Caris Life Sciences®, a leading biotechnology company focused on fulfilling the promise of precision medicine, today announced the presentation of data from three studies that demonstrate the utility of Caris Molecular Intelligence®, the company's panomic comprehensive tumor profiling service, in facilitating comparisons between tumors that share histological features. In separate presentations at the annual meeting of the American Society of Clinical Oncology (ASCO) in Chicago, Ill., the studies showed how multi-platform tumor profiling via Caris Molecular Intelligence revealed a similar genetic landscape with limited differences in brain metastases in patients with non-small-cell lung cancer (NSCLC), breast cancer and melanoma when compared to the corresponding primary tumors, while also highlighting molecular similarities and differences between specific histological subtypes in endometrial, renal and ovarian cancers.
Caris Molecular Intelligence is a multi-platform tumor profiling service that includes gene sequencing (Next-Generation Sequencing [NGS] and Sanger), protein expression analysis (Immunohistochemistry [IHC]), gene copy number and translocation analysis (Chromogenic or Fluorescence in situ Hybridization [CISH or FISH]). Investigators used these methods to compare the molecular profiles of brain metastases from common tumor types, as well as the profiles of histologically similar cancers, as a means to identify potential treatment strategies that specifically target the biomarkers detected in individual patients' malignancies.
"Multi-platform profiling is a tool that can enhance understanding of tumor biology, not only by elucidating patterns of gene expression in brain metastases and primary tumor sites, but also by helping to identify areas of overlap as well as molecular differences in cancers that are histologically similar," said Joanne Xiu, Ph.D., Research Scientist and Molecular Science Liaison at Caris Life Sciences, and a co-investigator in all three studies. "The data presented at ASCO show that detection of molecular similarities between different tumor types may dictate similar treatment strategies, while subtle differences between tumors with similar histological features may call for different therapeutic approaches."
Brain Metastases Study
In a June 1 poster discussion (abstract #2060), researchers presented data demonstrating the genetic similarity of brain metastases in patients with NSCLC, breast cancer and melanoma, compared to primary tumors. Using Caris Molecular Intelligence, they profiled 5,391 NSCLC (including 293 brain metastases and 5,098 primary lung tumors), 3,595 breast cancer (99 brain metastases, 3,496 primary breast tumors) and 761 melanoma (101 brain metastases, 660 primary skin malignancies) unpaired samples. The investigators found no significant differences in 48 genes between brain metastases and the primary tumor sites, except for PIK3CA in breast cancer, which was mutated less frequently in brain metastases than in the primary breast samples (10% vs. 26%, respectively; p = 0.02). In contrast, expression of topoisomerase 2A (TOPO2), TOPO1 and thymidylate synthase (TS), as well as amplification of epidermal growth factor receptor (EGFR), were more prevalent in brain metastases than in all three primary tumor sites.
"The limited differences in brain metastases are more consistent with a linear progression model of cancer metastasis, by which the metastatic capabilities of tumor cells develop at primary sites following the accumulation of alterations, than with the so-called parallel progression model, which argues that tumor cells disseminate early and accumulate changes independently at the secondary site," noted lead investigator Santosh Kesari, M.D., Ph.D., Professor of Neurosciences at the University of California San Diego (UCSD) School of Medicine, and Director of Neuro-oncology at Moores UCSD Cancer Center. "Our data also suggest that both primary tumors and brain metastases would respond to similar chemotherapeutics, particularly those that effectively penetrate the blood-brain barrier."
Clear-cell Carcinoma Study
Another poster presented at ASCO (abstract #5595) suggests that clear-cell uterine carcinomas (CCUCs) and clear-cell ovarian carcinomas (CCOCs) are molecularly similar to each other but significantly different from clear-cell renal carcinomas (CCRCs), even though all three clear-cell subtypes appear similarly under a microscope. "Clear-cell uterine carcinoma accounts for approximately 5% of endometrial carcinomas and exhibit aggressive clinical behavior with poor outcomes," explained Robert DeBernardo, M.D., Gynecologic Oncologist in the Department of Gynecologic Oncology and Director of Minimally Invasive Surgery in the Obstetrics/Gynecology & Women's Health Institute at the Cleveland Clinic. "Clear-cell ovarian cancers are a subtype of epithelial ovarian cancers that are chemo-resistant, with a poorer prognosis than other subtypes. Seventy percent of renal cell carcinomas are clear-cell and respond to inhibitors of tyrosine kinase and mTOR. We profiled these clear-cell carcinomas to determine if they rely on similar molecular pathways, and to identify subsets of patients that may benefit from different therapies."
In their analysis of 139 CCUCs, 409 CCOCs and 94 CCRCs, Dr. DeBernardo and colleagues reported that CCRCs had fewer mutations in the mammalian target of rapamycin (mTOR) pathway (PIK3CA: 25% in CCUC, 40% in CCOC, 4% in CCRC) and the MAP kinase (MAPK) pathway (KRAS: 14%, 11% and 0%, respectively). They also observed VHL mutations in 47% of CCRCs but not in any of the CCUCs or CCOCs. Estrogen receptor (ER) and progesterone receptor (PR) expression were more common in CCUCs and CCOCs but rare or nonexistent in CCRCs (ER: 35%, 8% and 0%, respectively; PR: 22%, 13% and 2%, respectively), while androgen receptor (AR) expression was more common in CCRCs (7%, 5% and 26%, respectively). All three clear-cell carcinoma types had some immune-positivity for the programmed cell death protein 1 (PD-1; 73%, 47% and 68%, respectively) or its ligand, PD-L1 (13%, 6% and 29%, respectively).
"Our data suggest that blockade of the mTOR and/or MAP kinase pathways may be important in clear-cell uterine and ovarian carcinomas," Dr. DeBernardo commented. "For patients with clear-cell renal carcinoma, anti-angiogenic agents are more likely to be of benefit. The frequency of PD-1 and PD-L1 expression suggests that immunotherapies warrant further investigation in selected patients with clear-cell carcinoma. More studies are needed to correlate these markers with sensitivity to chemotherapy."
Study Comparing Ovarian Carcinosarcoma, Ovarian Serous Carcinoma and Endometrial Carcinosarcoma
In a third ASCO poster presentation (abstract #5560), Dr. DeBernardo and colleagues reported that ovarian carcinosarcomas (OCS), a rare and aggressive malignancy with limited treatment options, share molecular changes similar to that of serous ovarian carcinoma (SOC). They used Caris Molecular Intelligence to evaluate 110 OCS, 141 endometrial carcinosarcoma (ECS, another tumor type with similar histology to OCS) and 1,587 SOC samples to explore the potential overlap in treatment paradigms. TP53 was the most commonly mutated gene in all three malignancies, appearing in 76.4% of OCS, 68.8% of ECS and 69% of SOC. Alteration of the PI3K/AKT/mTOR and MAPK pathways were similar in OCS and SOC but was less frequently altered in ECS; these alterations included mutations in PIK3CA (7.6% and 2.3% vs. 22.2%, p < 0.001), FBXW7 (0% and 0.6% vs. 12.1%, p < 0.001), PTEN (3.7% and 0.8% vs. 12%, p < 0.001) and KRAS (5.2% and 5.0% vs. 13.5%, p < 0.001). For the homologous recombination pathway, SOC and ECS were more likely than OCS to have BRCA1 (20% and 18% vs. 9%) and BRCA2 mutations (18% and 27% vs. 12%). However, the differences were not statistically significant. No difference in alteration of RB, NOTCH, angiogenesis and FGFR pathways was noted among the three cohorts. ER (14.6% and 25.1% vs. 53.1%, p < 0.001) and AR (18.8% and 12.2% vs. 32.4%, p < 0.001) were expressed less frequently in OCS and ECS than SOC, respectively. On the other hand, expression of PR was more frequent in in OCS and SOC than in ECS (26.5% and 30.5% vs. 20.9%, p < 0.001).
"Both SOC and OCS have significantly lower activity of the PI3K/AKT/mTOR and MAP kinase pathways and higher progesterone receptor expression than ECS," Dr. DeBernardo noted. "Treatments that are active against SOC may be considered when treating patients with OCS."
Mapping tumor microenvironments: Uncovering spatial subclones for improved cancer treatment