut.ee/gprofiler/gost). Statistical analyses were performed with GraphPad PRISM 8 software.

Results: Nonsynonymous SNVs were identified in 56 genes. Integration of RNA-seq with WES data revealed that about half of these genes contained somatic mutations. Among them, the most frequently transcribed mutated genes were TET2, JAK3, NCOR1, PDCD11, RHOA, and TP53. Nearly all the remaining genes had germline-restricted mutations, and included ARID1A, ATM, ATR, CREBBP, POLD1, and POT1 genes, which are involved in DNA repair, homologous recombination, and chromatin remodeling, and the CROCC gene, implicated in centrosome cohesion. Monitoring of the mutated genes, identified within an enlarged panel of CTCL associated genes, revealed their reduction in almost 70% of SS patients as well as a significant decline of total number of mutations (SNVs) during ECP treatment. Several mutated genes persisted post-therapy, representing novel candidates associated with ECP resistance that could also have a potential prognostic relevance. Notably, these genes mainly converge on pathways related to DNA repair (ATR, ATRIP, POLD1, TP53, TP53BP1/2) which might represent novel targets to be explored in combination with ECP.

Conclusions: This is the first evaluation in SS of expressed mutations in a large panel of CTCL-driver genes. Also innovative is the monitoring of mutated genes in patients' malignant lymphocytes during ECP, a first-line treatment of CTCL, which highlights novel candidates associated with ECP resistance that might unmask novel pharmacological vulnerabilities to be exploited during ECP for a personalized treatment.

Keywords: RNA-seq; Sezary syndrome; candidates associated with therapy resistance and personalized treatment; cutaneous T-cell lymphoma; extracorporeal photopheresis; whole exome sequencing.

Figure 1

Comparison between the mutation frequencies of 56 CTCL driver genes resulting from our RNA-seq and WES data and those reported in the literature. Genes common in at least two authors are highlighted in bold. Genes without mutations detected by our WES are highlighted in italics. RNA-seq samples: n=35; WES somatic samples: n=35; WES germline samples: n=17. The reported frequencies for CTCL literature refer to the Sezary samples of each study. >5; >10: Frequency reported in more than 5% or 10% of samples investigated. *: Extension cohort (N=68 patients); NR: Frequency not reported; -: chromosome region frequently lost in SS; +: chromosome region frequently gained in SS.

Figure 2

Validation of CROCC SNVs identified by RNA-Seq. (A) Chromosome localization (1p36.13) and gene structure of CROCC showing SNVs detected by RNA-seq (modified from NCBI). (B) Chromatograms showing the sequencing of the nucleotides surrounding the highlighted SNVs (relative peak indicated by the arrow) in CROCC gene. Sequencing of the tumor sample (top, frequencies of the SNVs according to RNA-seq in brackets) and of the matched normal cells represented by granulocytes or CD4^- T-cells (bottom), whose purity was confirmed by cytofluorimetric analysis (% CD4⁺ <4%). G, germline.

Figure 3

Mutation tracking in patient’s malignant lymphocytes during ECP treatment. (A) Histograms showing the number of mutated genes in T1 and T2 samples calculated on the total number of patients studied (n=16). (B) Dot plot showing mutation number (SNVs) at T1 and T2. *p<0.05 according to Paired Student’s t-test. (C) Histogram plots related to naïve and pretreated patients showing the difference in the number of genes affected by SNVs between T1 and T2. Patient number is on the x axis. D_{SNVS =} difference in the number of mutated genes between T1 and T2 samples. In the boxes, dot plots showing SNV number at T1 and T2 detected in naïve and pretreated patients, respectively. *p<0.05 according to Paired Student’s t-test.

Figure 4

Persistent and common genes emerging after ECP treatment have a prognostic relevance for SS patients. (A) Histogram plot showing the number of genes affected by SNVs at T1 and T2. Hatched bars represent persistent mutated genes, i.e. genes that present SNVs at both T1 and T2. (B) Dot plot showing the positive Pearson’s correlation between tumor burden at T1, expressed as percentage of clonal CD4+TCRVβ+ T-cells calculated within total CD4+ T-cells, and percentage of persistent mutated genes calculated on total mutated genes at T2 for each individual. (C) Kaplan-Meier analysis comparing OS between patients bearing more than 50% of persistent mutated genes at T2 (red, n=7) and patients for whom the percentage was less than or equal to 50% (green, n=9). Significance was calculated by the log-rank test. Patients with >50% of mutations showed an increased risk of death (HR: 3.57. 95% CI: 1.043-12.26). (D) Kaplan-Meier analysis comparing survival between patients with <20.25% of 15-commonly mutated genes at T2 (1^st quartile) and all other patients with ≥20.25% of mutations at T2 (2^nd-4^th quartile). Significance was calculated by the log-rank test. Patients with ≥20.25% of mutations showed an increased risk of death (HR: 3.4. 95% CI: 1.13-10.45).

Figure 5

Pathways most affected by mutated genes found in naïve and pretreated patients during the disease course. Figure showing the top 10 pathways for each gene list queried: genes mutated mostly at T1, at both T1 and T2 (persistent) and mostly at T2. Pathways in bold are those significantly over-represented. Padj = False Discovery Rate (FDR).