Background Sentinel lymph node biopsy for melanoma determines treatment and prognostic factors and improves disease-specific survival. To risk-stratify patients for sentinel lymph node biopsy consideration, Memorial Sloan Kettering Cancer Center and Melanoma Institute Australia developed nomograms to predict sentinel lymph node positivity. We aimed to compare the accuracy of these 2 nomograms. Methods A multi-institutional study of patients with melanoma receiving sentinel lymph node biopsy between September 2018 and December 2022 was performed. The accuracy of the 2 risk prediction tools in determining a positive sentinel lymph node biopsy was analyzed using receiver operating characteristic curves and area under the curve. Results In total, 532 patients underwent sentinel lymph node biopsy for melanoma; 98 (18.4%) had positive sentinel lymph node. Increasing age was inversely related to sentinel lymph node positivity (P < .01); 35.7% of patients ≤30 years had positive sentinel lymph node compared with 9.7% of patients ≥75 years. When we analyzed the entire study population, accuracy of the 2 risk prediction tools was equal (area under the curveMemorial Sloan Kettering Cancer Center: 0.693; area under the curveMIA: 0.699). However, Memorial Sloan Kettering Cancer Center tool was a better predictor in patients aged ≥75 years (area under the curveMemorial Sloan Kettering Cancer Center: 0.801; area under the curveMelanoma Institute Australia: 0.712, P < .01) but Melanoma Institute Australia tool performed better in patients with a higher mitotic index (mitoses/mm2 ≥2; area under the curveMemorial Sloan Kettering Cancer Center: 0.659; area under the curveMelanoma Institute Australia: 0.717, P = .027). Both models were poor predictors of sentinel lymph node positivity in young patients (age ≤30 years; area under the curveMemorial Sloan Kettering Cancer Center: 0.456; area under the curveMelanoma Institute Australia: 0.589, P = .283). Conclusion The current study suggests that the 2 risk stratification tools differ in their abilities to predict sentinel lymph node positivity in specific populations: Memorial Sloan Kettering Cancer Center tool is a better predictor for older patients, whereas Melanoma Institute Australia tool is more accurate in patients with a higher mitotic index. Both nomograms performed poorly in predicting sentinel lymph node positivity in young patients.

BACKGROUND:Despite differences in tumour behaviour and characteristics between duodenal adenocarcinoma (DAC), the intestinal (AmpIT) and pancreatobiliary (AmpPB) subtype of ampullary adenocarcinoma and distal cholangiocarcinoma (dCCA), the effect of adjuvant chemotherapy (ACT) on these cancers, as well as the optimal ACT regimen, has not been comprehensively assessed. This study aims to assess the influence of tailored ACT on DAC, dCCA, AmpIT, and AmpPB. PATIENTS AND METHODS:Patients after pancreatoduodenectomy for non-pancreatic periampullary adenocarcinoma were identified and collected from 36 tertiary centres between 2010 - 2021. Per non-pancreatic periampullary tumour type, the effect of adjuvant chemotherapy and the main relevant regimens of adjuvant chemotherapy were compared. The primary outcome was overall survival (OS). RESULTS:The study included a total of 2866 patients with DAC (n = 330), AmpIT (n = 765), AmpPB (n = 819), and dCCA (n = 952). Among them, 1329 received ACT, and 1537 did not. ACT was associated with significant improvement in OS for AmpPB (P = 0.004) and dCCA (P < 0.001). Moreover, for patients with dCCA, capecitabine mono ACT provided the greatest OS benefit compared to gemcitabine (P = 0.004) and gemcitabine - cisplatin (P = 0.001). For patients with AmpPB, no superior ACT regime was found (P > 0.226). ACT was not associated with improved OS for DAC and AmpIT (P = 0.113 and P = 0.445, respectively). DISCUSSION:Patients with resected AmpPB and dCCA appear to benefit from ACT. While the optimal ACT for AmpPB remains undetermined, it appears that dCCA shows the most favourable response to capecitabine monotherapy. Tailored adjuvant treatments are essential for enhancing prognosis across all four non-pancreatic periampullary adenocarcinomas.

The objective of the study was to use a deep learning model to differentiate between benign and malignant sentinel lymph nodes (SLNs) in patients with breast cancer compared to radiologists' assessments. Seventy-nine women with breast cancer were enrolled and underwent lymphosonography and contrast-enhanced ultrasound (CEUS) examination after subcutaneous injection of ultrasound contrast agent around their tumor to identify SLNs. Google AutoML was used to develop image classification model. Grayscale and CEUS images acquired during the ultrasound examination were uploaded with a data distribution of 80% for training/20% for testing. The performance metric used was area under precision/recall curve (AuPRC). In addition, 3 radiologists assessed SLNs as normal or abnormal based on a clinical established classification. Two-hundred seventeen SLNs were divided in 2 for model development; model 1 included all SLNs and model 2 had an equal number of benign and malignant SLNs. Validation results model 1 AuPRC 0.84 (grayscale)/0.91 (CEUS) and model 2 AuPRC 0.91 (grayscale)/0.87 (CEUS). The comparison between artificial intelligence (AI) and readers' showed statistical significant differences between all models and ultrasound modes; model 1 grayscale AI versus readers, P = 0.047, and model 1 CEUS AI versus readers, P < 0.001. Model 2 r grayscale AI versus readers, P = 0.032, and model 2 CEUS AI versus readers, P = 0.041. The interreader agreement overall result showed kappa values of 0.20 for grayscale and 0.17 for CEUS. In conclusion, AutoML showed improved diagnostic performance in balance volume datasets. Radiologist performance was not influenced by the dataset's distribution.

4167 Background: Cancer arising in the periampullary region can be anatomically classified in pancreatic ductal adenocarcinoma (PDAC), distal cholangiocarcinoma (dCCA), duodenal adenocarcinoma (DAC) and ampullary adenocarcinoma (AAC). Based on histopathology, the AAC is currently subdivided in the intestinal (AmpIT) and pancreatobiliary (AmpPB) subtype. Despite close anatomical resemblance, it is unclear how the ampullary subtypes relate to the remaining periampullary cancers in tumor characteristics and behavior. Methods: This is an international multicenter cohort study, including patients after curative intent resection for periampullary cancer retrieved from 45 centers (from Europe, USA, Asia and Canada) between 2010 and 2021. Pre-operative CA19-9, differences in pathology and postoperative pancreatic fistula (POPF, grade B/C), 8-year overall survival (OS) and disease-free interval (DFI) were compared between DAC, AmpIT, AmpPB, dCCA and PDAC. Results: Overall, 3809 patients were included of which 348 DAC, 774 AmpIT, 848 AmpPB, 1036 dCCA and 803 PDAC. The best 8-years overall survival was found in patients with AmpIT and DAC (49.8% and 47.9%), followed by AmpPB (34.9%, p<0.001), dCCA (26.4%, p=0.020) and last, PDAC (12.9%, P<0.001). The best 8-years DFI was measured for AmpIT (62.7%) and the worst for both dCCA and PDAC (24.9% and 21.8%). This pattern correlated with the height of the pre-operative CA19-9 but not with the pathology, in which the largest and most progressed tumors were found in DAC. The incidence of POPF was lower in the pancreas-related tumor types (AmpPB 18.5%, PDAC 8.3%) compared to the non-pancreas related types (DAC 27.3%, AmpIT 25.5%, dCCA 27.6%). Conclusions: Despite the close anatomic relation of the five periampullary cancers, this study shows that there are prognostic and clinically relevant differences in terms of preoperative blood markers, pathology, complications, long-term survival, and recurrence. More characteristics are shared between DAC and AmpIT and between AmpPB and dCCA than between the two ampullary subtypes. Instead of using collective definitions for “periampullary cancers”, this study emphasizes the need for individual evaluation of each histopathological periampullary subtype with the ampullary subtypes as individual entities in future studies.

BackgroundAlthough melanoma survival rates have improved in recent years, survivors remain at risk of recurrence, second primary cancers, and keratinocyte carcinomas. The National Comprehensive Cancer Network recommends skin examinations by a physician every 3 to 12 months. Regular thorough skin self-examinations (SSEs) are recommended for survivors of melanoma to promote the detection of earlier-stage, thinner melanomas, which are associated with improved survival and lower treatment costs. Despite their importance, less than a quarter of survivors of melanoma engage in SSEs. ObjectivePreviously, our team developed and evaluated a web-based, fully automated intervention called mySmartSkin (MSS) that successfully improved SSE among survivors of melanoma. Enhancements were proposed to improve engagement with and outcomes of MSS. The purpose of this paper is to describe the rationale and methodology for a type-1 hybrid effectiveness-implementation randomized trial evaluating the enhanced MSS versus control and exploring implementation outcomes and contextual factors. MethodsThis study will recruit from state cancer registries and social media 300 individuals diagnosed with cutaneous malignant melanoma between 3 months and 5 years after surgery who are currently cancer free. Participants will be randomly assigned to either enhanced MSS or a noninteractive educational web page. Surveys will be collected from both arms at baseline and at 3, 6, 12, and 18 months to assess measures of intervention engagement, barriers, self-efficacy, habit, and SSE. The primary outcome is thorough SSE. The secondary outcomes are the diagnosis of new or recurrent melanomas and sun protection practices. ResultsMultilevel modeling will be used to examine whether there are significant differences in survivor outcomes between MSS and the noninteractive web page over time. Mixed methods will evaluate reach, adoption, implementation (including costs), and potential for maintenance of MSS, as well as contextual factors relevant to those outcomes and future scale-up. ConclusionsThis trial has the potential to improve outcomes in survivors of melanoma. If MSS is effective, the results could guide its implementation in oncology care and nonprofit organizations focused on skin cancers. International Registered Report Identifier (IRRID)RR1-10.2196/52689

4005 Background: If 5FU/Capecitabine sensitized radiotherapy (RT) is beneficial in the adjuvant (adj) management of PA after adj chemotherapy (chemo) is controversial. NRG/RTOG 0848 was designed to address this issue. Methods: This was a 2 step NCTN randomized (rndmzd) trial. Step 1 rndmzd patients (pts) to 5 cycles of gemcitabine +/- Erlotinib. Step 2 rndmzd pts to a 6th cycle of the same chemo +/- 5FU/Capecitabine with 50.4 Gy in 28 fractions RT (chemo+CRT). Step 1 eligibility included: R0/R1 resection, M0, ECOG PS 0-1, CA19-9≤180. Step 2 eligibility included > 4 cycles chemo (gem, gem combo, (m)FOLFIRINOX). RT included real time 3D/IMRT treatment (RX) plan review, scoring, and approval. At Step 2, pts stratified by nodal status (+ vs -), CA19-9 (≤90 vs > 90-180), surgical margins (R0 vs R1), and adjuvant chemo. Primary endpoint was OS. Secondary endpoints are DFS and AEs (CTCAEv4). Assuming 17 months median OS (chemo) and hypothesized 22.5 months (chemo+CRT), sample size was 354 pts (HR = 0.76, 80% power, 1-sided α = 0.05, 316 OS events). Due to lower than projected event rate, trial was amended to report at the earlier of (a) 316 observed OS events or (b) 5 years of follow-up time from Step 2 accrual closure (265 OS events, 72% power, same α). OS and DFS were estimated by Kaplan-Meier and arms compared using log-rank test. Multivariable analyses (MVA) used Cox proportional hazards models. Results: Accrual began 11/2009; closed 10/2018. 354 pts rndmzd (174 chemo, 180 chemo+CRT). Median follow-up for all & alive pts = 2 & 7 years, respectively, with 270 OS events. Median age 63, 45% female, 81% white, 13% AA. 83% R0, 26% node negative, 96% CA19-9 < 90. 13% of chemo+CRT pts did not receive RT. AEs were comparable (grade 4: 10% [chemo] vs 11% [chemo+CRT] and 1 grade 5 AE in each arm). Univariate OS/DFS results shown in Table. In initial MVA, RX, CA19-9, surgical margins were not statistically significantly associated with OS or DFS, but nodal status (OS, DFS) and race (OS) were. In further analyses, significant interactions were found between RX and nodal status for both OS and DFS. Node negative pts treated with chemo+CRT had better outcome than chemo pts; node positive pts did not (Table). Conclusions: Chemo+CRT did not improve OS overall, but did improve DFS. Both OS and DFS were improved with Chemo+CRT in node negative pts. Chemo+CRT did not increase Gr 4 or 5 AEs compared to chemo. Clinical trial information: NCT01013649 . [Table: see text]

Background: The autologous tumor lysate, particle-loaded, dendritic cell (TLPLDC) vaccine is produced from dendritic cells (DC) loaded ex vivo with autologous tumor lysate (TL). TLPLDC has been shown to decrease recurrence in resected Stage III/IV melanoma patients in a Phase IIb trial. The TL particle only (TLPO) vaccine is produced by loading of yeast cell wall particles with autologous TL and direct injection allowing for in vivo DC loading. We have compared the TLPO and TLPLDC vaccines in an embedded Phase I/IIa trial of a larger Phase IIb trial of the TLPLDC vaccine. Methods: Patients rendered clinically disease-free after surgery were randomized 2:1 to receive the TLPO or TLPLDC vaccine and followed for recurrence and death. Patients had scheduled intradermal inoculations at 0, 1, 2, 6, 12, and 18 months after enrollment. Kaplan-Meier and log-rank analysis were used to compare disease-free survival (DFS) and overall survival (OS) in an intention-to-treat (ITT) analysis. Results: Sixty-three patients were randomized, 43 TLPO and 20 TLPLDC. Patients randomized to the TLPO arm were more likely to be female (37.2% vs. 10.0 %, p = 0.026), but otherwise no significant clinicopathological differences were identified. No differences in related adverse events (AE) were found between treatment arms. At a median follow-up of 20.5 months, the DFS (60.8% vs. 58.7 %, p = 0.714) and OS (94.6% vs. 93.8 %, p = 0.966) were equivalent between the TLPO and TLPLDC groups, respectively. No statistical differences were found in subgroup analyses between vaccine types, which accounted for receipt of immunotherapy and the use of G-CSF pre-blood draw. Conclusions: In a randomized, double-blind Phase I/IIa trial, there were no differences in DFS or OS in resected Stage III/IV melanoma patients receiving adjuvant TLPO versus TLPLDC vaccines. Given manufacturing advantages, further efficacy testing of TLPO is warranted in a Phase III trial.

Melanoma, with increasing annual incidence worldwide, is one of the deadliest cutaneous malignancies often treated with immunotherapy, which has led to commonly encountered immune-related adverse events (irAEs) including bullous pemphigoid (BP). The relation between an individual's HLA inheritance and risk in development of BP is well studied. The development of BP as an irAE in melanoma patients receiving immunotherapy is also well reported and considered to be related to the expression of BP180 in malignant melanoma. In presented work, the association of enriched presence of certain HLA alleles and haplotypes that increase risk of developing BP as an irAE has been investigated in seven cutaneous cancer patients who presented with immunotherapy-associated BP, diagnosed by skin biopsy and positive serology for anti-BP 180 autoantibody. Their HLA typing revealed an enriched presence of certain BP-related HLA alleles, and haplotypes, particularly HLA-DQB1*03:01, which was present in all seven patients, whilst its occurrence in the general population is very low. Additionally, some patients had more than one and in a few patients had the whole haplotype of BP related HLA. Not only can such enriched presence be considered a role in risk stratification to initiate or continue immunotherapy, but it may shed a light to existing disease genotypes that can become unmasked to phenotypic clinically presenting BP due to immunotherapy rather than just a random AE.

Supplemental Figure 1: PLX8394 does not alter CRAF or c-Src phosphorylation; Supplemental Figure 2. Treatment of xenograft-derived tumor pieces in an ex vivo explant system; Supplemental Figure 3: Patient sample extended data; Supplemental Figure 4: Parallel ex vivo dosing and homodimerization of vemurafenib and PLX8394 treatments; Supplemental Figure 5: PBRT characterization; Supplemental Figure 6: Targeting AKT and PDGFRβ in PBRT cell lines does not resensitize to PLX8394; Supplemental Figure 7: BRAF V600E splice variant analysis of PBRT cell lines and signaling and growth response of second-line therapies; Supplemental Methods.

Supplementary Figures and Legends.

BACKGROUND:The tumor lysate, particle-loaded, dendritic cell (TLPLDC) vaccine is made by ex vivo priming matured autologous dendritic cells (DCs) with yeast cell wall particles (YCWPs) loaded with autologous tumor lysate (TL). The tumor lysate, particle only (TLPO) vaccine uses autologous TL-loaded YCWPs coated with silicate for in vivo DC loading. Here we report the 36-month prespecified analyses of this prospective, randomized, double-blind trial investigating the ability of the TLPO and TLPLDC (±granulocyte-colony stimulating factor (G-CSF)) vaccines to prevent melanoma recurrence in high-risk patients.METHODS:Patients with clinically disease-free stage III/IV melanoma were randomized 2:1 initially to TLPLDC versus placebo (n=124) and subsequently TLPO versus TLPLDC (n=63). All patients were randomized and blinded; however, the placebo control arm was replaced in the second randomization scheme with another novel vaccine; some analyses in this paper therefore reflect a combination of the two randomization schemes. Patients receiving the TLPLDC vaccine were further divided by their method of DC harvest (with or without G-CSF pretreatment); this was not randomized. The use of standard of care checkpoint inhibitors was not stratified between groups. Safety was assessed and Kaplan-Meier and log-rank analyses compared disease-free (DFS) and overall survival (OS).RESULTS:After combining the two randomization processes, a total of 187 patients were allocated between treatment arms: placebo (n=41), TLPLDC (n=103), or TLPO (n=43). The allocation among arms created by the addition of patients from the two separate randomization schemes does not reflect concurrent randomization among all treatment arms. TLPLDC was further divided by use of G-CSF in DC harvest: no G-CSF (TLPLDC) (n=47) and with G-CSF (TLPLDC+G) (n=56). Median follow-up was 35.8 months. Only two patients experienced a related adverse event ≥grade 3, one each in the TLPLDC+G and placebo arms. DFS was 27.2% (placebo), 55.4% (TLPLDC), 22.9% (TLPLDC+G), and 60.9% (TLPO) (p<0.001). OS was 62.5% (placebo), 93.6% (TLPLDC), 57.7% (TLPLDC+G), and 94.6% (TLPO) (p=0.002).CONCLUSIONS:The TLPO and TLPLDC (without G-CSF) vaccines were associated with improved DFS and OS in this clinical trial. Given production and manufacturing advantages, the efficacy of the TLPO vaccine will be confirmed in a phase 3 trial.TRIAL REGISTRATION NUMBER:NCT02301611.

Cell culture-Inhibitor-Sequencing-Additional siRNA-Statistical analysis

objective of the work described here was to evaluate the efficacy of lymphosonography in identi-fying sentinel lymph nodes (SLNs) in patients with breast cancer undergoing surgical excision. Of the 86 individ-uals enrolled, 79 completed this institutional review board-approved study. Participants received subcutaneous 1.0-mL injections of ultrasound contrast agent (UCA) around the tumor. An ultrasound scanner with contrast-enhanced ultrasound (CEUS) capabilities was used to identify SLNs. Participants were administered with blue dye and radioactive tracer to guide SLN excision as standard-of-care. Excised SLNs were classified as positive or negative for the presence of blue dye, radioactive tracer and UCA, and sent for pathology. Two hundred fifty-two SLNs were excised; 158 were positive for blue dye, 222 were positive for radioactive tracer and 223 were positive for UCA. Comparison with blue dye revealed accuracies of 96.2% for radioactive tracer and 99.4% for lympho-sonography (p > 0.15). Relative to radioactive tracer, blue dye had an accuracy of 68.5%, and lymphosonography achieved 86.5% (p < 0.0001). Of 252 SLNs excised, 34 were determined to be malignant by pathology; 18 were positive for blue dye (detection rate = 53%), 23 for radioactive tracer (detection rate = 68%) and 34 for UCA (detection rate = 100%) (p < 0.0001). Lymphosonography was similar in accuracy to radioactive tracer and higher in accuracy than blue dye in identifying SLNs. All 34 malignant SLNs were identified by lymphosonogra-phy. (E-mail: flemming.forsberg@jefferson.edu)(c) 2022 World Federation for Ultrasound in Medicine & Biology. All rights reserved.

ObjectivesThis study evaluated the efficacy of lymphosonography in the identification of sentinel lymph nodes (SLNs) in post neoadjuvant chemotherapy patients with breast cancer scheduled to undergo surgical excision.MethodsSeventy‐nine subjects scheduled for breast cancer surgery with SLN excision completed this IRB‐approved study, out of which 18 (23%) underwent neoadjuvant chemotherapy before surgery. Subjects underwent percutaneous Sonazoid (GE Healthcare) injections around the tumor area for a total of 1.0 mL. Lymphosonography was performed using CPS on an S3000 HELX scanner (Siemens Healthineers) with a linear probe. Subjects received blue dye and radioactive tracer as part of their standard of care. Excised SLNs were classified as positive or negative for the presence of blue dye, radioactive tracer and Sonazoid. The results were compared between methods and pathology findings.ResultsSeventy‐two SLNs were surgically excised from 18 subjects, 29 were positive for blue dye, 63 were positive for radioactive tracer and 57 were positive for Sonazoid. Comparison with blue dye showed that both radioactive tracer and lymphosonography achieved an accuracy of 53% (P > .50). Comparison with radioactive tracer showed that blue dye had an accuracy of 53%, while lymphosonography achieved an accuracy of 67% (P < .01). Of the 72 SLNs, 15 were determined malignant by pathology; the detection rate was 47% for blue dye (7/15), 67% for radioactive tracer (10/15) and 100% for lymphosonography (15/15) (P < .001).ConclusionsLymphosonography achieved similar accuracy as radioactive tracer and higher accuracy than blue dye for identifying SLNs. The 15 SLNs positive for malignancy were all identified by lymphosonography.

Background: Immunotherapy is a lifesaving treatment but not without side effects; almost 60% of patients experience immune-related adverse events (irAEs), and cutaneous adverse events are some of the most commonly encountered.1 Currently, the risk factors for developing irAEs are poorly understood, and there is no mechanism to predict who and who will not develop these potentially fatal irAEs. It is postulated that patients with preexisting autoimmune diseases are more likely to have flares on immunotherapy rather than developing a de novo autoimmune condition.2 Studies have suggested that HLA inherence may play a role in irAEs; this area remains under investigation.3-5 Herein we report six patients with irAE for whom retrospective HLA-typing revealed multiple disease-related-HLA alleles, and none had a prior history of such immunologic condition. The patients developed a range of irAE including bullous pemphigoid, pemphigus, thyroiditis, vitiligo, and hepatitis all of which may have been predicted by their HLA subtypes. The authors propose that the irAEs may be associated with the patients' genetic propensity toward such auto-immune conditions which became unmasked by immunotherapy. Notably, two of the thyroiditis patients share HLA-A*03:01, HLA Bw4 and all three thyroiditis patients shared DRB4*01:03, while the two BP patients share HLA DPA1*01:03, potentially representing at risk alleles. The presented work may highlight the value of using HLA subtyping as a risk stratifying tool and may help providers with a risk-benefit analysis for the decision to start immunotherapy. Our work is one step towards a collective endeavor towards what precision medicine may one day provide.