NEW YORK – Studies presented at the American Society of Clinical Oncology's annual meeting this week suggest that investigators are getting better at identifying therapeutic opportunities for children with cancer based on molecular profiling.
Even though a minority of patients ultimately end up getting these drugs, in at least one of these studies, some patients had "extraordinary responses" to their molecularly matched treatments.
At the meeting, Alanna Church, associate director of the Laboratory for Molecular Pediatric Pathology at Boston Children's Hospital, presented interim data from the GAIN/iCat2 Consortium study on the clinical impact of molecular tumor profiling in pediatric, adolescent, and young adult patients with extra-cranial solid malignancies.
The study, which began in 2015, is enrolling children and young adult patients up to 30 years of age with high-risk, relapsed or refractory solid tumors. At the meeting, the researchers reported on the first 389 participants. The largest tumor cohort comprised osteosarcoma patients, representing 19 percent of the study participants. Altogether, sarcoma patients — including those with osteosarcoma, rhabdomyosarcoma, Ewing's sarcoma, and other sarcomas — represented 65 percent of all enrolled participants.
Targeted NGS was performed on one or more tumor samples from each patient, and selected patients also had RNA sequencing performed on their tumors. The researchers used established guidelines from ASCO, the Association for Molecular Pathology, and the College of American Pathologists to review the significance of the genomic alterations detected in each patient in terms of their impact on cancer diagnosis and prognosis, and used evidence tiers developed within iCat to evaluate the significance of genomic alterations for predicting response to targeted therapies.
Molecular tumor profiling was successful in 345 patients, or 89 percent of the participants. Around two-thirds of these patients had sarcoma, Church said, and 299 patients with molecular profiling results had one or more alterations of clinical significance. Genomic alterations with diagnostic, prognostic, or therapeutic significance were present in 208 patients, 51 patients, and 240 patients, respectively.
Of the 240 patients with tumors harboring genomic alterations with a therapeutic impact, 23 had molecular findings supported by iCat Tier 1 evidence, the category with the highest threshold of evidence. Of the 240 patients, 205 were eligible to receive targeted therapy based on having a molecular alteration of therapeutic significance and sufficient follow-up. The treatments that were recommended spanned a wide range of therapeutic classes, Church said. The most common treatment recommendations were for PARP, WEE1, and BET inhibitors, while the most commonly recommended therapies supported by high-tier evidence were mTOR, AKT, ERK, EZH2, PI3K, PD-1, PD-L1, FGFR, and MEK inhibitors.
Ultimately, however, only 31 patients, out of 205 patients eligible to receive a therapy based on the test results and out of 345 patients tested in total, got a drug in the study. Seven patients responded to the therapies they received, and six had "extraordinary responses," Church said, adding that all such extensive responders had a therapeutically actionable gene fusion.
One of these "extraordinary responders" was a 14-year-old boy with a diffuse inter-abdominal tumor, she said. He was diagnosed with a progressive grade 3 neuroendocrine carcinoma, and sequencing revealed a KHDRBS2-BRAF fusion. Investigators matched the patient to the MEK inhibitor trametinib (Novartis' Mekinist), and he had a partial response to the treatment, which he received for five months.
The researchers also considered the diagnostic value of each molecular alteration they identified in patients. They found diagnostically significant alterations in 208 patients, of which 78 percent were fusions. Seventeen patients enrolled in the study experienced a real-time impact on their diagnosis based on a molecular alteration identified by sequencing, she said. Many of these patients had been evaluated using more traditional techniques, including eight tumors evaluated with FISH assays. Of the 17 patients with a real-time change in diagnosis, 13 were attributable to fusions.
As an example of the diagnostic value of molecular profiling, Church presented the case of another 14-year-old boy who had lower leg pain. After an excisional biopsy, pathologists diagnosed him with an aneurysmal bone cyst, which is a benign tumor. The mass recurred after a year, and a second biopsy revealed concerning features. However, pathologists disagreed on the diagnosis — one said it was a recurrence of the previously seen benign tumor, while another diagnosed a more aggressive, telangiectatic osteosarcoma.
A targeted NGS panel was negative for EWSR1 fusions, as was FISH testing for the USP6 fusion, which didn't help doctors reach a definitive diagnosis. But additional sequencing then identified a completely novel TP53-USP6 fusion, connecting exon 1 of TP53 to exon 1 of USP6.
"Taken together, these data suggest that this fusion simultaneously inactivates TP53 and activates USP6, with both alterations potentially contributing to the oncogenesis of this unusual hybrid tumor," Church said. "A third opinion [from the pathology department] incorporating the TP53-USP6 fusion rendered a diagnosis of telangiectatic osteosarcoma, and as a result, the patient received chemotherapy." At the time of data cutoff in the study, this boy had no evidence of disease, two years after diagnosis.
Even though the majority of molecularly profiled pediatric patients in the GAIN/iCat2 Consortium study did not receive matched therapies, Church highlighted the therapeutic and diagnostic value of the approach in patients who did. "In conclusion, molecular tumor profiling has a significant impact on diagnosis and treatment recommendations for young patients with solid tumors," she said. "These results emphasize the importance of fusion detection for patients with sarcomas and rare tumors."
In a second presentation, Williams Parsons, the deputy director of cancer and hematology centers at Texas Children's Hospital, said the latest analysis from the National Cancer Institute-Children's Oncology Group's Pediatric Molecular Analysis for Therapy Choice, or Pediatric-MATCH, study showed a 28 percent treatment match rate, which may be a reflection of the increased number of targetable events in recurrent or refractory pediatric cancers. In May 2019, 24 percent of pediatric patients in the study were eligible to receive targeted treatment.
The Pediatric-MATCH trial assigns patients with relapsed or refractory solid tumors, lymphomas, and histiocytic disorders to different arms, within which they receive targeted treatments based on the genetic alterations detected in their tumors. Around 1,000 patients had their tumors profiled using DNA and RNA-based gene sequencing panels.
Among these patients, who were enrolled between July 2017 and October 2020 and had testing completed, 310 patients had one actionable mutation and treatment protocol slots were available for 284 patients, or 28 percent. This exceeds the pre-study expectation that 10 percent of patients would be recommended to a therapeutic arm.
Ultimately, 131 patients — 46 percent of the 284 patients who were eligible for treatment or 13 percent of the 1,000 patients who were genomically tested for the trial — ended up getting a therapy in the Pediatric-MATCH trial.
The most common reasons patients provided for not enrolling on a treatment protocol were that the patient was receiving another treatment, had poor clinical status, had a lack of measurable disease, or had an ineligible diagnosis for that treatment arm.
Parsons noted that patients with central nervous system tumors were more likely to have a targetable alteration and get on a treatment arm compared to those with other kinds of tumors. Among patients with central nervous system tumors, 71 percent had an actionable mutation, 67 percent matched to a protocol, and 30 percent were enrolled in one of the treatment protocols, he said.
He also noted that the match rate was high for atypical teratoid rhabdoid tumors — rare, fast-growing tumors of the brain and spinal cord — which are known to have SMARCB1 mutations or SMARCA4 mutations, though the latter are less common. About 80 percent of the AT/RT patients in the study were found to have an actionable mutation.
"The numbers for the rest [of the patients] are variable, ranging down to ependymoma, where, as we'd expect based on our understanding of the genomics of that tumor, only a small fraction had an actionable mutation detected," Parsons said.
While Parsons discussed the ability of Pediatric-MATCH investigators to identify potential therapeutic opportunities for patients based on the molecular markers identified in their tumors, he did not report patients' therapeutic outcomes during his presentation.
He did emphasize that a prior history of molecular testing was associated with higher match and reiterated that poor clinical status was a common reason for a patient not enrolling in a treatment protocol. This suggested that early molecular screening could facilitate enrollment to biomarker-selected trials of targeted therapies, Parsons said.