Search

9579 Background: Recently, trials have shown the benefit of adjuvant aPD-1 therapy in macroscopic stage III melanoma patients. This treatment has been incorporated in daily clinical practice, however, a substantial part of patients still does not benefit from this therapy, as they develop recurrences. The aim of this study is to evaluate the results of adjuvant aPD-1 therapy and the potency of the IFNγ signature as a prognostic or predictive marker, as it has proven to be predictive of response in neoadjuvant trials. Methods: Patients participating in an ongoing biobank study and naïve for systemic therapy were included, between 10-2017 and 06-2020, after complete resection of macroscopic stage III melanoma. Approval and reimbursement of adjuvant therapy in the Netherlands started in 12-2018, resulting in 2 cohorts of similar high risk patients: prior to availability of adjuvant aPD-1 (cohort A) and thereafter (cohort B). Data cut-off for clinical data was January 1st 2021. Transcriptome sequencing was performed on samples of stage III melanoma by CeGaT GmbH, IFNγ signature was determined on these data with the median as cut-off. Clinical data were compared between cohort A and B as intention-to-treat population, including patients with a recurrence before adjuvant therapy start (n=10). Results: In total, 99 patients were included: 50 in cohort A and 49 in cohort B. Majority of included patients had thick primary melanomas (Breslow >2mm in 59.6%) and stage IIIC/IIID disease (83.3%) according to AJCC 8th edition. At a median follow-up of 20.6 months (95% confidence interval [CI] 16.6-24.7), median recurrence-free survival (RFS) was 6.1 months (95%CI 3.9-8.4) versus 22.8 months (95%CI 8.7-36.9), significantly in favor of cohort B (p=0.011). Median overall survival (OS) was not reached in both patient groups, but was overall significantly different (p=0.040), favoring cohort B. RNA sequencing was performed in 25 patients who received adjuvant therapy and in 24 who did not, excluding patients with an early recurrence (

9563 Background: Talimogene Laherparepvec (T-VEC) is a genetically modified herpes simplex type 1 virus and known as an effective oncolytic immunotherapy for injectable cutaneous, subcutaneous, and nodal melanoma lesions in stage IIIB-IVM1a patients. Recently, Stahlie et al. published (Cancer Immunol Immunother '21) a model for predicting a complete response (CR) to T-VEC based on 3 easily accessible tumor characteristics identified using univariable and multivariable logistic regression analysis. The aim of this study was to externally validate this model in an independent, American patient cohort. Methods: A total of 76 patients with stage IIIB-IVM1a melanoma treated with T-VEC at Moffitt Cancer Center were included. A second nomogram was built incorporating the same predictive factors: tumor size (diameter of largest metastasis in mm), type of metastases (cutaneous, subcutaneous and nodal) and number of metastases (cut-off: 20). Predictive accuracy was assessed through calculation of overall performance, discriminative ability, and calibration. Outcomes and previously published outcomes were compared. Statistical analyses were done using R software. Results: Overall performance of the validation dataset nomogram was calculated with the Brier score and found to be 0.195, demonstrating good overall performance and similar to the original model Brier score of 0.182. Discriminative power, assessed by calculating the area under the receiver operating characteristic (ROC) curve was similar for both models, 0.767 and 0.755 for the NKI and Moffitt, respectively, resulting in a fair discriminative ability. The calibration curve showed mostly slight underestimation for predicated probabilities >0.37 and slight overestimation

TPS9592 Background: Merkel cell carcinoma (MCC) is a rare but highly aggressive human skin cancer often caused by the Merkel cell polyomavirus or extended exposure to sunlight. Since the approvals of avelumab globally and subsequently pembrolizumab (US only), anti–PD–(L)1 antibody therapies have become the standard of care for advanced/metastatic MCC patients in recent years. Still, a significant proportion of MCC patients do not respond to or relapse on anti–PD–(L)1 antibody monotherapy. Recent preclinical data suggest that the small molecule, selective class I histone deacetylase inhibitor (HDACi) domatinostat can overcome critical mechanisms of MCC resistance to checkpoint inhibitors. These escape mechanisms include the epigenetic downregulation of the antigen processing and presentation machinery, hence treatment with domatinostat is thought to favorably modulate the tumor environment allowing a reintroduction of anti–PD–(L)1 therapy for an improved and sustained clinical benefit. Methods: The study is a phase II, multicenter, single arm clinical trial of the orally administered HDACi domatinostat in combination with the anti–PD–(L)1 antibody avelumab for patients with advanced unresectable/metastatic MCC that are progressing on previous anti–PD–(L)1 therapy. ClinicalTrials.gov Identifier: NCT04393753. Key Inclusion Criteria are: histologically confirmed MCC, an ECOG performance status ≤ 1, MCC in an advanced, unresectable stage III or metastatic stage IV, and progressing on previous anti–PD–(L)1 antibody monotherapy within the last 12 weeks before planned first administration of study medication. Key Exclusion Criteria are: history of serious anti–PD–(L)1 therapy related adverse reactions prohibiting further avelumab treatment, more than one line of previous systemic anti neoplastic therapy other than anti–PD–(L)1 antibody monotherapy (excluded: palliative radiation therapy of single lesions within 2 weeks before planned administration of study medication), significant active or chronic disease (infections, immunodeficiencies, cardiovascular, psychiatric disorders). A total of 40 patients will be enrolled in up to 46 clinical study sites in Europe and USA. Anti-tumor activity will be primarily assessed by the objective response rate according to RECIST v1.1 as an exploratory analysis. Secondary objectives include additional efficacy assessments, safety, quality of life and pharmacokinetics of domatinostat in combination with avelumab. Correlative aims include evaluating biomarkers for association with clinical benefit. The first patient was enrolled on Oct. 16, 2020, 21 of 46 clinical sites are active and 4 out of 40 planned patients have been enrolled as of Feb. 15, 2021. Clinical trial information: NCT04393753.

9575 Background: MCC is a rare and aggressive neuroendocrine malignancy of the skin. Postoperative radiotherapy (PORT) is recommended by current guidelines to reduce recurrences and improve survival in patients with locoregional MCC. However, evidence supporting these recommendations is conflicting and deviations from the protocol occur frequently, due to the generally elderly and frail patient population. We aim to evaluate the influence of PORT on survival in stage I-III MCC patients treated in the Netherlands. Methods: All patients with stage I-III MCC treated in three referral centers between 2013 and 2018 were included retrospectively. Recurrence free survival (RFS) and disease specific survival (DSS, including death from unknown causes) were compared between patients with and without PORT. Prognostic factors for DSS were analyzed using Kaplan-Meier curves, logrank test and cox regression. Since sentinel node biopsies (SN) are frequently omitted in this patient population, analyses were performed in patients with clinical (SN not performed) stage I/II (c-I/II-MCC), pathologic (SN negative) stage I/II (p-I/II-MCC) and stage III MCC (III-MCC), separately. Propensity score matching (PSM) was performed to assess possible confounding by indication. Results: In total 219 patients were included, of whom 54 had p-I/II-MCC, 82 had c-I/II-MCC and 83 had III-MCC. Median follow up time was 53.4 (IQR 32.8-62.4), 28 (11.8-43.3) and 30.8 (19.5-50.0) months, respectively. PSM identified no confounding by indication, analyses were therefore performed in the unmatched cohort. Majority of recurrences were regional in p-I/II-MCC (77.8%) and c-I/II-MCC (74.2%), and distant in III-MCC (61.7%). RFS was significantly different across all stages (p

e22079 Background: Stage IIB/IIC (pT3b-T4N0) patients are known to have high-risk primary tumors, even higher risk than some stage IIIA/B melanomas (AJCC Staging System 8th edition), however they follow the same routine to sentinel lymph node biopsy (SLNB) as more low-risk tumors. A priori the risk of finding SLNB or other metastases is much higher for these thick and/or ulcerated primary melanomas compared to the thinner ones. Guidelines are not conclusive regarding the use of preoperative imaging in these cases. Recently, a trend to more frequently use cross-sectional imaging has been noticed. However, others have previously shown that preoperative ultrasound was the most sensitive. The aim of this pilot study was to assess the value of ultrasound (US) and Positron Emission Tomography/Computerized Tomography (PET/CT) prior to SLNB for stage IIB/C (pT3b-T4N0) melanoma patients. Methods: Starting 2019-04, all patients with a pT3b melanoma or higher (8th AJCC) were included. All patients underwent US and PET/CT before their planned lymphoscintigraphy and routine SLNB. Suspected metastases were confirmed with cytologic puncture. Results: A total of 20 patients were screened. Seven patients (35%) had metastases detected by imaging: one by PET/CT, three by US and three by both imaging modalities. Three of these metastases were detected by US as well as PET/CT. All metastases were nodal. For all seven patients treatment was altered to lymph node dissection with adjuvant therapy. Of the 13 patients in whom no metastases were identified by imaging, six (46%) still had a positive sentinel node (SN). Conclusions: This study showed that this select group of patients had a high risk of metastases prior to SLNB and that all recurrences except one, were detected by ultrasound. This suggests that nodal staging with US is sufficient and can replace the need for SLNB when metastases is proven with cytology. Despite negative imaging, SLNB cannot be foregone for pT3b-pT4N0 melanoma patients, as many still have an involved SN. Cross-sectional imaging can be reserved for patients after positive cytology or SN to confirm the absence of distant visceral metastases.

10015 Background: Early results of the OpACIN-neo study testing 3 different dosing schedules of neoadjuvant IPI + NIVO demonstrated that 2 cycles IPI 1mg/kg + NIVO 3mg/kg (IPI1NIVO3, arm B) was the most favorable schedule with 20% grade 3-4 immunotherapy-related adverse events (irAEs) and a pathologic response rate (pRR) of 77%. After a median follow-up (FU) of 8.3 months, none of the 64 patients (pts) with a pathologic (path) response ( < 50% viable tumor cells) versus 9/21 (43%) without a path response had relapsed. Here, we present the updated 2-year RFS, EFS and long-term toxicity data. Methods: In the phase 2 multi-center OpACIN-neo trial, 86 stage III melanoma pts with resectable and RECIST 1.1 measurable lymph node metastasis were randomized between 3 different dosing schedules of neoadjuvant IPI + NIVO: arm A: 2x IPI3+NIVO1 Q3W (n = 30), arm B: 2x IPI1+NIVO3 Q3W (n = 30), and arm C: 2x IPI3 Q3W followed by 2x NIVO3 Q2W (n = 26). Lymph node dissection was scheduled at week 6. Primary endpoints were toxicity, radiologic RR and pRR; RFS and EFS were secondary endpoints. Results: After a median FU of 24.6 months, the median RFS and EFS was not reached in any of the 3 arms. In total, 2 pts progressed before surgery, 12 pts relapsed (11 pts without path response and 1 pt with pCR) and 5 pts died (4 due to melanoma and one pt due to toxicity). Estimated 24-months RFS was 84% (95% CI 76-92%) for the total population, 97% (95% CI 93-100%) for pts with a path response and 36% (95% CI 17-74%) for pts without a path response. Estimated 24-months EFS for the total population was 82% (95% CI 74-91%). RFS and EFS did not differ between the arms. Of the 81 pts alive, 55 (68%) have ongoing irAEs; only 2 (3%) pts have ≥ grade 3 irAEs. Most frequent ongoing irAEs were vitiligo (35%), fatigue (14%), sicca syndrome (11%), rash (10%), arthralgia (7%) and endocrine toxicities (20%). 17 pts need hormone replacement therapy: 11 (14%) thyroid hormone and 7 (9%) hydrocortisone. No difference between treatment arms was observed. Ongoing surgery-related AEs were observed in 31 (38%) pts of which lymphedema was seen most frequently (17 pts; 21%). Conclusions: Extended follow-up data shows that 2 cycles of neoadjuvant IPI + NIVO without adjuvant therapy induces durable RFS. While almost no ongoing high-grade irAEs were observed, the majority of pts have low-grade ongoing toxicities. These outcomes strongly support the need to test 2 cycles of neoadjuvant IPI1+NIVO3 versus adjuvant anti-PD-1 in a randomized phase 3 trial. Clinical trial information: NCT02977052.

TPS10087 Background: Previous OpACIN and OpACIN-neo studies, investigating neoadjuvant ipilimumab (IPI) plus nivolumab (NIVO), demonstrated high pathologic response rates (74-78%) and favorable long-term outcomes in patients (pts) achieving pathologic response; at 36 and 18 months follow-up, respectively, only 1/71 (1.4%) pts with response has relapsed. In contrast, pts without pathologic response (pNR) have a poor prognosis; 15/23 (65.2%) have relapsed so far. This emphasizes the need for baseline biomarkers predictive of non-response and new neoadjuvant treatment combinations for these pts. In our previous studies, baseline interferon-gamma (IFN-γ) signature low pts were less likely to respond to neoadjuvant IPI plus NIVO. The DONIMI study tests the combination of NIVO +/- IPI with domatinostat (DOM), a class 1 histone deacetylase inhibitor, according to the IFN-γ signature in the tumor. Based on the signature previously described by Ayers et al. we have developed a neoadjuvant IFN-γ signature algorithm that will be used for the first time to classify pts in this prospective trial. Methods: The aim of this two-center investigator-initiated phase 1b study is to assess the safety and feasibility of neoadjuvant NIVO +/- DOM +/- IPI in 45 stage III melanoma pts with RECIST 1.1 measurable de-novo or recurrent disease. IFN-γ signature high pts (n = 20) will be randomized (stratified by center) to Arm A (2 cycles NIVO 240mg q3wk) or Arm B (2 cycles NIVO 240mg q3wk + DOM 200mg twice daily (BID), d1-14, q3wk). IFN-γ signature low pts (n = 25) will be randomized to Arm C (2 cycles NIVO 240mg q3wk + DOM 200mg BID, d1-14, q3wk) or Arm D (2 cycles NIVO 240mg q3wk + IPI 80mg q3wk + DOM 200mg once daily (OD), d1-14, q3wk). Based on safety data of the first 5 pts in arm D, the remaining pts will be treated with either a higher dosing scheme (200mg BID, d1-14, q3wks), a lower dosing scheme (100mg OD, d1-14, q3wks) or the same dosing scheme (200mg OD, d1-14, q3wks). The primary endpoint is safety and feasibility. A treatment arm will be declared as not feasible if 2/5 or 3/10 pts cannot adhere to the preplanned time of surgery (week 6 +/- 1week) due to treatment-related adverse events. Biopsies (week 0, 3), blood samples (week 0, 3, 6, 12) and feces (week 0, 3, 6) will be collected for translational research. The first patient was enrolled on January 23th, 2020. Clinical trial information: NCT04133948.

10064 Background: Neoadjuvant ipilimumab and nivolumab induces high pathologic response rates of 74-78% (OpACIN and OpACIN-neo trial), thus the role of Therapeutic Lymph Node Dissections (TLND) in patients with major pathologic responses (MPR: pathological (near) complete response) is now unclear. In the PRADO trial, TLND was omitted in patients with MPR in their index lymph node ((ILN), the largest LN marked prior to neoadjuvant therapy). We sought to determine if less extensive surgery is associated with better Health Related Quality of Life (HRQoL). These are the first results of the comparison of HRQoL between patients undergoing a TLND or less extensive ILN excision. Methods: HRQoL was assessed with the European Organisation for Research and Treatment of Cancer QoL questionnaire-C30 (QLQ-C30). A generalized estimation equation was used to assess the difference in HRQoL outcomes between patients who underwent TLND (pathological non- and partial-responders, pNR/pPR) versus those who did not (pathological (near)complete responders, pNCR/pCR). Differences were adjusted for age, gender and follow-up (FU, in weeks), but not for pathological responses (pNR, pPR, pNCR & pCR). Differences in QLQ-C30 scores were classified as clinically important according to published guidelines. Results: A total of 49 patients from the PRADO study had reached at least 24 weeks FU, and were included in the first explorative analysis. The median age of this study population was 58 years (range, 22-84). Questionnaire completion rates were high: 94% at baseline, 100%, 90%, 88% at week 6, 12 and 24, respectively. Sixteen (33%) patients underwent TLND versus 33 (67%) who had ILN excision only. Over a FU period of 24 weeks, patients who underwent TLND scored significantly lower on global (68 vs 78, adjusted difference (diff) = -9.53, p = .005), physical (84 vs 94 diff = -11.1, p = < .001), emotional (69 vs 83, diff = -11.7, p = .001), role (70 vs 85, diff = -13, p = .004), and social functioning (81 vs 91, diff = -8.9, p = .016) and had a higher symptom burden of fatigue (35 vs 23, diff = 11.1, p = .004), insomnia (38 vs 18, diff = 16.6, p = .002) and financial impact (12 vs 4, diff = 7.9, p = .027) than patients undergoing ILN excision only. These differences were indicated as clinically relevant. Conclusions: First results from PRADO suggest that reducing the extent of surgery following neoadjuvant immunotherapy might result in better HRQoL of high-risk stage III melanoma patients. Clinical trial information: NCT02977052.

e22089 Background: Talimogene laherparepvec (T-VEC) is a genetically modified herpes simplex type 1 virus, which is used as an oncolytic immunotherapy in stage IIIB-IVM1a melanoma patients. It is known to be an effective therapy for injectable cutaneous, subcutaneous and nodal melanoma lesions, as approved by the European Medicines Agency (EMA). Combination therapy is not yet approved by EMA pending the results of the phase 3 Masterkey-265 trial. The objective of the current study was to identify prognostic factors for achieving a complete response (CR) that can be used to select patients for treatment with T-VEC monotherapy. Methods: Patients with stage IIIB-IVM1a melanoma, treated with T-VEC at the Netherlands Cancer Institute between 2016-12 and 2019-05 with a follow-up time > 6 months, were included. Data was collected on baseline characteristics, responses and adverse events (AEs). Durable response rate (DRR) was defined as the percent of patients with a CR or partial response (PR) maintained continuously > 6 months. Univariable analyses were conducted and a prediction model was developed to identify prognostic factors associated with complete response. Results: For this study, a total of 71 patients were included with a median follow-up of 16.1 months. The median age was 70 years (range: 35-90). As best response, 47 patients (66%) had a CR and 10 patients (14%) had a PR, resulting in an overall response rate of 80%. Twenty-one patients (30%) stopped treatment because of progressive disease and sixteen patients (23%) developed a recurrence during follow-up after achieving a PR or CR. Median duration of CR was 11 months. The durable response rate was 42%. Grade 1-2 AEs occurred in almost every patient. Tumor size, type of metastases, previous treatment with systemic therapy and stage (8Th AJCC) were independent prognostic factors for achieving a CR and for progression-free survival. Achieving a CR was associated with a reduced risk of death. The prediction model includes tumor size, type of metastases (only cutaneous vs. subcutaneous (+/- cutaneous) vs. nodal (+/- cutaneous/subcutaneous)) and number of lesions as predictors. Conclusions: This study shows that intralesional T-VEC monotherapy for stage IIIB-IVM1a melanoma is able to achieve high complete and durable response rates. The prediction model shows that use of T-VEC in patients with less tumor burden is associated with better outcomes, suggesting T-VEC should perhaps be used earlier in the course of the disease.

10070 Background: In the context of multiple in-transit melanoma metastases without nodal involvement, a variety of treatment modalities have historically been employed including surgery, laser, isolated limb perfusion/infusion, intralesional interleukin-2, T-VEC and electrochemotherapy. Unfortunately, most patients treated with these modalities experience subsequent disease progression. While checkpoint inhibitors (CPI) are a standard of care for bulky unresectable stage III and for stage IV melanoma, patients with isolated in-transit metastases were rarely included in registration studies. There are anecdotal reports of lower response rates in these patients despite them having disease characteristics that would usually be associated with a good response. Methods: We report data from 11 retrospective patient cohorts treated at cancer centres across Europe who received CPI between 2016 and April 2019. All patients had multiple in-transit metastases without clinical or radiological evidence of nodal or distant disease. Disease response was assessed using CT, PET-CT or MRI depending on clinical indication. All patients had at least one prior resection of loco-regional relapsed disease and were deemed not curable by further surgery. Results: Sixty three patients meeting criteria were identified, 40 females and 23 males. Median age was 72 years and 54 (86%) patients had a normal lactate dehydrogenase (LDH). 19 (30%) patients had a BRAF mutation. At treatment initiation, the majority 55 (87.3%) received single agent PD-1 inhibitor, 7 (11.1%) combination ipilimumab + nivolumab and 1 (1.6%) received single agent anti-CTLA 4. The overall response rate was 62% for the full population. The response rate with anti-PD1 monotherapy was 59%. With a median FU of 23 months, the median PFS was 26 months, median OS not reached. OS estimates with 95% CI: 12 month - 93% (87-100%), 24 month - 88% (80-98%), 36 month - 80% (67-95%). Conclusions: The results show a high response rate to CPI in patients with in-transit metastases and support early treatment with CPI following identification of in-transit metastases not curable with surgery whilst disease characteristics remain favourable.

e22039 Background: To date, international consensus concerning the use of PET/CT as a surveillance tool in the follow-up of high-risk melanoma patients after complete resection of disease is lacking. Moreover, with the rise of adjuvant therapy it seems appropriate to investigate the role of this imaging modality to exclude newly developed metastases after resection and prior to starting treatment. The aim of this study was to investigate the use of PET/CT as surveillance tool in the follow-up and prior to adjuvant therapy in asymptomatic patients with complete resection of stage IIIB and IIIC melanoma. Methods: Prospectively two cohorts were set up with stage III melanoma patients with complete resection of disease. In the first cohort (stage IIIB/C AJCC 7th) surveillance PET/CT was performed 6-monthly for two years if patients stayed asymptomatic with normal serum S100B, with a final scan at three years. In the second cohort (stage IIIB/C/D AJCC 8th) patients underwent one screening PET/CT after resection and prior to starting adjuvant treatment. Results: Eighty patients entered follow-up in cohort 1. Of these, the majority did not undergo surveillance scans, because they required treatment for newly detected clinical metastases. Thirty-five patients remained asymptomatic and were included in surveillance cohort one (105 scans) with a median follow-up of 33 months. Twelve patients (34%) developed a recurrence, seven (20%) of which were detected on the first scan at six months. Seven recurrences involved stage IIIC patients, five stage IIIB patients. Sensitivity and specificity were 92% and 100% respectively. Forty-two patients were included in cohort 2. Recurrence was suspected on nine scans, four (10%) of which were true positive. One patient proceeded to undergo a node dissection and then started adjuvant therapy. The other three patients had progressed to stage IV and therefore started radiotherapy and/or systemic immunotherapy. Five (12%) scans were false positive, the suspected lesions were not related to the preceded surgery. The number of scans needed to find one asymptomatic recurrence were 8.8 and 10.5 in cohort one and two, respectively. Conclusions: This study shows that PET/CT is a useful surveillance tool for detecting recurrence in asymptomatic high-risk resected stage III melanoma patients, especially within the first six months after surgery and therefore should be considered when monitoring these patients during follow-up as well as prior to starting adjuvant therapy.

TPS9605 Background: Adjuvant (adj) immune checkpoint inhibition (ICI) improves relapse free survival (RFS) in stage III melanoma patients (pts). However, preclinical and translational data suggest that neo-adjuvant (neoadj) treatment might be favorable due to broader immune activation. The phase 1b OpACIN study comparing neoadj to adj IPI plus NIVO demonstrated a high pathological response rate (pRR) of 78% complicated by 90% gr 3-4 immune-related adverse events (irAEs). The phase 2 OpACIN-neo trial tested safety and efficacy of three different schemes of neoadj IPI+NIVO and identified two cycles of IPI 1mg/kg + NIVO 3mg/kg as well tolerated (20% gr 3-4 irAEs), with a high pRR of 77%. In both trials, none of the pts with a pathologic response have relapsed after a median follow-up of 30 and 8.3 months. In stage IV melanoma, long-term benefit is observed in patients achieving CR with ICI, even after cessation of therapy. This raises the question of whether a therapeutic lymph node dissection (TLND) can be omitted when a deep pathologic response with neoadj IPI+NIVO is achieved. Methods: The aim of this international multi-center investigator-initiated phase 2 PRADO extension study is to confirm the pRR and toxicity of 2 cycles of neoadjuvant IPI 1mg/kg + NIVO 3mg/kg (the preferred OPACIN-neo regimen) and to test response-driven subsequent therapy i.e. omitting surgery and adjuvant ICI based on the pathological response. 100-110 pts with stage IIIB/C melanoma and a measurable lymph node (≥15mm according to RECIST 1.1) will receive two cycles of IPI 1mg/kg + NIVO 3mg/kg after marker placement into the largest lymph node metastasis. After six weeks, pts will undergo resection of the index lymph node. For pCR/near pCR, pts will not undergo TLND; For pPR, pts will undergo TLND; and for pNR, pts will undergo TLND and start adjuvant NIVO or targeted therapy +/- radiotherapy for 52 weeks. Primary endpoints are pRR of marked lymph node and RFS at 24 months. Baseline biopsies, blood samples (week 0, 6, 12) and faeces (week 0, 6) will be collected for translational research analyses. The first patient in this trial was included in October 2018; 22 patients have been enrolled. Clinical trial information: NCT02977052.

9587 Background: The aim of this study is to evaluate the potency of short-term neoadjuvant cytoreductive therapy with dabrafenib and trametinib (BRAF and MEK inhibitor respectively) to allow radical surgical resection in patients with unresectable BRAF-mutated, locally advanced stage III or oligometastatic stage IV melanoma. Methods: A total of 25 patients with BRAF-mutated, unresectable locally advanced stage III or oligometastatic stage IV (≤3 metastases) melanoma will be treated with dabrafenib and trametinib for 8 weeks. Response evaluation by positron emission tomography/computed tomography (PET/CT) will occur at 2 and 8 weeks. If sufficient downsizing occurs, surgical resection will be performed. Biopsies for translational research will be taken at baseline and 2 weeks. The dissection specimen will be stored at 8 weeks. Results: Currently 20 patients have been included. Of these, 2 patients showed PD upon treatment and did not proceed to surgery. In 17/18 (94%) patients resection was possible after neoadjuvant treatment, of which 16 (94%) were R0 resections. Median follow-up time is 28 months with a median recurrence free survival of 9 months in patients undergoing surgery. The 1-year overall survival (OS) was 94% and 2-year OS 82%. Median OS was not reached. Metabolic response rates (RR) on PET/CT at 8 weeks were: 4 (20%) CR, 14 (70%) PR, 0 (0%) SD, 2 (10%) PD. Pathologic RR differed: 7 (35%) CR, 7 (35%) PR, 3 (15%) SD, 0 (0%) PD and in 3 patients (15%) no pathologic response was measured, since no resection was performed. Most patients (85%) experienced any toxicity, of which 50% was grade 1, 20% grade 2 and 3 patients (15%) experienced grade 3 toxicity. The most common reported toxicity was fever. Conclusions: Neoadjuvant dabrafenib and trametinib shows to be a potent cytoreductive treatment, allowing radical resection of metastases in 16/20 (80%) patients with prior unresectable locally advanced melanoma. Patients with no recurrence remained disease-free for a prolonged period of time. If there was recurrent disease, this usually occurred within months after surgery and this may present an opportunity for further tailored adjuvant therapy. Clinical trial information: NTR4654.

9570 Background: To date no adequate biomarker for Merkel Cell carcinoma (MCC) has been identified. The introduction of immunotherapy (IT) for metastatic MCC increases the need for a biomarker. Serum Neuron-specific enolase (NSE) has already been tested and is commonly used as a biomarker for several small cell malignancies. However, the role of NSE in MCC is still unclear. Aim: To investigate the role of NSE as a biomarker in MCC. Methods: A prospective cohort of MCC patients treated from 2016 to July 2018 was analyzed. Kaplan Meier curves with log rank test, Cox regression and mixed models were used to analyze NSE. A separate evaluation was performed for patients treated with IT. Results: A total of 78 patients (42 males, median age 71 years, stage I&II, III and IV MCC in 37(47%), 39(50%) and 2(3%) patients at time of diagnosis with 474 NSE levels (median 15 ; IQR 12,6-22 ng/ml were included. Baseline NSE (n=36) had no influence on survival or progression. During follow-up (FU) NSE levels correlated with tumorload (p=0,01) with increase of 15 ng/ml per class (no tumorload, localized MCC, nodal and distant metastases, respectively). NSE level during FU was able to detect progression (AUC 0,89). Several cut off values were evaluated. A NSE of 18,2 ng/ml was considered the most optimal level for clinical use (sensitivity 91%, specificity 78%, PPV 48%, NPV 98% to detect progression). During IT (n=16; 195 NSE values) all complete responders (n=7) had a normalized NSE (

9503 Background: Pathological complete response (pCR) to neoadjuvant systemic therapy (NST) correlates with survival, and is recognized as a path to regulatory approval in several cancers. Recent trials have reported that neoadjuvant immunotherapy (IT) and targeted therapy (TT) regimens achieve high pCR rates and impressive recurrence-free survival in stage III melanoma, however, the relationship between pCR, relapse-free (RFS) and overall survival (OS) in larger datasets of melanoma patients (pts) remains unknown. Methods: We pooled data from 6 modern NST clinical trials of anti-PD-1 based immunotherapy or BRAF/MEK targeted therapy conducted across institutions participating in the INMC. Pts with RECIST measurable, surgically resectable clinical stage III melanoma who underwent surgery were included. NST regimens included nivolumab (as monotherapy or in combination with ipilimumab), pembrolizumab or dabrafenib+trametinib. Baseline disease characteristics, treatment regimen, pCR and RFS were examined. Results: 184 pts with clinical stage III melanoma (AJCCv7: 100 IIIB, 84 IIIC) completed NST (133 IT, 51 TT) and underwent surgery. Median age was 57y (range 18-87). A pCR was observed in 41% of patients; 51 (38%) with IT and 24 (47%) with TT. Median follow-up post-surgery is 13 mo (95% CI 12-16); 10 mo with IT and 22 mo with TT. 44 (24%) pts have recurred (17 loco-regional, 21 distant, 6 both sites at first recurrence), 18 (14%) after IT and 26 (51%) after TT. 12-month RFS was improved with IT vs TT (83% vs 65%, p < 0.001). For those with pCR, 7% have recurred, 0/51 (0%) after IT, 7/17 (41%) after TT. For those without pCR, 34% have recurred, 18/82 (22%) after IT and 19/27 (70%) after TT. 12-month RFS was improved in those with pCR vs without pCR (95% vs 62%, p < 0.001), including in those with IT (100% vs 72%, p < 0.001) and TT (88% vs 43%, p < 0.001). 16 (9%) patients have died including two who had a pCR, both from TT. Conclusions: Neoadjuvant IT and TT are active regimens in resectable clinical stage III melanoma patients and are associated with high pCR rate. The ability to achieve pCR correlates with improved RFS and remarkably no patient with pCR from immunotherapy has recurred to date.

2517 Background: Several studies suggested that patients (pts) with an immune-related adverse event (irAE) during immunotherapy have better outcomes than those without. It remains uncertain whether these observations can be explained by guarantee-time bias or the role of irAE as an indicator of drug activity. Here, we investigated the association between irAEs and recurrence-free survival (RFS) in the double-blind EORTC 1325/KEYNOTE-054 trial that compared pembrolizumab and placebo in high-risk stage III melanoma pts. Methods: Eligible pts included adults with complete resection of cutaneous melanoma metastatic to lymph node(s), classified as stage IIIA (lymph node metastasis > 1 mm), IIIB or IIIC (without in-transit metastasis) and with no active autoimmune disease that required systemic treatment in past 2 years. We used a Cox model adjusted for sex, age, and stage with a time-varying covariate taking a value zero before the irAE onset and a value one afterwards to estimate the association between the occurrence of irAEs and RFS. Results: Consistent with the main analysis in the ITT population (n = 1019, Eggermont et al, NEJM, 2018), RFS was longer in the pembrolizumab than in the placebo arm (HR = 0.56, 98.4% CI: 0.43-074) among pts who started the treatment (n = 1011). The incidence of irAE on study was 37.3% in the pembrolizumab (n = 509) and 9.0% in the placebo arm (n = 502) and, in each treatment group, it was similar in males and females. The occurrence of an irAE was significantly associated with a longer RFS in the pembrolizumab arm (HR = 0.61, 95% CI: 0.39-0.95, P = 0.03). This was true for both males and females. However, in the placebo arm, no association was observed (HR = 1.39, 95% CI: 0.83-2.32, P = 0.21). Compared to the placebo arm, the reduction in the hazard of recurrence or death in the pembrolizumab arm was greater (P = 0.028) after an onset of an irAE (HR = 0.37, 95% CI: 0.24-0.57) than without/before an irAE (HR = 0.61, 95% CI: 0.49-0.77). Conclusions: In the EORTC 1325/KEYNOTE-054 study conducted in high-risk stage III melanoma pts, the occurrence of an irAE was strongly associated with a longer RFS in those treated with pembrolizumab, but not with placebo. Clinical trial information: NCT02362594.

9593 Background: With the approval of adjuvant therapy for stage III melanoma, accurate staging in melanoma patients is important more than ever to prevent over- or undertreatment. Sentinel node biopsy (SNB) is an accurate staging tool, yet the presence of capsular nevi (CN) can lead to a false positive diagnosis. We compared positive SNB and CN patient outcomes and aimed to evaluate the cause of false positive SNB and discern diagnostic pitfalls in their evaluation. Methods: Retrospective analysis of AJCC 7th Edition stage IIIA melanoma patients (N1-2a, non-ulcerated primary tumor) who were treated at our institute between 2000 and 2015. SNB slides were reviewed for this study by an expert melanoma pathologist. Baseline characteristics were assessed for SN+ and CN+ patients. Concordance rates for SNB evaluation before and after revision were documented and diagnostic pitfalls were discerned. Results: 169 patients were diagnosed, 10 could not be reviewed due to lack of evaluable slides. Of these 159 cases, 14 patients originally diagnosed with metastatic melanoma were shown to have capsular nevi (8.8%). Another 2 patients were shown to have melanophages that were incorrectly interpreted as metastases (1.3%). Thus, 10.1% was considered false positive after revision. In 14 patients the SN tumor burden was originally reported > 1 mm, but turned out to have < 1 mm SN tumor burden. 4 patients originally reported as SN tumor burden < 1 mm before revision turned out to have > 1 mm SN tumor burden. These 32 patients (20%) might have potentially been over- or undertreated in the current era of adjuvant therapy for stage III melanoma. Conclusions: False positive sentinel node results in melanoma are real, they can occur for a number of reasons, but distinguishing metastatic melanoma from benign capsular nevi and melanophages can be a diagnostic challenge. We plead for an expert pathologists’ review in any case, but certainly when using the SNB+ results to determine treatment consequences for SN+ melanoma patients.

Purpose Prognosis in patients with sentinel node (SN) –positive melanoma correlates with several characteristics of the metastases in the SN such as size and site. These factors reflect biologic behavior and may separate out patients who may or may not need additional locoregional and/or systemic therapy. Patients and Methods Between 1993 and 2008, 1,080 patients (509 women and 571 men) were diagnosed with tumor burden in the SN in nine European Organisation for Research and Treatment of Cancer (EORTC) melanoma group centers. In total, 1,009 patients (93%) underwent completion lymph node dissection (CLND). Median Breslow thickness was 3.00 mm. The median follow-up time was 37 months. Tumor load and tumor site were reclassified in all nodes by the Rotterdam criteria for size and in 88% by the Dewar criteria for topography. Results Patients with submicrometastases (< 0.1 mm in diameter) were shown to have an estimated 5-year overall survival rate of 91% and a low nonsentinel node (NSN) positivity rate of 9%. This is comparable to the rate in SN-negative patients. The strongest predictive parameter for NSN positivity and prognostic parameter for survival was the Rotterdam-Dewar Combined (RDC) criteria. Patients with submicrometastases that were present in the subcapsular area only, had an NSN positivity rate of 2% and an estimated 5- and 10-year melanoma-specific survival (MSS) of 95%. Conclusion Patients with metastases < 0.1 mm, especially when present in the subcapsular area only, may be overtreated by a routine CLND and have an MSS that is indistinguishable from that of SN-negative patients. Thus the RDC criteria provide a rational basis for decision making in the absence of conclusions provided by randomized controlled trials.

TPS9606Background: The outcome of high risk stage III melanoma patients (pts) is poor, with a 5 year overall survival (OS) rate of < 50%. Adjuvant (adj) high dose IPI significantly improves 5 year progression free survival (PFS) and OS and adj NIVO improves the median PFS even more. In stage IV pts, the combination of IPI and NIVO improves response rates (RR) and PFS compared to monotherapy, but at cost of higher toxicity. Neo-adjuvant (neoadj) treatment may be a favorable approach as immune checkpoint inhibition (ICI) is of greatest value at the moment of TCR triggering and therefore dependent on the amount of antigen. The phase Ib OpACIN study compared neoadj versus adj IPI plus NIVO. The pathological RR (pRR) was 80% in the neoadj arm, and to date after a median follow-up of 24 months, none of the responders has relapsed, while 4/10 pts have relapsed in the adj arm. Moreover, pts in the neoadj arm expanded more tumor-resident TCR clones than adj treated pts. Neoadj IPI+NIVO was feasible, but toxicity w...

Purpose We have shown that ultrasound (US) -guided fine needle aspiration cytology (FNAC) can accurately identify the sentinel node (SN). Moreover, US-guided FNAC before the surgical SN procedure could identify up to 65% of all SN metastases. Herein we analyzed in detail the different US morphologic patterns of SN metastases. Patients and Methods From July 2001 to December 2007, a total of 650 patients with melanoma scheduled for sentinel lymph node dissection were examined. We present the first 400 with sufficient follow-up (mean 40, median 39 months). Several morphologic characteristics were scored. In case of suspicious/clearly malignant US patterns a FNAC was performed. The final histology was considered the gold standard. Results Median Breslow was 1.8 mm. The sensitivity and positive predictive value of the most important factors were: peripheral perfusion (PP) present (77% and 52%, respectively), loss of central echoes (LCE; 60% and 65% respectively), and balloon shape (BS; 30% and 96% respectively). Together these factors have a sensitivity of 82% and PPV of 52% (P < .001). PP identified more patients with lower volume disease. PP and combined BS and LCE were independent prognostic factors for survival (hazard ratio, 2.19; P < .015; and hazard ratio, 5.50; P < .001, respectively). Conclusion Preoperative US and FNAC can identify 65% of SN metastases and thus reduce the need for surgical SN procedures. Peripheral perfusion is an early sign of involvement and of crucial importance to achieve a high identification rate. Balloon shape and loss of central echoes are late signs of metastases. We recommend US evaluation to identify those patients, who can directly proceed to a complete lymph node dissection after a positive US-guided FNAC of the SN.

Purpose Sentinel node (SN) status is the most important prognostic factor for overall survival (OS) for patients with stage I/II melanoma, and the role of the SN procedure as a staging procedure has long been established. However, a less invasive procedure, such as ultrasound (US) -guided fine-needle aspiration cytology (FNAC), would be preferred. The aim of this study was to evaluate the accuracy of US-guided FNAC and compare the results with histology after SN surgery was performed in all patients. Patients and Methods Four hundred consecutive patients who underwent lymphoscintigraphy subsequently underwent a US examination before the SN procedure. When the US examination showed a suspicious or malignant pattern, patients underwent an FNAC. Median Breslow thickness was 1.8 mm; mean follow-up was 42 months (range, 4 to 82 months). We considered the US-guided FNAC positive if either US and/or FNAC were positive. If US was suggestive of abnormality, but FNAC was negative, the US-guided FNAC was considered negative. Results US-guided FNAC identified 51 (65%) of 79 SN metastases. Specificity was 99% (317 of 321), with a positive predictive value of 93% and negative predictive value of 92%. SN-positive identification rate by US-guided FNAC increased from 40% in stage pT1a/b disease to 79% in stage pT4a/b disease. US-guided FNAC detected SN tumors more than 1.0 mm in 86% of cases, SN tumors of 0.1 to 1.0 mm in 46% of cases, and SN tumors less than 0.1 mm in 23% of cases. Estimated 5-year OS rates were 92% for patients with negative US-guided FNAC results and 51% for patients with positive results. Conclusion US-guided FNAC of SNs is highly accurate. Up to 65% of the patients with SN-positive results in our institution could have been spared an SN procedure.

Purpose This study analyzes (1) the value of tyrosinase reverse-transcriptase polymerase chain reaction (RT-PCR) of aspirates obtained by ultrasound-guided fine-needle aspiration cytology (US-FNAC) of sentinel nodes (SNs) in patients with melanoma before sentinel lymph node biopsy (SLNB) and (2) the value of RT-PCR of blood samples of all SLNB patients. Patients and Methods Between 2001 and 2003, 127 patients with melanoma (median Breslow depth, 2.1 mm) underwent SLNB. FNAC was performed in all SNs of all patients pre- and post-SLNB. The aspirates were partly shock-frozen for RT-PCR and were partly used for standard cytology. Peripheral blood was collected at the time of SLNB and at every outpatient visit thereafter. Results Thirty-four (23%) of 120 SNs were positive for melanoma. SN involvement was predicted by US-FNAC with a sensitivity of 82% and a specificity of 72%. Additional tyrosinase RT-PCR revealed the same sensitivity of 82% and a specificity of 72%. At a median follow-up time of 40 months from first blood sample, peripheral-blood RT-PCR was a significant independent predictor of disease-free survival (DFS) and overall survival (OS; P < .001). Conclusion US-FNAC is highly accurate and eliminates the need for SLNB in 16% of all SLNB patients. RT-PCR of the aspirate or excised SN does not improve sensitivity or specificity. RT-PCR of blood samples predicts DFS and OS.

DOI: 10.1200/JCO.2015.63.1333 Dedicated to Christo, to my children Michael (with his girlfriend Helene), Lukas, and Christina Helena, and to Nina. Next to me sounds the buzzing of my Lympha Press machine, which substitutes for the constant visits of the physiotherapist who performs the lymph drainage. This gives me more freedom, and we have more privacy at home. I can use the machine whenever I need it, and my 5-year-old daughter Christina can be around me without problems. Today, the little one is at school. This gives me time to work a bit, but as I lie on the bed like a large insect in Michelin tire–like blow-up pants, with several liters of ascites in my abdomen, it is difficult to use the laptop. Thus I decide to get up to speed with reading my journals, which I have not done for months. As an oncologist, I read Journal of Clinical Oncology (JCO) with feelings of pride and empathy. I am proud of my own articles in JCO, but during the last few months, I have not been able to touch it. Why? If you change sides of the bed—moving from physician to patient—you begin to change your style of reading. I realized today that I was more interested in the Art of Oncology articles than in the Original Report articles. I realized that except for stories of pediatric oncologists who lose children to cancer, few if any of the Art of Oncology articles seemed to tell the story of an oncologist with cancer. Does this mean it is unprofessional to get cancer? Or that it is unprofessional to write about that experience? Original Reports seem to always follow the same template: an Introduction tells why a particular cancer is so horrible or that there is no effective treatment for a specific stage of that cancer. In the Results section, the focus is usually on rates of response and relapse-free survival, all of which are appropriate to read and write about when seen through the eyes of an oncologist. However, these topics are seen differently through the eyes of a patient: what does an increase in relapse-free survival from 1.5 to 5.5 months mean if overall survival is the same in the end? When I looked at my own malignant lymph node on a rainy Sunday in October 2012 on the screen of my own high-end ultrasound machine, I immediately assumed a certain professional detachment. I saw that all the criteria I had developed for determining whether lymph nodes are malignant could be applied to my own. The only logical next step was to perform an ultrasound-guided fineneedle aspiration on myself. To prevent any colleagues from finding my personal pictures in the ultrasound machine, I gave myself the fictive name of a character from a German children’s book I had always loved. I saved the pictures and film clips because they were exemplary, and I would be able to use them for my next presentations and courses. Unfortunately, the cytology revealed undifferentiated cancer. From that point until the time of this writing, I have wrestled with a number of issues: why did the doctors I visited 6 months before my self-diagnosis not pick up on my complaints of abdominal pain, fatigue, and a yellow complexion? Why did they not want to do genetic testing when I asked them to do so years ago, because seven relatives (all maternal side) had developed breast cancer? I was refused because my mother and sister did not have breast or ovarian cancer. Now, in the final stage of my ovarian cancer, I have made peace with all of this. After 12 hours of massive debulking surgery, weeks of parenteral nutrition on the intensive care unit as a result of vomiting, and courses of chemotherapy, I had again a small glimmer of hope. First results of several anti-cell death protein 1 (PD-1) and anti-cell death protein ligand 1 (PD-L1) treatments for different types of cancer were being presented and published, and they seemed encouraging. Perhaps as a result of my professional background, I was aware of these developments and was able to gain access to a clinical trial with anti-PD-1 in Paris with the help of my Dutch mentor who nowadays resides there. My tumor even expressed PD-L1, and I was allowed to be included in the study as one of the last patients. Despite all my efforts, my disease progressed on treatment. It might seem contradictory, but as I write this, it is now a happy time. My eldest two children, Michael and Lukas (26 and 21 years old), have been catapulted into adult roles as a result of my cancer diagnosis. They patiently wait with me at the doctor’s office, make fun of me when things get too serious, and give me good advice, like when I should cut my too-long hair. We are even able to talk about JOURNAL OF CLINICAL ONCOLOGY A R T O F O N C O L O G Y VOLUME 33 NUMBER 35 DECEMBER 1

e21549 Background: Data on isolated limb perfusion (ILP) in elderly melanoma patients is scarce. We aimed to evaluate the efficacy and safety of ILP in our institutional cohort of melanoma patients. Methods: Retrospective analysis of AJCC stage IIIB/C melanoma patients who underwent ILP for locally advanced (unresectable) melanoma in-transit metastases of the limb between 2000 and 2016. Normothermic ILP (37-38 °C) was performed with either Melphalan or Melphalan and Tumor Necrosis Factor (TNF). Baseline characteristics, local toxicity (Wieberdink) and systemic toxicity, locoregional progression-free survival (PFS) and melanoma-specific survival (MSS) were assessed and prognostic factors for response to ILP, melanoma recurrence and survival were analyzed using univariable and multivariable analysis. Results: 88 patients were included. The overall response rate to ILP was 81%, with a complete response (CR) rate of 47%. Median overall locoregional PFS was 6 months, while patients with a CR had a median PFS of 16 months. Median overall MSS was 38 months. Two patients (2.3%) suffered Wieberdink IV toxicity (compartment syndrome), while no patients required amputation because of severe toxicity. Based on the median age of 70, we split patients into younger and elderly groups. Toxicity rates, response rates and locoregional PFS did not differ significantly between younger and elderly patients. CR was prognostic for improved locoregional PFS and MSS. Moreover, patients > 70 and patients with stage IIIC disease had a higher risk of melanoma-specific death. Conclusions: ILP has good and potential durable responses, but also less frequent and less severe toxicity than recently developed targeted or immune therapies, which can be reserved for progression to stage IV. This study showed that ILP is an effective and safe procedure for elderly patients ( > 70) with locally advanced melanoma of the limb.

9586 Background: The combination of IPI+NIVO induces high response rates and improved overall survival in late stage melanoma. T cell checkpoint inhibition is of greatest value at the moment of TCR triggering and therefore dependent on the amount of antigen present, indicating that adjuvant immunotherapy may work most efficiently, when initiated prior to surgery. Methods: Two-arm Phase 1b feasibility trial of 20 high risk AJCC stage IIIB and IIIC melanoma patients with palpable nodal disease receiving the combination of IPI 3mg/kg and NIVO 1mg/kg, either adjuvant four courses after surgery, or split two courses neo-adjuvant and two courses adjuvant. Results: In this update all 20 patients are evaluable. Neo-adjuvant application of IPI+NIVO was feasible and no surgery-associated adverse events were attributed to (neo-)adjuvant therapy. 18/20 patients had to stop early due to grade 3/4 toxicities. Neo-adjuvant IPI+NIVO reduced tumor load in 8/10 patients (3 pCR, 4 near-pCRs [minimal remaining micrometastases], 1 pPR [75% reduction], 1 SD and 1 PD). To date, after a median follow-up of 45 weeks (range 13-74), none of the responders in the neoadjuvant arm has relapsed. Relapse was observed for both non-responders within the neo-adjuvant arm and for 3 patients within the adjuvant arm. TCR sequencing and MHC tetramer-based analysis to compare the induction and expansion of tumor (neo-)antigen-specific T cell responses between both treatment arms are underway and will be presented. Conclusions: The combination of IPI+NIVO in the (neo-)adjuvant treatment setting for high risk stage III melanoma patients is feasible and induces very frequent responses. At the same time, severe grade 3/4 toxicity is more frequent than expected from stage IV melanoma patient study data. These results indicate that IPI+NIVO is a promising combination for neo-adjuvant treatment in stage III melanoma. Adjusted combination schemes are currently tested in the phase 2 OpACIN-neo trial, with the aim of reducing toxicity while preserving efficacy. Clinical trial information: NCT02437279.

9067 Background: Ultrasound (US) guided Fine Needle Aspiration Cytology (FNAC) is a common diagnostic tool in the work-up of other cancers, i.e. breast and thyroid. The results in melanoma (mel) we...

9097 Background: FNAC is used to determine nature of suspected lesions that are either clinically apparent (palpable) or discovered by imaging (non-palpable) with the same high sensitivity (Voit et...

e20035 Background: The surgical Sentinel Node (SN) is standard of care worldwide for the staging of AJCC stage I/II melanoma patients. It remains unclear if SN followed by early lymph node dissection in case of metastasis can lead to a survival benefit. Unlike other cancers, US-guided-FNAC has not been proven effective in melanoma. Aim was to analyze sensitivity (sens), specificity (spec), positive (PPV) and Negative (NPV) predictive values with use of the Berlin Morphology criteria, a low threshold for performing repeated FNAC and overnight cytology reports. Methods: Between 2001 and 2010 over 1,000 stage I / II consecutive melanoma pts have undergone US-FNAC prior to SN. All patients underwent lymphoscintigraphy prior to US-FNAC. Peripheral Perfusion (PP), Loss of Central Echoes (LCE), Balloon Shaped (BS) are the Berlin Morphology Criteria, which were registered. FNAC was performed (3-4x) in case any of these factors were seen. SN tumor burden was measured according to the Rotterdam Criteria. All patients underwent SN or LND in case of positive FNAC. Results: Mean/med Breslow thickness 2.59/1.57 mm (0.2 – 44 mm). Mean/med follow-up 39 / 32 months (0 – 115). Ulceration (ulc) present in 24 %. SN pos rate was 20.8 % (208 / 1000). 34.2% underwent FNAC. 8.9% was FNAC positive; this was 26% of all FNAC. Sens was 51%. Spec, PPV and NPV were 99%, 95% and 89%. Sensitivity was highest for T4 (76%) and in ulc (63%) tumors. PP, LCE, BS had sens of 51%, 37%, 24%. Sens of US-FNAC increased with increasing SN tumor burden (12% in < 0.1 mm to 61% in > 1 mm). PP was an early sign of metastasis (58% in < 0.1 mm mets). Threshold for pos FNAC was 0.4 mm in maximum diameter. 5-yr survival correlated to US-FNAC status (93% in neg, 51% in pos). 5-yr survival of pts with PP was 82% vs. 46% in BS vs. 92% in neg pts (P=0.001). Conclusions: US-FNAC according to the Berlin Morphology criteria can correctly identify half of the positive SNs, prior to the surgical SN procedure. PP is an early sign of metastasis, BS is a late sign. US-FNAC can significantly reduce the amount of unnecessary surgical SN procedures. US-FNAC sensitivity correlated with increasing T-stage, increasing SN tumor burden and ulc. US-FNAC can accurately predict survival.

e19022 Background: Melanoma of unknown primary site (MUP) is unique, not completely understood entity with nodal metastases as the most common clinical manifestation. The aim of this multicentric study was to assess genetic alterations in MUP with clinically detected nodal metastases in terms of clinicopathological features and prognosis. Methods: We analyzed contemporary series of 37 MUP patients (median age 51 years) after therapeutic lymphadenectomy - LND (period: 1996-2010, 20 – axillary, 16 - inguinal, 1 – other basin) not treated with BRAF inhibitors and performed molecular characterization of BRAF/NRAS mutational status in nodal metastases using direct sequencing of respective coding sequences. Median follow-up time was 37 months. Results: BRAF mutations were detected in 23 (63%) cases (21 V600E - 91%, 2 others - 9%), and mutually exclusive NRAS mutations in 3 (8%) cases (Q61K, Q61R, Q13R). Presence of BRAF mutation correlated with younger age of patients (median 47 vs 60 years for BRAF+ vs. BRAF-, p

8535 Background: US-guided-FNAC prior to surgical SN staging is emerging as a possible cost-effective addition to the staging of melanoma patients (pts). Formerly, sensitivity (sens) rates of lymph node US in melanoma were disappointing (20–40%). The introduction of the Berlin Morphology Criteria has significantly improved sens rates for US-FNAC (J Clin Oncol 2010;28(5):847-52). The aim of the current study was to report on 1000 patients the sens, specificity (spec), positive (PPV) and negative (NPV) predictive value rates of US-FNAC from our prospective database with prolonged follow-up. Methods: Since 2001, >1000 stage I / II consecutive melanoma pts have undergone US-FNAC prior to SN. All patients underwent lymphoscintigraphy. Peripheral Perfusion (PP), Loss of Central Echoes (LCE), Balloon Shaped (BS) were the Berlin Morphology Criteria which were registered. FNAC was performed in case of presence of any of these factors. SN tumor burden was measured according to the Rotterdam Criteria. All patients underwent SN or LND in case of positive FNAC. Results: Mean/median Breslow thickness was 2.56 / 1.57 mm (0.2 – 44 mm).Mean/median follow-up was 39 / 32 months (0 – 115). Ulceration was present in 24 %. SN positivity rates were 20 % (202 / 1000). Sens was 51 %. Spec, PPV and NPV were 99%, 91% and 89%. Sensitivity was highest for T4 tumors (77%). PP, LCE, BS had sens of 69%, 24%, 25%. SN tumor burden > 1 mm in largest diameter according to the Rotterdam Criteria was identified by US-FNAC in 86%. Threshold for positive FNAC was 0.4 mm in maximum diameter. Conclusions: The new criterion of Periferal Perfusion is of key importance to achieve the high sensitivity of US-FNAC according to the Berlin Morphology Criteria (J Clin Oncol 2010; 28:847-852) to identify lymph node metastases. Especially for T4 patients and in patients with advanced SN tumor burden it can reduce significantly the need for surgical SN staging. The EORTC Melanoma Group will launch the prospective validation study, USE FNAC, in 2012.