Long-term survivors in stage IV melanoma: a regional population-based study in France

ARTICLE

Auteur(s) : Tatiana Tchen¹, Fabienne Léonard¹, Christian Derancourt^1,2, Géraldine Perceau¹, Céline Goutorbe¹, Ziad Reguiaï¹, Philippe Bernard¹, Florent GrangeFlorent Grange¹

¹Department of Dermatology, Hôpital Robert Debré, Avenue du Général Koenig, Reims 51100, France
²Department of Dermatologie, Hôpital Pierre Zobda Quitman, BP 632, 97261 Fort de France Cedex, Martinique

accepté le 12 Août 2008

Stage IV melanoma was defined in the 6th edition of the American Joint Committee on Cancer (AJCC) by the presence of distant metastases [1]. The prognosis is poor, with a median survival time of 6 to 8 months [2]. However, there are rare cases of patients with unusual long-term survival. These patients, referred to as long survivors (LSs), have been reported in studies with various designs. Most reports derive from phase 2 clinical trials [3-7], in which investigators often base their arguments on a few LSs to emphasize the possible efficacy of assessed therapies. However, such hypotheses have rarely been confirmed in randomized trials and the role of specific therapies in long-term survival remains uncertain. In observational studies of a large number of patients with stage IV melanoma, 10 patients out of 635 (1.6%) [8], 5 out of 153 (3.3%) [9], and 10 out of 503 (2%) [10] were found to be alive 5 years after entering stage IV.

Lastly, survival analyses using the Kaplan-Meier method were performed in reference center-based large series of patients with stage IV melanoma. These studies reported survival rates of 8% at 1 year [11], 11% at 2 years [12] and 6 to 7% at 5 years [2, 13]. In one of the larger studies to date, Balch et al. evaluated the 5-year survival rate of 1158 patients with stage IV melanoma at around 10%. In this study, the site of distant metastases and the lactic dehydrogenase level were found to have significant value in predicting survival.

Although these earlier studies provide different evaluations of the proportion of LSs among patients with stage IV melanoma referred to reference centers, details of the clinical histories of these exceptional cases are often lacking and few studies have been performed on a population basis. Our aim was to identify LSs among patients with stage IV melanoma living in a defined geographic region over a specific period of time, to evaluate their clinical history, spontaneous evolution and responses to various therapies, and to find out their possible clinical and biological characteristics.

Materials and methods

Geographical setting

The crude incidence of melanoma in Champagne-Ardenne was punctually estimated at around 13/100,000 in 1998 [14] but more accurate and long-term assessments are not available due to the absence of cancer registries. Data on mortality, however are available through exhaustive reviewing of death certificates (Center of Epidemiology on Medical Causes of death, Inserm) [15]. Between 1 January 1994 and 31 December 2003, 316 deaths related to melanoma were reported in the Champagne-Ardenne region, in 170 male and 146 female patients, respectively. Our aim was i) to identify patients in this region who had developed a stage IV disease during this 10-year period and had a subsequent survival time of 4 years or longer, and ii) to study their characteristics, clinical history and outcome.

Pre-selection and inclusion

• 1) Visceral and/or distant metastases of melanoma (AJCC stage IV) occurring within the study period irrespective of previous history.
• 2) Clinical evidence or histological confirmation of distant metastases.
• 3) Survival time of 4 years or longer after diagnosis of stage IV disease.

Patients with uncertain stage IV melanoma and/or suspicion of distant metastases on imaging studies without histological assessment were excluded.

Data collection

Complete remissions (CRs) were defined by the absence of any detectable disease on clinical examination and appropriate imaging studies (most often Computed Tomography (CT) scans), lasting for at least 3 months. In addition, patients alive in June 2006 were investigated for biological features of auto-immunity, i.e. antithyroid, anticardiolipin and antinuclear antibodies. Blood samples were tested by enzyme-linked immunosorbent assays. A positive result was defined as a titration of greater than 1/100 for antinuclear antibodies, > 150 UI/mL for antithyroid antibodies, > 15 U GPL for Ig G anticardiolipin antibodies.

Follow-up and statistical analysis

Results

These 10 LSs represent 3.5% of the 316 patients who died of melanoma in the Champagne-Ardenne region during the 10-year study period.

We first present the detailed clinical history of 3 patients and then summarize the data in the entire series (tables 1 and 2).

Case reports

Patient 4: a 62-year-old woman was treated for a thick nodular melanoma of the right leg with regional lymph node metastases. She successively underwent an excision of the primary tumor, two regimens of cisplatin and dacarbazin and a complete lymph node dissection. She presented with a contra-lateral dissemination 18 months later, including a tumor of the left buttock, and metastases of iliac lymph nodes and the iliac bone. A treatment with 9 infusions of fotemustin and/or dacarbazin followed by radiotherapy resulted in a slow regression of all detectable disease within 10 months, as she developed a progressive vitiligo. She was alive without evidence of disease 60 months after the diagnosis of stage IV disease.

Patient 6: a 45-year-old man with no known history of melanoma underwent two pulmonary lobectomies at 4 months’ interval for isolated tumors of both lungs. The diagnosis of carcinoma and sarcoma were discussed on histological examination, without definitive conclusion. He presented 1 year later with acute abdominal pains due to intra-abdominal metastases. A large surgical resection was performed, including duodenopancreatectomy, splenectomy, cholecystectomy and total colectomy. Immunohistochemical examination revealed duodenal and multiple latero-aortic lymph node metastases of melanoma. A comprehensive immunohistochemical review of previous lung tumors also concluded to melanoma, with Protein S100 and HMB45 positive tumor cells identical to those observed in the duodenum and nodes. Further regular clinical and CT-scan examinations were negative for 3 years, when small pulmonary and abdominal images were reported by the radiologist to be suggestive of lung and node metastases, before they spontaneously disappeared within 3 months. The patient received 3 infusions of dacarbazin at this time. He was alive without detectable disease 41 months after the diagnosis of lung tumors.
Table 1 Clinical history and outcome in 10 long-term survivors with stage IV melanoma

Table 2 Complete remissions (CR) in 10 long-term survivors with stage IV melanoma

Data in the entire series

Sites and number of metastases at stage IV diagnosis are shown in table 1. Histological confirmation of distant metastases was obtained in all cases except in patient 4, who had clinical evidence of dissemination. All but 1 patient had only 1 metastatic site. Five patients had a single detectable metastasis and 5 had multiple ones. The AJCC stage IV at diagnosis was M1a, M1b and M1c in 1, 4 and 5 cases respectively. Initial therapies after diagnosis of stage IV disease were surgery; surgery and chemotherapy; surgery, chemotherapy and interferon; chemotherapy alone; and chemotherapy and radiotherapy in 4, 2, 1, 1 and 1 cases, respectively. Further metastatic sites and therapies during the course of the disease are summarized in table 1.

Patient 9 never achieved a CR and had a stable disease under maintenance chemotherapy until he died of acute leukaemia 55 months after diagnosis of lung metastases. The other 9 patients had 1 to 4 (median: 2) apparent CRs during the course of their disease. A total of 18 probable CRs was documented in the entire series (table 2). CRs occurred for single metastasis in 8 cases and for multiple metastases in 10 cases. Sites of regressive metastases included the lung (6), liver (2), skin (2), small intestine (2), bones (2), pancreas (1), brain (2), lymph nodes (3), muscles (1) and adrenal gland (1).

Treatments leading to CRs were of any type, including surgery alone (7 cases); surgery and chemotherapy (3 cases); surgery, chemotherapy and interferon (1 case); chemotherapy alone (1 case); radiotherapy and chemotherapy (2 cases); and radiotherapy alone (1 case). Three apparent CRs occurred spontaneously. Patient 1 had a spontaneous regression of a histologically confirmed hepatic metastasis. The last 2 spontaneous CRs (patients 1 and 6) remain uncertain because they only concerned small suspicious images detected on systematic CT-scans without histological confirmation. The duration of CRs varied from 3 months to 139 months (mean: 26.4).

Three patients developed another cancer after the diagnosis of stage IV melanoma. Patient 9 died of a possibly chemo-induced acute leukaemia, after having received a total of 10 infusions of dacarbazine and 16 infusions of fotemustin. Patient 7 was treated for breast cancer but finally died of melanoma. Patient 3 had a lung epidermoid carcinoma and achieved a CR of both malignancies but finally died of a heart attack 11.5 years after stage IV melanoma diagnosis.

At the end point on 31 March 2007, 4 patients were alive without disease, 1 had a progressive disease, 3 had died of melanoma and 2 had died of other causes. The survival time from the diagnosis of stage IV disease varied from 52 to 139 months (mean: 70.4; median: 65). The survival time from the first diagnosis of melanoma varied from 66 to 262 months (mean: 132.9; median: 109).

Results of auto-immune investigation are shown in table 1. Overall, 4 of 8 patients still alive in June 2006 had clinical or biological features of auto-immunity.

Discussion

Few reports have focused on characteristics and clinical history of LSs [9, 10, 12]. Eton et al. [12] described 36 patients with survival greater than 2 years, 10 of whom had exclusive extravisceral metastases. He found that LSs were more frequently female patients, and were characterized by frequent nodal, cutaneous or single visceral metastases, complete surgical resections and achievement of complete remissions. Ahmann et al. [10] described 10 patients with a survival of more than 5 years among 503 patients included in 25 clinical trials. Only 3 of these 10 patients had experienced a complete response to chemotherapies investigated within these trials, suggesting that long-term survival may have been independent of these therapies.

Our patients were characterized by a long delay from the primary tumor to distant metastases and by the frequent occurrence of prolonged CRs, which even included spontaneous regressions. The mean delay of nearly 70 months to stage IV disease was much longer than the 3 years generally observed in the natural course of progressive melanoma [11].

CRs frequently concerned single or lung metastases, but were also observed for multiple tumors and metastases of any site, including the liver, brain, pancreas and bones (table 2). These CRs most often followed exclusive or combined surgery, but were also achieved after any kind of treatment, including chemotherapy, radiotherapy and interferon. We also observed spontaneous regressions. While this was histologically documented in only one case, our data suggest that LSs may have small visceral metastases that do not increase in size and may even eventually disappear as a result of spontaneous immune control. However, only exceptional cases of spontaneous regression of visceral metastases have been reported to date [16].

All these data on a population basis suggest that unusual prolonged survival in a small minority of patients with stage IV melanoma might depend less on treatment choices or location of metastases than on the patients themselves and their spontaneous antitumor control. A long free interval between the diagnosis of primary melanoma and distant metastases may reflect a slow tumoral kinetic, secondary to a state of relative equilibrium between tumor cells and immune regulatory cells. The long delay between the occurrence of regional nodal metastases, and progression to stage IV, and the long duration of CRs observed in our patients, may result from the same persistent immune control.

Several investigators have demonstrated that melanoma is an immunogenic tumor [17]. Both cellular and humoral immune responses to melanoma-associated antigens have been observed.

Therefore, immunotherapy with various cancer vaccines or adoptive immunotherapy have been evaluated at different stages of the disease [18-21]. In stage IV, prolonged survival was mainly observed when vaccines were used following complete resection of metastases (Hsuech [22] and Tagawa [23]), which suggests a better efficacy of this approach in the context of slowly growing tumors.

It has also been demonstrated that vitiligo is associated with an improved prognosis in melanoma patients of different stages. In stage IV, a recent prospective study showed that, among patients treated with interleukin 2 as maintenance therapy, those who developed vitiligo had a longer survival than those without vitiligo [24]. In this context, vitiligo could be an auto-immune response to melanoma-associated antigens shared by normal melanocytes and melanoma cells. Other auto-immune reactions have been suspected to be associated with a favorable prognosis in melanoma. Gogas et al. [25] found a significant association between the development of auto-immune manifestations (including vitiligo, and/or antinuclear, anticardiolipin and antithyroid antibodies) and improved overall and relapse-free survival in patients treated with adjuvant high-dose interferon for stage II and III melanoma. In a recent study of 283 patients in stage IV, Vercambre-Darras et al. [26] identified 15 LSs (5.3%) who were still alive more than 3 years after stage IV diagnosis. Among 10 of these patients who were investigated for the presence of auto-antibodies, 5 (50%) had antithyroid antibodies, as compared to 10% in a control group of non-LS/stage IV patients. The authors suggested that spontaneous auto-immunity in stage IV melanoma could be linked to a better anti-tumor control and a longer survival. In our series, 4 of 8 patients still alive in June 2006 had clinical or biological features of auto-immunity, as defined by Gogas et al. [25] Despite the absence of a control group, this result seems in accordance with these recent hypotheses. Our population-based clinical study suggests that long survival in stage IV melanoma probably depends on a prolonged equilibrium between disseminated tumor cells and the host antitumor response. Although we (and others) have observed persistent complete responses for up to 10 years following stage IV diagnosis [14], an escapement to anti-tumor control eventually occurs in many cases, with faster proliferation, poorer responses to therapies and death long after diagnosis. Further genetic and immunological studies and specific clinical trials in these rare long-survivor patients could lead to a better comprehension of the disease and to the development of new immunological therapies.

Acknowledgements

References

2 Barth A, Wanek LA, Morton DL. Prognostic factors in 1521 melanoma patients with distant metastasis. J Am Coll Surg 1995; 18: 193-201.

3 Legha SS. Durable complete responses in metastatic melanoma treated with interleukin-2 in combination with interferon alpha and chemotherapy. Semin Oncol 1997; 24(1 suppl 4): S39-S43.

4 Samuel LM, Harvey VJ, Mitchell PL, et al. Phase II trial of procarbazine, vincristine and lomustine (POC) chemotherapy in metastatic cutaneous malignant melanoma. Eur J Cancer 1994; 30A: 2054-6.

5 Margolin KA, Doroshow JH, Akman SA, et al. Phase II trial of cisplatin and alpha-interferon in advanced malignant melanoma. J Clin Oncol 1992; 10: 1574-8.

6 Berd D, Mastrangelo MJ. Combination chemotherapy of metastatic melanoma. J Clin Oncol 1995; 13: 796-7.

7 Kokoschka EM, Klocker J, Loicht U, et al. Interferon-alpha-carboplatin combination for metastatic melanoma: phase II study. Eur J Cancer 1994; 30A(14): 2182-3.

8 Ryan L, Kramar A, Borden E. Prognostic factors in metastatic melanoma. Cancer 1993; 71: 2995-3005.

9 Petit T, Janser JC, Petit JC. Complete remission seven years after treatment for metastatic malignant melanoma. Cancer 1996; 78: 571.

10 Ahmann DL, Creagan ET, Hahn RG, et al. Complete responses and long-term survivals after systemic chemotherapy for patients with advanced malignant melanoma. Cancer 1989; 63: 224-7.

11 Balch CM, Soong SJ, Murad TM, et al. A multifactorial analysis of melanoma IV. Prognostic factors in 200 melanoma patients with distant metastases (stage III). J Clin Oncol 1983; 1: 126-34.

12 Eton O, Legha SS, Moon TE, et al. Prognostic factors for survival of patients treated systemically for disseminated melanoma. J Clin Oncol 1998; 16: 1103-11.

13 Brand CU, Ellwanger U, Stroebel W, et al. Prolonged survival of 2 years or longer for patients with disseminated melanoma. An analysis of related prognostic factors. Cancer 1997; 79: 2345-53.

14 Bernard P, Derancourt C, Arnoult-Caudoux E, et al. Association des Dermatologues de Champagne-Ardenne. Etude prospective de l’incidence des cancers cutanés dépistés en pratique dermatologique dans la région Champagne-Ardenne. Ann Dermatol Venereol 2001; 28: 883-7.

15 Grange F. Epidémiologie du mélanome cutané: données descriptives en France et en Europe. Ann Dermatol Venereol 2005; 132: 975-82.

16 Satzger I, Schenck F, Kapp A, et al. Spontaneous regression of melanoma with distant metastases- report of a patient with brain metastases. Eur J Dermatol 2006; 16: 454-5.

17 Turk MJ, Guevara-Patino A, Rizzuto GA, et al. Concommitant tumor immunity to a poorly immunogenic melanoma is prevented by regulatory T cells. J Exp Med 2004; 200: 771-82.

18 Van Baren N, Bonnet MC, Dréno B, et al. Tumoral and immunologic response after vaccination of melanoma patients with ALVAC virus encoding MAGE antigens recognized by T cells. J Clin Oncol 2005; 23: 9008-21.

19 Salcedo M, Bercovici N, Taylor R, et al. Vaccination of melanoma patients using dendritic cells loaded with an allogenic tumor cell lysate. Cancer Immunol Immunother 2006; 55: 819-29.

20 Khammari A, Nguyen JM, Pandolfino MC. Long-term follow-up of patients treated by adoptive transfer of melanoma tumor-infiltrating lymphocytes as adjuvant therapy for stage III melanoma. Cancer Immunol Immunother 2007; 56: 1853-60.

21 Quereux G, Pandolfino MC, Knol AC, et al. Tissue prognostic markers for adoptive immunotherapy in melanoma. Eur J Dermatol 2007; 17: 295-301.

22 Hsueh EC, Essner R, Foshaq LJ, et al. Prolonged survival after complete resection of disseminated melanoma and active immunotherapy with cancer vaccine. J Clin Oncol 2002; 20: 4549-54.

23 Tagawa ST, Cheung E, Banta W, et al. Survival analysis after resection of metastatic disease followed by peptide vaccines in patients with stage IV melanoma. Cancer 2006; 106: 1353-7.

24 Boasberg PD, Hoon DS, Piro LD, et al. Enhanced survival associated with vitiligo expression during maintenance biotherapy for metastatic melanoma. J Invest Dermatol 2006; 126: 2658-63.

25 Gogas H, Ionnovich J, Dafni U, et al. Prognostic significance of auto-immunity during treatment of melanoma with interferon. New Engl J Med 2006; 354: 709-18.

26 Vercambre-Darras S, Dubucquoi S, Fajardy I, et al. Does spontaneous autoimmunity improve survival in visceral metastatic melanoma? Br J Dermatol 2007; 157: 413-5.