Oops, you're using an old version of your browser so some of the features on this page may not be displaying properly.

MINIMAL Requirements: Google Chrome 24+Mozilla Firefox 20+Internet Explorer 11Opera 15–18Apple Safari 7SeaMonkey 2.15-2.23

Poster Display session 3

5346 - Evaluating polygenic risk score prediction model for melanoma prognosis

Date

30 Sep 2019

Session

Poster Display session 3

Topics

Tumour Site

Melanoma

Presenters

Miriam Potrony

Citation

Annals of Oncology (2019) 30 (suppl_5): v533-v563. 10.1093/annonc/mdz255

Authors

M. Potrony1, N. Calbet-llopart2, M. Combalia2, G. Tell-Martí1, J.A. Puig-Butille3, A. Barreiro2, S. Podlipnik2, C. Carrera1, J. Malvehy1, S. Puig1

Author affiliations

  • 1 Dermatology Department, Melanoma Unit, Hospital Clínic of Barcelona, Institut D'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), CIBERER, 08036 - Barcelona/ES
  • 2 Dermatology Department, Melanoma Unit, Hospital Clínic of Barcelona, Institut D'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), 08036 - Barcelona/ES
  • 3 Molecular Biology Core, Biochemistry And Molecular Genetics Department, Hospital Clinic y Provincial de Barcelona, 08036 - Barcelona/ES

Resources

Login to access the resources on OncologyPRO.

If you do not have an ESMO account, please create one for free.

Abstract 5346

Background

Melanoma is the most aggressive of common skin cancers. We aimed to create a polygenic risk score (PRS) and evaluate its capability to predict melanoma prognosis better than staging.

Methods

The cohort included 1126 melanoma patients (567 males, 559 females); 57%, 24% and 19% patients stage I, II and III at diagnosis, respectively. The mean age at diagnosis was 54 yo (range 12-97). We genotyped 252 candidate SNPs by OpenArray. After quality control, we selected SNP associated with disease-free survival (DFS) and melanoma-specific survival (MSS) (log Rank P < 0.05), in the whole cohort and independently by sex. We performed cross-validation using 2/3 for training and 1/3 for validation. If the model was consistent in the three comparisons (concordance rate > 0.75), we created a PRS based on the weight of each SNP in MSS or DFS modulation. We compared the score including PRS and clinical data (age, sex, staging), with the clinical score alone or the staging score alone. ROC curves were calculated for each score to assess the capability to predict DFS and MSS.

Results

We identified 29 SNPs associated with DFS survival in the whole cohort. The score with PRS had a higher prediction capability (AUC 0.844), compared to clinical score (AUC 0.770) or staging alone (AUC 0.741). Male-specific analyses revealed 8 male-specific SNPs. The male-PRS improved also the prediction capability (AUC 0.831), compared to clinical (AUC 0.760) or staging alone score (AUC 0.735). Female-specific analyses revealed 21 female-specific SNPs. The female-PRS improved also the prediction capability (AUC 0.868), compared to clinical (AUC 0.767) or staging alone score (AUC 0.742). Using an optimal PRS + clinical score cut-off, we improved the classification of patients into low and high-risk groups within each stage and comparison (Table). Similar results were obtained regarding MSS.Table:

1366P 5-year DFS rate (%)

SexStageALLLow-risk groupHigh-risk group
ALLI92.594.962.7
II69.787.750.6
III59.390.250.3
MALEI91.492.080.0
II68.087.449.6
III56.386.348.1
FEMALEI93.596.169.8
II72.294.147.9
III62.694.142.7

Conclusions

We have identified a potential PRS that improves classification of melanoma patients within prognostic groups.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Instituto de Salud Carlos III.

Disclosure

All authors have declared no conflicts of interest.

This site uses cookies. Some of these cookies are essential, while others help us improve your experience by providing insights into how the site is being used.

For more detailed information on the cookies we use, please check our Privacy Policy.

Customise settings
  • Necessary cookies enable core functionality. The website cannot function properly without these cookies, and you can only disable them by changing your browser preferences.