Adjuvant Therapy of Melanoma

January 1, 2002

In 2001, the American Joint Committee on Cancer Melanoma Staging Committee proposed and created a new staging system for melanoma. This new system will become official in 2002, with the publication of the sixth

ABSTRACT: In 2001, the American Joint Committee on Cancer Melanoma Staging Committee proposed and created a new staging system for melanoma. This new system will become official in 2002, with the publication of the sixth edition of the AJCC Cancer Staging Manual. The new system identifies significant prognostic variables in patients with melanoma and validates them in an analysis of 17,600 patients, making it possible to precisely determine the patient’s chance for survival. In light of physicians’ ability to determine with more precision which patients are at high risk for melanoma recurrence, they face the dilemma of which, if any, surgical adjuvant therapy to choose. Alpha-interferon is the only agent approved for adjuvant therapy of melanoma in the United States, but its questionable benefits and substantial side effects make it hard to justify recommending it to patients. Discussion of trials of high- and low-dose interferon is presented here. The author’s group has conducted trials of granulocyte-macrophage colony-stimulating factor (GM-CSF [Leukine]) as surgical adjuvant treatment of patients at high-risk for melanoma recurrence. One of the most important activities of GM-CSF is its ability to activate macrophages and cause them to become cytotoxic for human melanoma cells, at doses low enough to avoid the toxicity associated with other cytokines. The author presents promising trial results, discusses GM-CSF in other malignancies, and includes discussion of tumor vaccines, biochemotherapy, and other agents being studied as adjuvant therapy of melanoma. It is hoped that these newer approaches will result in therapies that are more effective and less toxic than interferon. [ONCOLOGY 16(Suppl 1):40-48, 2002]

The American Joint Committee on Cancer (AJCC) MelanomaStaging Committee proposed a new staging system for melanoma[1,2]; the finalversion and validation of this new staging system were published in 2001.[3,4]The new staging system has been approved by the AJCC Executive Committee, theInternational Union Against Cancer (UICC) TNM Committee, the World HealthOrganization Melanoma Program, and the European Organization for Research andTreatment of Cancer (EORTC) Melanoma Group, and will become official withpublication of the sixth edition of the AJCC Cancer Staging Manual in 2002.

This system represents an important advance in that it identifies significantprognostic variables in patients with melanoma and validates them in an analysisof 17,600 patients, making it possible to precisely determine the patient’schance for survival. The important prognostic indicators regarding the primarytumor are now identified as thickness and ulceration (level is an importantindication only for T1—primary tumors ≤ 1.0 mm in thickness). For patientswith nodal involvement, the number of metastatic nodes, whether the tumor burdenis microscopic or macroscopic, and ulceration of the primary tumor are importantin prognosis. For patients with metastatic disease, the site of metastasis(skin, subcutaneous, and/or lymph node vs lung vs other sites) is relevant, asis whether or not the lactate dehydrogenase (LDH) level is elevated.

The new system takes these factors into account: it incorporates ulcerationof the primary as a prognostic indicator (not previously included), includespatients with satellites and in-transit lesions in stage III, includes thenumber of positive nodes (rather than gross dimensions) and tumor burden in thesubstaging within stage III, and adds the site of distant metastases and thepresence of elevated LDH as a prognostic indicator for patients with stage IVdisease.

Under the new staging system, survival curves correlate well with stage ofdisease (see Figure 1) and subgroupings within each stage allow furtherrefinement of prognosis (see Table 1). This allows the physician to determinewith precision which patients are at high risk for recurrence, and to counselpatients about their prognosis and therapeutic options. Based on this importantinformation, the dilemma being faced by the treating physicians and patients athigh risk for recurrence is what, if any, surgical adjuvant therapy to choose.

Interferon

Alpha-interferon has direct antitumor activity, and also produces immunemodulatory effects. In patients with metastatic disease, the response rate toalpha-interferon is approximately 15% to 20%. It is the only drug approved inthe United States for adjuvant therapy of melanoma. It was approved as surgicaladjuvant treatment of patients with a T4 primary (> 4.0 mm in thickness) ornode-positive (stage III) melanoma by the US Food and Drug Administration (FDA)in 1996. It has not been tested or approved as adjuvant therapy for patientswith stage IV disease.

High-Dose Interferon Studies

The Eastern Cooperative Oncology Group (ECOG) completed a prospectiverandomized controlled study of interferon (IFN) alfa-2b vs observation assurgical adjuvant therapy in 287 patients with melanoma (protocol E1684).[5] Thedose used in this study was the maximum tolerated dose: 20 million IU/m²/dintravenously (IV) 5 days per week for 4 weeks followed by 10 million IU/m²three times per week subcutaneously (SC) for 48 weeks. The results showed thatthere was a significant prolongation of relapse-free survival and overallsurvival in the group receiving IFN alfa-2b. Patients randomized to treatmentwith IFN alfa-2b had a 9% improvement in continuous disease-free survival (from26% to 37%) and an 11% improvement in survival (from 37% to 46%) compared withpatients receiving no further treatment.

Dose modification was required in the majority of patients. Toxicity—includingconstitutional symptoms (fever, chills, flu-like symptoms, fatigue, malaise,diaphoresis), myelosuppression, hepatotoxicity, and neurologic symptoms—wasconsiderable: 67% of the patients had severe (grade 3) toxicity, 9% hadlife-threatening toxicity, and there were two deaths due to hepatotoxicity. Itis important to consider this toxicity in context. These are patients who havehad surgical excision of their melanoma and are clinically free of disease.Although they are at high risk for recurrence, some of them may already becured. This is different from the situation in which one is administeringchemotherapy for widespread metastatic disease, where there is no choice and thetoxicity must be accepted as part of the treatment.

In follow-up, a large, prospective, randomized trial was done by the samegroup in a similar patient population (protocol E1690). This study included 642patients randomized to three arms: high-dose interferon as described above,low-dose interferon, or observation. The results show that time to diseaseprogression was prolonged in the patients receiving high-dose interferon, butoverall survival was the same in all three arms, indicating that there was nosurvival benefit in patients receiving high-dose interferon.[6]

There has also been a third prospectively randomized trial of high-doseinterferon as adjuvant therapy of high-risk melanoma. This trial, describedbelow, compared outcome following this therapy with that in patients randomizedto receive a therapeutic melanoma vaccine. The vaccine was based on GM2, aganglioside derived from bovine brain, which is overrepresented on melanomacells and thus can be used in a vaccine in an effort to stimulate an immuneresponse to melanoma.

A double-blind randomized trial compared a ganglioside vaccine with bacilleCalmette-Guérin (GM2/BCG) to BCG alone as adjuvant therapy in 122 patients withstage III melanoma.[7] The results showed improved survival in patients with GM2antibodies as compared with patients who did not have these antibodies. In asubsequent phase I trial, it was shown that conjugating the GM2 ganglioside withkeyhole limpet hemocyanin (KLH) and administering it with the saponin adjuvantQS-21 resulted in serologic responses against GM2 that were strikingly superior,quantitatively and qualitatively, to any seen with previously tested GM2vaccines.[8]

These results formed the basis for the trial mentioned above, which wasconducted in a combined protocol by the Eastern Cooperative Oncology Group(protocol E1694), the Southwest Oncology Group (SWOG, protocol S9512), andCancer and Leukemia Group B (CALGB protocol C509801). The study included 880patients who were randomized to either high-dose interferon or the GM2/KLHvaccine with QS-21 (called GMK), 774 of which were eligible for efficacyanalysis.[9] The results showed that patients receiving the high-dose interferonenjoyed benefit in overall and disease-free survival as compared to patients whoreceived the vaccine. However, the study has been criticized for severalreasons: (1) the median follow-up was short for this patient population (16months); (2) there was no control group given observation alone; (3) thedifference in overall survival in the two arms was not great (78% for IFNalfa-2b and 73% for GMK); and (4) the vaccine regimen was changed between thepositive phase II trial and the phase III trial (QS-21 was substituted for BCG).This might have resulted in an adverse effect on the outcome because ofresultant increased titers of IgG antimelanoma antibody, which might have had ablocking effect on the cell-mediated immune response to the tumor.

Other groups have also studied interferon as adjuvant therapy of melanoma.The North Central Cancer Treatment Group (NCCTG) compared high-dose IFN alfa-2a(20 million IU/m² intramuscularly three times per week for 12 weeks) vsobservation in 262 patients with high-risk melanoma.[10] There was no benefit inthe treated group. The World Health Organization (WHO) evaluated IFN alfa-2a in444 patients with node-positive melanoma using a dose of 3 million IU/d SC threetimes per week for 3 years and found no impact on disease-free or overallsurvival.[11]

The EORTC conducted a trial (protocol 18871) in parallel with the GermanCancer Society (DKG, protocol 80-1) in which 830 patients with stage II (> 3mm) or stage III melanoma were randomized to either 1 million IU of IFN alfa-2bSC every other day for 1 year, IFN-gamma, mistletoe extract (Iscador, availablein the United States as Iscar), or observation.[12] This is a mature study, witha median follow-up of 5.9 years. The results show no difference in disease-freeor overall survival for the patients treated with IFN alfa-2b as compared withobservation-only controls (patients treated with Iscador actually had a worseoutcome than the controls or interferon-treated patients).

The EORTC has launched a follow-on trial (protocol 18952) in patients withstage IIB or III melanoma.[12] Patients are randomized to either (1) IFN alfa-2bat 10 million IU 5 days/wk for 4 weeks SC, then three times/wk for 1 year; (2)IFN alfa-2b at 10 million IU 5 days/wk for 4 weeks SC, then 5 million IU threetimes/wk for 2 years; or (3) observation. So far, 1,418 patients have beenenrolled in the study. At the 2001 meeting of the American Society of ClinicalOncology, Dr. Eggermont reported that with a median follow-up of 1.9 years,there was an advantage (P = .04) for the 2-year regimen with very littletoxicity compared with high-dose interferon. The authors conclude that it is tooearly to make a determination regarding the impact on overall survival and/or todraw conclusions regarding this important study.

Low-Dose Interferon Studies

Other investigators have also studied low-dose alpha-interferon as surgicaladjuvant treatment for stage II melanoma. Low-dose alpha-interferon wasevaluated in a large, randomized study of adjuvant therapy in 499 patients withstage II melanoma (primary melanoma greater than 1.5 mm in thickness withclinically negative nodes).[13] Results in this patient population are promisingeven though the low-dose regimen does not seem to be effective in higher-riskpatients; there was significant benefit in progression-free survival in thepatients randomized to low-dose interferon. In a second randomized trial oflow-dose interferon in 311 patients with resected primary stage II melanoma,prolonged disease-free survival was also demonstrated.[14]

Summary

While interferon is the only agent approved in the United States for adjuvanttherapy of patients at high risk for recurrence of melanoma, the benefit, ifany, is marginal for patients with node-positive disease and side effects aresignificant. For patients with stage II disease, there have been two largerandomized trials with encouraging results using a low-dose regimen and onenegative study. The low-dose regimen has been evaluated in at least threestudies in patients with positive nodes and has not been shown to be of benefit.Patients with positive nodes must weigh the risks and potential benefits of thehigh-dose interferon regimen; patients with stage II disease might consider thelow-dose regimen.

The Role of GM-CSF

Both granulocyte-macrophage colony-stimulating factor (GM-CSF [Leukine]) andgranulocyte colony-stimulating factor (G-CSF [Neupogen]) are growth factors thatact on bone marrow progenitor cells to increase production of neutrophils,monocytes, eosinophils, and dendritic cells. Only GM-CSF also causes activationof monocytes, macrophages, and dendritic cells; G-CSF does not exhibit thisactivity.

GM-CSF, a multifunctional molecule, plays a vital role in various functionsof the immune system. It has many activities that suggest it has potential inadjuvant therapy of cancer (see Table 2[15-24]). One of the most important ofthese activities relative to cancer therapy is the ability of GM-CSF to activatemacrophages and cause them to become cytotoxic for human melanoma cells.[16,17]Activated macrophages distinguish tumor cells from normal cells and kill onlythe tumor cells.[15] Moreover, the in vivo administration of GM-CSF causesmonocyte activation at low doses in human subjects,[18,19] thus avoiding thetoxicity associated with the high doses of other cytokines that are necessaryfor therapeutic activity.

Two Melanoma Trials

We conducted a trial of GM-CSF as surgical adjuvant treatment of patients atvery high risk for recurrence of melanoma.[25] The population included patientswith stage III disease with more than four positive nodes or a mass greater than3 cm (14 patients), or with stage IV disease (34 patients). All 48 patients wererendered clinically disease-free by surgery prior to enrollment. Treatment withGM-CSF commenced within 60 days from when the patient was rendered clinicallytumor-free and continued until recurrence or for 1 year. Patients who had atreatable recurrence could elect to continue treatment with GM-CSF.

GM-CSF was administered SC in 28-day cycles, such that a dose of 125 mg/m2was delivered daily for 14 days followed by 14 days rest. Results were comparedwith those in a historical control population, matched for important prognosticvariables. Patients with stage III disease were matched on the basis of numberof positive nodes, and patients with stage IV melanoma were matched to controlsbased on the presence of visceral or nonvisceral metastases. These matchingcriteria represent the most important prognostic factors, respectively, forstage III and IV malignant melanoma.[26,27] GM-CSF was well tolerated; only onesubject discontinued the drug due to an adverse event (grade 2 injection sitereaction).

Overall and disease-free survival were significantly prolonged in patientswho received GM-CSF compared to matched historical controls. The major impactwas on survival, with a lesser effect on disease-free survival. At 2 years,survival in the control group was 15%, and that in the treatment group was 64%.The median survival was 37.5 months in the study patients vs 12.2 months in thematched controls (P < .001) (see Figure2).

In view of these encouraging results, we initiated a new trial of adjuvanttherapy with GM-CSF in patients with malignant melanoma. Patients with stage IIIor IV melanoma, surgically excised, are eligible. They must be tumor-free byclinical and laboratory examination, and must start the study drug within 90days of their last positive surgery. The treatment regimen has been modifiedsuch that the study drug is given for 3 years, rather than 1 year, because inthe first study patients were experiencing recurrences when the drug was stoppedafter 1 year. An interim analysis of the study was done for presentation at the2001 annual meeting of the American Society of Clinical Oncology in SanFrancisco.[28] As of the time of the analysis, 38 patients had entered thetrial. These patients represented a group at very high risk for recurrence (see Table 3). Over half of the patients with stage III disease had stage IIICdisease, and of the nine patients with stage IV disease, 89% had visceralmetastases and 67% had failed first-line therapy. The results are very similarto those published for the first study. Administration of GM-CSF resulted in adramatic impact on survival (see Figure 3) and had a lesser impact on time todisease progression (see Figure 4).

Further analysis of these results showed that over half of the diseaserecurrences were solitary nodules that could be surgically excised (see Table4). We postulate that therapy with GM-CSF may be changing the biologic behaviorof metastatic melanoma. We envision that macrophages activated by GM-CSF caneradicate small metastatic tumor nodules that would have appeared as systemicmetastases, but they cannot overcome a nodule that was larger at the time oftreatment and would subsequently appear as a localized metastasis. Thispostulate is supported by the observation that small tumors are infiltrated witha relatively large number of macrophages, whereas large tumors are not.[29-31]

We conclude that GM-CSF appears to offer benefit as surgical adjuvant therapyin patients with high-risk (stage III and IV) melanoma. It can be safelyself-administered in an outpatient setting and side effects are minimal,consisting mainly of injection site reactions. The major impact is on survival,not time to disease recurrence. GM-CSF appears to change the biologic behaviorof melanoma such that when it recurs, it recurs as a solitary metastasis thatcan be surgically excised rather than as widespread disease, which requiressystemic chemotherapy. For this reason, it is important to recognize that arecurrence of melanoma in the face of GM-CSF therapy does not necessarilyrepresent failure of the treatment. In this circumstance, it is reasonable totreat the recurrence with surgery, if possible, and maintain GM-CSF therapy.This represents new thinking for the oncologist, who has been trained thatrecurrence of disease in the face of therapy represents failure of that therapy.

Case Report #1

R.E. is a 58-year-old woman who had a primary melanoma of the left calf inJuly 1991. In November 1992, she had a metastasis of the melanoma to theinguinal lymph nodes, which was surgically excised; there were six positivenodes, the largest being 6 cm in diameter. In September 1993, she had ametastasis to the ipsilateral iliac lymph node, which was surgically excised;there was one positive node, 5 cm in diameter. In March 1995, she had anadditional metastasis to an iliac node, 3 cm in diameter. This was surgicallyexcised and she was given postoperative radiation therapy. In November 1995, shehad a periaortic lymph node metastasis, 2.5 cm in diameter. This was surgicallyexcised and followed by radiation therapy. Thus, R.E. had had a series of bulkylymph node recurrences progressing up the draining lymph node chain, finallyreaching a periaortic node. These were treated with surgery and radiation.

The patient was at extremely high risk for further recurrence and death frommelanoma. In April 1996 she started GM-CSF therapy. She experienced no furtherrecurrences and completed the treatment specified in the protocol 1 year later,in April 1997. She continued on GM-CSF off protocol for another 1.5 years untilJanuary 1999, when treatment was discontinued. She has not experienced anyfurther recurrences and remains well and tumor-free as of April 2001, more than5 years since her last surgery for melanoma.

Case Report #2

S.S. is a 66-year-old woman who, in November 1994, was found to have asubcutaneous metastasis of a melanoma on her back from an unknown primary. Thelesion was surgically excised. In March 1999, she experienced sharp pain in herright side and was found to have a large metastasis to the right lobe of theliver and a smaller metastasis to the left lobe of the liver. An MRI-guidedneedle biopsy confirmed the diagnosis of melanoma. The patient was treated withtwo cycles of biochemotherapy consisting of cisplatin (Platinol), vinblastine,temozolomide (Temodar), interleukin (IL)-2, and interferon. Follow-on scansshowed that there was no measurable change in the size of the tumors and that nonew sites had appeared. The biochemotherapy was discontinued.

In August 1999, a repeat CT scan showed that there was a slight increase inthe size of the tumor. S.S. underwent a right hepatic lobectomy, and the tumorthat was removed weighed approximately 7 pounds; no involvement of the left lobeof the liver could be detected at surgery. Thus, this patient had stage IV, M1cdisease, having had hematogenous dissemination of her melanoma to the liver. Shewas at very high risk for recurrence and death from her disease. She started GM-CSFtherapy in November 1999 and has had no further recurrences of melanoma. Shecompleted 1 year of therapy per the protocol and remains on GM-CSF off protocol.She is well and tumor-free as of September 2001, 2 years after surgical excisionof the liver metastasis of her melanoma.

ECOG Trial of GM-CSF in Melanoma

A prospective randomized trial of GM-CSF as surgical adjuvant therapy forpatients with high-risk melanoma has been initiated by ECOG (protocol E4697). Inthis study, which will include 600 patients, patients who are human leukocyteantigen (HLA-A2) positive are randomized to either placebo, multiepitope peptidevaccine, multiepitope peptide vaccine with GM-CSF, or GM-CSF. Patients who areHLA-A2 negative are randomized to placebo or GM-CSF.

GM-CSF in Other Malignancies

There have been a number of studies showing the efficacy of GM-CSF asadjuvant therapy in other malignancies. ECOG conducted a randomizedplacebo-controlled phase III study (protocol E1490) in adult patients (55 to 70years of age) with acute myelogenous leukemia (AML).[32] A total of 117 patientswith AML were randomized to GM-CSF or placebo following induction withdaunorubicin (DaunoXome) and cytarabine. There was no difference in disease-freesurvival in patients who achieved a complete response, but median overallsurvival was more than doubled in the patients randomized to GM-CSF as comparedto the placebo group (10.6 months for the group on GM-CSF vs 4.8 months forthose on placebo).

Another prospective randomized placebo-controlled trial of GM-CSFadministered during and after induction treatment for AML in elderly patientswas conducted by the Groupe Ouest-Est Leucémies Aiguës Myéloblastiques (GOELAM).[33]In this study, 240 patients with AML (55 to 75 years old) were randomized to Escherichia coli-derived GM-CSF or placebo following induction with idarubicin(Idamycin) and cytarabine. Disease-free survival was improved in patients on GM-CSFwho achieved complete response, and there was overall survival benefit that wasof borderline statistical significance (P = .082). The 2-year overall survivalwas 39% in the patients randomized to GM-CSF and 27% in those randomized toplacebo.

In contrast, two studies showed that G-CSF does not prolong survival inpatients with AML.[34,35] SWOG conducted a study (protocol 9031) in 211 patientsover 55 years old and reported no survival benefit (P = .71).[34] A second studywas conducted by the same group that had reported positive results in the studyof GM-CSF (GOELAM).[35] They conducted a randomized trial of G-CSF in 194patients in complete response after intensive consolidation chemotherapy inpatients with AML and found no difference in disease-free or overallsurvival.[35]

Edmonson et al studied 15 adult patients with advanced sarcoma treated withchemotherapy followed by GM-CSF.[36] They found that the median survival timewas 27 months (double the usual survival expectation) and that there were fivelong-term survivors. Three of these five had excision of solitary recurrences(lung and omentum)—a finding similar to our observations in patients withmelanoma, referenced above.

In patients with locally advanced breast cancer, a study was done evaluatingprolonged neoadjuvant chemotherapy with GM-CSF (E coli-derived).[37] A totalof 42 patients were evaluated; results showed a 98% response rate with 50%complete response. Disease-free survival at 3 years was 57% and overall survivalat 3 years was 79%. On the basis of these encouraging findings a randomizedtrial has been initiated in Europe of GM-CSF plus neoadjuvant therapy vsneoadjuvant therapy alone. The trial also compares GM-CSF with G-CSF.

Promising results have also been reported in patients with advanced prostatecancer.[38] Small et al evaluated 36 patients with hormone-refractory prostatecancer. Ten of the first 22 patients (45%) demonstrated oscillating levels ofprostate-specific antigen (PSA) coinciding with GM-CSF administration (onepatient had a cerebral vascular accident before being treated and was excludedfrom the analysis). In the next 13 patients, all but one experienced a declinein PSA concurrent with the administration of GM-CSF. Only one patient hadclinical evidence of a response other than the PSA data.

The principles described herein pertaining to the activity of GM-CSF incancer therapy have been applied in other settings as well. DeGast et aldeveloped a combined chemobiotherapy protocol with temozolomide, subcutaneousGM-CSF, low-dose IL-2, and alpha-interferon. They reported promising results inpatients with melanoma, and also in patients with metastatic renal cellcarcinoma, in whom the biologic agents were administered without temozolomide.[39]O’Day also applied these principles in a maintenance biotherapy regimen formelanoma patients that involved decrescendo IL-2 pulses, maintenancesubcutaneous IL-2, and GM-CSF following response to biochemotherapy.[40] He hasreported promising results in these patients. In 23 patients treated followingbiochemotherapy, median survival was 14.1 months as compared with 9.1 months inpatients treated with biochemotherapy without maintenance (P = .0027).

Other Agents

Tumor vaccines are being extensively studied as surgical adjuvant therapy ofmelanoma. The approach, aimed at enhancing immune responses to the patient’stumor, has considerable popular and scientific appeal. There have been a numberof phase II studies using historical controls that suggested potential benefitof tumor vaccines in the surgical adjuvant setting. However, the threeprospective randomized phase III trials of tumor vaccines in the surgicaladjuvant setting were all negative.[41-43] As of now, no vaccines have beenapproved for use as adjuvant therapy of melanoma in this country; they areavailable only in the context of clinical trials.

Biochemotherapy is also being studied as surgical adjuvant therapy ofmelanoma. This involves a regimen of combination chemotherapy and biotherapy,which appears to have a higher response rate than single-agent or combinationregimens previously described.[44] This combination includes cisplatin,vinblastine, and dacarbazine (DTIC-Dome), in combination with IFN alfa-2a andIL-2. It is hoped that this regimen of biochemotherapy might be effective in thesurgical adjuvant setting even though previous trials of dacarbazine alone havebeen negative. The combination is now being used in the surgical adjuvantsetting even though it is associated with substantial toxicity.

In phase II trials, other agents have been reported promising, but these havenot yet been tested in prospective randomized phase III trials. Retsas et alevaluated the efficacy of adjuvant therapy with the chemotherapeutic agentvindesine in patients with malignant melanoma metastatic to regional lymphnodes.[45] They evaluated a total of 169 evaluable patients, 87 of whichreceived adjuvant vindesine following resection of palpable metastatic lymphnodes and 82 had no systemic treatment after surgery. The disease-free interval,time to dissemination from lymph node metastases, survival time after lymph nodedissection, and overall survival time after the initial diagnosis of melanomawere all statistically superior for the patients who received adjuvant vindesinetherapy.

Lissoni et al tested the pineal hormone melatonin administered to patientsfollowing surgical excision of lymph node relapsed patients with melanoma.[46] Atotal of 30 patients were randomized to receive no treatment or adjuvant therapywith melatonin (20 mg/d orally in the evening) every day until diseaseprogression. They reported benefit in terms of disease-free survival in thepatients given melatonin.

Conclusions

The only agents that have been reported to be of benefit in prospectiverandomized trials are high-dose alpha-interferon[5] and levamisole,[47] but theefficacies of both have been challenged by reports of negative trials.[6,48]Alpha-interferon is the only agent approved for adjuvant therapy of melanoma inthe United States, but as the benefits are not great and the side effects aresubstantial, it is hard to justify recommending this therapy to patients. Amacrophage activator (GM-CSF), tumor vaccines, chemotherapy, and melatonin offerthe potential for greater efficacy and less toxicity, but the efficacy of theseagents will remain unknown until appropriate randomized clinical trials arecompleted. It is hoped that these newer approaches will result in therapies thatare more effective and have lower toxicity than interferon, the only currentlyapproved agent.

References:

1. Buzaid AC, Ross MI, Balch CM, et al: Critical analysis of the currentAmerican Joint Committee on Cancer Staging System for cutaneous melanoma andproposal of a new staging system [see comments]. J Clin Oncol 15:1039-1051,1997.

2. Balch CM, Buzaid AC, Atkins MB, et al: A new American Joint Committee onCancer staging system for cutaneous melanoma. Cancer 88:1484-1491, 2000.

3. Balch CM, Buzaid AC, Soong SJ, et al: Final version of the American JointCommittee on Cancer Staging System for cutaneous melanoma. J Clin Oncol19:3635-3648, 2001.

4. Balch CM, Soong SJ, Gershenwald JE, et al: Prognostic factors analysis of17,600 melanoma patients: Validation of the American Joint Committee on CancerMelanoma Staging System. J Clin Oncol 19:3622-3634, 2001.

5. Kirkwood JM, Strawderman MH, Ernstoff MS, et al: Interferon alfa-2badjuvant therapy of high-risk resected cutaneous melanoma: The EasternCooperative Oncology Group Trial EST1684 [see comments]. J Clin Oncol 14:7-17,1996.

6. Kirkwood JM, Ibrahim JG, Sondak VK, et al: High- and low-dose interferonalfa-2b in high-risk melanoma: First analysis of Intergroup trialE1690/S9111/C9190. J Clin Oncol 18:2444-2458, 2000.

7. Livingston PO, Wong GY, Adluri S, et al: Improved survival in stage IIImelanoma patients with GM2 antibodies: A randomized trial of adjuvantvaccination with GM2 ganglioside. J Clin Oncol 12:1036-1044, 1994.

8. Livingston PO, Adluri S, Helling F, et al: Phase 1 trial of immunologicaladjuvant QS-21 with a GM2 ganglioside-keyhole limpet haemocyanin conjugatevaccine in patients with melignant melanoma. Vaccine 12:1275-1280, 1994.

9. Kirkwood JM, Ibrahim JG, Sosman JA, et al: High-dose interferon alfa-2bsignificantly prolongs relapse-free and overall survival compared with theGM2-KLH/QS-21 vaccine in patients with resected stage IIB-III melanoma: Resultsof intergroup trial E1694/S9512/C509801. J Clin Oncol 19:2370-2380, 2001.

10. Creagan ET, Dalton RJ, Ahmann DL, et al: Randomized, surgical adjuvantclinical trial of recombinant interferon alfa-2a in selected patients withmalignant melanoma. J Clin Oncol 13:2776-2783, 1995.

11. Cascinelli N: Evaluation of efficacy of adjuvant rIFN-alpha-2a inmelanoma patients with regional node metastases (abstract 1296). Proc Am SocClin Oncol 14:410, 1995.

12. Eggermont A: The EORTC melanoma group experience: IFN-alpha2b adjuvanttherapy in stage II-III melanoma phase III trial experience in > 2000 pts.Proc Am Soc Clin Oncol 20:158, 2001.

13. Grob JJ, Dreno B, Delaunay M, et al: Long-term results of adjuvanttherapy with low dose IFN-a2A in resected primary melanoma thicker than 1.5 mmwithout clinically detectable node metastases (abstract 1983). Proc Am Soc ClinOncol 17:514a, 1998.

14. Pehamberger H, Soyer HP, Steiner A, et al: Adjuvant interferon alfa-2atreatment in resected primary stage II cutaneous melanoma. Austrian MalignantMelanoma Cooperative Group [see comments]. J Clin Oncol 16:1425-1429, 1998.

15. Fidler IJ, Kleinerman ES: Lymphokine-activated human blood monocytesdestroy tumor cells but not normal cells under cocultivation conditions. J ClinOncol 2:937-943, 1984.

16. Grabstein KH, Urdal DL, Tushinski RJ, et al: Induction of macrophagetumoricidal activity by granulocyte-macrophage colony-stimulating factor.Science 232:506-508, 1986.

17. Thomassen MJ, Barna BP, Rankin D, et al: Differential effect ofrecombinant granulocyte macrophage colony-stimulating factor on human monocytesand alveolar macrophages. Cancer Res 49:4086-4089, 1989.

18. Chachoua A, Oratz R, Hoogmoed R, et al: Monocyte activation followingsystemic administration of granulocyte-macrophage colony-stimulating factor. JImmunother Emphasis Tumor Immunol 15:217-224, 1994.

19. Wing EJ, Magee DM, Whiteside TL, et al: Recombinant human granulocyte/macrophagecolony-stimulating factor enhances monocyte cytotoxicity and secretion of tumornecrosis factor alpha and interferon in cancer patients. Blood 73:643-646, 1989.

20. Young JW, Szabolcs P, Moore MA: Identification of dendritic cellcolony-forming units among normal human CD34+ bone marrow progenitors that areexpanded by c-kit-ligand and yield pure dendritic cell colonies in the presenceof granulocyte/macrophage colony-stimulating factor and tumor necrosis factoralpha [published erratum appears in J Exp Med 183:1283, 1996]. J Exp Med182:1111-1119, 1995.

21. Szabolcs P, Moore MA, Young JW: Expansion of immunostimulatory dendriticcells among the myeloid progeny of human CD34+ bone marrow precursors culturedwith c-kit ligand, granulocyte-macrophage colony-stimulating factor, and TNF-alpha.J Immunol 154:5851-5861, 1995.

22. Szabolcs P, Avigan D, Gezelter S, et al: Dendritic cells and macrophagescan mature independently from a human bone marrow-derived, post-colony-formingunit intermediate. Blood 87:4520-4530, 1996.

23. Kumar R, Dong Z, Fidler IJ: Differential regulation of metalloelastaseactivity in murine peritoneal macrophages by granulocyte-macrophagecolony-stimulating factor and macrophage colony-stimulating factor. J Immunol157:5104-5111, 1996.

24. Dong Z, Kumar R, Yang X, et al: Macrophage-derived metalloelastase isresponsible for the generation of angiostatin in Lewis lung carcinoma. Cell88:801-810, 1997.

25. Spitler LE, Grossbard ML, Ernstoff MS, et al: Adjuvant therapy of stageIII and IV malignant melanoma using granulocyte-macrophage colony-stimulatingfactor [see comments]. J Clin Oncol 18:1614-1621, 2000.

26. Balch CM, Soong SJ, Murad TM, et al: A multifactorial analysis ofmelanoma. IV. Prognostic factors in 200 melanoma patients with distantmetastases (stage III). J Clin Oncol 1:126-134, 1983.

27. Balch CM, Soong SJ, Murad TM, et al: A multifactorial analysis ofmelanoma. III. Prognostic factors in melanoma patients with lymph nodemetastases (stage II). Ann Surg 193:377-388, 1981.

28. Spitler LE, Jacobs M, Weber R, et al: Adjuvant therapy of stage III andIV melanoma using granulocyte-macrophage colony-stimulating factor (GM-CSF)(abstract 1429). Proc Am Soc Clin Oncol 20:358a, 2001.

29. Bugelski PJ, Kirsh RL, Poste G: New histochemical method for measuringintratumoral macrophages and macrophage recruitment into experimentalmetastases. Cancer Res 43:5493-5501, 1983.

30. Bugelski PJ, Kirsh RL, Sowinski JM, et al: Changes in the macrophagecontent of lung metastases at different stages in tumor growth. Am J Pathol118:419-424, 1985.

31. Bugelski PJ, Corwin SP, North SM, et al: Macrophage content ofspontaneous metastases at different stages of growth. Cancer Res 47:4141-4145,1987.

32. Rowe JM, Andersen JW, Mazza JJ, et al: A randomized placebo-controlledphase III study of granulocyte-macrophage colony-stimulating factor in adultpatients (> 55 to 70 years of age) with acute myelogenous leukemia: A studyof the Eastern Cooperative Oncology Group (E1490). Blood 86:457-462, 1995.

33. Witz F, Sadoun A, Perrin MC, et al: A placebo-controlled study ofrecombinant human granulocyte-macrophage colony-stimulating factor administeredduring and after induction treatment for de novo acute myelogenous leukemia inelderly patients. Groupe Ouest Est Leucemies Aigues Myeloblastiques (GOELAM).Blood 91:2722-2730, 1998.

34. Godwin JE, Kopecky KJ, Head DR, et al: A double-blind placebo-controlledtrial of granulocyte colony-stimulating factor in elderly patients withpreviously untreated acute myeloid leukemia: A Southwest Oncology Group study(9031). Blood 91:3607-3615, 1998.

35. Harousseau JL, Witz B, Lioure B, et al: Granulocyte colony-stimulatingfactor after intensive consolidation chemotherapy in acute myeloid leukemia:Results of a randomized trial of the Groupe Ouest-Est Leucemies AiguesMyeloblastiques. J Clin Oncol 18:780-787, 2000.

36. Edmonson JH, Long HJ, Kvols LK, et al: Can molgramostim enhance theantitumor effects of cytotoxic drugs in patients with advanced sarcomas? AnnOncol 8:637-641, 1997.

37. Honkoop AH, Luykx-de Bakker SA, Hoekman K, et al: Prolonged neoadjuvantchemotherapy with GM-CSF in locally advanced breast cancer. Oncologist4:106-111, 1999.

38. Small EJ, Reese DM, Um B, et al: Therapy of advanced prostate cancer withgranulocyte macrophage colony-stimulating factor. Clin Cancer Res 5:1738-1744,1999.

39. De Gast GC, Klupen H-J, Vyth-Dreese FA, et al: Phase I trial of combinedimmunotherapy with subcutaneous GM-CSF, low dose IL-2 and IFNa in progressivemetastatic melanoma or renal cell carcinoma (abstract 1720). Proc Am Soc ClinOncol 18:446a, 1999.

40. O’Day SJ: Maintenance biotherapy with decrescendo interleukin-2 and GM-CSFfor patients with metastatic melanoma following modified concurrentbiochemotherapy. Immune-Enhancing Cytokines 2:6-10, 2000.

41. Hersey P, Coates A, McCarthy WH, et al: Vaccinia melanoma lysates (VMCL)in immunotherapy of stage IIb and III melanoma. Third International Conferenceon the Adjuvant Therapy of Malignant Melanoma. Royal College of Physicians,London, 1999, pp 1-15.

42. Wallack MK, Sivanandham M, Balch CM, et al: A phase III randomized,double-blind multi-institutional trial of vaccinia melanoma oncolysate-activespecific immunotherapy for patients with stage II melanoma. Cancer 75:34-42, 1995.

43. Kirkwood JM: ECOG studies of vaccines, past and present. Melanoma Res11:S48-S49, 2001.

44. Legha SS, Ring S, Eton O, et al: Development of a biochemotherapy regimenwith concurrent administration of cisplatin, vinblastine, dacarbazine,interferon alfa, and interleukin-2 for patients with metastatic melanoma. J ClinOncol 16:1752-1759, 1998.

45. Retsas S, Quigley M, Pectasides D, et al: Clinical and histologicinvolvement of regional lymph nodes in malignant melanoma. Adjuvant vindesineimproves survival. Cancer 73:2119-2130, 1994.

46. Lissoni P, Brivio O, Brivio F, et al: Adjuvant therapy with the pinealhormone melatonin in patients with lymph node relapse due to malignant melanoma.J Pineal Res 21:239-242, 1996.

47. Quirt IC, Shelley WE, Pater JL, et al: Improved survival in patients withpoor-prognosis malignant melanoma treated with adjuvant levamisole: A phase IIIstudy by the National Cancer Institute of Canada Clinical Trials Group [seecomments]. J Clin Oncol 9:729-735, 1991.

48. Spitler LE: A randomized trial of levamisole versus placebo as adjuvanttherapy in malignant melanoma. J Clin Oncol 9:736-740, 1991.