Meredith McKean | Director, Melanoma and Skin Cancer Research, Sarah Cannon Research Institute (SCRI); Medical Oncologist, SCRI Oncology Partners, Nashville, Tennessee, USA
Citation: EMJ Oncol. 2026; https://doi.org/10.33590/emjoncol/2N7783N8
Your recent work has explored strategies targeting the IL-2 pathway alongside immunotherapy. What potential does this approach hold for improving responses in patients with advanced melanoma?
Historically, IL-2 was an effective strategy for patients; but its use was limited by toxicities.
We needed very fit patients, who were able to tolerate the cytokine release syndrome, which could potentially involve significant drops in blood pressure, hypotension, and hypoxia, in addition to fevers, chills, and other side effects. That really limited the number of patients that we were able to offer IL-2 to, so it’s been exciting to see a number of different approaches to modifying IL-2, to try to make it more tolerable while also trying to maintain efficacy.
I think we’ve come full circle in immuno-oncology, from novel targets to looking back at molecules and cytokines that we know can be effective, while trying to rein in that toxicity and improve tolerability to see if more patients may be able to benefit.
Your work has also looked at uveal melanoma, which has a very different biology from cutaneous melanoma. What have we learned from studying these rarer melanomas that could change how we approach them therapeutically?
Unfortunately, uveal melanoma is very different from cutaneous melanoma. We know there are classic mutations that all patients can be profiled by, looking at BAP1, SF3B1, and EIF1AX. These tumours also have low tumour mutation burden, so, inherently, patients with uveal melanoma are far less likely to respond to immune therapies. We’ve had some success with tebentafusp but that’s only an option for patients who are HLA-A*02:01 positive, limiting the patient population.
There’s been a lot of work looking at regulators of transcription, in addition to targeted therapies that may be more applicable to all patients with uveal melanoma. We’re hopeful that these approaches could offer additional options to patients.
We’re also seeing work done targeting preferentially expressed antigen in melanoma (PRAME), which is a cancer-testis antigen expressed in both cutaneous melanoma and uveal melanoma. Its high expression in uveal melanoma does parallel that seen in cutaneous melanoma. Some development efforts in cutaneous melanoma have carried over to uveal melanoma with some of the newer drugs targeting PRAME.
Combination immunotherapy can improve outcomes but may also increase toxicity. How are clinicians balancing treatment efficacy with quality of life when deciding between different immunotherapy strategies?
As we continue to advance immunotherapy, we often see a corresponding increase in treatment-related side effects. Unfortunately, we don’t have predictive biomarkers to understand who’s going to respond to just single-agent programmed cell death protein 1 (PD-1) versus someone who needs combination therapy with a LAG-3 inhibitor versus an anti-CTLA-4 monoclonal antibody.
Additionally, we don’t have predictors of response or toxicity. In the absence of those markers, treatment decisions are guided by a combination of disease characteristics and individual patient factors, including age, overall health, comorbidities, and the extent of disease.
While combination immunotherapy approaches generally produce higher response rates, they also come with an increased risk of side effects. The goal is to carefully balance those risks and benefits for each patient. Given their success in improving outcomes, combination approaches have increasingly become the standard treatment strategy for many patients with metastatic disease.
You have a particular interest in early-phase drug development and novel treatment combinations. What makes a new therapeutic approach compelling enough to move from the laboratory into clinical trials?
It has certainly been exciting to see some molecules that are IL-2 based, as well as interferon molecules that we know can be effective in melanoma, re-emerge with next-generation engineering designed to enhance their efficacy.
I think if it’s a known target or a known cytokine that has been re-engineered, there’s always a lot of enthusiasm for those molecules.
Otherwise, for novel molecules, you’re evaluating the preclinical evidence. Does it mechanistically make sense? Has it been explored in other tumour types?
Targeted therapies have also seen a lot of development, looking at next-generation BRAF inhibitors and molecules targeting atypical BRAF and RAS.
A lot of it is based on preclinical models, but also what you anticipate the toxicity is for targeted therapy, because that’s a very different toxicity profile compared to immune therapy. Those are certainly all areas of unmet need. And so, if you are able to see molecules that look like they’re able to achieve reasonable pharmacokinetics (PK) and pharmacodynamics (PD) levels, some efficacy, and a relatively tolerable safety profile, it’s worth trying to move them forward.
A review you co-authored highlighted AI-assisted skin screening and teledermatology. Where do you think AI can genuinely improve melanoma care, and where should we be cautious?
I think it will be really interesting. A lot of work was done with mole mapping and how you can assess the evolution of lesions. For melanomas, we talk about the ABCDEs, so looking at asymmetry of the lesion, borders for irregularity, variation in colour, a diameter greater than 6 mm, and then the evolution of the lesion over time. There will be a role for AI technology there, to be able to look for these changes in lesions, especially for patients who have a high number of lesions or a family history of melanoma.
The challenges that we see with AI, however, are determining where to set the threshold for intervention. We never want to miss a malignancy, but at the same time, we don’t want to biopsy every suspicious lesion and subject patients to unnecessary procedures. Striking that balance remains a significant challenge, which is why I don’t believe we’re yet at a point where AI can independently drive clinical decision-making.
It will probably be a combination, ideally using AI to identify the high-risk lesions, but it will ultimately still rest in the hands of an experienced clinician to evaluate that patient’s skin, perhaps using a dermatoscope, along with a pathology review for biopsies, to really determine which lesions are concerning.
Teledermatology is becoming increasingly relevant to skin cancer detection. How can remote assessment complement, rather than replace, the expertise of dermatologists and oncologists?
There is such a high demand for dermatologists to evaluate lesions on patients’ skin. So, I think being able to use teledermatology and reach out to patients who are in rural areas, or potentially patients who can’t take a day off work to go to a dermatologist, is going to be valuable. Hopefully, this will be an option for patients that we otherwise aren’t capturing through in-office visits and help with patient compliance.
When we recommend a patient who is at high risk or maybe has had a melanoma to go back to their dermatologist every 3 months, that’s a pretty high volume of visits in addition to all the other appointments they’re going to. So, anything that we can do to help fit medical care into a patient’s life is going to be helpful.
As precision oncology continues to develop, how do you see biomarkers helping us identify which patients are most likely to benefit from specific melanoma treatments?
Molecular profiling is going to be incredibly helpful because, historically, all of our staging systems for skin cancers have been primarily based on pathology assessments under the microscope or imaging, looking at depth of invasion, location, and involvement of surrounding structures. Now, being able to use molecular tools to really understand the genetics of the tumours, and how can we use that to add to staging to predict which lesions are high risk or recurrence, is going to be really valuable.
I think it’s really one of the next-generation changes that we’re seeing. Using ctDNA to be able to follow patients to better understand the biology of the disease is also another big change that we’re seeing.
It’s going to be up to us as clinicians to really understand where to use these technologies and which patients might benefit most.
Looking to the next 5–10 years, what advances in melanoma and skin cancer treatment are you most excited about, and what unanswered questions still need to be addressed?
I am certainly very excited about these newer technologies, being able to detect recurrences earlier with things like ctDNA. I’m also very excited about additional personalised approaches, such as mRNA-4157, the individualised neoantigen therapy for patients with high-risk melanoma. So, I’m looking forward to seeing those results from the Phase III trial.
We’re developing additional personalised approaches for patients with mutations outside of the classic BRAF mutations. We know that’s only a small portion of the patients with cutaneous melanoma, so how can we develop more targeted approaches based on each patient’s individual disease?
Some of the other big challenges in the field are for patients who have a history of melanoma or skin cancer. We don’t have great options, or really any options, to prevent another primary lesion. We see patients who come in with a number of different primary squamous cell carcinomas that are just being cut out, and we don’t have any primary prevention for those patients.
So, what else can we do for them? We know that immune therapy alone probably isn’t the answer, but primary prevention will continue to be a big area of need for patients . Primary and secondary prevention is another place where developing effective therapies with minimal toxicity is needed.
