Purpose Circulating tumor DNA (ctDNA) allows noninvasive disease monitoring across a range of malignancies. In metastatic melanoma, the extent to which ctDNA reflects changes in metabolic disease burden assessed by 18F-labeled fluorodeoxyglucose positron emission tomography (FDG-PET) is unknown. We assessed the role of ctDNA analysis in combination with FDG-PET to monitor tumor burden and genomic heterogeneity throughout treatment. Patients and Methods We performed a comprehensive analysis of serial ctDNA and FDG-PET in 52 patients who received systemic therapy for metastatic melanoma. Next-generation sequencing and digital polymerase chain reaction were used to analyze plasma samples from the cohort. Results ctDNA levels were monitored across patients with mutant BRAF, NRAS, and BRAF/NRAS wild type disease. Mutant BRAF and NRAS ctDNA levels correlated closely with changes in metabolic disease burden throughout treatment. TERT promoter mutant ctDNA levels also paralleled changes in tumor burden, which provide an alternative marker for disease monitoring. Of note, subcutaneous and cerebral disease sites were not well represented in plasma. Early changes in ctDNA and metabolic disease burden were important indicators of treatment response. Patients with an early decrease in ctDNA post-treatment had improved progression-free survival compared with patients in whom ctDNA levels remained unchanged or increased over time (hazard ratio, 2.6; P = .05). ctDNA analysis contributed key molecular information through the identification of putative resistance mechanisms to targeted therapy. A detailed comparison of the genomic architecture of plasma and multiregional tumor biopsy specimens at autopsy revealed the ability of ctDNA to comprehensively capture genomic heterogeneity across multiple disease sites. Conclusion The findings highlight the powerful role of ctDNA in metastatic melanoma as a complementary modality to functional imaging that allows real-time monitoring of both tumor burden and genomic changes throughout therapy.

8520 Background: Activation of CDK4 by amplification, increased expression of Cyclin D1 (CCND1) or reduced expression of the CDK inhibitor p16 (CDKN2A) can contribute to transformation of melanocytes indicating that CDK4 can act as an oncogene in melanoma.To explore if CDK4 may be a viable target for the treatment of human melanoma we have analyzed the frequency and clinico-pathological associations of genomic alterations of the CDK4 pathway in primary human melanoma and examined the genomic predictors of sensitivity to the highly selective CDK4/6 inhibitor PD 0332991 (991) in a panel of melanoma cell lines. Methods: A series of 167 primary melanomas with clinical, molecular and pathological annotation, including median follow up of 6.6 years, were analyzed for copy number variation (CNV)- gain of CDK4 or CCND1 (average gene copy >2.4) or loss of CDKN2A (average gene copy <1.4), by fluorescence in situ hybridization. A panel of 39 cell lines were treated with 991 in vitro and GI50s calculated. The mean GI50 of the melanoma cell lines was used to define sensitivity. Gene expression profiling and mutation or CNV in CDKN2A were used to identify predictors of sensitivity. Results: 75% of primary melanomas had at least one CNV (75%, 70% and 82% for BRAF, NRAS or wild-type BRAF/NRAS mutation status respectively). 55% showed loss of CDKN2A. 28% of melanomas had two or more CNVs. Melanomas with two or more CNVs involving CCND1 had worse overall survival (HR 5.56, p=0.02). Low CDKN2A mRNA expression or mutation or loss of CDKN2A predicted sensitivity to 991 with 30/33 mutant/loss lines being sensitive compared to only 2/6 wild type lines (p<0.006). Expression of CDK4, CCND1 or other cyclins or CDK-inhibitors did not predict sensitivity to 991. Conclusions: Genomic alterations in the CDK4 pathway are frequent in melanoma and are associated with worse survival, particularly when melanomas harbor two or more CNVs involving CCND1. Mutation, loss or low expression of CDKN2A in melanoma cell lines predicted sensitivity to the CDK4 inhibitor 991. Taken together these data support evaluation of CDK4 inhibitors in melanoma and suggest that CDKN2A maybe a genomic predictor of sensitivity to these agents.

the vaccination. Here, we evaluated attitude towards and effects of COVID-19 vaccination in patients with breast or gynecological cancer. The aim was to improve counseling of our patients in clinical routine.Methods: Since March 15 th 2021, patients who received one of the approved COVID-19 vaccines were routinely interviewed about immediate (0-2 days) and late side effects (within two weeks after vaccination). Clinical parameters such as current therapy, time interval between therapy administration and vaccination, and changes in the therapy schedule due to the vaccination were documented. Furthermore, the willingness of non-vaccinated patients to be vaccinated was assessed. The collected data were anonymously analyzed as a part of routine quality assurance.Results: By May 10 th 2021, 111 out of 217 (51.1%) interviewed patients had received at least one shot of COVID-19 vaccine and 21 patients both shots. More than half of the vaccinated patients were >55y (60.2%; mean: 60.7y, range 30-92y); 69% with UICC/ FIGO stage III/IV cancer. 74.6% received Conmirnaty (BioNTech/ Pfizer), 18.9% Vaxzevria (AstraZeneca) and 6.5% Covid-19 Vaccine Moderna. After the first shot, 33.3% of the patients described no side effects, 49.1% reported a local reaction (swelling or pain), 23.4% flu-like symptoms, 10.8% headache and 3.6% nausea. 11 patients had symptoms that lasted longer than two days. In 11 cases, COVID-19 vaccination had an impact on delivery of the systemic therapy (n¼10 postponements of therapy and n¼1 dose reduction). 61.3% of the non-vaccinated patients (in total n¼118) were already registered to get vaccinated; 32.8% chose to postpone vaccination for personal reasons; 5% refused vaccination.Conclusions: Breast and gynecological cancer patients appear to tolerate COVID-19 vaccination well under systemic therapy and only in few cases the vaccination interfered with the treatment schedule. Updated results will be presented at the ESMO Congress.