- Email: [email protected]
Hepatic arterial Fotemustine chemotherapy in patients with liver metastases from cutaneous melanoma is as effective as in ocular melanoma
EJSO 33 (2007) 627e632
www.ejso.com
Hepatic arterial Fotemustine chemotherapy in patients with liver metastases from cutaneous melanoma is as effective as in ocular melanoma R. Siegel a, A. Hauschild b, C. Kettelhack a, K.C. Ka¨hler b, A. Bembenek a, P.M. Schlag a,* a
Department of Surgery and Surgical Oncology, Charite´dUniversita¨tsmedizin Berlin, Campus Buch, Robert-Ro¨ssle-Tumour-Hospital at Helios Klinikum Berlin-Buch, D-13125 Berlin, Germany b Department of Dermatology, University Hospital Schleswig-Holstein, Campus Kiel, Kiel, Germany Accepted 14 November 2006 Available online 28 December 2006
Abstract Aim: Hepatic metastases from melanoma are associated with poor prognosis. Systemic chemotherapy and biological treatments remain unsatisfactory. This study investigated the impact of hepatic arterial chemotherapy in patients with ocular and cutaneous melanoma. Methods: In a retrospectively analysed observational study, 36 consecutive patients with hepatic metastases from ocular or cutaneous melanoma were assigned for surgical hepatic port-catheter implantation. Fotemustine was delivered weekly for a 4-week period, followed by a 5-week rest and a maintenance period every 3 weeks until progression. Overall survival, response and toxicity were analysed and compared. Results: After port-catheter implantation 30/36 patients were finally treated (18 with ocular and 12 with cutaneous melanoma). A median of 8 infusions per patient were delivered (range 3e24). 30% thrombocytopenia grade 3, 7% neutropenia grade 3 but no nausea or vomiting grade 3 were encountered. Nine out of 30 patients achieved partial remission, 10/30 stable disease; 11/30 patients were progressive. Median survival for all treated patients was 14 months. Partial remission and stable disease were associated with a survival advantage compared to progressive disease (19 vs. 5 months). No significant difference in survival was observed for ocular versus cutaneous melanoma. Serum LDH was a significant predictor of both response and survival. Conclusions: Hepatic arterial Fotemustine chemotherapy was well tolerated. Meaningful response and survival rates were achieved in ocular as well as cutaneous melanoma. Careful patient selection in consideration of extra-hepatic involvement is crucial for the effectiveness of this treatment. Independent from the primary melanoma site, it is debatable if patients with highly elevated serum-LDH may benefit from this approach. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Hepatic arterial therapy; Fotemustine; Liver metastasis; Cutaneous melanoma; Ocular melanoma
Introduction Ocular melanoma is characterised by a high prevalence for liver metastases and an unusual propensity for late metastases.1,2 This pattern of dissemination differs from the one of cutaneous melanoma. In cutaneous melanoma, the main metastatic sites are the regional lymph nodes, subcutaneous tissue, and the lungs. Hepatic metastases occur in only 14e20% of metastatic cutaneous melanoma. Both ocular and cutaneous melanomas have high metastasis-related * Corresponding author. Tel.: þ49 30 9417 1400; fax: þ49 30 9417 1404. E-mail address: [email protected] (P.M. Schlag). 0748-7983/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.ejso.2006.11.019
mortality rates, with a median survival of only 5 to 7 months.3e5 Chemotherapy alone or in combination with interferon and/or interleukin-2 is ineffective in most cases of liver metastasis from ocular or cutaneous melanoma. In disseminated malignant melanoma, no prospective randomised study has proven an advantage of a distinct therapy regimen over another.6,7 The cytotoxic agent Fotemustine, a third-generation nitrosourea, has been administered as a systemic treatment first to patients with visceral metastases from cutaneous melanoma, resulting in a response rate of 13%.8 In patients with ocular melanoma, the response rate was 8% with a median survival of 12 months.9 Still, Dacarbazine is currently the standard therapy in metastatic cutaneous melanoma with response rates ranging
628
R. Siegel et al. / EJSO 33 (2007) 627e632
from 7% to 12% and a median survival time from 6 to 8 months.10e12 Recently, the results of a phase-III study comparing Fotemustine with Dacarbazine were published, confirming the activity of Fotemustine as a single agent in the first line treatment of metastatic cutaneous melanoma. Overall response rate was higher in the Fotemustine arm (in the intention to treat population) and a trend in favour of Fotemustine in terms of overall survival was evident.13 In addition to the systemic approach, certain centres have established regional treatment strategies for hepatic metastases. Complete resection is only possible in highly selected patients with limited disease. Debulking surgery that does not achieve a R0-resection did not improve the outcome of patients.14,15 In ocular melanoma, chemoembolising using Cisplatin and polyvinyl sponge resulted in a marked response, but has not become a widely accepted treatment modality because of important treatment-related morbidity.16 Intra-arterial chemotherapy with Cisplatin, Dacarbazine, and Vinblastine has been attempted in individuals, but never proved to be effective.17 Because of its high hepatic extraction rate at the first pass, Fotemustine was employed by the group of Leyvraz, who treated 31 patients by intra-arterial administration with this drug. This treatment resulted in a response rate of 40% with four patients in complete remission and a median survival of 14 months. The toxicity was acceptable with each neutropenia and thrombopenia grade IIIeIV in only 21% of the patients.18 Leyvraz’ data have recently been updated with 70 patients, resulting in a total of 101 patients from seven centres.19 Based on the results of the study from Leyvraz et al., intra-arterial therapy for liver metastases from ocular melanoma has been established in our institution. However, no comparable studies for local therapy of liver metastases from cutaneous melanoma were available. Because of the limited options in metastatic cutaneous melanoma and the unsatisfactory results obtained by systemic Fotemustine therapy, we applied the intra-arterial approach to patients with hepatic metastases from cutaneous melanoma as well. We conducted this study to compare the effectiveness of hepatic arterial Fotemustine infusion between liver metastases from ocular and cutaneous melanoma. In the following, we will review our results and discuss prognostic factors that are associated to survival. Methods
eligible if the primary tumour was locally controlled and no progressive extra-hepatic dissemination was evident. Pre-treatment evaluation included computed tomography (CT) or magnetic resonance imaging (MRI) of the abdomen, chest X-ray, whole blood count and biochemistry assessment (e.g. lactate dehydrogenasedLDH). Because of the palliative intention of treatment, abnormal results on liver function test were not an exclusion criterion. Catheter placement A totally implantable catheter connected to a port chamber was placed into the hepatic artery via the gastroduodenal artery. An angiography demonstrating the arterial road mapping was performed prior to catheter placement. To prevent misperfusion of the stomach and duodenum, collateral vessels including the right gastric artery were occluded. Prophylactic cholecystectomy was performed to avoid chemical cholecystitis. The port was placed in a subcutaneous pocket over the right rib cage. Patent blue injection via the catheter was performed to document perfusion of the whole liver and the absence of misperfusion. During surgery, liver metastases were histologically proven, unless diagnosis was already confirmed by prior biopsy. Prior to Fotemustine application, an angiogram via the port-catheter and/or a technetium scan was done to verify catheter position and liver perfusion. Hepatic arterial chemotherapy According to the protocol of Leyvraz, Fotemustine (MuphoranÒ) 100 mg/m2 was administered intra-arterially over 4 h.16 Anti-emetic treatment was given prophylactically using Ondasetron 8 mg intravenously and Methylprednisolone 125 mg intravenously. The induction phase consisted of the Fotemustine administration once a week for four times followed by a 5-week rest period and a maintenance phase with administration of one course of Fotemustine every 3 weeks until progression or unacceptable toxicity. The maintenance period was conducted only if partial response or stable disease was observed upon restaging. Follow-up and assessment of response were performed using CT or MRI after completion of the induction phase and subsequently controlled after every three or four cycles of the maintenance period. Assessment of response and toxicity was done according to WHO criteria.20 All patients were followed until death.
Patient selection Statistical methods Between June 1999 and December 2005, 36 consecutive patients were enrolled within the treatment protocol. Data was extracted retrospectively from a prospective clinical database. All patients with liver metastases from ocular or cutaneous melanoma considered not resectable were evaluated for hepatic intra-arterial treatment. Patients were
For statistical methods we used the Statistical Package for the Social Sciences Software program (SPSS, version 12.0, Chicago, IL, USA). Two-sample comparisons were conducted by the ManneWhitney test. Disease-specific survival rates were compared using the KaplaneMeier
R. Siegel et al. / EJSO 33 (2007) 627e632
method. Survival curves were tested for difference using the log rank test. A p value of less than 0.05 was considered significant.
Table 1 Characteristics of patients receiving hepatic arterial Fotemustine infusion Ocular melanoma (n ¼ 18)
Cutaneous melanoma (n ¼ 12)
Total (n ¼ 30)
p
Sex Male Female
9 9
7 5
16 14
0.722
Age, years Median Range
58 41e73
60 45e85
59 41e85
0.95
Characteristics
Results Catheter placement Thirty-six patients (18 men and 18 women) were enrolled to receive a port-catheter system into the hepatic artery. The median age at the time of surgery was 59 years (range 40e85 years). Twenty-two patients had a previously diagnosed ocular melanoma as their primary tumour, 14 patients were diagnosed with a cutaneous melanoma. In 31/36 patients open surgery was performed for the implantation of the port-catheter system, in 5/36 patients the catheter was implanted by minimal access surgery. Catheter related morbidity comprised 2 of 36 patients (one patient developed a severe pancreatitis, another patient presented with a port site infection). Three of 36 patients died due to tumour progress and liver failure within 30 days after surgery.
629
Time interval from diagnosis of melanoma to liver metastases, months Median 61 6 33 0.002 Range 1e265 0e93 0e265 Time interval from liver metastases to port-catheter implantation, months Median 2 2 2 0.884 Range 6e63 0e29 0e63 LDH, U/L Median Range
402 97e6050
338 171e2057
Intrahepatic fotemustine applications, n Median 9 7 Range 3e24 3e12
399 97e6050
0.285
8 3e24
0.305
LDH, lactate dehydrogenase; U/L, units per litre.
Fotemustine delivery and toxicity After implantation of the port-catheter system Fotemustine was administered according to the protocol in 30 of 36 patients. Six patients were excluded and did not receive Fotemustine because of rapid progressive disease and liver failure (n ¼ 5; 3/5 died within 30 days after surgery, see surgery-related mortality above) and a revised histopathologic diagnosis (revealing a hepatic manifestation of lymphoma; n ¼ 1). Eighteen of 30 patients suffered from ocular melanoma metastatic to the liver, whereas 12/30 patients had metastases from cutaneous melanoma. Characteristics of all 30 patients receiving intra-arterial Fotemustine chemotherapy are listed in Table 1. As suspected by the different pattern of metastases, the median time interval between diagnosis of the primary tumour and the detection of liver metastases differed between the two melanoma entities. Among the other baseline characteristics, there were no statistically significant differences. The 30 patients received a median of 8 intra-arterial Fotemustine applications, with a range from 3 to 24 (median 9 cycles in ocular vs. 7 cycles in cutaneous melanoma). The treatment could be administered on an outpatient basis. The toxicity comprised grade IIIeIV thrombocytopenia in 30%, grade IIIeIV neutropenia in 7%, nausea/vomiting in 17% (maximum grade IeII) and abdominal pain in 7% (maximum grade IeII). Therapeutic activity Two patients (one from both melanoma entities) were rapidly progressing under treatment and received only three
courses of the induction period. All other patients completed the induction phase and were evaluated for best response after four courses. Overall, 9/30 patients achieved partial remission (PR), 10/30 patients stable disease (no changedNC), and progressive disease (PD) was observed in 11/30 patients. The benefit in survival for patients with stable disease and partial remission was almost identical (NC with a median of 19 months vs. median 20 months for PR). Thereby response (PR and NC) to hepatic arterial therapy resulted in a survival advantage compared to patients with progressive disease (5 months). These results were still reproducible for the different melanoma subgroups. For cutaneous melanoma, the median overall survival was similar in patients with NC and PR (19 months for NC vs. 14 months for PR). This was different in ocular melanoma where median survival was significantly longer in patients with partial remission (38 months for PR vs. 14 months for NC; p ¼ 0.008). The median survival of all 30 patients with intra-arterial treatment was 14 months when calculated from the date of port-catheter implantation with 6 patients being still alive at the time of analysis. Patients with hepatic metastases from ocular melanoma achieved a prolonged survival (median 22 months) compared to patients with metastatic cutaneous melanoma (median 12 months), but this difference was statistically not significant (Fig. 1; p ¼ 0.306). Baseline serum LDH levels revealed a significant correlation to both response rates and survival. Patients presenting with a markedly elevated LDH level (>452 U/Ldtwice the upper normal limit) at the time of port-catheter implantation had a median survival of 5 months compared to median 33 months in patients with LDH levels below 452 U/L
R. Siegel et al. / EJSO 33 (2007) 627e632
630
1,0
1,0
Ocular Melanoma
LDH ≤ 452 U/L (2x UNL) LDH > 452 U/L (2x UNL)
Cutaneous Melanoma P = 0.306
0,8
Proportion Surviving
Proportion Surviving
0,8
0,6
0,4
0,2
P = 0.0006
0,6
0,4
0,2
0,0
0,0 0
10
20
30
40
50
Months after Implantation of Hepatic Port Catheter Figure 1. KaplaneMeier overall survival for all patients receiving intrahepatic Fotemustine (ocular melanoma, n¼18; cutaneous melanoma, n¼12) since placement of an intra-arterial catheter. Median overall survival for the whole group was 14 months, for patients with ocular melanoma 22 months and for patients with cutaneous melanoma 12 months ( p¼0.306). ‘‘þ’’ show censored observations.
( p ¼ 0.0006; Fig. 2). Analysing the subgroups, LDH is still very powerful for the prediction of response and survival, independent from the melanoma entity (Table 2). Discussion Little progress has been made in the treatment of metastatic cutaneous melanoma, resulting in a 5-year survival rate of only 7e19%.5 By increasing the tumour cells’ exposure to the cytotoxic agent and by reducing the exposure of the surrounding healthy tissues at the same time, intraarterial chemotherapy has a two-fold advantage in patients with liver metastases. The potential of this method of drug delivery for response and survival was recently re-emphasised by Kemeny et al. for patients with hepatic metastases from colorectal cancer.21 Since the report of Leyvraz et al., hepatic intra-arterial Fotemustine is accepted as an effective and well-tolerated regional treatment for liver metastases from ocular melanoma. Another regional approach, isolated hepatic perfusion with Melphalan (with or without TNF), following the concept of isolated limb perfusion for melanoma and sarcomas, has also been shown to induce a potent anti-tumour activity in ocular melanoma metastatic to the liver. However, the complex procedure and a considerable morbidity restricted this approach.22 Based on the experience from Leyvraz et al. and the results from systemic chemotherapy with Fotemustine in disseminated cutaneous melanoma, we applied hepatic arterial Fotemustine infusion in patients with liver metastases from both ocular and cutaneous melanoma.
10
0
20
30
40
50
Months after Implantation of Hepatic Port Catheter Figure 2. KaplaneMeier overall survival (for all 30 patients receiving intrahepatic Fotemustine) as it relates to LDH serum level at the time of port-catheter placement. ‘‘þ’’ show censored observations. LDH, lactate dehydrogenase; U/L, units per litre; 2 UNL, twice upper normal limit.
Response and survival We are aware of the limitations of this small series of patients. But by achieving partial remission or stable disease in almost two third of our patients, intra-hepatic Fotemustine infusion produced higher response and stabilisation rates compared to systemic application of any cytotoxic agent in malignant melanoma. Although the overall survival in patients with hepatic metastases from ocular melanoma (22 months) was higher than in patients with cutaneous melanoma (12 months), this difference was statistically not significant nor was a difference found in terms of response between the two melanoma entities. This is may be due to the limited number of patients. Although Table 2 Correlation of LDH serum levels and treatment response LDHa
Ocular melanoma Cutaneous melanoma Total
Partial response
Stable diseasedno change
Progressive disease
U/L
No. of pts.
No. of pts. (%)
No. of pts. (%)
No. of pts. (%)
<452 >452 <452 >452 <452 >452
10 8 8 4 18 12
5 0 3 1 8 1
4 2 4 0 8 2
1 6 1 3 2 9
(100) (0) (75) (25) (89) (11)
(67) (33) (100) (0) (80) (20)
(14) (86) (25) (75) (18) (82)
LDH, lactate dehydrogenase; U/L, units per litre. 452 U/L ¼ 2 upper normal limit. a LDH serum level at the time prior to port-catheter placement.
R. Siegel et al. / EJSO 33 (2007) 627e632
no definite conclusion may be drawn from the results achieved in these non-randomised series of 12 patients with cutaneous melanoma, our results with a median survival of 14 months for PR and 19 months for NC justify further studies with arterial Fotemustine treatment in melanoma patients with metastases confined to the liver. In cutaneous melanoma, the extra-hepatic progression of disease represents the main limitation of loco-regional therapy. Therefore, intra-arterial Fotemustine treatment could be considered preferentially in cases where extra-hepatic distant metastasis is not present or in patients with stable/ responding disease to systemic chemotherapy, where only hepatic metastases are progressive. Fotemustine related toxicity Compared to intravenous systemic Fotemustine therapy, our data highlight the favourable tolerance of Fotemustine delivered intra-arterially. Within the phase III study comparing Fotemustine with Dacarbazine, myelosuppression was the most common adverse event with grade 3 to 4 neutropenia in 51% of patients receiving Fotemustine and grade 3 to 4 thrombocytopenia in 43% of all patients in the Fotemustine arm.13 This exceeded by far the adverse events observed in our study with less than 7% patients presenting with grade 3 or 4 neutropenia and 30% with grade 3 or 4 thrombocytopenia. Our toxicity profile is comparable to the observed adverse events in the trial of Leyvraz, describing grade 3 or 4 neutropenia as well as thrombopenia in 21% each, and grade 3 or 4 nausea and vomiting in 7%. Lack of severe toxicities ensured the feasibility and acceptance of intra-arterially Fotemustine therapy on an outpatient basis. Catheter-related morbidity and mortality Complications related to the catheter have to be considered when comparing systemic and loco-regional therapy. The mortality after catheter implantation, before commencing chemotherapy, is a crucial issue and may arise from the preoperative reduced general performance status of patients with rapidly progressive disease and limited liver functions. The catheter related morbidity in our study was low with 6% and occurred after some weeks of treatment. This is in the range of the reports from the large series of 101 patients with ocular melanoma metastatic to the liver by Peters et al.19 Conclusion Hepatic arterial Fotemustine chemotherapy is well tolerated and seems to improve outcome of this poor prognosis population. In order to prevent perioperative mortality by catheter insertion and early therapy cessation, it is essential to select patients properly and define prognostic markers. The LDH value is a well-known negative prognostic factor
631
in melanoma,19,23,24 and was confirmed within our series of patients receiving intra-arterial Fotemustine. In addition to LDH, the protein S100b may be evaluated as another prognostic marker to assess the effect of intra-arterial therapy. It is still debatable whether patients presenting with a high serum LDH (e.g. more than twice upper normal limit, >452 U/L) will suffer from rapid progressive disease in most cases and therefore are unlikely to benefit from intra-hepatic Fotemustine application. This question may be addressed by a randomised study comparing best supportive care only vs. intra-arterial therapy for patients below a certain serum LDH level. However, because of ethical reasons and the limited number of eligible patients the realisation of studies with this design is at least questionable. Currently, European patients with liver metastases from ocular melanoma should preferentially be enrolled into the randomised phase III - EORTC Melanoma Group protocol 18021-88021, comparing intravenous versus intraarterial Fotemustine chemotherapy.25 For patients with cutaneous melanoma and liver metastases alone, further studies are warranted.
References 1. Pyrhonen S. The treatment of metastatic uveal melanoma. Eur J Cancer 1998;34(Suppl 3):S27–30. 2. Egan KM, Seddon JM, Glynn RJ, Gragoudas ES, Albert DM. Epidemiologic aspects of uveal melanoma. Surv Ophthalmol 1988;32:239–51. 3. Leiter U, Meier F, Schittek B, Garbe C. The natural course of cutaneous melanoma. J Surg Oncol 2004;86:172–8. 4. Gragoudas ES, Egan KM, Seddon JM, et al. Survival of patients with metastases from uveal melanoma. Ophthalmology 1991;98:383–9. 5. Balch CM, Soong SJ, Gershenwald JE, et al. Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 2001; 19:3622–34. 6. Eigentler TK, Caroli UM, Radny P, Garbe C. Palliative therapy of disseminated malignant melanoma: a systematic review of 41 randomized clinical trials. Lancet Oncol 2003;4:748–59. 7. Keilholz U, Punt CJ, Gore M, et al. Dacarbazine, cisplatin, and interferon-alfa-2b with or without interleukin-2 in metastatic melanoma: A randomized phase III trial (18951) of the European Organisation for Research and Treatment of Cancer Melanoma Group. J Clin Oncol 2005;23:6747–55. 8. Kleeberg UR, Engel E, Israels P, et al. Palliative therapy of melanoma patients with Fotemustine. Inverse relationship between tumor load and treatment effectiveness. Melanoma Res 1995;5:195–200. 9. Becker JC, Kaempgen E, Schadendorf D, Wagner S, Broecker EB. Treatment of disseminated uveal melanoma with sequential Fotemustine, Interleukin-2 and Interferon alpha2a. Proc ASCO 2000;572a. 10. Middleton MR, Grob JJ, Aaronson N, et al. Randomized phase III study of temozolomide versus dacarbazine in the treatment of patients with advanced metastatic malignant melanoma. J Clin Oncol 2000;18: 158–66. 11. Eggermont AMM, Kirkwood JM. Re-evaluating the role of dacarbazine in metastatic melanoma: what have we learned in 30 years? Eur J Cancer 2004;40:1825–36. 12. Bedikian AY, Millward M, Pehamberger H, et al. Bcl-2 Antisense (oblimersen sodium) plus dacarbazine in patients with advanced melanoma: The Oblimersen Melanoma Study Group. J Clin Oncol 2006; 24:4738–45.
632
R. Siegel et al. / EJSO 33 (2007) 627e632
13. Avril MF, Aamdal S, Grob JJ, et al. Fotemustine compared with Dacarbazine in patients with disseminated malignant melanoma: A phase III study. J Clin Oncol 2004;22:1118–25. 14. Rajpal S, Moore R, Karakousis CP. Survival in metastatic ocular melanoma. Cancer 1983;52:334–6. 15. Salmon RJ, Levy C, Plancher C, et al. Treatment of liver metastases from uveal melanoma by combined surgery-chemotherapy. Eur J Surg Oncol 1998;24:127–30. 16. Mavligit GM, Charnsangavej C, Carrasco CH, Patt YZ, Benjamin RS, Wallace S. Regression of ocular melanoma metastatic to the liver after hepatic arterial chemoembolization with cisplatin and polyvinyl sponge. JAMA 1988;260:974–6. 17. Bedekian AY, Legha SS, Mavligit G, et al. Treatment of uveal melanoma metastatic to the liver: a review of the M.D. Anderson Cancer Center experience and prognostic factors. Cancer 1995;76: 1665–70. 18. Leyvraz S, Spataro V, Bauer J, et al. Treatment of ocular melanoma metastatic to the liver by hepatic arterial chemotherapy. J Clin Oncol 1997;15:2589–95. 19. Peters S, Voelter V, Zografos L, et al. Intra-arterial hepatic Fotemustine for the treatment of liver metastases from uveal melanoma: experience in 101 patients. Ann Oncol 2006;17:578–83.
20. World Health Organization. Handbook for reporting results of cancer treatment. Geneva, Switzerland: World Health Organization; 1979. WHO publication no. 48. 21. Kemeny NE, Niedzwiecki D, Hollis DR, et al. Hepatic arterial infusion versus systemic therapy for hepatic metastases from colorectal cancer: a randomized trial of efficacy, quality of life, and molecular markers (CALGB 9481). J Clin Oncol 2006;24:1395–403. 22. Alexander HR, Libutti SK, Bartlett DL, Puhlmann M, Fraker DL, Bachenheimer LC. A phase I-II study of isolated hepatic perfusion using melphalan with or without tumor necrosis factor for patients with ocular melanoma metastatic to the liver. Clin Cancer Res 2000;6: 3062–70. 23. 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. 24. Sirott M, Bajorin D, Wong G, et al. Prognostic factors in patients with metastatic malignant melanoma, a multivariate analysis. Cancer 1993; 72:3091–8. 25. EORTC. EORTC protocol 18021e88021: Intravenous versus intraarterial fotemustine chemotherapy in patients with liver metastasis from uveal melanoma: a randomized phase III study of the EORTC Melanoma Group and the EORTC Ophthalmic Oncology Task Force; 2003.
www.ejso.com
Hepatic arterial Fotemustine chemotherapy in patients with liver metastases from cutaneous melanoma is as effective as in ocular melanoma R. Siegel a, A. Hauschild b, C. Kettelhack a, K.C. Ka¨hler b, A. Bembenek a, P.M. Schlag a,* a
Department of Surgery and Surgical Oncology, Charite´dUniversita¨tsmedizin Berlin, Campus Buch, Robert-Ro¨ssle-Tumour-Hospital at Helios Klinikum Berlin-Buch, D-13125 Berlin, Germany b Department of Dermatology, University Hospital Schleswig-Holstein, Campus Kiel, Kiel, Germany Accepted 14 November 2006 Available online 28 December 2006
Abstract Aim: Hepatic metastases from melanoma are associated with poor prognosis. Systemic chemotherapy and biological treatments remain unsatisfactory. This study investigated the impact of hepatic arterial chemotherapy in patients with ocular and cutaneous melanoma. Methods: In a retrospectively analysed observational study, 36 consecutive patients with hepatic metastases from ocular or cutaneous melanoma were assigned for surgical hepatic port-catheter implantation. Fotemustine was delivered weekly for a 4-week period, followed by a 5-week rest and a maintenance period every 3 weeks until progression. Overall survival, response and toxicity were analysed and compared. Results: After port-catheter implantation 30/36 patients were finally treated (18 with ocular and 12 with cutaneous melanoma). A median of 8 infusions per patient were delivered (range 3e24). 30% thrombocytopenia grade 3, 7% neutropenia grade 3 but no nausea or vomiting grade 3 were encountered. Nine out of 30 patients achieved partial remission, 10/30 stable disease; 11/30 patients were progressive. Median survival for all treated patients was 14 months. Partial remission and stable disease were associated with a survival advantage compared to progressive disease (19 vs. 5 months). No significant difference in survival was observed for ocular versus cutaneous melanoma. Serum LDH was a significant predictor of both response and survival. Conclusions: Hepatic arterial Fotemustine chemotherapy was well tolerated. Meaningful response and survival rates were achieved in ocular as well as cutaneous melanoma. Careful patient selection in consideration of extra-hepatic involvement is crucial for the effectiveness of this treatment. Independent from the primary melanoma site, it is debatable if patients with highly elevated serum-LDH may benefit from this approach. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Hepatic arterial therapy; Fotemustine; Liver metastasis; Cutaneous melanoma; Ocular melanoma
Introduction Ocular melanoma is characterised by a high prevalence for liver metastases and an unusual propensity for late metastases.1,2 This pattern of dissemination differs from the one of cutaneous melanoma. In cutaneous melanoma, the main metastatic sites are the regional lymph nodes, subcutaneous tissue, and the lungs. Hepatic metastases occur in only 14e20% of metastatic cutaneous melanoma. Both ocular and cutaneous melanomas have high metastasis-related * Corresponding author. Tel.: þ49 30 9417 1400; fax: þ49 30 9417 1404. E-mail address: [email protected] (P.M. Schlag). 0748-7983/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.ejso.2006.11.019
mortality rates, with a median survival of only 5 to 7 months.3e5 Chemotherapy alone or in combination with interferon and/or interleukin-2 is ineffective in most cases of liver metastasis from ocular or cutaneous melanoma. In disseminated malignant melanoma, no prospective randomised study has proven an advantage of a distinct therapy regimen over another.6,7 The cytotoxic agent Fotemustine, a third-generation nitrosourea, has been administered as a systemic treatment first to patients with visceral metastases from cutaneous melanoma, resulting in a response rate of 13%.8 In patients with ocular melanoma, the response rate was 8% with a median survival of 12 months.9 Still, Dacarbazine is currently the standard therapy in metastatic cutaneous melanoma with response rates ranging
628
R. Siegel et al. / EJSO 33 (2007) 627e632
from 7% to 12% and a median survival time from 6 to 8 months.10e12 Recently, the results of a phase-III study comparing Fotemustine with Dacarbazine were published, confirming the activity of Fotemustine as a single agent in the first line treatment of metastatic cutaneous melanoma. Overall response rate was higher in the Fotemustine arm (in the intention to treat population) and a trend in favour of Fotemustine in terms of overall survival was evident.13 In addition to the systemic approach, certain centres have established regional treatment strategies for hepatic metastases. Complete resection is only possible in highly selected patients with limited disease. Debulking surgery that does not achieve a R0-resection did not improve the outcome of patients.14,15 In ocular melanoma, chemoembolising using Cisplatin and polyvinyl sponge resulted in a marked response, but has not become a widely accepted treatment modality because of important treatment-related morbidity.16 Intra-arterial chemotherapy with Cisplatin, Dacarbazine, and Vinblastine has been attempted in individuals, but never proved to be effective.17 Because of its high hepatic extraction rate at the first pass, Fotemustine was employed by the group of Leyvraz, who treated 31 patients by intra-arterial administration with this drug. This treatment resulted in a response rate of 40% with four patients in complete remission and a median survival of 14 months. The toxicity was acceptable with each neutropenia and thrombopenia grade IIIeIV in only 21% of the patients.18 Leyvraz’ data have recently been updated with 70 patients, resulting in a total of 101 patients from seven centres.19 Based on the results of the study from Leyvraz et al., intra-arterial therapy for liver metastases from ocular melanoma has been established in our institution. However, no comparable studies for local therapy of liver metastases from cutaneous melanoma were available. Because of the limited options in metastatic cutaneous melanoma and the unsatisfactory results obtained by systemic Fotemustine therapy, we applied the intra-arterial approach to patients with hepatic metastases from cutaneous melanoma as well. We conducted this study to compare the effectiveness of hepatic arterial Fotemustine infusion between liver metastases from ocular and cutaneous melanoma. In the following, we will review our results and discuss prognostic factors that are associated to survival. Methods
eligible if the primary tumour was locally controlled and no progressive extra-hepatic dissemination was evident. Pre-treatment evaluation included computed tomography (CT) or magnetic resonance imaging (MRI) of the abdomen, chest X-ray, whole blood count and biochemistry assessment (e.g. lactate dehydrogenasedLDH). Because of the palliative intention of treatment, abnormal results on liver function test were not an exclusion criterion. Catheter placement A totally implantable catheter connected to a port chamber was placed into the hepatic artery via the gastroduodenal artery. An angiography demonstrating the arterial road mapping was performed prior to catheter placement. To prevent misperfusion of the stomach and duodenum, collateral vessels including the right gastric artery were occluded. Prophylactic cholecystectomy was performed to avoid chemical cholecystitis. The port was placed in a subcutaneous pocket over the right rib cage. Patent blue injection via the catheter was performed to document perfusion of the whole liver and the absence of misperfusion. During surgery, liver metastases were histologically proven, unless diagnosis was already confirmed by prior biopsy. Prior to Fotemustine application, an angiogram via the port-catheter and/or a technetium scan was done to verify catheter position and liver perfusion. Hepatic arterial chemotherapy According to the protocol of Leyvraz, Fotemustine (MuphoranÒ) 100 mg/m2 was administered intra-arterially over 4 h.16 Anti-emetic treatment was given prophylactically using Ondasetron 8 mg intravenously and Methylprednisolone 125 mg intravenously. The induction phase consisted of the Fotemustine administration once a week for four times followed by a 5-week rest period and a maintenance phase with administration of one course of Fotemustine every 3 weeks until progression or unacceptable toxicity. The maintenance period was conducted only if partial response or stable disease was observed upon restaging. Follow-up and assessment of response were performed using CT or MRI after completion of the induction phase and subsequently controlled after every three or four cycles of the maintenance period. Assessment of response and toxicity was done according to WHO criteria.20 All patients were followed until death.
Patient selection Statistical methods Between June 1999 and December 2005, 36 consecutive patients were enrolled within the treatment protocol. Data was extracted retrospectively from a prospective clinical database. All patients with liver metastases from ocular or cutaneous melanoma considered not resectable were evaluated for hepatic intra-arterial treatment. Patients were
For statistical methods we used the Statistical Package for the Social Sciences Software program (SPSS, version 12.0, Chicago, IL, USA). Two-sample comparisons were conducted by the ManneWhitney test. Disease-specific survival rates were compared using the KaplaneMeier
R. Siegel et al. / EJSO 33 (2007) 627e632
method. Survival curves were tested for difference using the log rank test. A p value of less than 0.05 was considered significant.
Table 1 Characteristics of patients receiving hepatic arterial Fotemustine infusion Ocular melanoma (n ¼ 18)
Cutaneous melanoma (n ¼ 12)
Total (n ¼ 30)
p
Sex Male Female
9 9
7 5
16 14
0.722
Age, years Median Range
58 41e73
60 45e85
59 41e85
0.95
Characteristics
Results Catheter placement Thirty-six patients (18 men and 18 women) were enrolled to receive a port-catheter system into the hepatic artery. The median age at the time of surgery was 59 years (range 40e85 years). Twenty-two patients had a previously diagnosed ocular melanoma as their primary tumour, 14 patients were diagnosed with a cutaneous melanoma. In 31/36 patients open surgery was performed for the implantation of the port-catheter system, in 5/36 patients the catheter was implanted by minimal access surgery. Catheter related morbidity comprised 2 of 36 patients (one patient developed a severe pancreatitis, another patient presented with a port site infection). Three of 36 patients died due to tumour progress and liver failure within 30 days after surgery.
629
Time interval from diagnosis of melanoma to liver metastases, months Median 61 6 33 0.002 Range 1e265 0e93 0e265 Time interval from liver metastases to port-catheter implantation, months Median 2 2 2 0.884 Range 6e63 0e29 0e63 LDH, U/L Median Range
402 97e6050
338 171e2057
Intrahepatic fotemustine applications, n Median 9 7 Range 3e24 3e12
399 97e6050
0.285
8 3e24
0.305
LDH, lactate dehydrogenase; U/L, units per litre.
Fotemustine delivery and toxicity After implantation of the port-catheter system Fotemustine was administered according to the protocol in 30 of 36 patients. Six patients were excluded and did not receive Fotemustine because of rapid progressive disease and liver failure (n ¼ 5; 3/5 died within 30 days after surgery, see surgery-related mortality above) and a revised histopathologic diagnosis (revealing a hepatic manifestation of lymphoma; n ¼ 1). Eighteen of 30 patients suffered from ocular melanoma metastatic to the liver, whereas 12/30 patients had metastases from cutaneous melanoma. Characteristics of all 30 patients receiving intra-arterial Fotemustine chemotherapy are listed in Table 1. As suspected by the different pattern of metastases, the median time interval between diagnosis of the primary tumour and the detection of liver metastases differed between the two melanoma entities. Among the other baseline characteristics, there were no statistically significant differences. The 30 patients received a median of 8 intra-arterial Fotemustine applications, with a range from 3 to 24 (median 9 cycles in ocular vs. 7 cycles in cutaneous melanoma). The treatment could be administered on an outpatient basis. The toxicity comprised grade IIIeIV thrombocytopenia in 30%, grade IIIeIV neutropenia in 7%, nausea/vomiting in 17% (maximum grade IeII) and abdominal pain in 7% (maximum grade IeII). Therapeutic activity Two patients (one from both melanoma entities) were rapidly progressing under treatment and received only three
courses of the induction period. All other patients completed the induction phase and were evaluated for best response after four courses. Overall, 9/30 patients achieved partial remission (PR), 10/30 patients stable disease (no changedNC), and progressive disease (PD) was observed in 11/30 patients. The benefit in survival for patients with stable disease and partial remission was almost identical (NC with a median of 19 months vs. median 20 months for PR). Thereby response (PR and NC) to hepatic arterial therapy resulted in a survival advantage compared to patients with progressive disease (5 months). These results were still reproducible for the different melanoma subgroups. For cutaneous melanoma, the median overall survival was similar in patients with NC and PR (19 months for NC vs. 14 months for PR). This was different in ocular melanoma where median survival was significantly longer in patients with partial remission (38 months for PR vs. 14 months for NC; p ¼ 0.008). The median survival of all 30 patients with intra-arterial treatment was 14 months when calculated from the date of port-catheter implantation with 6 patients being still alive at the time of analysis. Patients with hepatic metastases from ocular melanoma achieved a prolonged survival (median 22 months) compared to patients with metastatic cutaneous melanoma (median 12 months), but this difference was statistically not significant (Fig. 1; p ¼ 0.306). Baseline serum LDH levels revealed a significant correlation to both response rates and survival. Patients presenting with a markedly elevated LDH level (>452 U/Ldtwice the upper normal limit) at the time of port-catheter implantation had a median survival of 5 months compared to median 33 months in patients with LDH levels below 452 U/L
R. Siegel et al. / EJSO 33 (2007) 627e632
630
1,0
1,0
Ocular Melanoma
LDH ≤ 452 U/L (2x UNL) LDH > 452 U/L (2x UNL)
Cutaneous Melanoma P = 0.306
0,8
Proportion Surviving
Proportion Surviving
0,8
0,6
0,4
0,2
P = 0.0006
0,6
0,4
0,2
0,0
0,0 0
10
20
30
40
50
Months after Implantation of Hepatic Port Catheter Figure 1. KaplaneMeier overall survival for all patients receiving intrahepatic Fotemustine (ocular melanoma, n¼18; cutaneous melanoma, n¼12) since placement of an intra-arterial catheter. Median overall survival for the whole group was 14 months, for patients with ocular melanoma 22 months and for patients with cutaneous melanoma 12 months ( p¼0.306). ‘‘þ’’ show censored observations.
( p ¼ 0.0006; Fig. 2). Analysing the subgroups, LDH is still very powerful for the prediction of response and survival, independent from the melanoma entity (Table 2). Discussion Little progress has been made in the treatment of metastatic cutaneous melanoma, resulting in a 5-year survival rate of only 7e19%.5 By increasing the tumour cells’ exposure to the cytotoxic agent and by reducing the exposure of the surrounding healthy tissues at the same time, intraarterial chemotherapy has a two-fold advantage in patients with liver metastases. The potential of this method of drug delivery for response and survival was recently re-emphasised by Kemeny et al. for patients with hepatic metastases from colorectal cancer.21 Since the report of Leyvraz et al., hepatic intra-arterial Fotemustine is accepted as an effective and well-tolerated regional treatment for liver metastases from ocular melanoma. Another regional approach, isolated hepatic perfusion with Melphalan (with or without TNF), following the concept of isolated limb perfusion for melanoma and sarcomas, has also been shown to induce a potent anti-tumour activity in ocular melanoma metastatic to the liver. However, the complex procedure and a considerable morbidity restricted this approach.22 Based on the experience from Leyvraz et al. and the results from systemic chemotherapy with Fotemustine in disseminated cutaneous melanoma, we applied hepatic arterial Fotemustine infusion in patients with liver metastases from both ocular and cutaneous melanoma.
10
0
20
30
40
50
Months after Implantation of Hepatic Port Catheter Figure 2. KaplaneMeier overall survival (for all 30 patients receiving intrahepatic Fotemustine) as it relates to LDH serum level at the time of port-catheter placement. ‘‘þ’’ show censored observations. LDH, lactate dehydrogenase; U/L, units per litre; 2 UNL, twice upper normal limit.
Response and survival We are aware of the limitations of this small series of patients. But by achieving partial remission or stable disease in almost two third of our patients, intra-hepatic Fotemustine infusion produced higher response and stabilisation rates compared to systemic application of any cytotoxic agent in malignant melanoma. Although the overall survival in patients with hepatic metastases from ocular melanoma (22 months) was higher than in patients with cutaneous melanoma (12 months), this difference was statistically not significant nor was a difference found in terms of response between the two melanoma entities. This is may be due to the limited number of patients. Although Table 2 Correlation of LDH serum levels and treatment response LDHa
Ocular melanoma Cutaneous melanoma Total
Partial response
Stable diseasedno change
Progressive disease
U/L
No. of pts.
No. of pts. (%)
No. of pts. (%)
No. of pts. (%)
<452 >452 <452 >452 <452 >452
10 8 8 4 18 12
5 0 3 1 8 1
4 2 4 0 8 2
1 6 1 3 2 9
(100) (0) (75) (25) (89) (11)
(67) (33) (100) (0) (80) (20)
(14) (86) (25) (75) (18) (82)
LDH, lactate dehydrogenase; U/L, units per litre. 452 U/L ¼ 2 upper normal limit. a LDH serum level at the time prior to port-catheter placement.
R. Siegel et al. / EJSO 33 (2007) 627e632
no definite conclusion may be drawn from the results achieved in these non-randomised series of 12 patients with cutaneous melanoma, our results with a median survival of 14 months for PR and 19 months for NC justify further studies with arterial Fotemustine treatment in melanoma patients with metastases confined to the liver. In cutaneous melanoma, the extra-hepatic progression of disease represents the main limitation of loco-regional therapy. Therefore, intra-arterial Fotemustine treatment could be considered preferentially in cases where extra-hepatic distant metastasis is not present or in patients with stable/ responding disease to systemic chemotherapy, where only hepatic metastases are progressive. Fotemustine related toxicity Compared to intravenous systemic Fotemustine therapy, our data highlight the favourable tolerance of Fotemustine delivered intra-arterially. Within the phase III study comparing Fotemustine with Dacarbazine, myelosuppression was the most common adverse event with grade 3 to 4 neutropenia in 51% of patients receiving Fotemustine and grade 3 to 4 thrombocytopenia in 43% of all patients in the Fotemustine arm.13 This exceeded by far the adverse events observed in our study with less than 7% patients presenting with grade 3 or 4 neutropenia and 30% with grade 3 or 4 thrombocytopenia. Our toxicity profile is comparable to the observed adverse events in the trial of Leyvraz, describing grade 3 or 4 neutropenia as well as thrombopenia in 21% each, and grade 3 or 4 nausea and vomiting in 7%. Lack of severe toxicities ensured the feasibility and acceptance of intra-arterially Fotemustine therapy on an outpatient basis. Catheter-related morbidity and mortality Complications related to the catheter have to be considered when comparing systemic and loco-regional therapy. The mortality after catheter implantation, before commencing chemotherapy, is a crucial issue and may arise from the preoperative reduced general performance status of patients with rapidly progressive disease and limited liver functions. The catheter related morbidity in our study was low with 6% and occurred after some weeks of treatment. This is in the range of the reports from the large series of 101 patients with ocular melanoma metastatic to the liver by Peters et al.19 Conclusion Hepatic arterial Fotemustine chemotherapy is well tolerated and seems to improve outcome of this poor prognosis population. In order to prevent perioperative mortality by catheter insertion and early therapy cessation, it is essential to select patients properly and define prognostic markers. The LDH value is a well-known negative prognostic factor
631
in melanoma,19,23,24 and was confirmed within our series of patients receiving intra-arterial Fotemustine. In addition to LDH, the protein S100b may be evaluated as another prognostic marker to assess the effect of intra-arterial therapy. It is still debatable whether patients presenting with a high serum LDH (e.g. more than twice upper normal limit, >452 U/L) will suffer from rapid progressive disease in most cases and therefore are unlikely to benefit from intra-hepatic Fotemustine application. This question may be addressed by a randomised study comparing best supportive care only vs. intra-arterial therapy for patients below a certain serum LDH level. However, because of ethical reasons and the limited number of eligible patients the realisation of studies with this design is at least questionable. Currently, European patients with liver metastases from ocular melanoma should preferentially be enrolled into the randomised phase III - EORTC Melanoma Group protocol 18021-88021, comparing intravenous versus intraarterial Fotemustine chemotherapy.25 For patients with cutaneous melanoma and liver metastases alone, further studies are warranted.
References 1. Pyrhonen S. The treatment of metastatic uveal melanoma. Eur J Cancer 1998;34(Suppl 3):S27–30. 2. Egan KM, Seddon JM, Glynn RJ, Gragoudas ES, Albert DM. Epidemiologic aspects of uveal melanoma. Surv Ophthalmol 1988;32:239–51. 3. Leiter U, Meier F, Schittek B, Garbe C. The natural course of cutaneous melanoma. J Surg Oncol 2004;86:172–8. 4. Gragoudas ES, Egan KM, Seddon JM, et al. Survival of patients with metastases from uveal melanoma. Ophthalmology 1991;98:383–9. 5. Balch CM, Soong SJ, Gershenwald JE, et al. Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 2001; 19:3622–34. 6. Eigentler TK, Caroli UM, Radny P, Garbe C. Palliative therapy of disseminated malignant melanoma: a systematic review of 41 randomized clinical trials. Lancet Oncol 2003;4:748–59. 7. Keilholz U, Punt CJ, Gore M, et al. Dacarbazine, cisplatin, and interferon-alfa-2b with or without interleukin-2 in metastatic melanoma: A randomized phase III trial (18951) of the European Organisation for Research and Treatment of Cancer Melanoma Group. J Clin Oncol 2005;23:6747–55. 8. Kleeberg UR, Engel E, Israels P, et al. Palliative therapy of melanoma patients with Fotemustine. Inverse relationship between tumor load and treatment effectiveness. Melanoma Res 1995;5:195–200. 9. Becker JC, Kaempgen E, Schadendorf D, Wagner S, Broecker EB. Treatment of disseminated uveal melanoma with sequential Fotemustine, Interleukin-2 and Interferon alpha2a. Proc ASCO 2000;572a. 10. Middleton MR, Grob JJ, Aaronson N, et al. Randomized phase III study of temozolomide versus dacarbazine in the treatment of patients with advanced metastatic malignant melanoma. J Clin Oncol 2000;18: 158–66. 11. Eggermont AMM, Kirkwood JM. Re-evaluating the role of dacarbazine in metastatic melanoma: what have we learned in 30 years? Eur J Cancer 2004;40:1825–36. 12. Bedikian AY, Millward M, Pehamberger H, et al. Bcl-2 Antisense (oblimersen sodium) plus dacarbazine in patients with advanced melanoma: The Oblimersen Melanoma Study Group. J Clin Oncol 2006; 24:4738–45.
632
R. Siegel et al. / EJSO 33 (2007) 627e632
13. Avril MF, Aamdal S, Grob JJ, et al. Fotemustine compared with Dacarbazine in patients with disseminated malignant melanoma: A phase III study. J Clin Oncol 2004;22:1118–25. 14. Rajpal S, Moore R, Karakousis CP. Survival in metastatic ocular melanoma. Cancer 1983;52:334–6. 15. Salmon RJ, Levy C, Plancher C, et al. Treatment of liver metastases from uveal melanoma by combined surgery-chemotherapy. Eur J Surg Oncol 1998;24:127–30. 16. Mavligit GM, Charnsangavej C, Carrasco CH, Patt YZ, Benjamin RS, Wallace S. Regression of ocular melanoma metastatic to the liver after hepatic arterial chemoembolization with cisplatin and polyvinyl sponge. JAMA 1988;260:974–6. 17. Bedekian AY, Legha SS, Mavligit G, et al. Treatment of uveal melanoma metastatic to the liver: a review of the M.D. Anderson Cancer Center experience and prognostic factors. Cancer 1995;76: 1665–70. 18. Leyvraz S, Spataro V, Bauer J, et al. Treatment of ocular melanoma metastatic to the liver by hepatic arterial chemotherapy. J Clin Oncol 1997;15:2589–95. 19. Peters S, Voelter V, Zografos L, et al. Intra-arterial hepatic Fotemustine for the treatment of liver metastases from uveal melanoma: experience in 101 patients. Ann Oncol 2006;17:578–83.
20. World Health Organization. Handbook for reporting results of cancer treatment. Geneva, Switzerland: World Health Organization; 1979. WHO publication no. 48. 21. Kemeny NE, Niedzwiecki D, Hollis DR, et al. Hepatic arterial infusion versus systemic therapy for hepatic metastases from colorectal cancer: a randomized trial of efficacy, quality of life, and molecular markers (CALGB 9481). J Clin Oncol 2006;24:1395–403. 22. Alexander HR, Libutti SK, Bartlett DL, Puhlmann M, Fraker DL, Bachenheimer LC. A phase I-II study of isolated hepatic perfusion using melphalan with or without tumor necrosis factor for patients with ocular melanoma metastatic to the liver. Clin Cancer Res 2000;6: 3062–70. 23. 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. 24. Sirott M, Bajorin D, Wong G, et al. Prognostic factors in patients with metastatic malignant melanoma, a multivariate analysis. Cancer 1993; 72:3091–8. 25. EORTC. EORTC protocol 18021e88021: Intravenous versus intraarterial fotemustine chemotherapy in patients with liver metastasis from uveal melanoma: a randomized phase III study of the EORTC Melanoma Group and the EORTC Ophthalmic Oncology Task Force; 2003.