Genomic Profiling Shows Clinical Utility in Select Sarcomas

Margaret A. Hay, MD, The Ohio State University

Margaret A. Hay, MD

Comprehensive genomic profiling can be successfully performed in the majority of patients with sarcoma, depending on the tumor location and tissue subtype, which can impact treatment decisions and patient outcomes, according to results of a study published in the Journal of Clinical Oncology Precision Oncology.

In the trial, comprehensive genomic profiling was successfully performed in 75.3% (n = 295) of patients with sarcoma. Bone sarcomas had a lower evaluation rate at 65.3% (n = 32) versus soft tissue sarcomas at 76.7% (n = 263; P = .0008). The location of the tumor biopsy was associated with improved profiling efficiency. Moreover, bone biopsy samples had a 52.8% (n = 19) evaluation rate versus 61.1% (n = 33) with lung biopsy samples and 80.1% (n = 109) with abdomen biopsy samples.

Additionally, comprehensive genomic profiling impacted treatment selection in 25% of evaluable patients (n = 7), which translated to an improvement in progression-free survival (P = .03).

In order to evaluate the success of comprehensive genomic profiling and its potential impact on treatment decisions, investigators compiled data from 2 trials. In the first trial, investigators collected information on patients with sarcoma who had previously undergone comprehensive genomic profiling using FoundationOne-Heme at The Ohio State University. In the second study, patients at The Ohio State University were asked to undergo comprehensive genomic profiling. To be eligible for enrollment in the second study, patients had to have been within 10 weeks of starting their current therapy and have an expected survival >3 months.

A total of 413 tumor samples were collected from 392 patients with sarcoma. Fifty-three percent of patients were female, and the median age was 60 years (range, 16-89). Complete DNA and RNA sequencing results were reported for 76% (n = 313) of tumor samples. More than 1300 alterations were identified, the most common of which included TP53 and RB1 alterations and MDM2 and CDK4 amplifications.

Fifty-six (13.6%) of the 413 samples had low tumor purity, failed RNA extraction, or suboptimal sequencing metrics. Investigators attributed these disruptions to RNA failure (32.1%; n = 18), failed RNA metrics (25%; n = 14), low tumor purity (14.3%; n = 8), noise copy alteration data (12.5%; n = 7), and contamination (8.9%; n = 5).

Forty-four (10.7%) samples failed testing completely as a result of DNA failures (9.1%; n = 4), RNA failures (27.3%; n = 12), and combined DNA/RNA failure (20.5%; n = 9).

Regarding the type of sarcoma, leiomyosarcomas (74.0%; n = 57) and liposarcomas (79.4%; n = 50) had higher passing rates with comprehensive genomic profiling versus bone-based sarcomas. Approximately half of chondrosarcomas were successfully profiled, though more than 25% failed initial profiling. Chordomas had the lowest initial passing rate at 43%.

The most common biopsy sites were the abdomen (n = 136) and soft tissue (n = 93).

Excisional biopsies, CT-guided, and ultrasound-guided biopsies were the most commonly used biopsy methods. Excisional biopsies had the highest pass rate (80%; n = 200), followed by ultrasound-guided biopsies (69.4%; n = 25), and CT-guided biopsies (62.5%; n = 45).

"Obtaining excisional biopsy specimens rather than imaged-guided biopsy specimens seems to lead to greater likelihood of comprehensive genomic profiling success," Hay and coinvestigators wrote.

In the second study, treating oncologists were blinded to the comprehensive genomic profiling results until patients experienced progression on imaging. If the recommended next line of therapy was impacted after reviewing the comprehensive genomic profiling results, a change was reported.

Of the 34 patients in the second study, 28 were evaluable. Fifty percent of patients were male, and the median age was 62 years (range, 24-80). Among the 25% of patients who received an alternative therapy after their comprehensive genomic profiling results were evaluated, 6 received the recommended treatment. An exploratory analysis showed that the median PFS in these patients was 124 days in the comprehensive genomic profiling—selection group versus 54 days in the non-comprehensive genomic profiling–selection group (P = .03).

"Sarcoma comprehensive genomic profiling appears to alter physician decision-making regarding treatment and may affect the ultimate outcome of the patient," Hay and coinvestigators concluded. "Careful sample acquisition with attention to nucleic acid handling will be key to ensuring usable results."

Hay MA, Severson EA, Miller VA, et al. Identifying opportunities and challenges for patients with sarcoma as a result of comprehensive genomic profiling of sarcoma specimens. JCO Precis Oncol. 2020;4:176-182. doi: 10.1200/PO.19.00227

"Our results support implementing standard biopsy collection protocols for bone-based sarcoma," Margaret A. Hay, MD, of The Ohio State University, and coinvestigators wrote.