Meta-analysis of commonly mutated genes in leptomeningeal carcinomatosis

Abstract

Background: Leptomeningeal carcinomatosis (LMC) is a rare type of cancer that settles at the meninges through metastasis of non-small cell lung cancer (NSCLC), breast cancer and melanoma. The molecular mechanism underlying LMC is not known, therefore molecular studies investigating the development of LMC are needed. Here, we aimed to identify commonly mutated genes in LMC caused by NSCLC, breast cancer, and melanoma using an in-slico approach and their interactions using integrated bioinformatic approaches/tools in this meta-analysis.

Methods: We conducted a meta-analysis using information from 16 studies that included different sequencing techniques of patients with LMC caused by three different primary cancers: breast cancer, NSCLC, and melanoma. All studies that assessed mutation information from patients with LMC were searched in PubMed, from their inception to February, 16 2022. Studies that performed NGS on LMC patients with NSCLC, breast cancer, or melanoma were included, while studies that did not apply NGS to CSF samples, did not provide information on altered genes, were reviews, editorials, or conference abstracts, or whose main goal was the detection of malignancies were all excluded. We identified commonly mutated genes in all three types of cancer. Next, we constructed a protein-protein interaction network, then performed pathway enrichment analysis. We searched National Institutes of Health (NIH) and Drug-Gene Interaction Database (DGIdb) to find candidate drugs.

Results: We found that TP53, PTEN, PIK3CA, IL7R, and KMT2D genes were commonly mutated genes in all three types of cancer via our meta-analysis that consisted out of 16 studies. Our pathway enrichment analysis showed that all five genes were primarily associated with regulation of cell communication and signaling, and cell proliferation. Other enriched pathways included regulation of apoptotic processes of leukocytes and fibroblasts, macroautophagy and growth. According to our drug search we found candidate drugs; Everolimus, Bevacizumab and Temozolomide, which interact with these five genes.

Conclusion: In conclusion, a total of 96 mutated genes in LMC were investigated via meta-analysis. Our findings suggested vital roles of TP53, PTEN, PIK3CA, KMT2D, and IL7R, which can provide insight into the molecular basis of LMC development and paving the door to the development of new targeted medicine and will encourage molecular biologists to seek biological evidence.

Keywords: Bioinformatic analysis; Leptomeningeal carcinomatosis; Pathway enrichment analysis; Protein-protein interaction.

Figure 1. PRISMA 2020 flow diagram of the study selection.

Figure 2. Systematic workflow of the methodology.

The meta-analysis was conducted using information obtained from 16 studies involving different sequencing techniques of patients with LMC caused by three different primary cancers (breast cancer, NSCLC, and melanoma). Then, a protein-protein interaction network was created, and pathway-enrichment analysis was performed. Candidate drugs were investigated via the Drug-Gene Interaction Database and interpreted with the genes obtained as a result of the analysis. LMC, leptomeningeal carcinomatosis; NSCLC, non-small cell lung cancer.

Figure 3. Venn diagram showing mutated genes in LMC patients according to their primary tumor sites.

When all the genes reported were plotted in a Venn diagram, five genes appeared to be in common. These five genes shown in the red box were TP53, PTEN, PIK3CA, KMT2D and IL7R were mutated commonly in all three primary tumor sites (NSCLC, breast cancer, melanoma). LMC, leptomeningeal carcinomatosis; NSCLC, non-small cell lung cancer.

Figure 4. Mutated genes and their interactions with each other in each primary cancer type.

Bar graph showing mutated genes in (A) NSCLC-LMC, (B) Breast-LMC and (C) Melanoma-LMC patients (n ≥ 2). (D) PPI network of mutated genes in NSCLC-LMC. Among 87 proteins 12 of them did not interact with any other proteins. (E) PPI network of mutated genes in breast-LMC. Among 11 proteins only one of them (ALDH2) did not make any interaction, while CLTCL1 and EPHB6 only interacted with each other. (F) PPI network of mutated genes in melanoma-LMC. NRAS and BRAF interacted with each other. PPI, proteinprotein interaction; LMC, leptomeningeal carcinomatosis; NSCLC, non-small cell lung cancer; n, patient number.

Figure 5. PPI network and pathway enrichment analysis of commonly mutated genes in all three cancer groups.

(A) Bar-chart showing mutated genes in at least two primary tumor sites (n ≥ 2). Five genes (TP53, PTEN, PIK3CA, IL7R, and KMT2D) were mutated in all three tumor sites. (B) PPI network of mutated genes appeared in all three cancer groups. All four proteins were interacted with each other, with the exception of IL7R. (C) Visualization of pathway enrichment analysis of mutated genes among all three cancer groups using Cytoscape. All five genes were involved in signaling and cell population proliferation pathways. TP53, PTEN, PIK3CA, and KMT2D had roles in pathways regarding growth. TP53, PTEN, and PIK3CA were involved in the AKT signaling pathway. TP53, PTEN, and IL7R were involved in lymphocyte and leukocyte apoptotic processes. Finally, TP53 and PIK3CA had roles in fibroblast apoptotic processes and regulation of macroautophagy. n, patient number; NSCLC, non-small cell lung cancer.

Figure 6. Food and Drug Administration (FDA) approved drugs.

The number of FDA approved drugs used for treatments of breast cancer, NSCLC, and melanoma are shown in green. The number of drugs that are common in two cancer types are shown in pink and common in three cancer types shown in purple. Drugs able to cross BBB are shown in yellow. Five drugs that can cross BBB are shared between breast cancer and NSCLC patients, while three are common between NSCLC and melanoma patients. There are no common drugs between three cancer groups that can penetrate BBB. BBB, blood-brain barrier; NSCLC, non-small cell lung cancer.