As part of the National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) trial conducted at more 1000 study sites, researchers have identified actionable alterations in nearly 38% of patients with cancer, and almost 18% of patients were assigned to treatments based on their molecular profiling results.1 The patient population included a large proportion of patients with less common cancers, including gastroesophageal cancer, sarcoma, and kidney cancer, in addition to patients with the most common cancers — colorectal, breast, and prostate cancers, as well as non-small-cell lung carcinoma (NSCLC).
The findings may pave the way to increasing enrollment in cancer clinical trials, which has become even more challenging during the coronavirus disease 2019 (COVID-19) pandemic.2 However, the new paper did not include patient outcome data, even though some of this information has been published in separate papers. The investigators intend to publish the data for the whole trial population in the future.
“The most exciting aspect of this analysis is the large proportion of patients enrolled with ‘rare’ cancer types and the new genetic landscape data that we present here,” lead study author Keith T. Flaherty, MD, director of clinical research at the Mass General Cancer Center in Boston, Massachusetts, wrote in an email.
The study included 6391 adult patients registered across 1117 sites. These were patients with any solid tumor, lymphoma, or myeloma that had progressed when a patient received standard treatment, or patients who had not received prior therapy for the tumor because no curative treatment for their condition was available.
The researchers required that patients have a tumor biopsy specimen collected. They then performed next-generation sequencing (NGS) and immunohistochemistry assays. The investigators performed the NGS assay on the Ion Torrent PGM or S5 (Thermo Fisher Scientific) and sequenced 143 genes and more than 4000 annotated variants. Immunohistochemistry was performed for PTEN expression, nuclear expression of MLH1 and MSH2, and Rb expression if matches to a cyclin inhibitor were found.
Of the 6391 enrolled patients, 5954 had tumor samples available, and 5540 of those samples were sequenced successfully. Of all these sequenced samples, 2079 came from patients with NSCLC or breast, colorectal, or prostate cancers, whereas the majority — 3461 — came from patients with less common cancer types.
Study participants had undergone a median of 3 prior treatments, whereas less than 25% of all patients had received 1 treatment or no prior therapy.
The investigators detected actionable alterations in 37.6% of patients. Moreover, 17.8% were assigned to receive specific treatments based on their molecular results. The authors noted that 26.4% could have been assigned to treatments if all the protocols had been reached.
In the population of patients that were assigned to treatments, the assignment rate was 17.4% for patients with NSCLC, 13.7% for those with colorectal cancer, 17.8% for those with breast cancer, and 23% for those with prostate cancer. The researchers reported assignment rates greater than 25% in patients with CNS cancer, urothelial cancer, cholangiocarcinoma, pancreaticobiliary cancer, cervical cancer, gastroesophageal cancer, melanoma, uterine cancer, and anal cancer. The lowest assignment rates, below 6%, were identified in patients with pancreatic cancer, small cell lung cancer, and lymphoma.
Of all actionable alterations, PIK3CA and PTEN were observed most often, at 11.8% and 6.3%, respectively. Other actionable alterations were seen in 3% of patients or less.
However, 37.6% of patients with the most frequently identified actionable alterations were excluded from treatment because they had other mutations that have been shown to confer resistance. One example was those patients with PIK3CA alterations who also had RAS or PTEN resistance-conferring alterations.