Skip to main content
Premium Trial:

Request an Annual Quote

Mt. Sinai Researchers Use Multi-Omics Approach to Personalize Relapsed Multiple Myeloma Treatment

NEW YORK (GenomeWeb) – A group led by Mount Sinai researchers have developed a DNA- and RNA-based sequencing computational method to identify genetic mutations in patients with relapsed multiple myeloma and potentially tailor treatments based on tumor susceptibility to certain drugs.

Many patients with relapsed multiple myeloma are not often matched with correct treatment options in a timely or personalized manner, and most patients have a median survival rate of six years. With standard diagnosis and treatment, relapses are usually inevitable and fatal for most patients.

Researchers have explored the use high-throughput DNA sequencing to match genetic mutations to cancer-killing drugs, allowing treatment for patients with multiple myeloma. Sequencing has revealed wide heterogeneity across patients and complex sub-clonal structures, indicating that using personalized therapeutic approach can help improve patients with myeloma. However, to this point patient RNA profiling has not been added to the mix.

"The drugs we can look at using DNA mutations are a smaller set than we could [potentially] observe for RNA," Mt. Sinai associate professor of oncology Samir Parekh explained. "Multiple myeloma is not as easily segmented into actionable items [for treatment] like other conditions such as lung cancer, forcing us to come up with a new strategy."  

In a study published earlier this month in JCO Precision Oncology, Parekh and his team performed a precision medicine trial of their approach using a group of 64 relapsed multiple myeloma patients. The researchers collected 4 to 10 cubic centimeters of bone aspirate, as well as peripheral blood samples, extracting tumor genomic DNA and RNA from BM CD138+ cells.

"We can separate the tumor from non-tumor fraction of BM CD138+ cells using surface biomarkers," Parekh said. "Based on the patient's disease status, they could have anything from a few hundred thousand to millions of cells in that amount of bone marrow."

The team then performed whole-exome and targeting RNA sequencing on the samples, which created a library of complementary DNA. The pipeline identified 21,166 somatic mutations in 10,403 genes in patients. The mutations were distributed in 14 different categories according to their nature and position in the genome.

Using RNA-seq data, Parekh and his team determined "gene expressions, outliers in gene expression, enriched pathways, drugs that matched the patient's condition, and expression patterns." He emphasized that the point of the study was to identify more than the expected mutations and develop a bigger picture of the genetic causes behind relapsed multiple myeloma.

Using the integrated sequencing workflow, the team was able to produce results in five to six weeks. The team was also able to treat 63 of 64 patients based on the pipeline, as well as identify a subset of patients for treatment based on genetic risk factors.

"Basically, 62 percent of patients were treated based on RNA analysis, which would have been missed if we had done a targeted panel based on DNA alone," Parekh explained.

In addition, the researchers used Washington University's Clinical Interpretations of Variants in Cancer Database to define 28 actionable genes in the population — including mutations in the KRAS, TP53, NRAS, BRAF, SATM and APC genes— that indicated a high level of sensitivity to cancer-killing drugs. Parekh and his team also found a total of 3,541 genes affected by copy number alterations in at least one patient, identifying 31 actionable alterations.

Finally, the team performed RNA-based drug repurposing by matching each patient's RNA profile with gene expression profiles induced by different drugs from the L1000 profiling method, which was introduced by a Broad Institute-led team in November 2017 in Cell.

In the current study, 26 out of 63 patients with recommended treatment generated by the team's pipeline received at least one of the suggested drugs. Of the 26, the team found 21 that were evaluable for response. Of these, 11 received a drug based on RNA profiling, eight got treatment based on DNA profiling, and two were treated based on both RNA and DNA profiling.

Of the 21 patients evaluable for response, one achieved "complete response," three "very good partial response," and 10 "partial response." Two patients had "minimal response," three "stable disease," and two "progressive disease."

Parekh acknowledged multiple real-world challenges in carrying out next-generation sequencing-based recommendations for patients with multiple myeloma. First, he noted that his team could not implement recommendations in 60 percent of patients due to insurance denial for specific drugs. In order to workaround the issue with insurance companies, Parekh's team worked with specialty pharmacies to expedite drug processing and approval.

"As sequencing-based drug repurposing and genetic info becomes more common, we will need to move from a histology approach to a histological [and] genetic-based sequencing process," Parekh explained.

In addition, Parekh noted that the lengthy amount of time needed to produce sequencing data in the study stymied potentially useful genetic applications for patients. During the average turnaround time of six weeks, multiple myeloma patients experienced rapid progression and needed immediate treatment.

In order to improve overall time to results, Parekh's team is now using rapid-run sequencing, which will provide results in three to four days. While the cost for rapid-run sequencing is much more expensive than standard tools, Parekh noted that it might become more cost-effective with greater use in the research space.

While the researchers have not filed a patent for the technology, Parekh said that his team is currently discussing with undisclosed companies "how to develop the tool further and apply the IP in other tumors." He emphasized that the technology could also be used for other applications in precision medicine, including identification and implementation of patient genetic information for different immunotherapies.

"For multiple myeloma, we are using genomic data to predict which mutations may be antigenic, and we can follow these new antigens by looking at T-cell responses," Parekh explained. "[The tool] is applicable for any tumor, and [could] be configured for rare cancers and malignancies."

There is an ongoing debate about using broad sequencing for precision medicine, as some patients whose tumors are analyzed for dozens of mutations don't appear to be surviving longer than patients using routine genomic testing. However, Parekh argued that his pilot study showed that researchers can find rapid actionable changes in relapsed multiple myeloma and "that patients do respond to these in silico treatments for several months or longer."

"The patients treated in this study have had more than 10 lines of relapses, with very low or no standard of care, and couldn't get onto standard clinical trials," Parekh explained. "However, we've analyzed their genome and gave combination therapy based on the results, and it appears to be a success. To go beyond that, we will need larger sample sizes and bigger studies."

In future studies, Parekh's team will explore the impact of clonal heterogeneity on therapy selection by expanding its analysis to include clonality assessment based on WES and single-cell RNA sequencing (scRNA-seq). The study's authors believe that using both tools "will provide a more integrated profile of a patient's tumor, allowing drug repurposing at the sub-clonal level."

"We are treating not just monolithic cases, but different clones of the disease, where each might have different sensitivity to chemotherapy," Parekh said. "We therefore need to actively plan a combination approach that would target each clone."

The researchers are therefore starting a next-generation clinical trial that will involve machine learning algorithms and "interactive learning techniques" into the precision medicine platform to examine the integrated tool's feasibility. The tool will detect other clonal variations as well as RNA expression levels in patients.

However, while scRNA-seq acts a more direct approach than WES to examine cell heterogeneity, the authors noted that the technique presents multiple challenges such as an "increased cost, a high sensitivity to sample purity, drop-outs, [as well] as data processing and interpretation."

While the team may potentially commercialize the WES and scRNAseq tool in the future, Parekh emphasized that the main goal will be to improve treatment of relapsed patients. At the same time, he acknowledged that if the group were "to expand with a big company to commercialize [the sequencing tool], more hospitals and patients would benefit from having access" to it.

"ScRNA-seq represents a promising and powerful technology that will likely complement current precision medicine strategies in the near future," Parekh added.