A melanoma subtype emerges from the shadows

Unprotected exposure to ultraviolet (UV) radiation from the sun or other sources is a key risk factor for cutaneous melanoma, the most aggressive form of skin cancer. Having fair skin also increases the risk. The countries with the highest incidence of melanoma, including Australia, New Zealand, and Denmark, are largely composed of populations of European descent (1).

Daniela Robles-Espinoza, a bioinformatician studying skin cancer at the National Autonomous University of Mexico, was familiar with the most prevalent UV-associated subtypes of cutaneous melanoma. She had studied their genetic basis at the University of Cambridge and the Wellcome Trust Sanger Institute as a graduate student and postdoctoral researcher. However, when she returned to Mexico, her native country, to establish her own lab, clinicians attending melanoma cases introduced her to a subtype she had not heard of before.

Daniela Robles-Espinoza sitting at a desk typing on a keyboard.

Daniela Robles-Espinoza studies the genomic profiles of acral lentiginous melanoma tumors in the Mexican population.

Credit: Jesús René C. Wong Ramírez

Colleagues told her that in Mexico, skin cancer was different. “People have many more different levels of pigmentation,” Robles-Espinoza recalled, and they spend fewer vacations in the sun. The most common subtype in Mexico is acral lentiginous melanoma (ALM).

In contrast to other melanomas, ALM is not primarily associated with UV exposure. Most ALM tumors do not have the UV mutation signatures present in common melanoma subtypes (2). And these tumors develop in skin regions typically unexposed to the sun, such as the palms of the hands, soles of the feet, and beds of the nails. While ALM is rare in populations of European-descent, in some countries in Latin America, Africa, and Asia, it is the most frequently diagnosed subtype of melanoma (3). This does not necessarily mean that individuals with darker skin face a greater risk. Instead, the high incidence of sun-induced melanomas in fair-skinned populations may overshadow the numbers of ALM cases in those countries (3).

While UV light exposure plays a minimal role in the development of most ALM tumors, the primary risk factors remain elusive. A few studies suggest that genetic factors may be involved, while other analyses have revealed associations between a history of physical trauma and ALM development (4-7). “There’s some anecdotal evidence that mechanical injury to the foot plays a role in the pathogenesis of acral melanoma. That obviously should be explored more deeply,” said Richard White, a cancer researcher at the University of Oxford. “If that is true, you could think about prevention mechanisms.”

Similar to its etiology, ALM treatment remains understudied. The standard of care aligns with that of other melanoma subtypes, yet some evidence suggests that patients with acral tumors respond less favorably to those therapies (8). “Given this, there’s a need to try and find drug targets,” said Robles-Espinoza.

In recent years, scientists like Robles-Espinoza and White have turned their attention to this mysterious subtype of melanoma. By studying patient tumors and conducting experiments in fish and mice, they are beginning to bring this type of melanoma out of the shadows.

A peculiar genomic landscape

The distinctive genomic profile of ALM tumors may partially explain why these cancers do not always respond to standard melanoma treatments. Unlike other subtypes of cutaneous melanoma, ALM tumors have fewer point mutations. Instead, they have many structural alterations, such as copy number variation, large genomic rearrangements, and even whole genome duplications (9,10).

The fish, to me, is a discovery tool that we can use to dissect mechanisms and get hints towards therapeutics.
- Richard White, University of Oxford

Additionally, the frequency of driver mutations common in other melanomas is significantly lower in ALM. For example, while mutations in BRAF, a gene involved in cell growth, are present in more than 40 percent of other melanoma tumors, only around 15-20 percent of patients with ALM have them (11). On the other hand, ALM tumors show higher cyclin-dependent kinase 4 (CDK4) amplifications compared to other UV-associated melanomas (11).

While therapeutic strategies targeting BRAF have brought hope to many patients diagnosed with melanoma, the low prevalence of such mutations in ALM accounts for the treatment’s limited success in these individuals (12). Immunotherapy, another primary treatment for sun-induced melanomas, proves less effective in patients with ALM, likely due to the genetic makeup of these tumors (13,14).

Robles-Espinoza and her team are working to better understand the genomic profiles of ALM tumors within the Mexican population. “There’s a lot of evidence supporting the study of cancer in different populations,” she explained. “There’s an underrepresentation of Latinos.”

About six years ago, she started a collaboration with Mexico’s National Institute of Cancer, which has allowed her to analyze and sequence samples from more than one hundred patients with ALM. It has been a slow process, Robles-Espinoza noted, especially because of the pandemic.

Although the results of these analyses are not yet published, Robles-Espinoza’s team has identified significant differences in the genomic profiles of these tumors compared to other melanomas. She explained, “We do see a lot of copy number changes ... which we were expecting, and we also see some genes that are mutated maybe more than in other subtypes of melanoma.” One of those genes is KIT, which could potentially serve as an ALM drug target.

A petri dish with acral melanoma cells on it.

Researchers processed acral melanoma tumor cells after surgery for inoculation into immunosuppressed mice to generate patient-derived xenograft models.

Credit: Christian Molina Aguilar

White’s team also found differences between the DNA of ALM and cutaneous melanoma arising throughout the body. While the latter were enriched with BRAF mutations, the acral melanomas in his analyses had significantly more CRKL amplifications (15). White has used zebrafish to study cancer genetic signatures for nearly two decades, so he modeled these mutations to investigate the mechanisms by which they cause ALM tumors.

Transgenic zebrafish with CRKL alterations developed tumors primarily in their fins, the evolutionary precursors of human extremities (15). The team found that acral melanocytes were more susceptible to these specific genomic changes than melanocytes in other sites. White’s team is currently exploring the clinical implications of these findings. They are interested in targeting CRKL itself, which is challenging, or targeting the positional gene programs that give acral melanocytes their identities. “We have actually completed a screen to look for things that suppress these CRKL-driven acral melanomas and have actually a number of really interesting molecules that we think could be therapeutically quite interesting,” he said.

White acknowledged that studying ALM in zebrafish has obvious limitations. Fish and humans differ in many aspects, including anatomy and their immune systems.

“The fish, to me, is a discovery tool that we can use to dissect mechanisms and get hints towards therapeutics,” he said. However, testing the efficacy of those treatments ultimately depends on the development of other models that are closer to humans.

Models and diversity

“There aren’t many models available,” noted Patricia Possik, a melanoma researcher at the Brazilian National Cancer Institute (INCA). There are a few cell lines and patient-derived xenograft (PDX) models, where tissue from a patient tumor is transplanted into a mouse.

Similar to Robles-Espinoza’s academic trajectory, Possik approached the study of ALM upon returning to her native Brazil to establish her lab after a postdoctoral stay abroad. She got in touch with the medical staff at INCA and realized that no one was studying ALM in the translational and fundamental science labs. The physicians she met were very interested in the topic and were collecting data and conducting epidemiological studies. “It was a perfect match, and they were very open to collaborating,” she said.

Eleven people posing for a photo in a laboratory.

Patricia Possik and her colleagues have developed a large collection of patient-derived xenograft models of acral lentiginous melanoma.

Credit: Possik lab

This collaboration led to the development of a collection of PDX models for ALM that includes 40 models to date. “It’s big compared to what’s available out there,” she said. After getting a fresh ALM tumor sample from the hospital, Possik and her colleagues implant a fraction of the cancerous tissue into immunocompromised mice and let the tumor grow. Cell lines derived from PDX models of ALM and those PDX models themselves have been useful for testing treatments in vitro and in vivo (3).

Possik is also currently characterizing her ALM models. “We are in that stage of putting the things together. We sequence them. We analyze different parameters in the models to see biological characteristics of the cells,” she explained. She is excited about wrapping up these analyses and sharing the results with her colleagues. Simultaneously, her team is selecting suitable targets to work with. “We’re doing drug screens,” she said, treating the PDX-derived cell lines with different drugs to find those that kill the ALM cells.

PDX models have many limitations. “One of them is that we use immunocompromised mice,” Possik said. While this is advantageous for establishing the tumor, it overlooks the immune system’s potential modulation of tumor biology.

There is a big effort in the community to develop models.
- Patricia Possik, Brazilian National Cancer Institute

According to White, tumor location is another limitation of PDX models. Researchers often transplant the tumor cells subcutaneously but not directly into the paws of the mice. “Part of the reason why the PDX models will never quite recapitulate acral melanoma is that it’s very hard to transplant the cells in the right location,” he said.

Furthermore, there is no genetic mouse model for ALM. According to White, one of the reasons for this is the lack of technical tools and the rarity of the disease, which has deterred efforts. However, Possik said, “There is a big effort in the community to develop models.”

In addition to the limited models for testing drugs, another constraint for developing treatments for ALM is the genetic heterogeneity of the tumors. This makes it extremely difficult to target the disease therapeutically, because every patient has a specific genetic alteration, White said.

To address this heterogeneity, Possik and Robles-Espinoza recommended that melanoma researchers study more and diverse populations. This approach can provide a comprehensive overview of potential targets, or it could reveal a target frequently altered in most individuals with ALM within a specific population.

“We need to study diverse samples ... for many different reasons,” Robles-Espinoza emphasized. “That can only be achieved by equitable collaborations between scientists in developing countries and scientists in high income countries. And we’re seeing more of that recently, but it still needs a lot of work.”