Merkel Cell Carcinoma Therapeutic Update

Opinion statement

Merkel cell carcinoma (MCC) is a rare and aggressive neuroendocrine tumor of the skin. Early-stage disease can be cured with surgical resection and radiotherapy (RT). Sentinel lymph node biopsy (SLNB) is an important staging tool, as a microscopic MCC is frequently identified. Adjuvant RT to the primary excision site and regional lymph node bed may improve locoregional control. However, newer studies confirm that patients with biopsy-negative sentinel lymph nodes may not benefit from regional RT. Advanced MCC currently lacks a highly effective treatment as responses to chemotherapy are not durable. Recent work suggests that immunotherapy targeting the programmed cell death receptor 1/programmed cell death ligand 1 (PD-1/PD-L1) checkpoint holds great promise in treating advanced MCC and may provide durable responses in a portion of patients. At the same time, high-throughput sequencing studies have demonstrated significant differences in the mutational profiles of tumors with and without the Merkel cell polyomavirus (MCV). An important secondary endpoint in the ongoing immunotherapy trials for MCC will be determining if there is a response difference between the virus-positive MCC tumors that typically lack a large mutational burden and the virus-negative tumors that have a large number of somatic mutations and predicted tumor neoantigens. Interestingly, sequencing studies have failed to identify a highly recurrent activated driver pathway in the majority of MCC tumors. This may explain why targeted therapies can demonstrate exceptional responses in case reports but fail when treating all comers with MCC. Ultimately, a precision medicine approach may be more appropriate for treating MCC, where identified driver mutations are used to direct targeted therapies. At a minimum, stratifying patients in future clinical trials based on tumor viral status should be considered as virus-negative tumors are more likely to harbor activating driver mutations.

Introduction

Merkel cell carcinoma (MCC) is a rare neuroendocrine skin cancer. MCC is most common on body areas with higher ultraviolet exposure in Caucasian patients of advanced age [1]. There is also a subset of cases associated with immunocompromised states and a worse prognosis [2, 3•]. The overall 5-year survival for node-negative disease is 64 %, but in those with regional nodal disease or metastatic disease at presentation, the 5-year survival drops to 39 and 18 %, respectively [4]. Factors important in staging MCC are the following: size (greater or less than 2 cm); invasion to underlying structures; lymph node involvement, with a distinction between clinical and pathological node status; and the presence of metastases [5]. Changes in the forthcoming eighth edition AJCC cancer staging handbook may also incorporate the recent observations that nodal MCC with an unknown primary has a better prognosis than comparable stage III disease with a known primary tumor [3•, 6]. In 2008, the discovery of Merkel cell polyomavirus (MCV), believed to be involved in malignant transformation of most cases of MCC, prompted many new directions of research [7]. With an estimated annual incidence of only 0.79 per 100,000 persons (~1600 cases in the USA) in 2011 [8], large prospective clinical trials are difficult to perform and are widely lacking in the literature. As a consequence, the evidence supporting management recommendations for MCC (Fig. 1) relies heavily on small case series and expert opinion [9].

Fig. 1.

Merkel cell carcinoma general treatment algorithm. 1 Definitive treatment of primary lesion should not occur prior to SLNB. 2 Reserved for patients who are poor surgical candidates. 3 The benefit of RT to SLNB-negative basin is unclear- consider observation for low-risk patients with a small primary tumor (<1 cm); in head and neck disease, the risk of false-negative SLNB is higher. 4 PET-CT is the preferred modality; when unavailable, CT and MRI can be used. 5 NCCN guidelines recommend a clinical trial as a first-line treatment for M₁ disease. 6 As clinically indicated in high-risk patients. SLNB sentinel lymph node biopsy, FNA fine needle aspiration, CLND completion lymph node dissection.

Diagnosis and work-up

Histological diagnosis

• Applying proper immunostains is critical for accurate diagnosis.

The accurate diagnosis of MCC often requires the evaluation of histology and immunostaining by an expert dermatopathologist. Although hematoxylin and eosin (H&E) staining is often diagnostic, a diagnosis of MCC should be confirmed with neuroendocrine and epithelial makers such as neuron-specific enolase (NSE), synaptophysin, chromogranin, neurofilament (NF), epithelial membrane antigen (EMA), cytokeratin (CK) AE1/AE3, and CK20 [10•]. Of these markers, CK20 is the most commonly used for MCC and typically demonstrates a characteristic paranuclear punctate positivity [11]. However, not every MCC tumor stains positively for CK20. As metastatic small cell lung cancer, melanoma, and lymphoma can resemble MCC on H&E, confirming negative staining for thyroid transcription factor 1 (TTF-1), S-100, and leukocyte common antigen (LCA), respectively, is advisable in most cases.

Sentinel lymph node biopsy

• Lymph nodes should be pathologically evaluated in almost every case of MCC.

In the past, sentinel lymph node biopsy (SLNB) was more controversial for patients with MCC. Now, it is clear that pathologically node-negative disease has a survival benefit over clinically node-negative disease, a fact that should be considered when offering SLNB [4, 12]. Micrometastatic lymph node disease was present in nearly 30 % of 721 cases included in a recent literature review, reinforcing the utility of SLNB [10•]. Ultimately, SLNB evaluated with immunostains may be beneficial for prognostication and directing treatment in practically all patients with clinically node-negative MCC regardless of tumor size, as is recommended in the most recent National Comprehensive Cancer Network (NCCN) guidelines [4, 9, 13]. More recent work suggests that when evaluating nodal positivity, the number of positive lymph nodes [14•] and the histological presence of sheets of tumor cells [15] may correlate with worse prognosis.

Imaging

• Consider imaging for staging in all patients with high-risk disease.

• Consider imaging in patients with negative pathological lymph nodes if the primary tumor possesses aggressive features.

• There are no head-to-head studies comparing 2-fluoro-[¹⁸F]-deoxy-2-D-glucose (FDG)-positron emission tomography (PET) with somatostatin receptor (SSTR)-PET, but both are superior to computed tomography (CT) alone and may be complementary.

The identification of distant metastases during staging can alter the risk-to-benefit assessment of highly aggressive local and regional treatment plans. According to the NCCN guidelines, patients with advanced MCC, including those with grossly positive lymph nodes, should be evaluated with imaging studies. Imaging can also be considered in early-stage MCC for patients whose tumors possess features suggesting poor prognosis and where surgery is expected to have high morbidity.

The ideal imaging modality for MCC has yet to be determined. There are many retrospective studies, as well as meta-analyses, supporting the use of FDG-PET/CT, citing high specificity and sensitivity, and often upstaging patients [16, 17]. CT alone has varying sensitivities for distant and nodal metastases [18, 19] and may be inadequate for sites such as bone and bone marrow [16]. The ability of FDG-PET/CT to detect low-tumor burden metastases is due to the high metabolic activity of MCC. However, this strength is also its weakness as it may fail to identify well-differentiated tumors and, conversely, lead to false-positive results by detecting unrelated foci of non-specific inflammation [20]. Despite advances in imaging, SLNB with immunostains remains the standard for detecting occult nodal metastases [16, 21].

More recently, there has been investigation on the use of somatostatin analogues as imaging modalities, which can be useful because neuroendocrine tumors such as MCC have been found to express SSTRs [22]. Classically performed using ¹¹¹In-diethylenetriaminepentaacetic acid (DTPA)-octreotide, SSTR-PET is evolving to include newer radiolabeled tracers bound to octreotide, including ⁶⁸Ga-DOTATATE and ⁶⁸Ga-DOTATOC, which, when combined with CT, can have high sensitivity for metastases [23]. Cost and availability may factor into the choice between FDG-PET/CT and SSTR-PET/CT, and further comparison studies of the two techniques are needed [24]. An advantage of SSTR is that it also identifies tumors that may be targeted with somatostatin analogues for treatment. One notable drawback, however, is that because of higher background uptake in certain organs, it has lower sensitivity in identifying disease in the liver, adrenal glands, pancreas, thyroid, and spleen [20].

The timing of imaging during post-treatment follow-up also lacks clear evidence. Since nearly all recurrences happen within the first 2 years, with a median of 9 months [12], early and repeated scans are recommended for patients at increased risk of recurrence every 3 to 6 months for the first 2 years [16]. While there are no prospective studies to evaluate imaging frequency, the benefits of more frequent imaging should be weighed against the cost and risk to the patient of repeated scans, including false-positive results. According to current NCCN guidelines, routine imaging should be considered for high-risk patients, which may include patients with clinically node-positive disease and higher. Until treatments for advanced MCC are able to improve disease-specific survival, early detection of distant metastatic disease in asymptomatic patients with MCC may be of limited value. The morbidity associated with conventional systemic therapy and the consequences of false-positive scans should be given consideration when ordering follow-up imaging.

Viral status

• Of MCCs, ~80 % are MCV positive.

• MCV-negative tumors are associated with more genetic mutations, some of which are potential targets of therapy.

• Establishing a standard clinical test for MCV status is needed.

MCV is a double-stranded, non-enveloped DNA virus that has been reported in roughly 80 % of MCC tumors [7], although perhaps significantly less common in tumors from Australian patients [25]. The virus is ubiquitous in the general population and is believed to be acquired in childhood as an asymptomatic primary infection [26]. In contrast to the common episomal infections, MCV DNA in MCC is clonally integrated into the tumor genome [7]. The current lack of a Clinical Laboratory Improvement Amendments (CLIA)-approved test for MCV presents a challenge when comparing and validating studies involving MCC viral status. Detecting MCV in tumor samples is most commonly done by quantitative PCR or immunostaining. There is currently no standard protocol for the PCR detection of MCV DNA in patient samples, and the appropriate threshold for calling a sample positive is a matter of debate. Antibodies have been raised against various viral proteins including small tumor antigen (sTAg), large tumor antigen (LTAg), and the capsid protein VP1. However, only CM2B4, a monoclonal antibody against LTAg, is commercially available. A study using Ab3, a novel LTAg antibody purported to be more sensitive than other antibodies, reported MCV positivity in an unexpected 97 % of MCC tumors [27]; however, these findings have not been replicated by other laboratories. Newer methods have recently been developed for MCV detection, including in situ hybridization and an antibody to sTAg, which may be more sensitive than CM2B4 staining alone [28]. With many test options having varying claims of sensitivity and specificity, comparative studies and establishing a gold standard for MCV detection are needed to more accurately stratify patients for clinical investigations.

Measuring levels of circulating antibodies against MCV proteins is a way to assess humoral immune responses against the virus. Antibodies against VP1 are prevalent in the general population, whereas antibodies against T antigens (TAgs) are more specific to patients with MCC [29]. Interestingly, high levels of VP1 antibodies at the time of MCC diagnosis may portend a better prognosis [30, 31]. In the patients that have them, antibodies against TAgs appear to reflect disease burden, falling after treatment and rising with known or occult recurrent disease [29, 30]. Future directions may see these titers being monitored, much like prostate-specific antigen in prostate cancer, and are already commercially available (http://www.merkelcell.org/sero/).

Recent studies using panel sequencing or whole exome sequencing have provided insights into the biological differences between MCV-positive and MCV-negative MCC. In MCV-positive MCCs, p53 mutations, which are often associated with ultraviolet damage, are uncommon [32]. In contrast, MCV-negative MCCs are much more likely to contain p53 mutations. Additionally, virus-negative MCCs contain, with varying frequencies, mutations in NOTCH, NF1, FGF receptor 2 (FGFR2), and the PI3K/AKT pathway, among others [33, 34•, 35•]. MCV-positive MCCs, however, carry a relatively low mutational burden and lack enrichment of UV signature mutations.

Established treatments

Surgery

• Excision should be done with 1- to 2-cm margins for local disease.

• SLNB should be done at the time of surgery for patients with clinically negative nodes.

Surgical management of early local disease is primarily by wide local excision with 1–2-cm margins down to fascia, with the goal to achieve clear surgical margins. However, no prospective data is available to correlate the margin size with recurrence risk. When tissue sparing is critical due to the anatomic location of the tumor, techniques with complete peripheral and deep margin control (e.g. Mohs micrographic surgery) may be considered, provided they do not interfere with SLNB when indicated. Positive SLNB, fine needle aspiration, or core needle biopsy should be followed by completion lymph node dissection (CLND) and/or RT, although prospective studies are lacking. Location of the lymph node bed is an important consideration when recommending treatment. RT to the inguinal nodal basin may have lower morbidity than CLND, whereas axillary CLND may be better tolerated than RT [36]. Advances in intensity-modulated radiation therapy (IMRT) may be able to reduce adverse events associated with RT; however, prospective trials will be needed to assess the efficacy and safety of IMRT in MCC.

Radiation therapy

• Adjuvant RT improves locoregional control with reduced recurrence rate but may not affect overall survival.

• RT does not offer an advantage in the setting of negative SLNB.

• RT may be helpful for locoregional control in inoperable cases.

RT is a useful treatment adjunct as MCC tumors are known to be radiation sensitive [37]. Adjuvant RT to the tumor bed for local control may be associated with lower rates of local recurrence, though the impact on survival is unclear. Regional directed RT is also used to improve MCC recurrence-free survival [3•, 36]. To optimize locoregional control, 5-cm RT field margins are recommended, with a minimum of 2 cm when anatomy is constraining [9, 38].

RT may also be useful in specific situations where tissue-sparing techniques are required. A single-institution retrospective study of 179 patients showed a reduced local recurrence rate after RT to excision sites with <10-mm surgical margins, even with microscopically positive margins, although overall survival was unchanged. They recommend at least 50 Gy following an excision with narrow margins, suggesting that narrow margins plus RT may have similar outcomes to wide margins without RT and may be potentially useful in patients for whom a wide excision may carry higher morbidity [39]. Even RT alone to an unresectable tumor can be beneficial for locoregional control [40, 41].

To date, most of the studies on the adjuvant use of RT for MCC remain retrospective with shortcomings including missing or mixed SLN status. A large prospective trial comparing adjuvant RT to observation in stage I patients did not characterize pathological node status and actually had to be stopped early due to declining enrollment because of the gaining popularity of SLNB [42]. They concluded that adjuvant RT decreased regional recurrence when pathological node status is unknown. A similar argument was made by Bichakjian et al. regarding another large retrospective review of adjuvant RT for head and neck MCC [43]. However, the evidence supporting the general use of regional RT in MCC antedates the routine use of SLNB. Recently, the premise that nodal RT is beneficial in SLNB-negative patients has come under scrutiny. A large retrospective study of 111 SLNB-negative patients found no recurrence or survival benefit from RT of the nodal basin, including head and neck sites [44•], despite a reported 17 % false-negative SLNB rate [10•]. With NCCN recommendations now including SLNB on all patients, additional studies are needed to confirm if sentinel node status should direct RT use.

Chemotherapy

• It is most commonly platinum-based ± etoposide.

• There is short-term control but no durable response, with high toxicity.

Most regimens employing chemotherapeutics are in the setting of advanced MCC. The literature is flush with case reports, retrospective series, and reviews, but a dearth of prospective studies. MCC is very sensitive to chemotherapy, but responses are not durable and tumors often recur within 4–15 months. Furthermore, the associated toxicity may actually decrease overall survival [45]. The most common regimens, recommended in the NCCN guidelines, employ a platinum-based product, such as carboplatin or cisplatin, often adding etoposide. The next most common regimen is cyclophosphamide, doxorubicin, and vincristine (CAV). In the largest most pertinent review of chemotherapeutics, 204 cases between 1989 and 1997 were reported to have response rates of 76 % to CAV and 60 % to carboplatin and etoposide regimens, though significant toxicities including death were noted [46]. Given the toxicity and lack of durable response, the potential short-term benefits should be weighed against the risks.

Palliation

• Single-fraction RT is a promising new technique.

• Brachytherapy in high dose has exceptional locoregional control but no change in overall survival.

Currently, treatment for advanced MCC remains largely palliative. One palliative option that was recently described as well tolerated for symptomatic tumor masses is single-fraction radiotherapy. Although retrospective in nature with many confounding variables, such as chemotherapy and immunotherapy, the authors reported 93 tumors treated with objective responses in 94 % and complete responses in 45 % [47•].

Another new and promising treatment for palliation of in-transit disease is high-dose brachytherapy. Of 152 tumors treated in 10 patients, locoregional control at a median of 34 months was 99 % [48]. Despite these impressive results, they reported no change in overall survival as 7 out of the 10 patients developed further metastatic disease outside of the field treated.

Emerging treatments

Immunotherapy

Immunotherapy is one of the most recent and rapidly expanding treatment modalities for all cancers, including MCC. Immune checkpoint inhibitors, in particular, have been able to achieve durable cures in patients with certain tumor types by reactivating antitumor cellular immune responses [49]. When studying the application of immunotherapy to MCC, it may be important to establish the viral status of the tumor. MCV-negative MCCs typically have more mutations and tumor neoantigens than MCV-positive tumors, all of which are potential targets for antitumor immune responses [34•], whereas MCV-positive MCCs express viral peptides that can be targeted by cellular immune responses [50]. Currently, it is unclear how these differing target immune landscapes will impact the use of immunotherapy for MCC.

Anti-programmed cell death receptor 1/programmed cell death ligand 1 antibodies

• Monoclonal antibodies targeting programmed cell death receptor 1/programmed cell death ligand 1 (PD-1/PD-L1) interaction show great promise for MCC.

The binding of PD-L1 to PD-1 on cytotoxic T cells inhibits their tumor-killing activity. PD-L1 is frequently expressed on MCC tumor cells and peritumoral immune cells [51]. There is currently an ongoing phase II trial with avelumab, an anti-PD-L1 antibody, as a second-line treatment for patients with advanced MCC who have failed chemotherapy [52]. Although the study’s results have not yet been reported, avelumab as a second-line treatment of metastatic MCC recently received orphan drug designation, fast-track designation, and breakthrough therapy designation from the Food and Drug Administration.

Pembrolizumab is an anti-PD-1 antibody approved for the treatment of melanoma and non-small cell lung cancer. At least one patient with MCC showed a response to pembrolizumab during a phase I trial [53]. A small ongoing phase II trial of pembrolizumab as a first-line treatment for advanced MCC recently reported a 71 % initial response rate among 17 evaluable patients [54••, 55]. It remains to be seen if this response rate will be generalizable and if responses will be durable. Nonetheless, it appears that MCC is exceptionally sensitive to PD-1 checkpoint inhibition.

Other immunotherapies

• Adjuvant ipilimumab is in an ongoing phase II trial.

• A phase I/II trial of tremelimumab/durvalumab will open soon.

• Adoptive T cell transfer is in an ongoing phase I/II trial.

• Intratumoral IL-12 plasmid vaccine is in an ongoing phase II trial.

Ipilimumab inhibits the checkpoint molecule cytotoxic T lymphocyte antigen 4 (CTLA-4) and is an approved immunotherapy for metastatic melanoma. Recently, adjuvant ipilimumab was reported to improve recurrence rates for resected melanoma [56]. To test if the drug will show similar benefit in MCC, an ongoing controlled phase II trial in Europe is investigating ipilimumab versus observation as an adjuvant therapy for patients with resected MCC [57].

Combining immunotherapies has improved efficacy in other cancers while also increasing the risk of immune-related adverse events [58, 59]. A new phase I/II trial may open soon, treating metastatic solid tumors, including MCC, with the CTLA-4 inhibitor tremelimumab plus durvalumab, an anti-PD-L1 antibody, in combination with poly-ICLC, a Toll-like receptor-3 ligand [60].

Tumor-infiltrating lymphocytes (TILs), particularly CD8+, are associated with better MCC survival [61–63]. Adoptive cell transfer attempts to increase the available antitumor T cells by harvesting, growing, and reinjecting the patient’s own cells. A phase I/II trial is currently underway using this method in conjunction with an interleukin-2, aldesleukin [64], to stimulate the host’s immune system [65]. This treatment has shown great promise in patients with metastatic melanoma, but it remains to be seen if the same results can be achieved in MCC [66]. At least one patient has shown regression in their MCC by combining local and systemic immunotherapies with autologous T cell transfer [67].

Intratumoral injection, to avoid peripheral toxicity, of an interleukin-12 (IL-12) plasmid vaccine followed by electroporation in mouse melanomas demonstrated increased IL-12, interferon-gamma, and tumor lymphocytes, as well as reduced tumor vascularity [68]. This was followed by a phase I trial of the same method in patients with metastatic melanoma with nearly half of the patients showing stabilization and partial or complete response with few side effects [69]. A phase II trial for this approach is ongoing for MCC with imminent anticipated completion [70].

Targeted therapies

Targeted therapies use drugs that inhibit molecules required for tumor growth and progression. With a few exceptions, disease progression due to drug resistance is a limitation of targeted therapies. Nonetheless, when activating mutations in a specific receptor, pathway, or oncogene can be targeted in a cancer, responses can be rapid and there can be fewer toxicities than conventional chemotherapy. In the case of MCC, the majority of tumors are MCV-positive and lacking in druggable driver mutations. In contrast, MCV-negative MCCs frequently have mutations predicted to be oncogenic. Thus far, no recurrent activated pathway has been identified in the majority of MCV-negative MCCs, making it unlikely that a single targeted therapy will be highly effective in treating MCC. However, exceptional responders are likely if drugs can be matched to the targets present in a given tumor.

PI3K/AKT inhibition

• Mutations are uncommon, mostly found in MCV-negative MCC.

• Idelalisib in a case report resulted in a complete response.

PI3K/AKT-activating mutations have been found independent of viral status [71] or nearly entirely in MCV-negative patients [72], although each study reported extremely small sample sizes as these mutations are uncommon with a prevalence of 4 and 10 %, respectively. Since these mutations are commonly found in other cancers, targeted inhibitors already exist. In a recent case report, a patient with metastatic MCC carrying a known PI3K mutation was treated with the PI3Kδ inhibitor, idelalisib, resulting in a rapid and complete response. Unfortunately, the patient died from other causes before long-term response could be measured [73]. This case is a good example of precision medicine, where patient’s unique tumor mutations are treated specifically with a targeted therapy. As the costs and delays of genotyping fall, this approach may become more common in MCC clinical trials and in patient care.

Mammalian target of rapamycin inhibitors

• Pathway activation is more common in MCV-negative MCC.

• MLN0128 is an ongoing dual mammalian target of rapamycin (mTOR) inhibitor phase II trial.

Also in the PI3K-related kinase pathway, AKT and mTOR have been found to be inappropriately activated in MCC, more commonly in MCV-negative tumors [74]. Specifically targeting both mTOR complex 1 and complex 2, the dual mTOR inhibitor, MLN0128, was able to inhibit MCC xenograft tumor growth in vivo [75]. A phase I dose escalation protocol of MLN0128 is expected to complete imminently [76], and a phase II trial for patients with advanced MCC has just begun recruiting patients [77].

Tyrosine kinase inhibitors

• Targets include VEGF receptor (VEGFR), PDGF receptor (PDGFR), and c-kit.

• Pazopanib case reported a partial response.

• Cabozantinib is in an ongoing phase II trial.

• Imatinib has likely no benefit.

Other potential therapeutic targets found in varying percentages of MCC tumors include kinases such as VEGFR, PDGFR, and c-kit [78]. Pazopanib is a multi-tyrosine kinase inhibitor that targets VEGFR, PDGFR, FGFR, and c-kit. There is one case report of a patient with MCC with a partial response to pazopanib [79], and there is an ongoing trial in the UK [80]. Another phase II trial is actively recruiting patients with advanced MCC for treatment with cabozantinib, which has activity against c-Met and VEGFR2, blocking angiogenesis [81]. In theory, this mechanism might be expected to have an effect on such a vascular tumor as MCC that is known to express VEGFR2 [78]. Based on the expression of c-kit in MCC, a phase II trial using imatinib, a tyrosine kinase inhibitor active against c-kit, as well as abl and PDGFR, was completed but failed to show benefit [82]. It appears that although MCC may express c-kit, activating mutations in the gene are generally absent. However, there is one case report of clinical remission with imatinib in a patient with MCV-positive MCC [83].

Octreotide

• Peptide receptor radionuclide therapy (PPRT) with ¹⁷⁷Lu-DOTATATE is in an ongoing phase II trial.

• Several other studies are ongoing for other somatostatin analogues.

An extension of using SSTR ligands for imaging is to use the same mechanism to target treatment. Neuroendocrine tumors found to express SSTRs are good candidates for treatment with PPRT. This method uses a radiolabeled somatostatin analogue, such as ¹⁷⁷Lu-DOTATATE, ⁶⁸Ga-DOTA-Tyr³-octreotide, or ⁶⁸Ga-DOTA-Tyr³-octreotate, that is taken up by the tumor and metabolized to slowly emit targeted localized radiation [84–86]. Earlier treatments used DTPA-octreotide, which lacked significant tumor response but was then replaced by ⁹⁰Y-DOTA-Tyr3-octreotide (⁹⁰Y-DOTATOC), which had better tumor responses, but more myelosuppression and renal toxicity. This was further refined and replaced by DOTA-Tyr3-octreotate (DOTATATE), which was attached to ¹⁷⁷lutetium as opposed to ⁹⁰yttrium, resulting in less penetration (2 versus 11 mm) into normal tissue and fewer side effects such as renal failure [87]. A phase II study of ¹⁷⁷Lu-DOTATATE to treat neuroendocrine tumors has been completed, although no results are available yet [88]. Other non-radioactive somatostatin analogues, such as pasireotide [89] and lanreotide [90], are also being investigated in trials to treat MCC.

Future directions

• Stratifying clinical trials by viral status for precision medicine should be considered.

• Circulating biomarkers to track disease status and monitor for occult disease is a growing trend.

• Tumor vaccines could be used to stimulate host immune responses.

• RT is being examined in combination with checkpoint inhibitors.

It is now clear that the mutational profile of MCV-positive and MCV-negative MCC is vastly different. Going forward, it would be prudent to stratify clinical trials by tumor viral status, especially for therapies that target driver mutations more likely to be found in MCV-negative tumors. For this reason, developing a validated test for tumor viral status is a priority to help guide treatment and allow for more accurate interpretation of clinical trials. Of course, high-throughput sequencing could be applied to simultaneously identify viral status and the mutational profile of MCC tumors to better direct therapy. Tumors like MCC that lack highly recurrent oncogene mutations are candidates for this type of precision medicine approach.

Circulating biomarkers for the diagnosis, prognosis, and monitoring of cancer is a growing trend. Based on their use in other neuroendocrine tumors, some institutions have attempted to use serum levels of neuron-specific enolase and chromogranin A to monitor disease activity in MCC, but a recent paper has questioned the utility of these tests in determining outcomes, disease progression, and MCC tumor burden [91]. In contrast, circulating tumor cells appear to hold promise as biomarkers for MCC [91, 92]. The further development of circulating tumor cells, MCV serology, or other biomarkers could potentially identify patients with the highest risk of recurrence, who would benefit the most from adjuvant therapies.

Immunotherapy with checkpoint blockade or with cytokine therapy can be effective in reversing immune tolerance and stimulating antitumor immune responses. Since ~80 % of MCCs express viral proteins, adding an MCV vaccine to other immunotherapies may further enhance the efficacy of these approaches. This approach could be effective in treating advanced MCC as well as preventing recurrence in the setting of adjuvant immunotherapy. In 2012, a vaccine to the MCV LTAg demonstrated a successful immune response in mice [93]. One year later, the same group published data on a vaccine against MCV sTAg, which was also effective in mice [94]. Thus, a vaccine approach may be feasible to enhance immune responses against MCV-expressing MCC tumors.

Studies have demonstrated a synergistic effect when combining immunotherapy and RT to treat melanoma [95, 96]. Judicious application of RT to tumors can cause the release of tumor antigens, induce inflammation, and reverse immune escape mechanisms such as the downregulation of major histocompatibility complex I (MHC-I). All of these effects can potentially enhance responses to checkpoint blockade immunotherapy. Reduced MHC-I is observed in MCC, particularly in MCV-positive tumors. Multiple treatments such as interferon, etoposide, and RT are being explored as a mechanism to restore MHC-I expression in MCC, and combining these modalities with checkpoint blockade is another potential area of future study [97].

Conclusions

As with any rare disease, future progress in the management of MCC will require multi-institutional collaborative efforts, as no single institution treats enough patients to effectively analyze the natural history of the disease or effectively compare treatments. As we learn more about MCC, treatment protocols can be refined and standardized. It seems likely that immunotherapy with checkpoint inhibitors will be found to benefit patients with metastatic MCC in the near future, and adjuvant immunotherapy may eventually prove useful. However, not every patient will respond to immunotherapy, and hence appropriate targeted therapies are also needed. Until a common therapeutic target or predictive biomarkers can be identified in MCC, it is likely that targeted therapies will require a precision medicine approach to be effective.