Merkel cell carcinoma (MCC) is a rare neuroendocrine dermal neoplasm, first described by Toker in 1972. 1 Since that time, more than 600 cases of MCC have been described in the literature. 2 The aggressive nature of this tumor has led many groups to conclude in their reports that multimodality therapy offers the best results in terms of overall, as well as disease- and recurrence-free, survival, although, admittedly, results have been quite variable.
Most institutions have insufficient experience with MCC to provide broadly generalizable recommendations regarding therapy. Nevertheless, inspection of the literature provides valuable concepts that may be applied to the care of individual patients. This article describes the experience of Naval Medical Center, San Diego with MCC of the head and neck, and analyzes the existing literature regarding lesions arising at all sites.
PATIENTS AND METHODS
This study is a retrospective outcomes analysis of all cases of head and neck MCC seen at Naval Medical Center, San Diego, between January 1, 1988 and June 30, 1998. The records of the Naval Medical Center, San Diego Tumor Registry were searched for patients with that diagnosis, and supplemental information was retrieved from the Radiation Oncology and Head & Neck Surgery Clinic charts. Data on the nine patients identified are presented in Table 1. The average age of the group at diagnosis was 68.8 years, and all but one were white men.
Two patients had undergone some form of treatment before arrival at our institution. Occasionally, certain data points were not available for review because of lack of documentation; when such data are not available, it is mentioned specifically.
Therapy and its results are listed in Table 2. Six patients underwent wide-local excisions, and four of these had concurrent modified radical neck dissections (MRND). One-patient underwent an auriculectomy only. Another patient had a superficial parotidectomy and concurrent MRND. One patient was treated at another institution, and the specifics of the surgical procedure and adjuvant therapies are not available. One patient sought treatment for distant metastatic disease, and two others were subsequently found to have regional nodal involvement after wide-local excisions.
The margins were free of tumor in the surgical specimens immediately after excision of the primary lesions in the eight patients treated at Naval Medical Center, San Diego. Margin status was not available for the patient treated elsewhere. One patient who sought treatment for T2N0 disease had tumor more than 1 mm from the resection margins. He was offered adjuvant radiation therapy (XRT) and chemotherapy, but refused further treatment. Subsequently, distant metastatic disease developed in this patient, who died less than 5 months from the initial presentation.
Five patients underwent lymphadenectomy, and of those, two patients were discovered to have nodal involvement. Distant metastases to the brain developed in one of these two patients, who received palliative XRT and chemotherapy, and who died of metastatic disease less than 2 years from diagnosis. The other patient was treated with XRT to both the parotid-neck region and primary site, but developed a recurrence between the two fields. A second treatment of XRT was given to the recurrence site. The patient then had recurrence in the original parotid field, at which time he received a platinum-based chemotherapy regimen. The patient had a complete response to chemotherapy.
Eight patients underwent radiation therapy; however, each patient received individualized treatment regimens based on their status and location of recurrence (Table 2).
One patient sought treatment for distant metastatic disease and received palliative radiation therapy. This patient died of metastatic disease within 2 years.
MCC is a rare and aggressive dermal neuroendocrine tumor, originally described by Toker in 1972 as “trabecular cell carcinoma.”1 Merkel cells, neuroendocrine cells that reside in the basal layer of the epidermis, are presumed to be the origin for this tumor. MCC is morphologically characterized by infiltrative proliferation of the neoplastic cells within the dermis. The cells are arranged in loosely cohesive sheets or, rarely, trabecular cords that infiltrate between collagen bundles. The cells are usually of intermediate size with scant cytoplasm and high nuclear:cytoplasmic ratios. They demonstrate the typical histologic features of neuroendocrine carcinomas including finely stippled chromatin, nuclear molding, and are frequent apoptotic bodies (Fig. 1). Mitotic figures may be readily identified, or in some cases, relatively infrequent. The expression of immunohistochemical markers such as chromogranin, synaptophysin, and neuron-specific enolase and the existence of membrane-bound neurosecretory granules on ultrastructural examination confirm the impression of neuroendocrine differentiation. MCC displays a distinctive and relatively specific pattern of perinuclear dotlike or crescentic positive reactions to immunohistochemical stains for cytokeratins or neurofilament protein (Fig. 2). 3,4 This pattern reflects the presence of cytoplasmic condensations of intermediate filaments in the perinuclear region. Positive staining for neurofilament protein or cytokeratin 20 may be a useful discriminator for differentiating MCC from metastatic small-cell lung cancer, a more common type of neuroendocrine carcinoma. 3–6 This can be particularly valuable if the only identifiable disease is node based, as in Case 6.
Initially thought to be of low-malignant potential, MCC is now known to have a very high rate of local tissue spread (20–75%), regional nodal metastases (31–80%), and distant metastases (26–75%.) Approximately one third of patients eventually die of their disease. 2 In this review, locoregional and/or distant failure developed in 4 of 9 (44%) patients.
Tumor location has been shown to be a statistically significant predictor of disease-free survival by univariate analysis, with lesions occurring on the trunk usually having a worse prognosis than those on the head and neck. 2,7–9 This may be related to the notion that MCC lesions are typically smaller when discovered on the head and neck region as opposed to other locations. 8 One group that reported a worse prognosis for head and neck lesions suggested that it may have been a result of less aggressive initial treatment, reflecting the narrow margins taken to avoid cosmetic defects in these areas. 10 Goepfert et al. 9 determined that distant metastases subsequently appeared in approximately 50% of patients in whom regional nodal metastases developed, and in 69% of those who had locoregional failure as opposed to 17% of those with locoregional control. Although these data may be attributable to aggressive tumor biology, they may otherwise be interpreted to suggest that locoregional recurrence may contribute to subsequent metastatic disease. It is impossible to determine which mechanism is more important in therapy for these lesions, but in either case, local control at the primary site and the regional lymphatics should be aggressively sought. Data from the literature specifically dealing with head and neck presentations are summarized in Table 3.
More data are available when studies of MCC at all sites are reviewed. A few groups have attempted to provide prognostic information by using a conventional staging system. Stages were assigned according to the absence (stage I) or presence (stage II) of positive lymph nodes within the draining nodal basin, or by the presence of systemic metastases 7,8,11 (stage III). The series differ in terms of which variables have a statistically significant effect on disease-free survival. Most groups report statistically significant data showing that higher disease stage 2,7,11–13 and nodal metastases 7,11,13 negatively affect disease-free survival.
Tumor size 7,11,13 and surgical margins 2,7,11 have not reproducibly been shown to exert a statistically significant effect on disease-free survival, with one exception. Allen et al. 8 demonstrated that tumor size and stage were statistically significant for disease-free survival only with stage I disease.
Surgical margins (Table 4) of at least 2 cm to 3 cm have been suggested by some groups to both decrease recurrence and lengthen survival; however, the data supporting this are not statistically significant. The 3-cm margin is based on studies showing a significant reduction in local recurrences by increasing the margins from 1 cm to 2 cm and from 2 cm to 3 cm. 7,12 The consensus recommendation is to follow these guidelines when performing surgical excision of MCC. 8 In areas where wide excisions cannot be performed without producing significant cosmetic defects, such as the face, simple excision plus postoperative radiation therapy should provide optimal local control. 14 Another approach proposed by O’Conner and Brodland 12 is Mohs’ micrographic surgery for primary tumor excision. They report an excellent local control rate of 92% (11 of 12) at 36 months compared with 31.7% (13 of 41) with wide local excision alone. The addition of XRT after Mohs’ surgery reduced regional metastasis to zero (0 of 4) compared with 50% (4 of 8) without XRT. In this study, we were not able to quantify the margin size in our cases, because this information was not specified in the records. The sample size in this study was not large enough to determine any statistically significant data in terms of margin status; however, the one patient who had tumor within 1 mm of the margin and refused adjuvant therapies did die of distant metastatic disease in 16 months (without local failure).
Regional nodal metastases are reported in 31% to 80% of the cases. 2 In one series by Goepfert and associates, 50% of patients with clinically negative nodes had microscopic disease in the surgical specimens. 9 Furthermore, there was a 75% regional failure rate in untreated necks. The high occurrence of clinically undetectable regional metastases has led most groups to recommend elective regional lymphadenectomy (Table 5) at the time of primary tumor excision. This allows optimal assessment of nodal status, thus guiding additional therapy, and may reduce early spread to the draining nodal chain. 13,15 Definitive control of the neck is important because Shaw and Rumball 16 noted that two thirds of patients with MCC in whom locoregional recurrence develops eventually die of their disease.
The extent of elective neck dissection depends on the nodal drainage basin of the primary tumor. In general, MCC metastases follow a distribution similar to that seen in other skin malignancies. Primary lesions situated near the midline, such as the anterior part of the neck and chin, should undergo bilateral neck dissection. Caution should be exercised before limiting the dissection, because once the carcinoma has spread to regional lymphatics, there is a propensity to involve multiple levels. 9 In the series by Goepfert et al., 9 31% of dissected necks had histologically positive nodes at multiple levels. Therapeutic lymphadenectomy is recommended for any palpable regional metastases.
As with melanoma, sentinel node biopsy has been proposed as an alternative to more aggressive lymphadenectomy techniques. 17 Patients with a positive sentinel lymph node biopsy would then undergo formal, and in this case therapeutic, lymph node dissection. The predictive power of sentinel node biopsy depends on relatively reliable lymph drainage patterns from a given source. The head and neck region has extremely rich and variable lymphatic distribution, which may limit the usefulness of this technique. Regardless of which technique is used, lymphadenectomy has been shown to decrease regional recurrence rates, but in most studies, has not been shown to demonstrate a significant survival benefit. 8 This is true for both elective and therapeutic neck dissection. Nonetheless, regional failure can be associated with devastating morbidity, and most series recommend lymphadenectomy as an adjunct to excision of the primary lesion. 9,18–21
Some authors have suggested that prophylactic radiotherapy (Table 5) be given to the regional nodal chain for similar reasons. Most authors suggest radiation fields that cover the operative bed and the draining lymphatics of the entire ipsilateral area of the neck with generous margins. 16,22 Dosages suggested are typically between 45 Gy and 50 Gy to both the primary tumor site and draining regional lymphatics. 22–26 Treatment of midline facial lesions includes radiation therapy to the draining nodal chains on both sides of the neck. The field to the primary site has been suggested by one group to include a 2-cm margin around the surgical scar when possible. 27
Few authors would dispute treating clinically positive lymph nodes with either radiation therapy, lymphadenectomy, or a combined modality strategy. Controversy, however, does exist as to which approach is more effective. The morbidity of combined treatment with both radiation and lymph node dissection is also not in dispute and should prompt further studies to determine which approach will maximize locoregional control and minimize side effects, such as lymphedema. However, with the small number of cases of MCC in the literature, it may be difficult to determine a statistically significant benefit from one approach over another. As one reviewer commented, “we are therefore stuck with erring on the side of being more aggressive and combining therapies.”2
Chemotherapy has been used in extensive inoperable local or recurrent disease and in treatment of distant metastases (Table 6). It has been suggested by some groups that the chemotherapeutic regimens used in treating small-cell lung carcinoma may be effective in treating MCC, because they share some neuroendocrine and histologic characteristics. 28 The agents most commonly used are doxorubicin and cyclophosphamide, although other regimens include vincristine, prednisone, dacarbazine, and etoposide, and platinum-based compounds. Approximately 60% of MCC have either a partial or complete response to chemotherapy, which in general lasts less than 6 to 12 months. 28,29 Fuen et al. 29 observed that although dramatic tumor regression is associated with initial chemotherapy, once the disease has relapsed or progressed, the course thereafter is usually one of rapid deterioration. No chemotherapy protocol has been shown to consistently improve survival. 29,30 In this series, one patient had a complete response to a platinum-based regimen, albeit with only 2 months of follow-up. The other patient in this study received an unknown regimen at another institution, had no response, and died approximately 1 year after treatment.
Head and neck MCC is a rare and aggressive dermal tumor of neuroendocrine origin that requires multimodality therapy, including surgery, XRT, and possibly adjuvant chemotherapy. The data available in the literature are varied in terms of which treatment regimens yield the best results of disease control and overall survival. There are trend data, which do not reach statistical significance, which indicate improved outcomes with tumor-free margins; therefore, most reports recommend a 2-cm to 3-cm tumor-free margin surrounding the primary lesion. This is often difficult to achieve in the head and neck, especially in the periorbital area.
Treatment of the regional draining lymph nodes is advocated in most series because of the high propensity for this tumor to metastasize via the lymphatics. Prophylactic lymph node dissection or radiotherapy have been shown to improve disease recurrence but do not statistically affect overall survival. Controversy exists as to whether radiation therapy, lymphadenectomy, or a combination of the two is the most effective, keeping in mind that morbidity increases with a multimodality approach.
Adjuvant chemotherapy is also recommended by most groups for the treatment of metastatic disease. It is postulated that MCC is pathologically similar to small cell lung carcinoma; therefore, similar regimens have been used for both of these disease processes. Currently, there is no consensus on which chemotherapy protocol is the most effective. The majority of the cases in the literature have shown poor long-term control with chemotherapy, generally with responses of less than 6 to 12 months. No chemotherapy protocol has been shown to improve survival.
Multimodality therapy protocols need larger case numbers to answer many of the uncertainties of treatment involved in MCC. Multiinstitutional studies need to be initiated to obtain a large enough population of patients to reach statistical significance in the outcomes data of the various treatment regimens.
1. Toker C. Trabecular carcinoma of the skin. Arch Dermatol 1972; 105: 107–10.
2. Ott M, Tanabe K, Gadd M, et al. Multimodality management of Merkel cell carcinoma
. Arch Surg 1999; 134: 388–33.
3. Metz KA, Jacob M, Schmidt U, et al. Merkel cell carcinoma
of the eyelid: histological and immunohistochemical features with special respect to differential diagnosis. Graefes Arch Clin Exp Ophthalmol 1998; 236: 561–6.
4. Schmidt U, Muller U, Metz KA, et al. Cytokeratin and neurofilament protein staining in Merkel cell carcinoma
of the small cell type and small cell carcinoma of the lung. Am J Dermatopathol 1998; 20: 346–51.
5. Chan JK, Suster S, Wenig BM, et al. Cytokeratin 20 immunoreactivity distinguishes Merkel cell (primary cutaneous neuroendocrine) carcinomas and salivary gland small cell carcinomas from small cell carcinomas of various sites. Am J Surg Pathol 1997; 21: 226–34.
6. Shah IA, Netto D, Schlageter MO, et al. Neurofilament immunoreactivity in Merkel-cell tumors: a differentiating feature from small-cell carcinoma. Mod Pathol 1993; 6: 3–9.
7. Yiengpruksawan A, Coit D, Thaler H, et al. Merkel cell carcinoma
: prognosis and management. Arch Surg 1991; 126: 1514–9.
8. Allen P, Zhang ZF, Coit D. Surgical management of Merkel cell carcinoma
. Ann Surg 1999; 229: 97–105.
9. Goepfert H, Remmler D, Silva E, et al. Merkel cell carcinoma
(endocrine carcinoma of the skin) of the head and neck
. Arch Otolaryngol 1984; 110: 707–12.
10. Kokoska ER, Kokoska MS, Collins BT, et al. Early aggressive treatment for Merkel cell carcinoma
improves outcome. Am J Surg 1997; 174: 688–93.
11. Bielamowicz S, Smith D, Abemayor E. Merkel cell carcinoma
. An aggressive skin neoplasm. Laryngoscope 1994; 104: 528–32.
12. O’Connor W, Brodland D. Merkel cell carcinoma
. Dermatol Surg 1996; 22: 262–7.
13. Victor NS, Morton B, Smith JW. Merkel cell cancer: is prophylactic lymph node dissection indicated? Am Surg 1996; 62: 879–82.
14. Suntharalingam M, Rudoltz M, Mendenhall W, et al. Radiotherapy
for Merkel cell carcinoma
of the skin of the head an neck. Head Neck 1995; 17: 96–101.
15. Smith D, Bielamowicz S, Kagan A, et al. Cutaneous neuroendocrine carcinoma. Am J Clin Oncol 1995; 18: 199–204.
16. Shaw JHF, Rumball E. Merkel cell tumour: clinical behaviour and treatment. Br J Surg 1991; 78: 138–42.
17. Messina J, Reintgen D, Cruse C, et al. Selective lymphadenectomy in patients with Merkel cell carcinoma
. Ann Surg Oncol 1997; 4: 389–95.
18. Pilotti S, Rilke F, Bartoli C, et al. Clinicopathologic correlations of cutaneous neuroendocrine Merkel cell carcinoma
. J Clin Oncol 1988; 6: 1863–73.
19. Wilder R, Harari P, Graham A, et al. Merkel cell carcinoma
: improved locoregional control with postoperative radiation therapy. Cancer 1991; 68: 1004–8.
20. Hasle H. Merkel cell carcinoma
: the role of primary treatment with radiotherapy
. Clin Oncol 1991; 3: 114–6.
21. Cotlar A, Gates J, Gibbs F. Merkel cell carcinoma
: combined surgery and radiation therapy. Am Surg 1986; 52: 159–64.
22. Morrison W, Peters L, Silva E, et al. The essential role of radiation therapy in securing locoregional control of Merkel cell carcinoma
. Int J Radiat Oncol Biol Phys 1990; 19: 583–91.
23. Silva EG, Mackay B, Goepfert H, et al. Endocrine carcinoma of the skin (Merkel cell carcinoma
). Pathol Annu 1984; 19: 1–30.
24. Mendenhall WM, Million RR, Mancuso AA, et al. Carcinoma of the skin. In: Million RR, Cassisi NJ, eds. Management of head and neck
cancer: a multidisciplinary approach. 2nd ed. Philadelphia:JB Lippincott,
25. Taylor BW, Brant TA, Mendenhall NP, et al. Carcinoma of the skin metastatic to parotid area lymph nodes. Head Neck 1991; 13: 427–33.
26. Amdur RJ, Parsons JT, Mendenhall WM, et al. Postoperative irradiation for squamous cell carcinoma of the head and neck
: an analysis of treatment results and complications. Int J Radiat Oncol Biol Phys 1989; 16: 25–36.
27. Meeuwissen J, Bourne R, Kearsley J. The importance of perioperative radiation therapy in the treatment of Merkel cell carcinoma
. Int J Radiat Oncol Biol Phys 1995; 31: 325–31.
28. Wynne CJ, Kearsley JH. Merkel cell tumor: a chemosensitive skin cancer. Cancer 1998; 62: 28–31.
29. Fuen L, Savaraj N, Legha S, et al. Chemotherapy for metastatic Merkel cell carcinoma
. Cancer 1988; 62: 683–5.
30. Mercer D, Brander P, Liddell K. Merkel cell carcinoma
: the clinical course [Letter]. Ann Plast Surg 1990; 25: 136.
31. Kroll MH, Toker C. Trabecular carcinoma of the skin: further clinicopathologic and morphologic study. Arch Pathol Lab Med 1982; 106: 404–8.
32. Sibley RK, Dehner LP, Rosai J. Primary neuroendocrine carcinoma of the skin: I. A clinicopathologic and ultrastructural study of 43 cases. Am J Surg Pathol 1985; 9: 95–108.
33. O’Brien PC, Denham JW, Leong AS-Y. Merkel cell carcinoma
: a review of behavior patterns and management strategies. Aust N Z J Surg 1987; 57: 847–50.