|Home | About | Journals | Submit | Contact Us | Français|
Merkel cell carcinoma (MCC) is a rare, clinically aggressive cutaneous neuroendocrine neoplasm with a high mortality rate. Though the etiology is not precisely known, Merkel cell polyomavirus (MCV) DNA has been found recently in a large percentage of MCC tumors. Other suggested risk factors include sun-exposure, immunosuppression and a history of prior malignancy. Work-up of patients with MCC most notably includes nodal staging via clinical exam or sentinel lymph node biopsy (SLNB). The prognosis for most patients with MCC is poor, and the rarity of MCC precludes the prospective, randomized clinical trials necessary to elucidate optimum treatment protocols. Most published data support the use of a multimodality approach centered around surgical excision with negative margins, SNLB to establish the presence or absence of nodal metastases, adjuvant radiothearpy (RT) to decrease the risk of recurrence, and systemic chemotherapy in the case of widespread disease.
Merkel cell carcinoma (MCC) was initially described by Toker in 1972 as a trabecular cancer of the dermis with high lymphatic metastatic risk and found mainly in white elderly patients. The past two decades have given rise to a growing interest in characterization of the disease, as well as the best approach to its management. Numerous submitted case reports demonstrate that MCC is a rare, clinically aggressive neuroendocrine tumor of the skin with a high propensity for local, regional, and distant spread. MCC is lethal in 33% of cases and therefore carries a worse prognosis than malignant melanoma. The number of cases of MCC reported annually in the U.S. is quickly rising, likely secondary to advances in diagnostic techniques, a growing elderly population, and a larger number of immunosuppressed patients. Treatment options vary from surgery alone to multimodality therapy including surgery, radiation therapy (RT), and chemotherapy (CT). This article will highlight recent developments in the biology and management of patients with MCC.
The Merkel cell (MC) was first described by Friedrich Sigmund Merkel in 1875 as a nondendritic, nonkeratinocyte epidermal “tastzellen” or “touch cell” that functions as a tactile skin receptor.[3, 4] Although the true function remains enigmatic, it is presumed that the normal MC may play a role in signal transduction and function as slowly adapting mechanoreceptors to sense touch and hair movement.[5, 6] Continued advances in identifying unique immunocytochemical staining patterns in MC has allowed the diagnostician to more readily identify cases of MCC, increasing both recognition of true disease prevalence and rapid delivery of appropriate treatment.[7–10] In addition, the presence of cytoplasmic neurosecretory granules, cytoplasmic processes, and intermediate filaments surrounding the nucleus, demonstrated on EM, can be used to confirm the identity of MC as the progenitor line in MCC.[11–13]
Although the specific etiology of MCC is not clearly known, probable risk factors have been indicated by published case reports, case series, and population-based studies. A history of sun exposure or concurrence of other sun-associated skin conditions, particularly squamous cell carcinoma, is common in patients with MCC.[14–16] Studies have also shown an increased incidence of MCC in patients on immunosuppressive agents for solid organ transplant,[17–19] in patients with HIV infection and low CD4 counts,[20, 21] in patients undergoing radiation or chemotherapy,[22, 23] and in patients chronically exposed to arsenic. More recently, autoimmune disease has been suggested as a risk factor for developing MCC.
The most notable recent discovery concerning the pathogenesis of MCC is the characterization of a new polyomavirus named Merkel cell polyomavirus (MCV) by Moore and Chang. Polyomaviruses have a double-stranded, circular, supercoiled DNA genome and have been shown to have oncogenic potential. Feng et al. discovered and sequenced the genome of MCV and identified MCV DNA in 8 of 10 MCC tumors. Since the original report, a multitude of other groups have found MCV DNA in MCC tissue samples.[28–41] Although MCV DNA can be helpful in distinguishing MCC from other cutaneous neoplasms and in identifying the primary tumor in metastatic disease, Tolstov et al. report 50% and 80% prevalence of MCV antibodies among people under 15 years and older than 50 years, respectively. Some authors suggest vaccination with MCV virus-like particles could potentially prevent initial MCV infection, although only small pilot animal studies have been performed thus far.
Conflicting nomenclature, ill-defined diagnostic methods and criteria, and a paucity of reported cases are issues that have been addressed as more aggressive and widespread study of the disease has been undertaken. Since a more standardized approach to staging (Table 1) and ICD-9 coding (Table 2) have been implemented, a more reliable population-based registry of MCC over large geographic areas has become available. Using the SEER cancer registry database of the National Cancer Institute, the estimated overall incidence of first primary MCC in the USA for the decade 1992–2001 is 0.32 per 100,000 person-years. Since the majority of MCC cases have been diagnosed since the establishment of a specific immunologic profile for MCC in 1992, calculation of the increasing incidence of first primary MCC in the general population should use 1992 as the year of reference. Demographically, MCC is found most frequently in the elderly, with only sporadic cases reported before the age of 50. The incidence of MCC in whites is more than 8 times that in blacks and almost double the incidence of other ethnic groups. The incidence of first primary MCC is higher in males than females in all ethnic groups, with a ratio of 2:1 in white and in black, and a ratio of 1.5:1 in all other ethnic groups.[44, 45]
MCC has been described in the literature as a classically painless subcutaneous mass with a cystic or nodular appearance, but sometimes presenting with a plaque-like appearance that may be surrounded by small satellite lesions. The tumor size ranges from 2–200mm, but is most often <20mm.[3, 46] Lesions can vastly range in color, most often presenting as red/pink, blue/violaceous, or skin-colored. They can exhibit overlying telangiectasia or a shiny surface, making lesions easily confused with basal cell carcinoma. Figures 1A–C illustrate various MCC at presentation. Tumors may exhibit rapid growth in the preceding weeks or months with a mean length of time of 6.2 months (2 weeks to 2 years) between the patient noticing the lesion and presenting for histologic diagnosis. While the majority of cases present in sun-exposed areas, it is important to note that the neoplasm can also present in areas of minimal sun exposure. Various sources cite the incidence of disease of the head and neck to account for approximately 50% of cases, followed in decreasing order by the extremities, trunk, and buttock.
Studies show 71–79% of patients present with clinically negative nodal involvement (stage I and II disease), 19–24% present with regional nodal metastasis (stage III disease), and 5% have evidence of distant metastatic disease (stage IV disease)., Secondary sites of involvement include skin (28%), lymph nodes (27%), liver (13%), lung (10%), bone (10%), and brain (6%).[3, 47]
The physical characteristics of MCC can resemble those of an epidermoid cyst, lipoma, dermatofibroma, amelanotic melanoma, basal cell carcinoma, squamous cell carcinoma, lymphoma, sarcoma, or metastatic carcinoma. A recent study of 195 patients by Heath et al. outlines the most common clinical features of MCC to aid in its diagnosis. The authors suggest the employment of the mnemonic AEIOU to list significant characteristics: Asymptomatic, Expanding rapidly (doubling in less than 3 months), Immune suppressed, Older than 50 years, and UV exposed skin site, with a subcategory of fair skin. In the 62 patients for which all 5 criteria were available, at least 3 criteria were fulfilled by 89% of patients, at least 4 by 32%, and all 5 were seen in 6% of patients. However, the key to establishing and accurate of and timely diagnosis of MCC is a low clinical index of suspicion and prompt biopsy.
Under light microscopy, MCC presents as an intradermal mass, separated from the epidermis by a narrow ‘Grenz zone” often with necrosis (apoptotic bodies) and patchy lymphocytic infiltrates. Figures 2A, B demonstrate some pathologic features of MCC, characterized by neurosecretory granules, cytoplasmic processes, and intermediate filaments surrounding the nucleus. In addition to the specific clinical and histologic features, MCC is confirmed by immunohistochemistry for proteins such as CK-20, which is detected in a characteristic paranuclear dot-like pattern in 89–100% of Merkel cell tumors but is rarely positive in SCLC (Figures 2C–D).
Initial staging of MCC should include evaluation of primary site for satellite lesions and dermal seeding. Draining nodes should be palpated. Full blood count, serum electrolytes, and alkaline phosphatase are useful in providing a baseline.
Initial imaging workup based on severity of disease based on clinical presentation. Patients with localized disease without physical or symptomatic evidence of regional or distant metastases should be evaluated for spread to regional nodal basins with sentinel lymph node biopsy (SLNB) or ultrasound. Patients with lesions suspicious for regional lymph node involvement can undergo fine needle aspiration cytology, often with ultrasound guidance for more accurate sampling. When lymph nodes are positive, full-body imaging is generally indicated to detect distant metastases. This can be accomplished with a variety of imaging modalities (Figures3A–D), with mounting evidence demonstrating the utility of fluorodeoxyglucose positron emission tomography (FDG-PET) imaging in the staging and management of MCC.[53–62] In a recent study of 18 patients, staging was adjusted in 33% of patients, and the treatment approach was altered in 43% of patients after results of FDG-PET were obtained. Figure 3C shows a patient with MCC of the mandible who was found to have metastatic disease to the mediastinum and liver on post-treatment follow-up FDG-PET scan, and figure 3D illustrates the detection of an involved iliac node from a small foot primary MCC by PET-CT scan.
High-resolution CT and MRI can be used for clinical problem solving in special circumstances. CT is useful for imaging of the thorax and abdomen to exclude metastatic lesions (Figure 3A) and primary SCLC, and can also be used to image deeper lymph nodes and nodular metastases in subcutaneous fat.[47, 53] MRI has been recommended for local imaging and regional lymph node staging for patients with MCC (Figure 3B), and in particular for evaluation of local tumor spread in regions not easily accessible by sonography and the deeper fascia of the head and neck region such as the sinonasal region and orbit.
Based on an analysis of prognostic factors in 5823 patient cases for the development of a consensus system by the AJCC, extent of disease at presentation is highly predictive of survival. This is in concordance with prior studies and staging systems.[49, 63, 64] In this study, it was shown that survival at 5 years (relative to age- and sex-matched control/population data) was 64% for local (Stage I or II), 39% for regional nodal (Stage III), and 18% for distant metastatic (Stage IV) disease.[50, 65, 66] In particular, extent of nodal involvement has been shown to have an impact on survival, with a 42% survival at 5 years in patients with only microscopic nodal involvement and 26% survival for those with clinically apparent nodal disease.[50, 65, 66] The impact of this distinction is reflected in the new staging system (Table 1).
The prognostic value of tumor thickness is an area of continued study and controversy. While some researchers have suggested that tumor thickness is positively associated with likelihood of nodal metastasis and poor survival, others have denied this association.[67–69] Other patient and tumor characteristics often considered to portend poor outcome include male sex, age greater than 60 years, presence of co-morbid conditions, immunocompromised state, positive resection margin, high mitotic rate, small cell size, absence of an inflammatory infiltrate, primary site of disease (head/neck, trunk, perineum, lower extremities), and no adjuvant radiotherapy.[7, 10, 44, 63] Prognostic factor analyses of the impact on survival in larger-scale studies are required to determine the true prognostic value of these characteristics.
The treatment of choice for MCC depends on many factors including the stage of the tumor, particularly with respect to nodal involvement, the location of the tumor, and the medical comorbidities of the patient. Though far from a consensus, most current recommendations include the addition of adjuvant RT to reduce local or nodal relapse and to improve disease-specific and overall survival. A summary of results and recommendations from recent retrospective reviews and meta-analyses is provided in Table 3.
The standard of surgical care for MCC has traditionally been wide local excision. Some experts have suggested surgery alone is adequate for early stage disease,[70–72] but a recent study of 95 patients from a database at the National Cancer Institute in Milan further evaluated the use of surgery alone in early stage MCC. The authors report an 80% 5-year disease specific survival rate when lymph nodes were pathologically negative, which dropped to 58% when regional nodal metastases were present. While it was initially thought that margins ≥3 cm were necessary for the best reduction of local recurrence,[74, 75] margins 2–3 cm wide and 2cm deep are now generally accepted.[51, 76–78] Allen et al reviewed records of 102 patients with MCC treated at Memorial Sloan-Kettering Cancer center over a 27 year period and found no association between width of margins and either recurrence rate or survival. Two additional studies showed no difference in locoregional control or overall survival regardless of margin status.[80, 81]
As with more common skin cancers, Mohs microsurgery is an attractive treatment option for tumors on the head and neck, where the recommended margin width and depth are either cosmetically undesirable or technically impossible. Boyer et al. studied recurrence rates and survival in patients with stage I MCC treated with Mohs surgery with or without adjuvant RT and concluded that the addition of adjuvant RT after Mohs surgery was not necessary as it did not provide additional benefit in terms of recurrence or survival. O’Connor et al. evaluated patients treated with wide local excision or Mohs surgery and found that, while standard surgical excision resulted in local persistence and regional metastasis in 31.7% and 48.8% of patients respectively, Mohs surgery resulted in only 6.8% and 33.3% local persistence and regional metastasis respectively. O’Connor et al suggested that there may be some benefit in adding adjuvant RT after Mohs surgery to further reduce the risk of nodal metastasis.
Another point of disagreement among surgeons is whether or not SLNB or LND is necessary for patients undergoing surgical excision of MCC. Even in patients with no clinical evidence of lymph node involvement, 25%-100% of primary draining lymph nodes may contain microscopic metastases[79, 83–85] that can lead to nodal relapse in up to 76% of patients.[51, 86–88] Since MCC is thought to spread in an orderly fashion similar to melanoma, SLNB can detect subclinical nodal metastatic disease before progression to stage IV disease[89–91]. Allen et al. showed data suggesting incidence of positive lymph nodes was independent of tumor size, so routine SLNB should be offered to all patients undergoing surgical excision for MCC. Fang et al. also describe the morbidity due to LND and offer data to suggest lymph node irradiation can provide comparable control rates to LND whether lymph node disease is detected clinically or by SLNB.
RT has been advocated by a few authors as definitive treatment, particularly for patients who are poor surgical candidates.[93, 94] While RT is typically added when patients present at a higher stage or with nodal metastases, Mortier et al. reported data on 26 patients with stage I MCC. There was no difference in disease-specific or overall survival between patients with stage I MCC treated with RT alone or surgery + RT. The 9 patients treated with definitive RT received a median dose of 60 Gy, and the majority received prophylactic nodal irradiation. More recently, Veness et al presented an analysis of 46 patients treated with definitive RT. Included patients were inoperable either due to advanced disease or because of medical comorbidities. In fact, 77% of the patients examined had nodal metastases at the time of treatment. A slightly lower mean radiation dose was used for the nodal basins (51–55 Gy), but the authors reported a 75% in-field control rate. Even though the numbers are small, these two studies show that RT may be a viable alternative to surgery for patients unable to withstand an operation, but adjuvant or concurrent RT to the nodal basins is recommended even in patients presenting with stage I disease.
There have been few studies examining the optimum RT dose for the treatment of MCC. The similarities between MCC and SCLC have led to the common use of doses of 45–60 Gy, depending on the presence or absence of gross disease. Morrison et al recommend 56–60 Gy for gross unresected disease and 46–50 Gy for adjuvant treatment of the primary tumor site and draining lymphatics after resection. A more recent study confirmed the role of RT in reducing local and nodal relapses and demonstrated a dose-response relationship of up to 50 Gy for subclinical disease and >55 Gy for gross disease.
The majority of retrospective studies conducted over the past 15 years have consistently demonstrated better local control after surgery when adjuvant RT is added.[80, 96–104] In one of the largest single institution studies, Gillenwater et al. analyzed 66 cases of MCC of the head and neck treated at M.D. Anderson Cancer Center. Although width of surgical margins did not significantly affect survival, the authors did report significantly lower local and regional recurrence rates in the 26 patients receiving post-operative RT, even in the 77% of patients with clinically negative lymph nodes. In contrast, Allen et al. published the largest single institution study to date and found no difference in local or regional recurrence upon the addition of adjuvant RT. The 5-year disease specific survival for the patients included in this study was 64%, with 97% in patients found to be node-negative on pathological staging and 52% in patients with positive nodes. The authors concluded that adjuvant RT to the nodal basin was not necessary if operative nodal staging had been performed. Although most major academic institutions have not treated a sufficient number of MCC patients over the past few decades to definitively assess the efficacy of adjuvant RT, more statistical power is gained through review of multiple single-institution case series. Lewis et al published a meta-analysis pooling the data from 132 studies yielding the inclusion of 1254 patients who carried a pathologic diagnosis of MCC, received surgical excision with negative margins and had at least one month of follow up. Regardless of stage at presentation, local recurrence was reported in 24.5% when treated with surgery only and 6.9% also treated with adjuvant RT. Out of the 566 patients presenting with stage I disease, local recurrence was reported in 44.8% of patients treated with surgery only and 13.2% of patients who received adjuvant RT. This calls into question the common clinical practice of offering adjuvant RT only when lymph nodes metastases are present. Finally, there was no survival benefit shown upon adding adjuvant RT after excision of stage I MCC, but the authors noted that 3 times as many patients would be necessary to obtain 80% power. More recently, Mojica et al. offered further evidence for the addition of adjuvant RT with data from 1665 patients in the SEER cancer registry database. The authors found adjuvant RT was associated with a significant improvement in overall survival with a median survival 45 months with surgery alone and 63 months with surgery + RT. Though these studies are no substitute for a randomized, prospective clinical trial, the majority of experts seem to agree that adjuvant RT benefits patients in terms of decreased recurrence and overall survival. The practice guidelines recently updated by the National Comprehensive Cancer Network currently recommend the routine use of adjuvant RT to the lymph nodes in all cases unless the patient has had a negative SLNB.
As is the case with RT doses and schedules, the majority of the chemotherapeutic regimens used to treat MCC are similar to those found to be effective in the treatment of SCLC. Carboplatin, cisplatin, etoposide, cyclophosphamide, 5-fluorouracil, vincristine and doxorubicin are some of the more common agents used. Traditionally, primary chemotherapy has been reserved for those patients with unresectable disease, whether due to poor surgical candidacy because of metastatic disease or due to medical comorbidities. Adjuvant chemotherapy has been suggested to be of benefit when added to surgery and adjuvant RT for the treatment of stage II disease. In a study of 40 patients with stage I and II MCC, Poulsen et al. showed no beneficial effect of chemotherapy, but reported wide confidence intervals and an underpowered study. In addition to the small number of MCC patients treated with chemotherapy, the advanced stage at which chemotherapy is added makes drawing definitive conclusions problematic. In a study of 35 patients, Kokoska et al. reported no benefit when chemotherapy was added to adjuvant therapy regimens. Likewise, Allen et al. reported no benefit in survival upon the addition of adjuvant chemotherapy. Conversely, several authors have reported chemotherapy to be beneficial in terms of recurrence and survival especially for patients with high risk MCC. [80, 95, 98, 108]
Two retrospective analyses sought to further characterize the utility of chemotherapy for MCC with larger groups of patients.[109, 110] Voog et al. reported an impressive 61% response rate (57% for metastatic disease and 69% for advanced disease). However, the authors found that while MCC is chemosensitive, it is rarely curable, and the 7.7% risk of toxic death due to neutropenia further complicates the treatment decision. Tai et al. analyzed a larger group of patients treated with chemotherapy and found an even more impressive response rate of 75.7%.
Recent developments have improved accuracy in diagnosing MCC, and the discovery of MCV DNA in MCC tumors has shed considerable light on the pathogenesis. Although the incidence has been shown to be increasing, there are still far too few cases of MCC presenting to any one institution for definitive conclusions to be drawn. Collaboration will be the key to future breakthroughs in terms of diagnosis, prognosis and treatment of this rare neoplasm. The consensus staging system and revised ICD-9 codes are two steps taken recently in standardizing the way MCC cases are reported. A definitive staging system may be a first important step in making possible a world-wide, multi-institutional prospective, randomized case-control study to definitively determine whether a multimodality regimen is superior to surgical resection alone in the treatment MCC. Such large scale trials will also be necessary to answer existing questions concerning the use of chemotherapy. Until such data becomes available, there is ample data to advocate for the use of adjuvant RT after surgical resection due to the high incidence of local, nodal and distant metastatic relapse. Although MCC is chemosensitive, there is less data on the efficacy of chemotherapy, and its role in adjuvant therapy continues to be debated.