Evaluation of Multiple Tissue Levels Frequently Upstages Patients With Clinically Localized Thin Primary Cutaneous Melanoma.

BACKGROUND: Breslow thickness (BT), ulceration, and microsatellitosis are critical prognostic parameters for cutaneous melanoma staging. These parameters can vary depending on the number of tissue levels examined from individual paraffin blocks. We sought to evaluate all prognostic histopathologic parameters in melanoma for their variations between levels, taken at regular intervals, in a single study. METHODS: We analyzed 40 consecutive cases of primary cutaneous (nonacral) melanoma through five hematoxylin and eosin sections, taken at 100 µm intervals, for staging and prognostic parameters. RESULTS: Examination of additional levels resulted in (a) an increase in BT in 47.5% (19 out of 40) of cases and (b) detection of ulceration in a further 5% (2/40). This resulted in upstaging for 20% (8 out of 40) of patients (15% because of BT, 2.5% because of ulceration, and 2.5% because of BT and ulceration). The upstaging effect was incremental, with approximately 5% of patients upstaged with each additional 100 µm interval (up to 400 µm). Incipient ulceration and epidermal consumption were infrequent (10% of cases); however, when present, ulceration was subsequently observed in half of cases. We encountered no cases where microsatellitosis was detected at deeper levels. CONCLUSION: The performance of additional tissue levels is a simple and inexpensive procedure that can improve the accuracy of staging for patients with thin (pT1) primary cutaneous melanomas. It may be pertinent for pathologists to consider additional levels for thin melanomas when a BT measurement is close to a staging threshold (e.g., within 0.1-0.3 mm for pT1a vs. pT1b, or pT1b vs. pT2a), or when incipient ulceration is encountered.

Implementing structured pathology reporting protocol for non-melanocytic skin cancers: practical considerations.

Non-melanocytic skin cancers (NMSCs) account for five times the incidence of all other cancers combined and cost US $6 billion annually. These are the most frequent specimens encountered in community pathology practice in many Western countries. Lack of standardised structured pathology reporting protocols (SPRPs) can result in omission of critical information or miscommunication leading to suboptimal patient management. The lack of standardised data has significant downstream public health implications, including insufficient data for reliable development of prognostic tools and health-economy planning. The Royal College of Pathologists of Australasia has developed an NMSC SPRP. A multidisciplinary expert committee including pathologists, surgeons, dermatologists, and radiation and medical oncologists from high volume cancer centres was convened. A systematic literature review was performed to identify evidence for including elements as mandatory standards or best practice guidelines. The SPRP and accompanying commentary of evidence, definitions and criteria was peer reviewed by external stakeholders. Finally, the protocol was revised following feedback and trialled in multiple centres prior to implementation. Some parameters utilised clinically for determining management and prognosis including tumour depth, lymphovascular invasion or distance to the margins lack high level evidence in NMSC. Dermatologists, surgeons, and radiation oncologists welcomed the SPRP. Pathologists indicated that the variety of NMSC specimens ranging from curettes to radical resections as well as significant differences in the biological behaviour of different tumours covered by the NMSC umbrella made use of a single protocol difficult. The feedback included that using a SPRP for low risk NMSC was neither clinically justified nor compensated adequately by the Australian Medicare Reimbursement Schedule. Following stakeholder feedback, the SPRP implementation was restricted to excision specimens of head and neck NMSC; and low-risk NMSC, such as superficial basal cell carcinoma, were excluded. Implementing NMSC SPRP fulfils an unmet clinical need. Unlike other cancers, NMSCs generate a range of specimen types and are reported in a wide range of pathology practices. Limiting use of SPRP to NMSC at higher risk of progression and providing formatted templates for easy incorporation into laboratory information systems were essential to successful deployment. In the future, further consideration should be given to implementing the SPRP to include all relevant specimens, including non-head and neck and low-risk NMSC specimens.

Multiple eruptive squamoproliferative lesions during anti-PD1 immunotherapy for metastatic melanoma: Pathogenesis, immunohistochemical analysis and treatment.

Treatment with anti-PD1 inhibitors may enhance the risk for developing low grade squamoproliferative skin tumors. Immunohistochemical (IHC) analysis of the immune tumor microenvironment (TME) allows exploration of the pathogenesis and relationship with the PD1/PDL1 axis. Patients with eruptive keratoacanthoma (KA)-like lesions were recruited from the Melanoma Institute Australia, a tertiary referral specialist melanoma treatment center from January 2015 to August 2017. Clinicopathologic evaluation and IHC features of tumor cells (PDL1 expression) and peritumoral microenvironment (PD1, FOXP3, PDL1, CD4:CD8 expressing cells) in 12 eruptive KA-like lesions, were compared with solitary KAs in age and sex matched non-anti-PD1 treated controls. Four patients with repeated episodes of eruptive KA-like and lichenoid lesions developing 2-7 months after commencing pembrolizumab for AJCC stage IV melanoma, were recruited. Eruptive KA-like squamoproliferative lesions occurred in sun exposed sites and in areas of resolving, concomitant or delayed lichenoid reactions. Histologically, the lesions were well-differentiated squamoproliferative lesions resembling infundibulocystic squamous cell carcinoma or KA. IHC of cases and controls revealed low PDL1 expression of both squamous tumor cells and the TME immune cells. The numbers of immunosuppressive FOXP3 positive Tregs and PD1-expressing T-cells were higher in the cases than the controls but the CD4:CD8 ratio (2:1) was similar. The patients best responded to acitretin and were managed surgically if they demonstrated neoplastic features. Accelerated squamoproliferative growth in actinically damaged keratinocytes associated with lichenoid eruptions may be unmasked in patients treated with anti-PD1 immunotherapy potentially contributed to by a local cutaneous immunosuppressed TME.

BRAF mutation testing for patients diagnosed with stage III or stage IV melanoma: practical guidance for the Australian setting.

Targeted therapy (BRAF inhibitor plus MEK inhibitor) is now among the possible treatment options for patients with BRAF mutation-positive stage III or stage IV melanoma. This makes prompt BRAF mutation testing an important step in the management of patients diagnosed with stage III or IV melanoma; one that can help better ensure that the optimal choice of systemic treatment is initiated with minimal delay. This article offers guidance about when and how BRAF mutation testing should be conducted when patients are diagnosed with melanoma in Australia. Notably, it recommends that pathologists reflexively order BRAF mutation testing whenever a patient is found to have American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) stage III or IV melanoma (i.e., any metastatic spread beyond the primary tumour) and that patient's BRAF mutation status is hitherto unknown, even if BRAF mutation testing has not been specifically requested by the treating clinician (in Australia, Medicare-subsidised BRAFV600 mutation testing does not need to be requested by the treating clinician). When performed in centres with appropriate expertise and experience, immunohistochemistry (IHC) using the anti-BRAF V600E monoclonal antibody (VE1) can be a highly sensitive and specific means of detecting BRAFV600E mutations, and may be used as a rapid and relatively inexpensive initial screening test. However, VE1 immunostaining can be technically challenging and difficult to interpret, particularly in heavily pigmented tumours; melanomas with weak, moderate or focal BRAFV600E immunostaining should be regarded as equivocal. It must also be remembered that other activating BRAFV600 mutations (including BRAFV600K), which account for ∼10-20% of BRAFV600 mutations, are not detected with currently available IHC antibodies. For these reasons, if available and practicable, we recommend that DNA-based BRAF mutation testing always be performed, regardless of whether IHC-based testing is also conducted. Advice about tissue/specimen selection for BRAF mutation testing of patients diagnosed with stage III or IV melanoma is also offered in this article; and potential pitfalls when interpreting BRAF mutation tests are highlighted.

Residual melanoma in wide local excision specimens after 'complete' excision of primary cutaneous in situ and invasive melanomas.

Wide local excision (WLE) to achieve adequate clearance margins is the standard initial definitive treatment for patients with biopsy-proven primary cutaneous melanoma. Residual melanoma in WLE specimens after prior complete excision-biopsy (CEB) is reported in 0-6.3% of cases. However, studies evaluating the prevalence, clinicopathological features and relevance of persistent disease in WLE specimens are limited. This study sought to determine the frequency of and clinicopathological characteristics associated with residual melanoma in WLE specimens performed after a CEB of primary cutaneous or acral melanoma (in situ or invasive) with clinically and histologically tumour-free margins, and assess its relevance. A review of the research database and pathology archives of a large Australian tertiary referral melanoma treatment centre was performed. Eligible patients were those for whom a definitive WLE was performed after CEB of a primary melanoma (in situ or invasive) with negative clinical and histological margins, between May 2013 and May 2015. All partial biopsies were excluded. Of 640 eligible patients, 510 (79.7%) had invasive melanoma and 130 (20.3%) had melanoma in situ. Residual disease was identified in 20 cases (20/640, 3.1%), of which three (15%) were melanoma in situ on CEB and 17 (85%) were invasive melanoma. On univariate analysis, the presence of residual disease in WLE specimens was associated with lentigo maligna (LM)/LM melanoma (LMM) subtype [odds ratio (OR) 10.33; 95% confidence interval (CI) 2.84-37.54; p=0.004], nodular melanoma (NM) subtype (OR 4.92; 95% CI 1.53-15.85; p=0.0076) and, for invasive tumours, higher tumour mitotic rate (mean 7.7, SD 7.51 vs 3.4, SD 4.83; OR 1.11; 95% CI 1.04-1.18; p=0.0014). Breslow thickness >4 mm was associated with a higher risk of residual disease (OR 7.30; 95% CI 1.88, 28.26; p=0.004). Cases with residual disease had primary tumours with a significantly larger diameter (median 14 mm, range 4-25) than those without residual disease (median 9 mm, range 2-60), (OR 1.07; 95% CI 1.03-1.11; p≤0.001) and were also more likely to be amelanotic (38% vs 14%), (OR 3.69; 95% CI 1.17, 11.60; p=0.026). Residual disease was associated with assessment of >3 slides of tissue (OR 6.98; 95% CI 1.54-31.62; p=0.0118) and complete blocking of the scar (OR 31.69; 95% CI 3.98-252.21; p=0.0011). Residual melanoma in WLE specimens is an infrequent occurrence. Risk factors for residual disease are LM/LMM and NM melanoma subtypes, higher mitotic rate, larger lesion diameter and amelanosis. Tumours with these features warrant more extensive pathological sampling. WLE after CEB for melanoma remains an important procedure to reduce local recurrence; however, limited pathological sampling of the WLE scar is probably appropriate for cases lacking high risk features.

Molecular Profiling of Noncoding Mutations Distinguishes Nevoid Melanomas From Mitotically Active Nevi in Pregnancy.

The accurate recognition of subtle melanomas and their distinction from benign mimics is an oft-recurring diagnostic problem, critical for patient management. Melanomas that bear resemblance to benign nevi (so-called nevoid melanomas, NMs) and benign mitotically active nevi in pregnancy (MANP) are 2 lesions particularly prone to error. Molecular data, including analysis of noncoding regions, in MANP and NM are very limited. This study sought to identify differences in clinical, pathologic, and molecular characteristics between MANP and NMs to facilitate correct diagnosis and reduce the risk of overtreatment or undertreatment. Clinicopathologic characteristics of NM (n=18) and MANP (n=30) were evaluated, and mutation data were analyzed using next-generation sequencing for available cases in each group (NM, n=8; MANP, n=12). All MANP showed innocent histopathologic characteristics apart from increased mitotic activity, frequently in both superficial and deep parts of the lesion (median dermal mitotic rate: 2/mm, range: 1 to 7/mm). All cases of NM demonstrated a characteristic nevoid silhouette, subtle atypical architectural and cytologic features, and variable mitoses (median mitotic rate: 3/mm, range: 1 to 5/mm). Median NM tumor thickness was 1.4 mm. Four of 10 NM patients with follow-up had metastatic disease, including 3 patients who developed widespread metastases, with 1 disease-related death. No other recurrences have been identified (follow-up period: 24 to 60 mo). None of the 15 MANP patients with available follow-up had a recurrence. Most NMs harbored hotspot mutations in NRAS (6/8, 75%). Noncoding mutations were significantly more common in NMs than in MANP (median: 4 vs. 0, P=0.0014). Copy number alterations were infrequent but, when present, were seen in NMs (3/8 NMs vs. 0/12 MANP). All NMs but only 1 of 12 MANP had >1 abnormality in the noncoding regions. Similar to conventional common acquired nevi, MANP mostly harbored driver BRAF mutations, while activating NRAS mutations, noncoding mutations, and copy number alterations were rare. NM and MANP have subtle but recognizable distinguishing histopathologic characteristics that are underpinned by molecular differences. Mutation analysis of targeted noncoding mutations may assist in the diagnosis of difficult lesions.

A Review of Key Biological and Molecular Events Underpinning Transformation of Melanocytes to Primary and Metastatic Melanoma.

Melanoma is a major public health concern that is responsible for significant morbidity and mortality, particularly in countries such as New Zealand and Australia where it is the commonest cause of cancer death in young adults. Until recently, there were no effective drug therapies for patients with advanced melanoma however significant advances in our understanding of the biological and molecular basis of melanoma in recent decades have led to the development of revolutionary treatments, including targeted molecular therapy and immunotherapy. This review summarizes our current understanding of the key events in the pathway of melanomagenesis and discusses the role of genomic analysis as a potential tool for improved diagnostic evaluation, prognostication and treatment strategies. Ultimately, it is hoped that a continued deeper understanding of the mechanisms of melanomagenesis will lead to the development of even more effective treatments that continue to provide better outcomes for patients with melanoma.

RAB27A promotes melanoma cell invasion and metastasis via regulation of pro-invasive exosomes.

Despite recent advances in targeted and immune-based therapies, advanced stage melanoma remains a clinical challenge with a poor prognosis. Understanding the genes and cellular processes that drive progression and metastasis is critical for identifying new therapeutic strategies. Here, we found that the GTPase RAB27A was overexpressed in a subset of melanomas, which correlated with poor patient survival. Loss of RAB27A expression in melanoma cell lines inhibited 3D spheroid invasion and cell motility in vitro, and spontaneous metastasis in vivo. The reduced invasion phenotype was rescued by RAB27A-replete exosomes, but not RAB27A-knockdown exosomes, indicating that RAB27A is responsible for the generation of pro-invasive exosomes. Furthermore, while RAB27A loss did not alter the number of exosomes secreted, it did change exosome size and altered the composition and abundance of exosomal proteins, some of which are known to regulate cancer cell movement. Our data suggest that RAB27A promotes the biogenesis of a distinct pro-invasive exosome population. These findings support RAB27A as a key cancer regulator, as well as a potential prognostic marker and therapeutic target in melanoma.

Primary anorectal melanoma: clinical, immunohistology and DNA analysis of 43 cases.

Primary melanoma involving the anorectal region is rare, accounting for <1% of all melanomas in most Western countries. It characteristically presents at an advanced clinical stage and is associated with poor clinical outcomes. Preliminary reports suggest that response rates to immunotherapies in patients with advanced stage mucosal melanoma are much lower than in cutaneous (or acral) melanoma patients but reasons for this are unclear. Comprehensive characterisation of the immune microenvironment in anorectal melanoma has not previously been performed. A single-institution cohort of 43 primary anorectal melanoma patients was examined to describe clinicopathological features and characterise the immune microenvironment to provide insights into the behaviour of this rare melanoma subtype. The tumours displayed multiple adverse prognostic attributes including deep thickness (median 11.5 mm), ulceration (81%) and high mitotic rate (median 12/mm2). The median overall survival was 24 months and the median recurrence-free survival was 9 months. Tumour-infiltrating lymphocytes (TILs) were absent or mild in most tumours (75%); PD-L1 positive staining (>1% of tumour cells) was present in 44% of cases, however in 86% of positive tumours the percentage of positive cells was ≤10%. Four tumours underwent whole genome sequencing; no ultraviolet signature was identified, and there was a lower mutational load but higher structural chromosomal variant load compared with cutaneous melanomas. Poor responses of anorectal melanomas to immunotherapy may be caused by lower immunogenicity of these tumours as characterised by low mutation burden (and therefore low neoantigenicity), low TILs infiltrates and low PD-L1 expression. Further investigation is required to determine whether TILs and PD-L1 expression predict response to immunotherapy in patients with mucosal melanoma.

Oral malignant gastrointestinal neuroectodermal tumour with junctional component mimicking mucosal melanoma.

Malignant gastrointestinal neuroectodermal tumour (GNET) is a recently characterised rare and aggressive tumour that typically arises in association with the small intestine of adults. We present a novel case of this entity and expand the spectrum of its reported morphological features. The patient was a 5-year-old female, the youngest reported patient affected by the condition, and presented with extra-abdominal disease. The histopathological features included the presence of a junctional component of the palatal tumour, which mimicked mucosal melanoma, a feature that has not been previously reported in GNET. Whole genome and RNA sequencing was performed that demonstrated the EWSR1-ATF1 translocation characteristic of GNET. Knowledge of this entity and its features, together with careful morphological assessment supplemented by judicious immunohistochemical and molecular studies should enable the correct diagnosis to be established.

Negative immune checkpoint regulation by VISTA: a mechanism of acquired resistance to anti-PD-1 therapy in metastatic melanoma patients.

Understanding the mechanisms of acquired resistance to anti-PD-1 will allow development of better treatment strategies for cancer patients. This study evaluated potential mechanisms of acquired resistance to anti-PD-1 in longitudinally collected metastatic melanoma patient biopsies. Thirty-four metastatic melanoma biopsies were collected from 16 patients who had initially responded to either anti-PD-1 (n=13) alone or combination of anti-PD-1 and ipilimumab (n=3) and then progressed. Biopsies were taken prior to treatment (PRE, n=12) and following progression of disease (PROG, n=22). Immunohistochemistry was performed on all biopsies to detect CD8, FOXP3, PD-1 and VISTA expression on T-cells and PTEN, ß-catenin, PD-L1, HLA-A, and HLA-DPB1 expression in the tumor. The majority of patients showed significantly increased density of VISTA+ lymphocytes from PRE to PROG (12/18) (P=0.009) and increased expression of tumor PD-L1 from PRE to PROG (11/18). Intratumoral expression of FOXP3+ lymphocytes significantly increased (P=0.018) from PRE to PROG (10/18). Loss of tumor PTEN and downregulation of tumor HLA-A from PRE to PROG were each identified in 5/18 and 4/18 PROG biopsies, respectively. Downregulation of HLA-DPB1 from PRE to PROG was present in 3/18 PROG biopsies, whereas nuclear ß-catenin activation was only identified in 2/18 PROG biopsies. Negative immune checkpoint regulation by VISTA represents an important potential mechanism of acquired resistance in melanoma patients treated with anti-PD-1. Downregulation of HLA-associated antigen presentation also occurs with acquired resistance. Augmentation of the VISTA immune checkpoint pathway may hold promise as a therapeutic strategy in metastatic melanoma patients, particularly those failing anti-PD-1 therapy, and warrants assessment in clinical trials.

SF3B1 and BAP1 mutations in blue nevus-like melanoma.

Blue nevi are melanocytic tumors originating in the cutaneous dermis. Malignant tumors may arise in association with or resembling blue nevi, so called 'blue nevus-like melanoma', which can metastasize and result in patient death. Identifying which tumors will behave in a clinically aggressive manner can be challenging. Identifying genetic alterations in such tumors may assist in their diagnosis and prognostication. Blue nevi are known to be genetically related to uveal melanomas (eg, both harboring GNAQ and GNA11 mutations). In this study, we analyzed a large cohort (n=301) of various morphologic variants of blue nevi and related tumors including tumors diagnosed as atypical blue nevi (n=21), and blue nevus-like melanoma (n=12), screening for all gene mutations known to occur in uveal melanoma. Similar to published reports, we found the majority of blue nevi harbored activating mutations in GNAQ (53%) or GNA11 (15%). In addition, rare CYSLTR2 (1%) and PLCB4 (1%) mutations were identified. EIF1AX, SF3B1, and BAP1 mutations were also detected, with BAP1 and SF3B1 R625 mutations being present only in clearly malignant tumors (17% (n=2) and 25% (n=3) of blue nevus-like melanoma, respectively). In sequencing data from a larger cohort of cutaneous melanomas, this genetic profile was also identified in tumors not originally diagnosed as blue nevus-like melanoma. Our findings suggest that the genetic profile of coexistent GNAQ or GNA11 mutations with BAP1 or SF3B1 mutations can aid the histopathological diagnosis of blue nevus-like melanoma and distinguish blue nevus-like melanoma from conventional epidermal-derived melanomas. Future studies will need to further elucidate the prognostic implications and appropriate clinical management for patients with tumors harboring these mutation profiles.

Activating cysteinyl leukotriene receptor 2 (CYSLTR2) mutations in blue nevi.

Blue nevi are common melanocytic tumors arising in the dermal layer of the skin. Similar to uveal melanomas, blue nevi frequently harbor GNAQ and GNA11 mutations. Recently, recurrent CYSLTR2 and PLCB4 mutations were identified in uveal melanomas not harboring GNAQ or GNA11 mutations. All four genes (GNAQ, GNA11, CYSLTR2, and PLCB4) code for proteins involved in the same signaling pathway, which is activated by mutations in these genes. Given the related functional consequences of these mutations and the known genetic similarities between uveal melanoma and blue nevi, we analyzed a cohort of blue nevi to investigate whether CYSLTR2 and PLCB4 mutations occur in tumors lacking GNAQ or GNA11 mutations (as in uveal melanoma). A targeted next-generation sequencing assay covering known activating mutations in GNAQ, GNA11, CYSLTR2, PLCB4, KIT, NRAS, and BRAF was applied to 103 blue nevi. As previously reported, most blue nevi were found to harbor activating mutations in GNAQ (59%, n=61), followed by less frequent mutations in GNA11 (16%, n=17). Additionally, one BRAF (1%) and three NRAS (3%) mutations were detected. In three tumors (3%) harboring none of the aforementioned gene alterations, CYSLTR2 mutations were identified. All three CYSLTR2 mutations were the same c.386T>A, L129Q mutation previously identified in uveal melanoma that has been shown to lead to increased receptor activation and signaling. In summary, our study identifies CYSLTR2 L129Q alterations as a previously unrecognized activating mutation in blue nevi, occuring in a mutually exclusive fashion with known GNAQ and GNA11 mutations. Similar to GNAQ and GNA11 mutations, CYSLTR2 mutations, when present, are likely defining pathogenetic events in blue nevi.