Researchers have identified two genetic mutations-ARID1A and PIK3CA-as potential targets for the development of new treatments for recurrent or metastatic cervical cancer. These findings were recently presented by study author Anjali Hari, MD, an obstetrician/gynecologist at the University of California, Irvine, during the Society of Gynecologic Oncology's 54th Annual Meeting on Women's Cancer.
"Cervical cancer continues to be a worldwide epidemic with over 600,000 new cases worldwide and over 300,000 deaths, with over 90 percent of these cases happening in low- and middle-income countries," Hari noted during her presentation. "Even in the United States, where there's more access to screening, there continues to be 14,000 new cases a year and 4,000 deaths each year.
"The Cancer Genome Atlas project, or the TCGA project, has provided important molecular characterization for cervical cancer," she said. "However, the TCGA is comprised predominantly of CIN3 and early-stage cervical cancer samples. And so, this shows us that there is a need to study tumor samples for more advanced disease."
GOG-0240 (NCT00803062) led to the approval of bevacizumab in more than 60 countries, according to Hari, who noted that, at the time of patient enrollment, tumor tissue was collected and banked for future analyses. "We were awarded an NIH Beau Biden Cancer Moonshot Grant to identify potential druggable targets among patients with metastatic (FIGO Stage IVB) and recurrent cervical carcinoma."
"The objectives of our study include the following: compare mutational frequencies observed in advanced specimens to those reported in the TCGA cervix, identify potential molecular targets for biologic therapy, characterize exceptional responders and poor responders to systemic chemotherapy with and without bevacizumab, and analyze mutational clusters and pattern recognition via bioinformatics and track that with overall survival," outlined Hari, who shared data on the first two objectives during her SGO presentation.
Study Specifics
In this new analysis, the investigators co-extracted DNA/RNA from FFPE samples after a central pathology review at the NRG Oncology Biospecimen Bank at Nationwide Children's Hospital in Columbus, Ohio.
"DNA/RNA analytes were shipped to the New York Genomic Center for whole genome and whole exome sequencing and to the University of North Carolina for RNA and microRNA sequencing. Individual matched germline sequences were used to identify somatic mutations," Hari and colleagues stated. "Mutational frequencies of the most frequently mutated genes were compared with the TCGA to identify potential molecular targets. Exploratory statistical analyses were carried out with clinical data."
Of the 452 patients who provided tumors samples, 112 had sufficient DNA and RNA for mutational analysis. Hari and colleagues identified more than 35,917 mutations. Over 90 percent of the mutations identified from DNA were present in RNA sequences when the expression level was sufficient.
"Similar to early-stage cases from the TCGA, PIK3CA was mutated in 25 percent (28/112) of advanced/recurrent GOG-0240 specimens," Hari reported. "However, a significantly higher frequency of ARID1A mutants were observed in GOG-240 samples (17%, or 19 of 112) compared with TCGA (5%)."
Among patients with PIK3CA, data showed a median overall survival of 15.4 months (HR: 1.0; 95% CI: 0.61-1.62) compared with 16.2 months in wild-type patients. Median progression-free survival (PFS) in patients with PIK3CA mutations was 7.5 months (HR: 0.85; 95% CI: 0.54-1.34) versus 6.1 months in wild-type.
Median overall survival in patients with ARID1A was 14.3 months (HR: 1.18; 95% CI: 0.67-2.06) compared with 17.1 months in wild-type disease. Hari and colleagues observed a median PFS of 5.3 months (HR: 1.0; 95% CI: 0.59-1.68) among patients with ARID1A versus 6.9 months for their wild-type counterparts.
"The PIK3CA mTOR pathway promotes metabolism, proliferation, cell survival, and angiogenesis. PIK3CA has been shown to be dysregulated in a wide span of human cancers," explained Hari, who noted that there are a number of clinical trials in other cancers, including endometrial, ovarian, HPV, and oropharyngeal cancer, studying PIK3CA inhibitors. However, this has yet to be explored in cervical cancer.
"ARID1A is a tumor suppressor that also plays a role in the mTOR pathway. And in other cancers, such as gastric and hepatocellular carcinoma, it has been shown that the loss of ARID1A can make tumors more susceptible to mTOR inhibition," she added.
Concluding her presentation, Hari noted, "The frequency of PIK3CA mutations among our recurrent advanced samples was similar to that seen among the early-stage samples represented in the TCGA. ARID1A mutations were more frequently seen in our advanced recurrent samples compared to those in the TCGA. Both ARID1A and PIK3CA are potential targets and could be considered for drug development through clinical trials in cervical cancer."
Catlin Nalley is a contributing writer.