Skip to main content
Premium Trial:

Request an Annual Quote

Melanoma Study Hints At Potential Benefits of Single-Cell Diagnostics on Tumor, Microenvironment


NEW YORK (GenomeWeb) – New research suggests features found in just a subset of individual cells from a tumor or its surroundings might offer insights into metastatic melanoma response to targeted therapies and immunotherapies.

A Broad Institute- and Dana-Farber Cancer Institute-led team used RNA sequencing to profile copy number changes and gene expression in thousands of individual tumor, stromal, endothelial, and immune cells isolated from fresh melanoma biopsies for 19 patients.

"The most granular way to profile something is looking on the single-cell level, so that was really the major motivation," Benjamin Izar, a clinician and medical oncology researcher with the Dana-Farber Cancer Institute, told GenomeWeb.

From these single-cell profiles, the researchers detected heterogeneity in the malignant cells and in the tumor microenvironment, along with alterations associated with targeted treatment resistance that became more abundant in post-treatment melanomas that they subsequently tested by bulk RNA sequencing.

"By doing this type of assay, you get everything at the same time, which will help us to learn about clinical sensitivity or resistance, but it will also help us to learn more about the tumor biology," said Izar, who presented the data at the American Society of Clinical Oncology meeting in Chicago earlier this month. The work also appeared in Science in April, as part of a special issue on cancer metastasis.

Based on their findings, the authors of the Science study argued that "[p]utative interactions between stromal-derived factors and immune cell abundance in melanoma core biopsies suggest that future diagnostic and therapeutic strategies should account for tumor cell composition rather than bulk expression."

"Furthermore," they wrote, "our data suggest potential biomarkers for distinguishing exhausted and cytotoxic T cells that may aid in selecting patients for immune checkpoint blockade."

Izar cautioned that the current proof-of-principle study highlights the types of analyses that can be done with individual cells in the tumor and microenvironment, though it remains to be seen whether this information can improve patient outcomes.

"Ultimately, we hope that we can implement something like this either as a predictive marker, as a biomarker, [or a] tool for [treatment] sensitivity or resistance and hopefully we'll be able to guide some treatment decisions," he said, noting that prospective clinical trials are needed to see if treatment predictions made with single cells are beneficial.

Therapeutics targeting the RAF and MEK pathways have flourished in melanoma in recent years, making it possible to treat even some advanced, metastatic melanomas.

For example, the US Food and Drug Administration approved the Roche/Genentech drug Zelboraf (vemurafenib), alone or in combination with the MEK inhibitor Cotellic (cobimetinib), for BRAF-mutated melanoma. The MEK inhibitor Mekinist (trametibib) and BRAF V600E or V600K mutation-targeting Tafinlar (dabrafenib), both from GlaxoSmithKline, also received accelerated approval as a combination therapy for BRAF mutation-positive, advanced melanoma.

Advanced melanoma has also been a hotbed for the development of immune system-based treatment strategies such as the checkpoint blockade immunotherapy drug Keytruda (pembrolizumab) from Merck.

But there are still questions about treatment for BRAF mutation-free melanoma or tumors that resist or overcome available treatments, Izar and his co-authors explained. And while the importance of tumor interactions with cells around them has long been appreciated, less is known how these fine-scale cell interactions might impact cancer susceptibility and treatment outcomes.

"In the tumor, there are not only tumor cells, but obviously lots of cells from the tumor microenvironment — immune cells, cancer-associated fibroblasts, et cetera," Izar said. "With current profiling methods, we really are not able to investigate the inherent variability of the cancer cells but also the cells the microenvironment has."

In an effort to inch closer to a more complete understanding of tumor-microenvironment interactions that might inform such advances, the researchers used their 'rapid translational workflow' to quickly disaggregate cells from fresh biopsy samples in a protocol designed to minimize transcriptional artifacts introduced by temperature and time.

These cells were sorted into immune and non-immune cells by flow cytometry and subjected to a single-cell RNA-seq with the Illumina NextSeq platform, following a modified Smart-seq2 protocol, aiming for between half a million and 1 million raw reads per cell.

In 4,645 individual tumor, immune, or stromal cells in samples from 19 melanoma patients with a wide range of past treatments and clinical characteristics, the team identified a median of nearly 4,700 gene transcripts in the malignant tumor cells and more than 3,400, on average, in the associated immune cells.

One of the tumors was a primary melanoma, while the remaining biopsies came from melanoma metastases to the lymph nodes, spleen, subcutaneous sites, muscle, or the gastrointestinal tract.

To keep the analysis as unbiased as possible, Izar explained, the researchers began by clustering cells into malignant and non-malignant groups with an algorithm focused on large-scale copy number variations. As they further classified the cells, they found that non-malignant cells were apt to cluster by cell type, while the aneuploidy-prone malignant cells were more variable, grouping by tumor of origin.

In the malignant cells, the team saw subsets with low MITF gene expression, generally associated with treatment response. Those same cells usually had enhanced expression of AXL, a marker of resistance to targeted RAF/MEK treatments.

The proportion of MITF-low, AXL-high cells varied by sample, but turned up even in an untreated melanoma, suggesting cells with inherent resistance may be lurking unnoticed in many tumors.

To determine whether cells with AXL expression become more prevalent after targeted treatment, the team compared RNA-seq profiles in bulk pre-treatment and post-relapse biopsy samples from half a dozen melanoma patients — five treated with Mekinist (trametibib) and Tafinlar (dabrafenib) and one treated with Zelboraf (vemurafenib) alone. In all of the patients, post-treatment samples showed a shift toward more pronounced AXL-high patterns.

Consistent with that result, when the researchers used flow cytometry to sort AXL-positive cells in 18 melanoma lines, they found that those with a greater proportion of cells expressing AXL was linked to lower response to a RAF inhibitor.

Beyond the malignant cells, the team got clues to the relationships between immune cells and cancer-associated fibroblasts, along with a look at expression features found in tumor-infiltrating cells with so-called T-cell exhaustion.

The results suggest "there's a significant heterogeneity among the T cells and significant co-expression of multiple co-inhibitory markers," Izar said. He and his colleagues also found that T-cell activation tended to correspond with so-called T-cell exhaustion — a T-cell state that seemed to be characterized by the secretion of immunosuppressive cytokines, rather than by overall inactivity.

"Many of the things we propose in the paper need to be validated in a functional fashion using the markers that we have now identified — some known and some unknown," Izar noted. "Then we can make more assumptions about the possibility of using those signatures to predict sensitivity or resistance to immune checkpoint inhibitors."

A range of markers have been proposed for gauging response to checkpoint blockade immunotherapy, with a handful of tests reaching complementary or companion diagnostic status.

Likewise, the FDA has given the regulatory nod to companion diagnostics such as the Roche cobas 4800 BRAF V600 Mutation test (used with Zelboraf) or BioMérieux's THxID BRAF test for BRAF V600E or V600K, paired with the GlaxoSmithKline drugs Mekinist (trametibib) and Tafinlar (dabrafenib), for finding tumors most apt to respond to the RAF/MEK-targeted treatments.

But those involved in the new study and others believe it may also be useful to identify subsets of cells in the tumor and microenvironment with inherent resistance to available treatments, both to understand tumor biology and to deliver durable treatments.

"It becomes more and more clear that the microenvironment and the 'non-malignant' component, including immune cells and cancer-associated fibroblasts, play a critical role not only for sensitivity or resistance to targeted therapies, but also to immune checkpoint inhibitors," Izar said.

"I think single-cell analysis is the wave of the future, because tumors are very, very heterogeneous. There's the tumor microenvironment that you have to consider, which is tumor cells, immune cells and stroma, and all of those have to be considered in terms of the pathogenesis of the tumor," said Bruce Patterson, who was not involved in the study. 

Patterson is CEO and founder of IncellDX, a Menlo Park, California-based molecular diagnostics company that focuses on single-cell analyses. The firm has been developing single-cell diagnostic assays for quantifying markers of interest — for example, transcripts from human papillomavirus in HPV-related cancer types — using intact individual cells and a flow cytometry-centered approach.

In his experience at IncellDX, Patterson has found that regulatory bodies open to the possibility of single-cell tests with clear markers in specific cells types that may be muddied in mixed samples.

For their part, Izar and his colleagues at the Broad Institute and Dana-Farber have filed for patents related to sample processing and analytical methods used in their study and are exploring options for partnering with companies to implement their single-cell approach in prospective clinical trials.

The team is also looking at ways to address limitations in the existing workflow, including complications related to the need for fresh tissue samples, while beginning to tackle other tumor types such as breast and ovarian cancer.