Cancer 80:60-5, 1997
Copyright
© 1997 American Cancer
Society. All rights reserved.
Published by John
Wiley & Sons, Inc.
Bettina Schlagenhauff, M.D., Waltraud Stroebel, Ph.D., Ulf Ellwanger, M.S., Friedegund Meier, M.D., Caroline Zimmermann, M.D., Helmut Breuninger, M.D., Gernot Rassner, M.D., Claus Garbe, M.D.
Section of Dermatologic Oncology, Department of Dermatology, Eberhard-Karls-University, Tuebingen, Germany.
Received November 25, 1996; revision received February 28, 1997; accepted February 28, 1997.
BACKGROUND. Metastatic melanoma of unknown primary origin accounts for approximately 2-6% of all melanoma cases. The prognostic significance of this diagnosis is still controversial.
METHODS. Of 3258 patients with malignant melanoma recorded during the period 1976-1996, 2.3% had metastases of unknown primary origin. Anatomic distribution, clinical stage, and survival probabilities were evaluated.
RESULTS. Thirty patients were classified as having cutaneous or subcutaneous in-transit metastases, and they showed a 5-year survival rate of 83%. Thirty-seven patients were classified as having lymph node metastasis, and their 5-year survival rate was 50%. Disseminated disease was diagnosed in only 8 patients, who had a median survival of 6 months. Comparison of survival probabilities for patients with in-transit metastases and unknown primary tumors with the probabilities for those with cutaneous primary tumors revealed a significant advantage for the former group. No significant differences were found for patients with lymph node metastasis when those with unknown primary tumors were compared with those who had cutaneous melanomas with regional lymph node metastasis.
CONCLUSIONS. The clinical disease course of patients with metastatic melanoma of unknown primary origin is similar to that of patients with primary cutaneous melanoma when the same clinical stages of the disease are compared. Based on the assumption that the majority of regional metastases develop from completely regressed primary cutaneous melanoma, recommendations for initial staging examinations in patients with unknown primary tumors are given in this article.
metastatic melanoma, unknown primary, survival rates, staging.
The diagnosis of metastatic melanoma of unknown primary origin is rather frequent, made in 2-15% of all melanoma cases.1-3 Most authors estimate that 2-6% of patients are diagnosed with metastatic melanoma of unknown primary site. Metastases from melanoma most often develop as cutaneous or subcutaneous nodules or as lymph node metastases. A varying percentage of cases have simultaneous visceral metastases at the time of diagnosis.
Metastatic melanomas of unknown primary origin were extensively investigated for the first time by Das Gupta et al. in a series of 47 patients identified in a collective of 992 melanoma cases (3.7%).3 This series, published in 1963, excluded all patients in whom moles had been excised, electrodesiccated, or cauterized. It was assumed that each of these lesions might have been a primary tumor unless the histologic slides were available for subsequent review and a melanoma was not diagnosed. In addition, patients were excluded if a scar from a previous local treatment in the skin area drained by the respective lymphatic basin was observed. Das Gupta et al. proposed to examine patients thoroughly and perform physical examinations, including ophthalmoscopy and examination of the anogenital region. It is noteworthy that this group of authors did not report any primary tumors found in these physical examinations.
The prognostic significance of metastatic melanoma of unknown primary origin has been interpreted differently by various authors. Das Gupta et al. reported that 30% of their patients survived 5 years.3 Of a group of 98 patients at the M. D. Anderson Cancer Center in Houston, Texas, one-third had visceral metastases and a median survival of 6 months. Approximately two-thirds had lymph node metastases and a 5-year survival rate of 33%.2 The largest group of patients, reported from the Memorial Sloan-Kettering Cancer Center in New York, comprised a total of 166 patients. Two-thirds of these patients had visceral metastases, and their median survival was also 6 months. Patients with lymph node metastases showed an unexpectedly high survival rate (5-year survival rate of 46%).4
The aim of the current study was to identify patients with metastatic melanoma of unknown primary origin in a more recent large group of patients followed by a dermatologic department. Cases with unknown primary tumor were compared with cases of metastatic melanoma that developed from primary cutaneous melanoma, and their survival probabilities were calculated. The prognostic significance of metastatic melanoma of unknown primary origin is discussed herein.
During the period 1976-1996, a total of 3258 melanoma patients were recorded in the melanoma registry of the Department of Dermatology at Eberhard-Karls-University in Tuebingen, Germany. Informed consent was obtained from all patients included in this study. In seventy-five cases, the disease was identified as metastatic melanoma of unknown primary origin. The diagnoses of these cases were carefully reviewed by applying the criteria proposed by Das Gupta et al.3 These criteria had been already prospectively applied during the period of registration. Only 3 cases were withdrawn from this study due to the previous removal of potential primary melanomas without histologic examination.
At the time of diagnosis, staging examinations were performed for all patients. These included inspection of the whole skin and the adjacent mucosae, palpation of the lymphatic basins and the abdomen, chest X-ray, and/or computed tomography (CT) and abdominal ultrasound and/or CT. The following technical examinations were carried out, for varying percentages of patients: ultrasound of the lymph nodes (for 61% of patients), CT of the brain (79%), bone scan (72%), otorhinolaryngologic examination (59%), ophthalmologic examination (67%), gastroscopy (41%), coloscopy or rectoscopy (46%), and gynecologic examination (50% of females). All patients had regular follow-up at intervals of 3-6 months. The mean time of follow-up was 44 months; the median was 22 months.
The descriptive evaluation of data was performed using the statistical package SPSS for Windows, version 6.0 (SPSS Inc., Chicago, IL). Survival probabilities were calculated by the Kaplan-Meier method, and statistical testing was performed with the log rank test.
The percentage of cases of metastatic melanoma of unknown primary origin among all melanoma cases registered was 2.3%.
Gender and Age
Thirty-five patients were men and 40 patients were women; the male-to-female ratio was 0.9:1. This ratio was identical to that of the entire collective with primary cutaneous melanoma (P = 0.4). The median age of patients with unknown primary was 53 years. Patients with primary cutaneous melanoma were slightly younger, with a median age of 51 years. It is noteworthy that patients diagnosed with lymph node metastasis in association with primary cutaneous melanoma also had a median age of 53 years. Patients diagnosed with visceral metastasis at the time of diagnosis, however, were clearly older, with a median age of 70 years of age (Fig. 1 (5K)).
Sites of Metastases
Thirty-eight patients presented primarily with cutaneous or subcutaneous metastasis, 34 with lymph node metastasis and 3 with visceral metastasis. The sites of initial metastases were as follows, in order of frequency: axillae (25.3%), head and neck region (21.3%), upper extremities (14.7%), trunk and lower extremities (each 13.3%), groin (6.7%), and other locations (5.3%). In female patients, metastases to the head and neck and the extremities were observed most often; in male patients, metastases to the axillae predominated, followed by head and neck and trunk (Table 1).
Staging at the Time of Diagnosis
After technical examinations, the patients were classified according to the AJCC staging system. In addition, Stage III was divided into IIIa (in-transit metastasis) and IIIb (lymph node metastasis). Thirty patients were diagnosed as Stage IIIa, 37 as Stage IIIb, and 8 patients as Stage IV (visceral metastasis). Most female patients were classified as Stage IIIa, most male patients as Stage IIIb (Table 2). Visceral metastases were detected in 6 patients by chest X-ray or CT, in 2 patients by abdominal ultrasound or CT, and in 4 patients by brain CT. In addition, visceral metastases were detected by gastroscopy in two patients (Table 3). These two patients complained of gastrointestinal symptoms, and the subsequently performed endoscopy showed the first manifestation of the disease.
No metastases of melanoma were found by bone scan, coloscopy or rectoscopy, ophthalmoscopy, or otorhinolaryngologic examination.
Clinical Course and Survival
Cutaneous or subcutaneous metastases were surgically removed with safety margins of 1-2 cm. No elective regional lymph node dissections were performed. In patients in whom suspicious lymph nodes were detected by physical examination or by ultrasound, a radical lymph node dissection was performed.
The overall 5-year survival rate was 59 %. The difference in survival was not statistically significant between males (51%) and females (67%) (P = 0.3). Moreover, there were no significant differences in survival according to 2 different age groups (older than 52 years and 52 years or younger, P = 0.4).
The most favorable survival rate was found for patients with Stage IIIa disease. Their 5-year survival rate was 83%, which was significantly higher than that of patients with primary cutaneous melanoma and subsequent in-transit metastasis (5-year survival rate, 50%; P = 0.02) (Fig. 2 (5K)). The survival of patients with lymph node metastasis tended to be better than that of patients with primary cutaneous melanoma and subsequent lymph node metastasis. However, the difference was not statistically significant. In this group, the 5-year survival rate was 50% for patients with unknown primary and 36% for patients with primary cutaneous melanoma (P = 0.14) (Fig. 3 (5K)). The median survival was 6 months for patients with unknown primary and visceral metastasis and 5 months for those with primary cutaneous melanoma and subsequent visceral metastasis (P = 0.14) (Fig. 4 (5K)).
In 2.3% of our patients with malignant melanoma, metastases were found without identification of a primary tumor. This percentage lies within the lower range reported in the literature.2,4-6 A nearly equal proportion of men and women were affected in our study, whereas most other series showed a clear preponderance of male patients.2,4,7,8 The majority of patients were diagnosed in the fifth decade of life; this was comparable to the peak incidences reported in the literature.1,2,5,8 The median age was identical to that of patients with primary cutaneous melanoma and simultaneous lymph node metastasis at the time of diagnosis. Anticipating the conclusions of our study, we adopted the staging system of primary cutaneous melanoma for the patients with metastatic melanoma of unknown primary origin. The precondition for the Stage III classification was that only one lymphatic drainage area was affected. Most patients were classified as Stage IIIa (in-transit metastasis) or Stage IIIb (lymph node metastasis) after staging examinations. Only eight patients had visceral metastasis; this was a rather low number compared with other series. This may be explained by the fact that most studies of metastatic melanoma with unknown primary have been performed in surgical departments, where patients with lymph node or visceral metastases are seen more frequently than patients with isolated cutaneous or subcutaneous metastases.2,4
It is noteworthy that in men the metastases were located predominantly in the axillae, the head and neck region, and the trunk, whereas in women metastases were found more frequently in the extremities. This indicates similarities according to the gender-related patterns of primary melanomas, a finding that was described in a previous study as well.2
The survival of patients with unknown primary melanoma was demonstrated to be similar to that of patients with known primary tumors when corresponding stages were compared.1,7 In this study, the 5-year survival rate of patients with Stage IIIa disease (in-transit metastases) was 83%, which seemed surprisingly high compared with the corresponding control group with known primary tumors (50%). The patients in our study did not receive any specific adjuvant therapy, which could explain this discrepancy. To the best of our knowledge, such a high rate of survival has not been reported in the literature previously.
In patients with lymph node metastasis, there was no significant difference in 5-year survival rates between patients with unknown primary tumors and those with primary cutaneous melanomas, although the patients with unknown primary tumors had slightly better outcomes. Other authors also found no difference in survival between these groups of patients.2,6,7,9-12 In some series, patients with unknown primary tumors had slightly better survival.1,4,10,13 However, one author showed a worse prognosis for those patients.14
Almost all reviewed series, including our own, showed a similar unfavorable prognosis for patients with Stage IV disease and unknown primary melanoma, with no significant difference observed in the control group with primary cutaneous melanoma. In one study, a shorter survival of patients with unknown primary tumors was found.15
Thus, patients with regional metastases (Stage IIIa and Stage IIIb) and unknown primary melanoma have a prognosis similar to those with primary cutaneous melanoma. For patients with in-transit metastasis, an even higher survival rate was found, although it has to be taken into account that the number of this subgroup of patients was low. Some authors speculated that patients with unknown primaries have altered immune reactions compared with patients who have primary cutaneous melanoma, but no consistent differences were found regarding humoral and cellular immune responses.16
To explain the development of melanoma metastases with unknown primary origin, two main possibilities have been considered: (1) Complete regression of the primary melanoma after metastasis has occurred. Partial spontaneous resolution of melanoma cells from the primary site is fairly common. More rarely, patients reported the spontaneous disappearance of a pigmented tumor.3,17 (2) The primary origin of the melanoma is in lymph nodes or other subcutaneous tissues or in viscera. Ectopic benign nevus cells have sometimes been found in lymph nodes and other tissues.3 Melanoma arising from nevus cells in a lymph node was also reported in one study.18
In conclusion, the clinical course of cutaneous/subcutaneous or lymph node metastasis and unknown primary tumor is very similar to that of regional metastasis in cutaneous melanoma. Presumably the majority of these metastases develop from a subsequently regressing melanoma of the skin located in the anatomic area drained by the affected lymphatic basin. Therefore, we suggest that the initial search for a primary tumor in these patients be restricted to an inspection of the entire skin and the adjacent mucosae as well as a detailed examination of the areas drained by the affected lymphatic basin. In patients with only visceral metastasis, a more extensive check-up should be performed. As in all patients with metastatic melanoma, the patient with melanoma metastases of unknown primary origin should initially be thoroughly evaluated to determine the stage of the disease (Table 4). The treatment should be chosen according to the same criteria used in determining treatment for metastatic cutaneous melanoma.
Presented at the 39th Conference of the German Dermatological Society, Karlsruhe, Germany, April 23-26, 1997, and the Fourth World Conference on Melanoma, Sydney, Australia, June 10-14, 1997.
Address for reprints: Bettina Schlagenhauff, M.D., Department of Dermatology, Eberhard-Karls-University, Liebermeisterstr. 25, 72076 Tuebingen, Germany.