|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology High-Risk Merkel Cell Carcinoma of the Skin Treated With Synchronous Carboplatin/Etoposide and Radiation: A Trans-Tasman Radiation Oncology Group Study—TROG 96:07From the Division of Oncology, Queensland Radium Institute; Royal Brisbane Hospital; Princess Alexandra Hospital; Department of Medical Oncology, Mater Medical Centre, Brisbane; the Department of Medical Oncology and the Department of Radiation Oncology, Peter MacCallum Cancer Centre, Melbourne; Department of Radiation Oncology, Newcastle Mater Misericordia Hospital, Newcastle, Australia. Address reprint requests to Michael Poulsen, MD, Mater Queensland Radium Institute, Raymond Tee, South Brisbane, QLD 4101 Australia; e-mail: michael_poulsen{at}health.qld.gov.au.
Purpose: The effectiveness of synchronous carboplatin, etoposide, and radiation therapy was prospectively assessed in a group of patients with high-risk Merkel cell carcinoma (MCC) of the skin. Patients and Methods: Patients were eligible if they had disease localized to the primary site and nodes, and were required to have at least one of the following high risk features: recurrence after initial therapy, involved nodes, primary tumor size greater than 1 cm, gross residual disease after surgery, or occult primary with nodes. Radiation was delivered to the primary site and nodes to a dose of 50 Gy in 25 fractions over 5 weeks and synchronous carboplatin (area under the curve, 4.5) and intravenous etoposide 80 mg/m2 days 1 to 3 was given in weeks 1, 4, 7, and 10. The median age of the group was 67 (range, 43–86) years, and there were 39 males and 14 females. Involved nodes (stage II) were present in 33 cases (62%). The sites involved were head and neck (22 patients), occult primary (13 patients), upper limb (eight patients), lower limb (eight patients), and trunk (two patients). Results: Fifty-three patients were entered between 1996 and 2001. The median potential follow-up was 48 months. There were no treatment related deaths. The 3-year overall survival, locoregional control, and distant control were 76%, 75%, and 76%, respectively. Tumor site and the presence of nodes were factors that were predictive for local control and survival. Multivariate analysis indicated that the major factor influencing survival was the presence of nodes; however, this was not a significant factor in locoregional control. Conclusion: High levels of locoregional control and survival have been achieved with the addition of chemotherapy to radiation treatment for high-risk MCC of the skin. The role of chemoradiotherapy for high-risk MCC warrants further investigation.
THE TRANS-TASMAN Radiation Oncology Group embarked on a phase II study of synchronous chemoradiotherapy in high-risk Merkel cell carcinoma (MCC) of the skin in 1996. High risk MCC was defined as primary tumors more than 1 cm in diameter, and/or the presence of in-transit or nodal metastases, and/or locally or regionally lymphatic recurrent tumor, but with no evidence of distant disease. The use of radiotherapy in locoregional control of MCC is considered standard treatment, although the optimal combination of surgery and radiotherapy is contentious. The rationale for investigating synchronous chemoradiotherapy was based on the tumor’s high metastatic potential, its inherent chemosensitivity and radiosensitivity, and its similarity to small-cell carcinoma of the lung. This approach has become standard practice in the management of small-cell carcinoma of the lung, which has similar histologic morphology and biologic behavior to MCC. The main aims of the study were to investigate the efficiency and safety of chemotherapy in MCC, and to determine whether this regimen could be delivered in a multi-institutional setting. This is the first report in the literature, using a standard approach to MCC of the skin, that incorporates chemotherapy and radiation. Most reports have been from individual institutions and suffer from a lack of power because of the small numbers and inconsistent treatment regimens. A preliminary report on the acute and late toxicity for this protocol was published after the first 40 patients had been treated.1 We now report the overall survival, locoregional control, and distant metastases control rates achieved with the regimen, as well as updating the toxicity in all 53 patients treated with this protocol.
Trial Design Patients were only eligible for this study if they had disease confined to the primary or nodes region, and they were at high-risk of recurrence. This was defined as having at least one of the following criteria: primary size greater than 1 cm, involved nodes, recurrence following initial surgery (providing the recurrence was outside of the previous radiation field), gross residual disease after surgery, or occult primary with nodes. In addition, the following eligibility criteria were required to be met: biopsy proven MCC confined to the primary and regional nodes, Eastern Cooperative Oncology Group performance status of 0 to 2, no previous malignancy in the past 5 years other than nonmelanomatous skin cancer, hemoglobin > 10g/dL, neutrophils > 2 x 109/L, platelets > 100 x 109/L, and normal renal function (glomerular filtration rate > 50 mL/min). The protocol was approved by each of the hospital ethics committees and written informed consent was obtained on all patients. Activation of the trial occurred in 1996 and closure occurred in 2001. The close out date for the analysis was July 8, 2002.
Pretreatment Evaluation
Treatment
Radiotherapy A dose of 50 Gy (Internation Commission on Radiation Units and Measurements 50) in 25 fractions over 5 weeks was prescribed to macroscopically involved areas or to the operative bed. Electron doses were specified at the 90% isodose line. A shrinking field technique was used after 46 Gy if field volumes were large. Clinically uninvolved areas were treated to a dose of 45 Gy in 25 fractions over 5 weeks (eg, clinically uninvolved nodal areas that are outside the surgical volume). A dose reduction to 45 Gy was recommended for lesions located below the knee or areas where 50 Gy was thought to be poorly tolerated (eg, around the inguinal nodes or adjacent to the orbit).
Chemotherapy Glomerular filtration rate was calculated according to the Cockcroft and Gault Formula or by measurement of Tc-99m DTPA clearance. The latter method, which was being used at one site, was abandoned as it resulted in higher doses of carboplatin and therefore, lack of uniformity in dose calculations between sites. Etoposide was given in a dose of 80 mg/m2/d intravenously days 1 to 3. Once the radiation was completed, a further two cycles of chemotherapy were given in weeks 7 and 10, resulting in a total of four cycles. It was specified that the pretreatment neutrophil count should be > 1.5x 109/L and platelet count should be > 100 x 109/L. Blood counts were required weekly during the chemoradiotherapy. If the neutrophil or platelet count had not recovered by week 4, then the remainder of the chemotherapy was to be given once every 4 weeks (weeks 1, 5, 9, 13). Providing the counts had recovered, the option of once every 3 weeks chemotherapy remained. A dose reduction of carboplatin to AUC 4 and etoposide dose reduction of 20% was recommended if there was any grade 3 or 4 nonhematologic toxicity (excluding skin, hair loss, nausea, vomiting), grade 4 neutropenia, or fever with grade 3 or 4 neutropenia. Weight change of greater than 5% or a change in creatinine greater than 10 µmol/L required recalculation of the chemotherapy doses.
Quality Assurance
Follow-Up
Outcome Variables
Statistical Methods
Patient Characteristics A total of 53 patients were registered from six institutions in Australia. No patients were excluded from the analysis. The median potential follow-up time was 48 (range, 11–70) months and no patients were lost to follow-up. Twelve patients registered in the study presented with recurrence and had received prior therapy. The remaining 41 all received chemoradiotherapy as their initial treatment. No patients had received prior radiation treatment. In total, the chemoradiotherapy was given as an adjuvant treatment in 38 patients (72%) and as definitive therapeutic treatment in 15 patients (28%). The distribution of baseline characteristics is summarized in Table 1  .
Survival Analyses The 3 year overall survival was 76% (95% CI, 63.5 to 88.7; Fig 1A ). Relapse free survival was 65% (95% CI, 50.8 to 78.5) and distant control was achieved in 76% (95% CI, 63.4 to 88.7) of patients (Fig 1B). The 3-year survival for those with occult and known primary sites were 83% (95%CI, 61.5 to 100) and 74% (95% CI, 58.7 to 88.8), and distant control rates were 91% (95% CI, 73.9 to 100) and 56% (95% CI, 55.9 to 86.9), respectively.
Univariate analyses were performed by splitting the groups according to age, nodal status, site, and presence of gross residual disease at the time of chemoradiotherapy. The 3-year survival for patients with involved nodes compared with no nodal involvement was 66% and 93% (P = .27). Patients who had gross residual disease at the time of chemoradiotherapy had a poorer outcome of 45%, compared to 90% (P = .07).
Locoregional Control
Acute Toxicity
Late Toxicity Grade 3 or higher late skin and subcutaneous toxicity was 15% and 13%, respectively, at 3 years. One patient developed acute arterial obstruction in the femoral vessels at the site of an inguinal radiotherapy field 4 years after registration in the study. The inguinal node had been resected before radiotherapy and radiotherapy/chemotherapy was delivered to the involved inguinal and iliac nodes as per protocol. This patient had experienced a grade 4 skin reaction in the groin and had severe late and subcutaneous changes present in the radiation field. She was also noted to have risk factors for peripheral vascular disease. The obstruction was successfully treated with a bypass procedure but she went on to have an amputation 18 months later when the graft occluded.
Multivariate Analysis
The presence of nodes was not significant when locoregional control was used as an end point. Primary sites in the lower limb had a significant risk of locoregional failure with a hazard ratio of 12.67 (P = .002). The results of the Cox’s proportional hazards modeling have been summarized in Table 3
Quality Assurance An audit of the first 20 patients indicated that the percentage of patients treated strictly according to protocol (with no variation in timing or dose) was 50% for chemotherapy and 55% for radiotherapy. Compliance with the recommended radiation dose was checked in all 53 patients. Patients were treated as per protocol (defined as ± 5%) in 83% of cases. Minor deviations in dose occurred in 9.5% of cases. Major dose deviations of an acceptable nature occurred in 7.5% of cases. There were no major deviations in dose of an unacceptable nature. All four chemotherapy doses were given in 46 patients. There were two patients who received only one cycle, three that received two cycles, and two patients that received three cycles. In total there were 145 gaps or time intervals between the chemotherapy cycles. One hundred and thirty one (90.3%) of these time intervals were as per protocol (< 27 days). Prolongations of 8 to 14 days beyond the recommended interval occurred in 12 (8.3%) occasions and on two occasions there were prolongations of greater than 28 days.
Unlike previous reports in the literature, this group of patients has been prospectively assessed and treated with a uniform protocol of chemoradiotherapy with curative intent. There has been a high level of compliance in the delivery of the radiation dose and 87% of the patients were able to complete all four courses of chemotherapy. Only 9.7% of patients had protracted intervals greater than the recommendations in the protocol. The extent of the surgical treatment of primary and nodal disease could not be rigorously standardized in this trial because of the heterogeneity of disease status and the variety of anatomic locations in this rare skin malignancy. Furthermore, surgery was frequently performed before referral to the oncologist, and there is no evidence that further radical surgery would confer any benefit to locoregional control over and above this radical chemoradiotherapy trial regimen. Conversely, delaying the chemoradiotherapy treatment to undertake further radical surgery (that was likely to reduce tolerance of the trial regimen) was considered potentially detrimental to tumor control by reducing treatment efficacy. Although the treatment protocol was standardized for the chemotherapy and radiation therapy, the patient population was relatively heterogeneous in terms of patient and tumor parameters. The study only included patients who were perceived to be at high risk of recurrence, but they had to be fit enough to tolerate chemoradiation. All reported Merkel cell carcinoma studies suffer from some degree of heterogeneity in the patient population, and this limits the ability to compare since any differences detected may be as a result of patient selection rather than the treatment itself. The heterogeneity in this study should be less than others as we have confined our study group to patients that have high-risk locoregional disease, but good performance status; that is, patients with low-risk disease, distant metastatic disease, or poor performance status are excluded. Nevertheless, it is acknowledged that there are real limitations in comparing these results with others in the literature. Prognostic factors have been well described in previous reports3–12 and a recent review by Goessling et al13 summarizes these. In comparing these results with those reported in the literature, it is important to match the proportion of patients with adverse prognostic variables, the most important being the presence of nodal metastases.3–7 Thirty-three patients (62%) registered on this study had involved nodes (stage II) at presentation. This contrasts with the literature where the reported incidence of nodes at presentation varied from 9% to 33%,3–5,8–11 and confirms the particularly poor prognostic profile of patients in this trial (being at high risk of disseminated disease). The 3-year survival of 76% compares favorably with others in the literature. Meeuwissen et al3 reported 68% 3-year survival for a group of 80 patients (33% had stage II disease) treated at the Queensland Radium Institute, and Morrison et al4 achieved 30% 5-year survival for a group of 54 patients (16% had stage II disease). Other reports claim 63% 5-year survival (20% had stage II disease);14 58% 3-year survival (26% had stage II disease);5 64% 5-year survival (31% had stage II disease);6 35% 3-year survival (9% had stage II disease);10 and 40% 5-year survival (18% had stage II disease).15 Eleven (21%) of 53 patients suffered locoregional failure, and the 3-year actuarial locoregional control was 75%. In a review of the literature, Haag et al16 reported that local recurrence developed in 26% to 44% of patients, and during the course of the disease, 55% to 66% of patients developed regional node metastases. Given that 28% of patients had gross disease at either the primary site or the nodes, and 36% had recurrence before registering on the study, these figures compare favorably with those in the literature. Distant control at 3 years for the whole group was 76%, or looking at the crude figure, nine (17%) of 53 patients developed distant metastases. Early reported series quote rates of distant metastatic disease to be 30%,14,16,17 with some as high as 70%.8 This does suggest that chemotherapy may have had a positive influence in reducing distant metastatic disease. However, great caution must be taken in interpreting this data, given the significant diversity of patients, stages of disease, and various methods of reporting in these series. Occult primaries have been reported in a number of MCC series. The incidence varies from 0% to 18%,3,4,8,15 and can be attributed to spontaneous regression of the primary18 or metastatic neuroendocrine carcinoma from a noncutaneous site. The incidence in this series was proportionally higher (25%) and may be as a result of several factors. Excluding patients with primaries less than 10 mm from entering the study could have caused a relative increase in numbers with an occult primary. The hazard ratio for occult primaries was 0.02, indicating that this was a favorable prognostic variable. This may have biased the results of this series favorably compared with some of the historic series. Although the superior outcome in the occult primary group is counterintuitive, a possible explanation could be that patients with occult primaries have had an immunologic response to their cancer. A detailed report on toxicity was made after the first 40 patients were treated.1 This suggested that the most threatening side effect was associated with neutropenia. The incidence of febrile neutropenia was highest after the second cycle of chemotherapy, which coincided with maximal radiation skin reactions. It is postulated that this provided a portal of entry for bacteria at a time when the white cell count was low. We are now embarking on a pilot study aimed at reducing the incidence of febrile neutropenia. Patients will be treated with radiotherapy to a dose of 45 to 50 Gy with synchronous weekly carboplatin in a dose of AUC 2.0 to a maximum of five courses. This will be followed by three cycles of carboplatin and etoposide in the same doses in TROG 96:07. If the toxicity of this schedule is acceptable, it may form the basis of an experimental arm of a phase III randomized trial. International collaboration will be required for such as a study. The acceptability of a radiation-alone arm may cause some difficulties in the future. However, this will be the only way to overcome the biases that occur when retrospective comparisons are made with historical controls. This prospective phase II study demonstrates that chemoradiotherapy with curative intent for high risk MCC is tolerable for the vast majority of patients. Results indicate that despite the high proportion of patients with nodal metastases and gross residual disease at the time of chemoradiotherapy, 76% of patients were alive at 3 years. This survival and favorable locoregional and distant control rates appear superior to reported retrospective series in the literature even though it is recognized that we are comparing a diverse range of patients with MCC. However, a phase III study is needed if the true magnitude of the benefit of synchronous chemotherapy is to be defined.
The authors indicated no potential conflicts of interest.
1. Poulsen M, Rischen D, Walpole E, et al: Analysis of toxicity of Merkel cell carcinoma of the skin treated with synchronous carboplatin/etoposide and radiation: a Trans-Tasman Radiation Oncology Group Study. Int J Radiat Oncol Biol Phys 51:156–163, 2001[Medline] 2. Calvert AH, Newell DR, Gumbrell LA, et al: Carboplatin dosage: Prospective evaluation of a simple formula based on renal function. J Clin Oncol 7:1748–1756, 1989[Abstract] 3. Meeuwissen J, Bourne R, Kearsley J, et al: The importance of post-operative radiotherapy in the treatment of Merkel cell carcinoma. Int J Radiat Oncol Biol Phys 31:325–331, 1995[CrossRef][Medline] 4. Morrison W, Peters L, Silvia E, et al: The essential role of radiation therapy is securing loco-regional control of Merkel cell carcinoma. Int J Radiat Oncol Biol Phys 19:583–591, 1990[Medline] 5. Fenig E, Brenner B, Katz A, et al: The role of radiation therapy and chemotherapy in the treatment of Merkel cell carcinoma. Cancer 80:881–885, 1997[CrossRef][Medline]
6. Yiengpruksawan A, Coit D, Thaler H, et al: Merkel cell carcinoma: Prognosis and management. Arch Surg 126:1514–1519, 1991 7. Victor S, Morton B, Smith J: Merkel cell carcinoma: Is prophylactic lymph node dissection indicated? Am Surg 62:879–882, 1996[Medline] 8. Boyle F, Pendlebury S, Bell D: Further insights in to the natural history and management of primary cutaneous neuroendocrine (Merkel cell) carcinoma. Int J Radiat Oncol Biol Phys 31:315–323, 1995[CrossRef][Medline] 9. Shaw JH, Rumball E: Merkel cell tumour: clinical behaviour and treatment. Br J Surg 78:138–142, 1991[Medline] 10. Kokoska E, Kokoska M, Collins, et al: Early aggressive treatment of Merkel cell carcinoma improves outcome. Am J Surg 174:688–693, 1997[CrossRef][Medline] 11. Savage P, Constenla D, Fisher C, et al: The natural history and management of Merkel cell carcinoma of the skin: a review of 22 patients treated at the Royal Marsden Hospital. Clin Oncol (R Coll Radiol) 9:164–167, 1997
12. Ott MJ, Tanabe K, Gadd M, et al: Multi-modality management of Merkel cell carcinoma. Arch Surg 134:388–392, 1999
13. Goessling W, McKee PH, Mayer RJ: Merkel cell carcinoma. J Clin Oncol 20:588–598, 2002 14. Pacella J, Ashby M, Ainslie J, et al: The role of radiotherapy in the management of primary cutaneous neuroendocrine tumours: experience of the Peter MacCallum Cancer Institute (Melbourne, Australia). Int J Radiat Oncol Biol Phys 14:1077–1084, 1988[Medline] 15. Wong KC, Zuletta F, Clarke S, et al: Clinical management and treatment outcomes of Merkel cell carcinoma. Aust N Z J Surg 68:354–358, 1998[Medline] 16. Haag M, Glass LF, et al: Merkel cell carcinoma: diagnosis and treatment. Dermatol Surg 21:669–683, 1995[CrossRef][Medline] 17. Ratner D, Nelson B, Brown M, et al: Merkel cell carcinoma. J Am Acad Derm 29:143–156, 1993[Medline] 18. Yanguas I, Goday J, Gonzales-Guemes M, et al: Spontaneous regression of Merkel cell carcinoma of the skin. Br J Dermatol 137:296–298, 1997[CrossRef][Medline] Submitted March 24, 2003; accepted September 3, 2003.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 tests were used to compare proportions. All time-to-event curves were calculated using the actuarial method of Kaplan and Meier and were compared using the log-rank test. Overall survival considered any death as an event (including those from intercurrent illnesses). Locoregional control was defined as the proportion of patients who did not develop failure in the primary site, regional nodes or in-transit areas, with censoring of events occurring at death without locoregional failure. Distant control was defined as the proportion of patients not developing hematogenous spread, with censoring occurring at death in the absence of distant failure. Cox’s proportional hazards model was used to derive hazards ratios for prognostic variables using the end points of overall survival, locoregional control, and distant control. 
