Prediction of melanoma metastasis

Aim 1: Identify patients with a low risk of nodal metastasis

Patients with AJCC stage IB melanoma or higher may undergo a sentinel lymph node biopsy (SLNB) in order to detect nodal metastasis. The SLNB is used for accurate disease staging, prognosis determination, and since the introduction of adjuvant therapy, also for treatment planning. The SLNB is an invasive procedure and associated with possible morbidity. Moreover, ~80% of all SLNB’s is negative. Identification of patients with low risk of nodal metastasis may lead to a reduction of SLNB’s.

Aim 2: Identify patients with a high risk of disease progression:

Most patients with melanoma (90%) are diagnosed without any nodal or distant metastasis. Although it seems controversial to the good prognosis of most early stage melanomas, death due to melanoma after diagnosis of an early stage melanoma concerns 41% of all melanoma deaths (i.e. >300 of 800 melanoma deaths in the Netherlands). We aim to develop prognostic models based on both patient and tumor characteristics in order to identify patients with early stage melanoma who are at high risk for disease progression. Those patients can then be treated with adjuvant targeted or immunotherapies to increase the probability of cure.

Prediction of nodal metastasis

SkylineDx developed a clinicopathological gene expression profile (CP-GEP) model in collaboration with the Mayo Clinics in the United States in order to predict which patients have a low risk of nodal metastasis.

We recently validated the CP-GEP model among >200 patients who underwent a SLNB in the Erasmus MC. We observed that the CP-GEP model can accurately identify those patients with a low risk of nodal metastasis. We also assessed the prognostic value of CP-GEP.

Prediction of distant metastasis

Erasmus MC, the Netherlands Comprehensive Cancer Organization (IKNL) and SkylineDx collaborate on the development of a new prediction model for distant metastases in patients with cutaneous melanoma in an early stage. This model includes clinical data (e.g. age, sex, immunosuppressive drug use), digital histopathological images, images of immune cell distribution and multi-omics data (RNAseq and DNAseq). These data will be integrated using artificial intelligence techniques to develop a single prediction model.

TKI Health Holland
KWF
SkylineDx
Netherlands Comprehensive Cancer Organisation (IKNL)

Collaborations within Erasmus MC

• Department of Pathology, including the Pathology Research and Trial Service (PARTS) and the Clinical Bio-Informatics Unit
• Department of Surgical Oncology
• Department of Medical Oncology, including Tumor Immunology

Collaborations outside Erasmus MC

• Netherlands Comprehensive Cancer Organisation (IKNL)
• SkylineDx

Validation of a clinicopathological and gene expression profile model for sentinel lymph node metastasis in primary cutaneous melanoma.
Mulder EEAP, Dwarkasing JT, Tempel D, van der Spek A, Bosman L, Verver D, Mooyaart AL, van der Veldt AAM, Verhoef C, Nijsten TEC, Grunhagen DJ, Hollestein LM. Br J Dermatol. 2021 May;184(5):944-951. doi: 10.1111/bjd.19499. Epub 2020 Nov 2

Using a Clinicopathologic and Gene Expression (CP-GEP) Model to Identify Stage I-II Melanoma Patients at Risk of Disease Relapse.
Mulder EEAP, Johansson I, Grünhagen DJ, Tempel D, Rentroia-Pacheco B, Dwarkasing JT, Verver D, Mooyaart AL, van der Veldt AAM, Wakkee M, Nijsten TEC, Verhoef C, Mattsson J, Ny L, Hollestein LM, Olofsson Bagge R. Cancers (Basel). 2022 Jun 9;14(12):2854. doi: 10.3390/cancers14122854.

• Dr. L (Loes) Hollestein, epidemiologist, PI
• Dr. A. (Antien) Mooyaart, pathologist
• Dr. L. (Luba) Pardo, genetic epidemiologist
• Dr. Y. (Yunlei) Li, bio-informatician
• Drs.C. (Catherine) Zhou, PhD-candidate
• Drs. T. (Thamila) Kerkour, PhD-candidate
• Drs. J. (Jasper) Ouwerker, PhD-candidate