This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Allen, P. J. Articles by Coit, D. G. Search for Related Content PubMed Articles by Allen, P. J. Articles by Coit, D. G. Related Articles Related Correspondence Social Bookmarking                            What's this?

In Reply:

Department of Surgery and Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY

We appreciate the opportunity to respond to the letters by Dr Veness and Dr Wilson regarding our recent report of 251 patients treated at Memorial Sloan-Kettering Cancer Center for Merkel cell carcinoma.¹

Both have correctly pointed out a consistent error in labeling of the columns in Tables 4 and 5. Correctly stated, among 185 patients with margin negative resection of their primary tumors, 160 did not receive radiation therapy (RT); 12 (8%) recurred locally. Among 73 patients who underwent lymphadenectomy for pathologically confirmed nodal metastases, 51 patients did not receive RT, eight (14%) recurred in the nodal basin. We certainly do not present these retrospective observations as Class I evidence for or against the routine use of adjuvant radiation. As such, we do not feel that further statistical analysis, as suggested by Dr Wilson, will contribute much more to our understanding of the impact of adjuvant radiation of locoregional control.

Dr Veness refers to the results of the recent Trans-Tasman Radiation Oncology Group (TROG) phase II trial of patients with Merkel cell carcinoma.² We applaud their prospective efforts to investigate the role of multimodality treatment of this rare entity, and commend their completion of a multi-institutional study of 53 patients with locoregional disease treated with resection, RT, and chemotherapy. Unfortunately, neither our retrospective review, nor the TROG study can provide a definitive answer to the question of optimal stage-specific treatment of this unusual disease.

We are aware of the dramatic differences between our recommendations for the treatment of this disease and those of Dr Veness. We do not routinely recommend adjuvant RT following complete excision and operative nodal staging. Our data demonstrate that local recurrence is low after margin negative excision (8%), and that regional nodal recurrence is low after lymph node dissection for positive nodes (8% to 15%, depending on the clinical status of the nodes). Adjuvant RT was not associated with decreased local or regional recurrence in our study. A randomized study designed to demonstrate a reduction in local or regional recurrence from 12% to 6% with the addition of RT would require 281 patients per arm (one-sided P = .05, power 0.8), a study that will clearly never be undertaken. In the absence of such prospective data, we use adjuvant RT selectively, for those patients deemed at high risk for locoregional failure (ie, margin positive excision with no better resection possible, multiple nodes positive, extranodal extension).

In our study we were unable to demonstrate a survival benefit associated with the use of adjuvant chemotherapy, in either node-negative or node-positive patients. In fact, among node-positive patients, those who received chemotherapy actually fared worse than those who did not, an effect attributed to factors of patient selection rather than to any adverse effect of the treatment itself. As such, we do not recommend adjuvant chemotherapy to node-negative patients, and again, only selectively to node-positive patients at high risk for systemic failure (ie, multiple positive nodes, extranodal extension).

Despite the dramatic differences in treatment approach, our published recurrence and survival results are nearly identical to those of the TROG study. In our study, 131 patients presented with locoregional disease and were treated with wide excision, operative nodal staging/treatment, without adjuvant chemotherapy or RT. In this surgery-only group of patients, the three-year recurrence-free survival rate was 66%. The three-year recurrence-free survival rate in the TROG study was 75%. The three-year survival rate for node-negative patients within our surgery-only group was 96% (n = 47), versus 93% (n = 20) in the TROG study. The three-year survival rate for node-positive patients in our surgery-only group was 75% (n = 37), versus 66% (n = 33) in the TROG study. While the treatment approaches were different, the outcomes were similar.

It is also important however, to compare apples with apples. Locoregional (local + regional) recurrence may be as high as 45% when patients do not undergo pathologic nodal staging. Clinically occult regional nodal disease was identified by sentinel lymph node biopsy or elective lymph node dissection in 25% of patients in our study. In patients who did not undergo pathologic staging the regional recurrence rate was approximately 40%. Clinical staging of the regional nodal basin is therefore inadequate. Unfortunately, many studies report regional recurrence rates in patients who were treated before the era of operative nodal staging.^3,4 Delivering RT to a clinically negative nodal basin may help treat the occult disease that will be present in up to a third of patients, but it will also result in the majority of patients experiencing the time, cost, and toxicity of the therapy for no benefit. In fact, although Dr Veness alludes to the safety of contemporary chemoradiotherapy, one patient (out of 53 total patients treated) in the TROG study required lower extremity amputation, at least in part attributable to treatment related toxicity. We therefore feel that our approach of pathologic nodal staging with sentinel lymph node biopsy is at least as rational as a policy of routine radiation of the nodal basin for all patients with clinically localized disease.

The TROG study enrolled patients with Merkel cell carcinoma who were considered at high risk for recurrence to undergo surgical excision followed by locoregional RT and chemotherapy. The high-risk selection factors included any one or more of the following: tumor size greater than 1 cm, positive nodes, recurrent disease, or gross residual disease after surgery. It was not possible to ascertain how many patients in the TROG study fulfilled each of those entry requirements. We do not consider patients with tumors greater than 1 cm and pathologically negative nodes as high risk. In our study, 55 patients with pathologically negative nodes experienced a 3-year survival rate of 96%, independent of primary tumor size. We do not feel that adjuvant therapy of any sort can be justified in this group of patients. We would agree that patients with nodal disease, particularly those with extensive nodal disease, and those patients who recur after appropriate initial therapy are at risk for further locoregional disease recurrence and death, and are appropriate candidates for evaluation of aggressive multimodality adjuvant therapy.

In light of the similarity between our results and those of the TROG study, we feel that the additional toxicity, time, and cost of routine adjuvant RT and chemotherapy are unwarranted in node-negative patients. We do not recommend adjuvant radiotherapy to the surgical bed after margin-negative excision, and we do not recommend adjuvant RT to the regional nodal basin after operative nodal staging. Based on current retrospective reviews, we cannot make any recommendation about the impact of adjuvant therapy in high-risk patients. We strongly support additional prospective studies to analyze the impact of adjuvant treatment in appropriately risk-stratified cohorts.

In the meantime, Dr Veness does a disservice in describing all patients with Merkel cell carcinoma as having "an incredibly aggressive primary tumor." The outcome in these patients, as with virtually all other malignancies, is highly dependent on stage at presentation. In the absence of data from prospective randomized trials, treatment recommendations should be stage specific, based not on bias, but rather on thoughtful analysis of the natural history of the disease. To suggest otherwise could put many patients with early stage disease at significant risk of over treatment.

Authors' Disclosures of Potential Conflicts of Interest

The authors indicated no potential conflicts of interest.

REFERENCES

1. Allen PJ, Bowne WB, Jaques DP, et al: Merkel cell carcinoma: Prognosis and treatment of patients from a single institution. J Clin Oncol 23:2300-2309, 2005[Abstract/Free Full Text]

2. Poulsen M, Rischin D, Walpole E, et al: High-risk Merkel cell carcinoma of the skin treated with synchronous carboplatin/etoposide and radiation: A trans-tasman radiation oncology group study—TROG 96:07. Amer Soc Clin Oncol 21:4371-4376, 2003

3. Morrison WH, Peters LJ, Silva EG, et al: The essential role of radiation therapy in securing locoregional control of Merkel cell carcinoma. Int J Radiat Oncol Biol Phys 19:583-591, 1990[Medline]

4. Veness MJ, Morgan GJ, Gebski V: Adjuvant locoregional radiotherapy as best practice in patients with Merkel cell carcinoma of the head and neck. Head Neck 27:208-216, 2005[Medline]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

Related Correspondence

• Merkel Cell Carcinoma: Improved Outcome With the Addition of Adjuvant Therapy
  Michael J. Veness
  JCO 2005 23: 7235-7236 [Full Text]
• Merkel Cell Carcinoma: Improved Outcome With the Addition of Adjuvant Therapy
  Lynn D. Wilson, Douglas Housman, and Benjamin D. Smith
  JCO 2005 23: 7236-7237 [Full Text]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Allen, P. J. Articles by Coit, D. G. Search for Related Content PubMed Articles by Allen, P. J. Articles by Coit, D. G. Related Articles Related Correspondence Social Bookmarking                            What's this?