Faculty Bibliography

The prognostic value of mitotic rate in melanoma is increasingly recognized, particularly in thin melanoma in which the presence or absence of a single mitosis/mm can change staging from T1a to T1b. Still, accurate mitotic rate calculation (mitoses/mm) on hematoxylin and eosin (H&E)-stained sections can be challenging. Antimonoclonal mitotic protein-2 (MPM-2) and antiphosphohistone-H3 (PHH3) are 2 antibodies reported to be more mitosis-specific than other markers of proliferation such as Ki-67. We used light microscopy and computer-assisted image analysis software to quantify MPM-2 and PHH3 staining in melanoma. We then compared mitotic rates by each method with conventional H&E-based mitotic rate for correlation with clinical outcomes. Our study included primary tissues from 190 nonconsecutive cutaneous melanoma patients who were prospectively enrolled at New York University Langone Medical Center with information on age, gender, and primary tumor characteristics. The mitotic rate was quantified manually by light microscopy of corresponding H&E-stained, MPM-2-stained, and PHH3-stained sections. Computer-assisted image analysis was then used to quantify immunolabeled mitoses on the previously examined PHH3 and MPM-2 slides. We then analyzed the association between mitotic rate and both progression-free and melanoma-specific survival. Univariate analysis of PHH3 found significant correlation between increased PHH3 mitotic rate and decreased progression-free survival (P=0.04). Computer-assisted image analysis enhanced the correlation of PHH3 mitotic rate with progression-free survival (P=0.02). Regardless of the detection method, neither MPM-2 nor PHH3 offered significant advantage over conventional H&E determination of mitotic rate.

Background: Manipulation of key epigenetic regulators in melanoma proliferation is emerging as a new therapeutic strategy. Bromodomain-containing proteins such as the extraterminal domain (BET) family are components of transcription factor complexes and determinants of epigenetic memory. We investigated the expression of BRD4, a BET family member in melanoma cell lines and tissues, and the effects of its inhibition with the small molecule compounds MS436 and MS417 in in vitro and in vivo models of melanoma. Methods: BRD2 and BRD4 expression were analyzed by immunohistochemistry. We tested the effects of pharmacological or RNAi-mediated inhibition of BRD4 in melanoma cells using crystal violet-based assays for proliferation/colony formation and flow-cytometry for cell cycle analysis. The molecular effects of BRD4 suppression were examined using RNA sequencing, Real-Time quantitative PCR and western blots for p27, p21, MYC, ERK1 and SKP2. In the in vivo xenograft experiments NOD/SCID/IL2R-/-mice were injected with melanoma cells and treated with MS417. Statistical significance was determined by unpaired t-test (GraphPad). Results: BRD4 was found significantly upregulated in primary and metastatic melanoma tissues compared to melanocytes and nevi (p<0.001). Treatment with BET inhibitors impaired melanoma cell proliferation in vitro and tumor growth and metastatic behavior in vivo, effects that were mostly recapitulated by individual silencing of BRD4. Rapidly after BET displacement, key cell cycle genes (SKP2, ERK1 and c-MYC) were downregulated concomitantly with the accumulation of CDK inhibitors (p21, p27), followed by melanoma cell cycle arrest. BET inhibitor efficacy was not influenced by BRAF or NRAS mutational status. Conclusions: Our results demonstrate for the first time a role for BRD4 in melanoma maintenance and support the role of BET proteins as novel targets in melanoma. Further investigation in the clinical setting is warranted

Background: Estimating the risk of recurrence in patients with melanoma is extremely challenging. Standard of care is AJCC staging system, but the accuracy and robustness of this method is still under development. We conducted a proof of concept study exploring the use of machine-learned Bayesian Belief Networks (ml-BBNs) using a miRNA profiled cohort of melanoma patients with extended follow up to create Bayesian Biological Systems Models (BBSMs). We sought to determine if ml-BBNs could describe the biological system and if we could use the model to identify new cases with higher risk of recurrence. Methods: Our study cohort consisted of 89 patients (42 of which recurred) with a median follow up time of 118 months, that were examined for 869 miRNAs. Prior to modeling we segmented the data into training data (72 cases/80%) and testing data (17 cases/20%) at random. We recursively trained ml-BBNs on the training set, using all miRNAs. We used the directed graph structure of the ml-BBNs to identify miRNAs that consistently had more connectivity and goodness of fit as determined by Bayesian Information Criteria (BIC) scoring. MiRNAs that were in the top 50 BIC-scoring nodes across all models were selected for use to train the recurrence BBSMs. To compensate for a small number, bootstrapping was used to increase the sample to 100 records. We then compared our test set cases to our recurrence model, and used a similarity scoring algorithm to evaluate the similarity of values in each test instance to our biological models. For comparison we also trained an ml-BBN using clinical data from the same cohort. We then evaluated the scores against known recurrence outcome using Receiver Operating Characteristic (ROC) curve analysis. Results: BIC-scoring analysis selected 35 miRNAs for use in BBSM modeling. Area Under the Curve (AUC) for detection of recurrence is 0.76 in the training set and 0.62 in the testing set, while the clinical data yielded an AUC of 0.5 in this cohort. Conclusions: Our data suggest !

Background: PP6C binds to regulatory units toaffect a number of important pathways including cell proliferation and DNA repair. Recently two independent groups reported for the first time the presence of somatic mutations in the PPP6C gene in ~10% of short term cultures and limited number of human melanoma tissues. However, the clinical or biological relevance of PPP6C mutations in melanoma patients is unknown. Our objectives were to examine the clinical relevance of PPP6C mutations in a well characterized cohort of melanoma specimens linked to extensive, prospectively-collected clinical information and to explore the functional consequence of different categories of mutations. Methods: Sanger dideoxy sequencing was performed on PCR-amplified DNA from macro-dissected FFPE tumors. Associations between PPP6C mutations and baseline characteristics, recurrence, survival, and BRAF/ NRAS mutational status were examined. The impact of mutations on binding PP6C regulatory units was assessed as well as the effect on additional downstream pathways. Results: 308 primary melanoma patients (118 Stage I, 92 Stage II, and 98 Stage III) were examined (median follow up: 5.3 years). 50 PPP6C mutations in 33 patients (10.7%) were identified with 11 tumors harboring more than one mutation. One mutation (R301C) was identified in 6 patients. PPP6C mutations occurred with similar frequencies across stages and showed no association with BRAF or NRAS mutations.Mutations were categorized into 3 groups: Mutations resulting in premature stop codon (n=9), those occurring in the active site (n=16) and others (n=8). 8/9 (89%) patients with stop mutations recurred and developed visceral metastases. Functional studies revealed that PPP6C mutants also behaved differently; some PPP6C mutations led to decreased binding to regulatory subunits, others, including the R301C mutation did not. Conclusions: Our data suggest that PPP6C mutation is an early event in melanoma progression and independent of BRAF or NRAS mutations. Data also!

Background: Although patient age at diagnosis is not currently included in guidelines for treatment of primary melanoma, several lines of evidence suggest that patient age is an important, yet understudied, factor when considering treatment options. Here, we attempt to address the limited knowledge of the impact of age on primary melanoma treatment. Methods: In a prospectively enrolled and followed-up cohort of melanoma patients at NYU, we used logistic regression models to evaluate the association between patient age at diagnosis, tumor baseline characteristics, including BRAF and NRAS mutation status, and likelihood of receiving and responding to adjuvant therapy. We examined adjuvant therapy effectiveness using recurrence and melanoma-specific survival as endpoints. Results: 444 primary melanoma patients were included in the study (median follow-up: 6.3 years; age range: 19-95 years). Age was categorized into three groups spanning the range of age at presentation: younger (19-45 years; 24%), middle (46-70 years; 50%), and older (71-95 years; 26%). Older patients were significantly more likely to have advanced stage, nodular subtype (P < 0.01, both variables), and BRAF wildtype tumors (P = 0.04). Controlling for these factors as well as gender, older patients experienced a higher risk of recurrence (HR older vs. younger 3.34, 95% CI 1.53-7.25; P < 0.01). Of the 128/444 (29%) patients who were eligible for adjuvant treatment (clinical stage IIB), only 67/128 (52%) received treatment. Using a propensity score that accounts for stage at presentation, patients in the middle age group were more likely to receive adjuvant therapy than those in the older group (OR 2.61, 95% CI 1.12-6.08; P = 0.03). In addition, a trend suggesting benefit from adjuvant therapy (defined as longer melanoma-specific survival) was observed only in the middle age group (P = 0.07). Conclusions: Our data suggest that older melanoma patients, despite having a significantly worse prognosis, are less likely to receive and bene!

Background: Patients with metastatic melanoma are eligible for BRAF inhibitor therapy if the BRAF V600E mutation can be identified in their tumor specimen. Patients lacking an available specimen for genotyping are unable to receive inhibitor therapy. We developed two mutation-specific genotyping platforms and tested their ability to detect BRAF and NRAS mutations in archived plasma and tumor samples to determine the potential utility of blood-based tumor genotyping in melanoma. Methods: We analyzed a group of 96 patients with stage III or IV melanoma, prospectively enrolled and followed in the NYU Melanoma Biorepository program. Each patient had a plasma sample and one or more tumor samples available for analysis. We used a combination of allele-specific PCR (Taqman) and SNaPshot assays to identify BRAF V600 and NRAS Q61 mutations in the tumor and plasma samples. Results: Among the 96 patients, 51 had stage III disease at the time of analysis; 45 had stage IV disease. Seventy-two patients had 2 or more tumor samples available for analysis, for a total of 204 tumors analyzed. In total, 52/96 (54%) patients had one or more BRAF or NRAS mutant tumors, including one patient with separate BRAF and NRAS mutant tumors (BRAF, n=35 (36%); NRAS, n=18 (19%)). We successfully amplified plasma DNA from 39/52 (75%) patients with tumor-associated mutations. Among those patients with amplifiable plasma DNA we detected mutations in 7 (18%) patients including 3 BRAF V600E, one V600K, 2 NRAS Q61K and one Q61L. Plasma-based mutations matched tumor-associated mutations in all 7 patients. All 7 patients had active disease at the time of blood draw. There were 32 patients with tumor-associated mutations in which a mutation could not be detected in the plasma. Only 15 of those 32 (47%) had active disease at the time of blood draw. There were no mutations detected in the plasma of the 44 patients whose tumors lacked BRAF or NRAS mutations. Conclusions: These data suggest that plasma-based detection of BRAF and NRAS mut!

Background: Small reported studies have provided some evidence implicating immune related genes in melanoma susceptibility and prognosis; however candidate selection of these prior efforts has been limited. In this study, we performed an analysis of germline variants in immuno-modulatory genes for their association with melanoma survival in a well characterized cohort of prospectively accrued melanoma patients. Methods: Germline DNA isolated from blood samples of 817 melanoma patients was genotyped for 94 SNPs tagging 55 immuno-modulatory genes using Sequenom iPLEX. Cox models were used to test associations between each SNP and recurrence-free and overall survival (RFS and OS), with adjustments for age, gender, subtype, thickness, ulceration, and anatomic site. ROC curves were constructed from different SNP/clinical covariate combinations and the area under the curve (AUC) was used to assess their utility in the classification of 3-year recurrence. Results: The SNP rs2796817 in TGFB2 had strong associations with both RFS (HR=3.8, CI 95%: 1.3-11, p=0.02) and OS (HR=5.5, CI 95%: 1.6-19, p=0.029). Other interesting associations with OS came from IRF8 (rs4843861, HR=0.62, CI 95%: 0.39-0.99, p=0.017), CCL5 (rs4796120, HR=7.6, CI 95%: 2.3-25, p=0.035), and CD8A (rs3810831, HR=2.4, CI 95%: 0.91-6.2, p=0.048). A multivariate model including stage, subtype, and one of the SNPs (rs3810831 from CD8A), was shown to improve the AUC when compared to a model including only stage and subtype (0.77 vs. 0.79). Conclusions: We identified several immune-related loci associated with melanoma RFS and OS. The strongest association, rs2796817, maps in TGFB2, which among other functions suppresses IL-2 dependent T-cell growth. In addition to other associations found in the study these findings provide evidence for the involvement of immuno-modulatory genes in melanoma prognosis and suggest further investigations of immune related genes in disease progression. This is currently underway in the second stage validation an!

T cells expressing antigen-specific T-cell receptors (TCRs) can mediate effective tumor regression, but they often also are accompanied by autoimmune responses. To determine the TCR affinity threshold defining the optimal balance between effective antitumor activity and autoimmunity in vivo, we used a unique self-antigen system comprising seven human melanoma gp100(209-217)-specific TCRs spanning physiological affinities (1-100 muM). We found that in vitro and in vivo T-cell responses are determined by TCR affinity, except in one case that was compensated by substantial CD8 involvement. Strikingly, we found that T-cell antitumor activity and autoimmunity are closely coupled but plateau at a defined TCR affinity of 10 microM, likely due to diminished contribution of TCR affinity to avidity above the threshold. Together, these results suggest that a relatively low-affinity threshold is necessary for the immune system to avoid self-damage, given the close relationship between antitumor activity and autoimmunity. The low threshold, in turn, indicates that adoptive T-cell therapy treatment strategies using in vitro-generated high-affinity TCRs do not necessarily improve efficacy.

Brain metastasis (B-Met) from melanoma remains mostly incurable and the main cause of death from the disease. Early stage clinical trials and case studies show some promise for targeted therapies in the treatment of melanoma B-Met. However, the progression-free survival for currently available therapies, although significantly improved, is still very short. The development of new potent agents to eradicate melanoma B-Met relies on the elucidation of the molecular mechanisms that allow melanoma cells to reach and colonize the brain. The discovery of such mechanisms depends heavily on pre-clinical models that enable the testing of candidate factors and therapeutic agents in vivo. In this review we summarize the effects of available targeted therapies on melanoma B-Met in the clinic. We provide an overview of existing pre-clinical models to study the disease and discuss specific molecules and mechanisms reported to modulate different aspects of melanoma B-Met and finally, by integrating both clinical and basic data, we summarize both opportunities and challenges currently presented to researchers in the field.

Objectives: Age is an understudied factor when considering treatment options for melanoma. Here, we examine the impact of age on primary melanoma treatment in a prospective cohort of patients. Methods: We used logistic regression models to examine the associations between age and initial treatment, using recurrence and melanoma-specific survival as endpoints. Results: 444 primary melanoma patients were categorized into three groups by age at diagnosis: 19-45 years (24.3%), 46-70 (50.2%), and 71-95 (25.5%). In multivariate models, older patients experienced a higher risk of recurrence (hazard ratio 3.34, 95% confidence interval, CI, 1.53-7.25; p < 0.01). No significant differences were observed in positive biopsy margin rates or extent of surgical margins across age groups. Patients in the middle age group were more likely to receive adjuvant therapy than those in the older group (odds ratio 2.78, 95% CI 1.19-6.45; p = 0.02) and showed a trend to longer disease-free survival when receiving adjuvant therapy (p = 0.09). Conclusion: Our data support age as an independent negative prognostic factor in melanoma. Our data suggest that age does not affect primary surgical treatment but may affect decisions of whether or not patients receive postoperative treatment(s). Further work is needed to better understand the biological variables affecting treatment decisions and efficacy in older patients.