The
n e w e ng l a n d j o u r na l
of
m e dic i n e
c or r e sp ondence
Sentinel-Node Biopsy in Melanoma To the Editor: In reporting the results of both the third interim analysis of the Multicenter Selective Lymphadenectomy Trial (MSLT-I) in 2006 and the current final analysis, Morton et al. (Feb. 13 issue)1 describe a disease-free survival advantage. This advantage is an inevitable consequence of the trial design and does not reflect any therapeutic advantage for patients. In both analyses, patients with positive sentinel nodes in the biopsy group underwent prophylactic lymphadenectomy. Therefore, inevitably, the site of first recurrence would be a nodal site more often in the observation group than in the biopsy group. This error was appealed and upheld in 2007 by the Clinical Investigations Branch of the National Cancer Institute (NCI) of the National Institutes of Health, which funded the trial.2 The authors were requested to provide a calculation of distant disease–free survival among all patients with intermediate-thickness tumors in the final update. This calculation is provided only for node-positive patients. After Morton and colleagues reported the results of the third interim analysis in 2006, I postulated that a proportion of positive sentinel nodes were prognostically false positive.3,4 The authors claimed to disprove this hypothesis in a this week’s letters 2148 Sentinel-Node Biopsy in Melanoma 2150 Management of Early Prostate Cancer 2152 Bleeding and Coagulopathies in Critical Care 2153 PICC Placement in the Neonate 2154 Pulmonary Fibrosis Associated with Aluminum Trihydrate (Corian) Dust
2148
graph showing that the cumulative incidence of nodal recurrences converged at 8 years.5 In the current article, the final results (Fig. 3A) are contradictory and confirm that the curves are parallel from 8 to 10 years; this reflects the presence and incidence of prognostic false positivity. J. Meirion Thomas, F.R.C.P., F.R.C.S. Royal Marsden Hospital London, United Kingdom [email protected] No potential conflict of interest relevant to this letter was reported. 1. Morton DL, Thompson JF, Cochran AJ, et al. Final trial re-
port of sentinel-node biopsy versus nodal observation in melanoma. N Engl J Med 2014;370:599-609. 2. Thomas JM. Concerns relating to the conduct and statistical analysis of the Multicenter Selective Lymphadenectomy Trial (MSLT-1) in patients with melanoma. J Plast Reconstr Aesthet Surg 2009;62:442-6. 3. Idem. Sentinel-node biopsy in melanoma. N Engl J Med 2007; 356:418. 4. Idem. Prognostic false-positivity of the sentinel node in melanoma. Nat Clin Pract Oncol 2008;5:18-23. 5. Morton DL, Cochran AJ, Thompson JF. Sentinel-node biopsy in melanoma. N Engl J Med 2007;356:419-20. DOI: 10.1056/NEJMc1402989
To the Editor: In the comparison of sentinelnode biopsy with observation of regional lymph nodes in patients with melanoma, Morton et al. report no significant difference in the rate of disease-specific survival. The question is whether delayed removal of metastases, as compared with immediate removal, allowed unimpeded growth that caused dissemination of disease and decreased survival. When a subgroup is retrospectively found to have a survival advantage, an offsetting subgroup, perhaps unrecognized, must of necessity have poorer survival with equivalent diseasespecific survival. This is critical in discussing a negative trial because of misinterpretation of contrived subgroup results showing advantages
n engl j med 370;22 nejm.org may 29, 2014
The New England Journal of Medicine Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
correspondence
to immediate node dissection. Surgery that prevents later disease improves disease-free survival but is irrelevant to melanoma-specific survival. Literature reviews1,2 do not show improved disease-specific survival in the comparison of various forms of regional node management (including dissection and observation only) in lung, esophageal, stomach, colon, rectal, or breast cancers, or in melanoma. The biologic explanation for this failure comes from laboratory studies and randomized clinical trials showing that organ-specific metastatic cells have no ability to grow in other organs.3 This explains curative resections of lymph-node, liver, and lung metastases after prolonged disease. The curability of metastatic disease is pattern-dependent, not timedependent. Blake Cady, M.D. Harvard Medical School Boston, MA [email protected] No potential conflict of interest relevant to this letter was reported. 1. Gervasoni JE Jr, Taneja C, Chung MA, Cady B. Biologic and
clinical significance of lymphadenectomy. Surg Clin North Am 2000;80:1631-73. 2. Gervasoni JE Jr, Sbayi S, Cady B. Role of lymphadenectomy in surgical treatment of solid tumors: an update on the clinical data. Ann Surg Oncol 2007;14:2443-62. 3. Cady B. Regional lymph node metastases; a singular manifestation of the process of clinical metastases in cancer: contemporary animal research and clinical reports suggest unifying concepts. Ann Surg Oncol 2007;14:1790-800. DOI: 10.1056/NEJMc1402989
and editors to fairly referee scientific disagreements” (Abrams J: personal communication). This process was vigorously undertaken by the reviewers and editors of the Journal. Reviewers thought the addition of distant disease–free survival for the entire population (which insignificantly favored the biopsy group [hazard ratio for distant metastasis, 0.89; 95% confidence interval, 0.70 to 1.13; P = 0.34]) did not warrant displacement of the presented results. Thomas’s statement that the curves showing the cumulative incidence of nodal metastasis are parallel between 8 and 10 years of follow-up is inaccurate. More than one fourth of the remaining minimal numerical difference between the curves evaporates during those 2 years. The criticism is, in any case, irrelevant since the rates are not statistically different, even with existing follow-up. His repetition of the illogical concept of “prognostic false positivity” therefore remains entirely without merit. Cady comments that a survival benefit in one subgroup may offset a disadvantage in another. In MSLT-I, a separate analysis involving a prospectively defined subgroup of patients with node-positive disease was justified by a strong biologic rationale. Survival among patients in this subgroup who underwent sentinel-node biopsy was substantially improved. Examination of the potentially offsetting group of node-negative patients (Fig. 3C of our article) reveals no suggestion of disadvantage in node-negative patients to counteract this benefit. It is likely that the considerable benefit to node-positive patients is simply diluted to statistical nonsignificance by the preponderance of node-negative patients, who derived prognostic, but not therapeutic, value. Regarding earlier literature reviews, we draw a different conclusion. Although trials of elective node dissection did not show a statistically significant benefit from early nodal intervention, they consistently showed an advantage to early dissection that approached significance. Lack of statistical significance in underpowered trials should not be considered a demonstration of equality. John F. Thompson, M.D.
The Authors Reply: Thomas suggests that prevention of melanoma recurrence is neither desirable in nor beneficial to patients. This curious position is inconsistent with a trial for which he was principal investigator, which used loco regional recurrence as the primary end point.1 He equates outcomes from early treatment of nodal metastases with delayed treatment, which MSLT-I clearly disproves. Early treatment, guided by sentinel-node biopsy, halves the extent of nodal involvement at surgery2 and decreases the rate of surgical complications.3 Early treatment also results in a clinically and statistically significant decreased risk of death from melanoma among patients with nodal metastases from intermediate-thickness melanomas. His appeal to the NCI Melanoma Institute Australia actually evoked a clear statement in 2007 from Sydney, NSW, Australia the chief of the Clinical Investigations Branch Alistair J. Cochran, M.D. that the NCI “does not interfere with the peer University of California at Los Angeles review process and depends on journal reviewers Los Angeles, CA
n engl j med 370;22 nejm.org may 29, 2014
The New England Journal of Medicine Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
2149
The
n e w e ng l a n d j o u r na l
of
m e dic i n e
2. Morton DL, Thompson JF, Cochran AJ, et al. Sentinel-node
Mark B. Faries, M.D. John Wayne Cancer Institute at Saint John’s Health Center Santa Monica, CA [email protected] Since publication of their article, the authors report no further potential conflict of interest. 1. Thomas JM, Newton-Bishop J, A’Hern R, et al. Excision mar-
gins in high-risk malignant melanoma. N Engl J Med 2004;350: 757-66.
biopsy or nodal observation in melanoma. N Engl J Med 2006; 355:1307-17. [Erratum, N Engl J Med 2006;355:1944.] 3. Faries MB, Thompson JF, Cochran A, et al. The impact on morbidity and length of stay of early versus delayed complete lymphadenectomy in melanoma: results of the Multicenter Selective Lymphadenectomy Trial (I). Ann Surg Oncol 2010;17: 3324-9. DOI: 10.1056/NEJMc1402989
Management of Early Prostate Cancer To the Editor: In reporting on 23.2 years of follow-up in the Scandinavian Prostate Cancer Group Study Number 4 (SPCG-4), Bill-Axelson et al. (March 6 issue)1 state that radical prostatectomy was superior to watchful waiting with respect to overall and cancer-specific mortality. The conclusions based on this randomized, controlled trial are completely different from those based on the American Prostate Cancer Intervention versus Observation Trial (PIVOT),2 which showed no difference between surgery and watchful waiting. The study designs were very similar. Both were randomized, controlled trials involving roughly 700 patients of a similar mean age, and the inclusion criteria were the same. The mean prostate-specific antigen (PSA) level was 10.1 ng per milliliter in PIVOT and 13 ng per milli liter in SPCG-4. PIVOT, as compared with SPCG-4, involved more patients with stage T1c disease (50% vs. 12%), and overall mortality in the watchful-waiting group was similar in the two trials (49.9% vs. 44.8%). Nevertheless, at approximately the same follow-up (10.0 years in PIVOT and 10.8 years in SPCG-43), cancer-specific mortality was less than half as high in the watchful-waiting group (8.4% in PIVOT vs. 19.5% in SPCG-4). It should be clarified whether adherence to the randomized treatment assignment (which was not negligible in PIVOT) or, much more likely, a considerable attribution bias can explain such significant differences in cancerspecific mortality. Bernardo Rocco, M.D. Andrea Conti, M.D. Elisa De Lorenzis, M.D. Ospedale Maggiore Policlinico di Milano Milan, Italy [email protected]
2150
No potential conflict of interest relevant to this letter was reported. 1. Bill-Axelson A, Holmberg L, Garmo H, et al. Radical prosta-
tectomy or watchful waiting in early prostate cancer. N Engl J Med 2014;370:932-42. 2. Wilt TJ, Brawer MK, Jones KM, et al. Radical prostatectomy versus observation for localized prostate cancer. N Engl J Med 2012;367:203-13. [Erratum, N Engl J Med 2012;367:582.] 3. Bill-Axelson A, Holmberg L, Filén F, et al. Radical prostatectomy versus watchful waiting in localized prostate cancer: the Scandinavian Prostate Cancer Group-4 randomized trial. J Natl Cancer Inst 2008;100:1144-54. DOI: 10.1056/NEJMc1403862
To the Editor: Bill-Axelson et al. report longterm follow-up results showing that the benefit of prostatectomy was greater than that of watchful waiting with respect to death from prostate cancer. “Watchful waiting” in this study was defined as doing nothing for cancer until the development of overt metastasis or the occurrence of any symptom related to cancer. However, in 2003, Klotz1 proposed “active surveillance,”2,3 and that concept has been widely adopted recently. As compared with watchful waiting, active surveillance consists of more aggressive policies such as initial selection of low-risk patients according to a specific criterion, strict and systematic monitoring, and early transfer to aggressive intervention even in patients who do not have symptoms.4 Therefore, in the current state of the art, high-risk patients should not be randomly assigned to a passive watchful-waiting group. Clinicians who want to reduce the rate of death from prostate cancer and avoid complications of surgery such as urinary incontinence and erectile dysfunction must be aware that the results of this study cannot be simply applied to daily clinical practice, since the treatment strategy
n engl j med 370;22 nejm.org may 29, 2014
The New England Journal of Medicine Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
n e w e ng l a n d j o u r na l
of
m e dic i n e
c or r e sp ondence
Sentinel-Node Biopsy in Melanoma To the Editor: In reporting the results of both the third interim analysis of the Multicenter Selective Lymphadenectomy Trial (MSLT-I) in 2006 and the current final analysis, Morton et al. (Feb. 13 issue)1 describe a disease-free survival advantage. This advantage is an inevitable consequence of the trial design and does not reflect any therapeutic advantage for patients. In both analyses, patients with positive sentinel nodes in the biopsy group underwent prophylactic lymphadenectomy. Therefore, inevitably, the site of first recurrence would be a nodal site more often in the observation group than in the biopsy group. This error was appealed and upheld in 2007 by the Clinical Investigations Branch of the National Cancer Institute (NCI) of the National Institutes of Health, which funded the trial.2 The authors were requested to provide a calculation of distant disease–free survival among all patients with intermediate-thickness tumors in the final update. This calculation is provided only for node-positive patients. After Morton and colleagues reported the results of the third interim analysis in 2006, I postulated that a proportion of positive sentinel nodes were prognostically false positive.3,4 The authors claimed to disprove this hypothesis in a this week’s letters 2148 Sentinel-Node Biopsy in Melanoma 2150 Management of Early Prostate Cancer 2152 Bleeding and Coagulopathies in Critical Care 2153 PICC Placement in the Neonate 2154 Pulmonary Fibrosis Associated with Aluminum Trihydrate (Corian) Dust
2148
graph showing that the cumulative incidence of nodal recurrences converged at 8 years.5 In the current article, the final results (Fig. 3A) are contradictory and confirm that the curves are parallel from 8 to 10 years; this reflects the presence and incidence of prognostic false positivity. J. Meirion Thomas, F.R.C.P., F.R.C.S. Royal Marsden Hospital London, United Kingdom [email protected] No potential conflict of interest relevant to this letter was reported. 1. Morton DL, Thompson JF, Cochran AJ, et al. Final trial re-
port of sentinel-node biopsy versus nodal observation in melanoma. N Engl J Med 2014;370:599-609. 2. Thomas JM. Concerns relating to the conduct and statistical analysis of the Multicenter Selective Lymphadenectomy Trial (MSLT-1) in patients with melanoma. J Plast Reconstr Aesthet Surg 2009;62:442-6. 3. Idem. Sentinel-node biopsy in melanoma. N Engl J Med 2007; 356:418. 4. Idem. Prognostic false-positivity of the sentinel node in melanoma. Nat Clin Pract Oncol 2008;5:18-23. 5. Morton DL, Cochran AJ, Thompson JF. Sentinel-node biopsy in melanoma. N Engl J Med 2007;356:419-20. DOI: 10.1056/NEJMc1402989
To the Editor: In the comparison of sentinelnode biopsy with observation of regional lymph nodes in patients with melanoma, Morton et al. report no significant difference in the rate of disease-specific survival. The question is whether delayed removal of metastases, as compared with immediate removal, allowed unimpeded growth that caused dissemination of disease and decreased survival. When a subgroup is retrospectively found to have a survival advantage, an offsetting subgroup, perhaps unrecognized, must of necessity have poorer survival with equivalent diseasespecific survival. This is critical in discussing a negative trial because of misinterpretation of contrived subgroup results showing advantages
n engl j med 370;22 nejm.org may 29, 2014
The New England Journal of Medicine Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
correspondence
to immediate node dissection. Surgery that prevents later disease improves disease-free survival but is irrelevant to melanoma-specific survival. Literature reviews1,2 do not show improved disease-specific survival in the comparison of various forms of regional node management (including dissection and observation only) in lung, esophageal, stomach, colon, rectal, or breast cancers, or in melanoma. The biologic explanation for this failure comes from laboratory studies and randomized clinical trials showing that organ-specific metastatic cells have no ability to grow in other organs.3 This explains curative resections of lymph-node, liver, and lung metastases after prolonged disease. The curability of metastatic disease is pattern-dependent, not timedependent. Blake Cady, M.D. Harvard Medical School Boston, MA [email protected] No potential conflict of interest relevant to this letter was reported. 1. Gervasoni JE Jr, Taneja C, Chung MA, Cady B. Biologic and
clinical significance of lymphadenectomy. Surg Clin North Am 2000;80:1631-73. 2. Gervasoni JE Jr, Sbayi S, Cady B. Role of lymphadenectomy in surgical treatment of solid tumors: an update on the clinical data. Ann Surg Oncol 2007;14:2443-62. 3. Cady B. Regional lymph node metastases; a singular manifestation of the process of clinical metastases in cancer: contemporary animal research and clinical reports suggest unifying concepts. Ann Surg Oncol 2007;14:1790-800. DOI: 10.1056/NEJMc1402989
and editors to fairly referee scientific disagreements” (Abrams J: personal communication). This process was vigorously undertaken by the reviewers and editors of the Journal. Reviewers thought the addition of distant disease–free survival for the entire population (which insignificantly favored the biopsy group [hazard ratio for distant metastasis, 0.89; 95% confidence interval, 0.70 to 1.13; P = 0.34]) did not warrant displacement of the presented results. Thomas’s statement that the curves showing the cumulative incidence of nodal metastasis are parallel between 8 and 10 years of follow-up is inaccurate. More than one fourth of the remaining minimal numerical difference between the curves evaporates during those 2 years. The criticism is, in any case, irrelevant since the rates are not statistically different, even with existing follow-up. His repetition of the illogical concept of “prognostic false positivity” therefore remains entirely without merit. Cady comments that a survival benefit in one subgroup may offset a disadvantage in another. In MSLT-I, a separate analysis involving a prospectively defined subgroup of patients with node-positive disease was justified by a strong biologic rationale. Survival among patients in this subgroup who underwent sentinel-node biopsy was substantially improved. Examination of the potentially offsetting group of node-negative patients (Fig. 3C of our article) reveals no suggestion of disadvantage in node-negative patients to counteract this benefit. It is likely that the considerable benefit to node-positive patients is simply diluted to statistical nonsignificance by the preponderance of node-negative patients, who derived prognostic, but not therapeutic, value. Regarding earlier literature reviews, we draw a different conclusion. Although trials of elective node dissection did not show a statistically significant benefit from early nodal intervention, they consistently showed an advantage to early dissection that approached significance. Lack of statistical significance in underpowered trials should not be considered a demonstration of equality. John F. Thompson, M.D.
The Authors Reply: Thomas suggests that prevention of melanoma recurrence is neither desirable in nor beneficial to patients. This curious position is inconsistent with a trial for which he was principal investigator, which used loco regional recurrence as the primary end point.1 He equates outcomes from early treatment of nodal metastases with delayed treatment, which MSLT-I clearly disproves. Early treatment, guided by sentinel-node biopsy, halves the extent of nodal involvement at surgery2 and decreases the rate of surgical complications.3 Early treatment also results in a clinically and statistically significant decreased risk of death from melanoma among patients with nodal metastases from intermediate-thickness melanomas. His appeal to the NCI Melanoma Institute Australia actually evoked a clear statement in 2007 from Sydney, NSW, Australia the chief of the Clinical Investigations Branch Alistair J. Cochran, M.D. that the NCI “does not interfere with the peer University of California at Los Angeles review process and depends on journal reviewers Los Angeles, CA
n engl j med 370;22 nejm.org may 29, 2014
The New England Journal of Medicine Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
2149
The
n e w e ng l a n d j o u r na l
of
m e dic i n e
2. Morton DL, Thompson JF, Cochran AJ, et al. Sentinel-node
Mark B. Faries, M.D. John Wayne Cancer Institute at Saint John’s Health Center Santa Monica, CA [email protected] Since publication of their article, the authors report no further potential conflict of interest. 1. Thomas JM, Newton-Bishop J, A’Hern R, et al. Excision mar-
gins in high-risk malignant melanoma. N Engl J Med 2004;350: 757-66.
biopsy or nodal observation in melanoma. N Engl J Med 2006; 355:1307-17. [Erratum, N Engl J Med 2006;355:1944.] 3. Faries MB, Thompson JF, Cochran A, et al. The impact on morbidity and length of stay of early versus delayed complete lymphadenectomy in melanoma: results of the Multicenter Selective Lymphadenectomy Trial (I). Ann Surg Oncol 2010;17: 3324-9. DOI: 10.1056/NEJMc1402989
Management of Early Prostate Cancer To the Editor: In reporting on 23.2 years of follow-up in the Scandinavian Prostate Cancer Group Study Number 4 (SPCG-4), Bill-Axelson et al. (March 6 issue)1 state that radical prostatectomy was superior to watchful waiting with respect to overall and cancer-specific mortality. The conclusions based on this randomized, controlled trial are completely different from those based on the American Prostate Cancer Intervention versus Observation Trial (PIVOT),2 which showed no difference between surgery and watchful waiting. The study designs were very similar. Both were randomized, controlled trials involving roughly 700 patients of a similar mean age, and the inclusion criteria were the same. The mean prostate-specific antigen (PSA) level was 10.1 ng per milliliter in PIVOT and 13 ng per milli liter in SPCG-4. PIVOT, as compared with SPCG-4, involved more patients with stage T1c disease (50% vs. 12%), and overall mortality in the watchful-waiting group was similar in the two trials (49.9% vs. 44.8%). Nevertheless, at approximately the same follow-up (10.0 years in PIVOT and 10.8 years in SPCG-43), cancer-specific mortality was less than half as high in the watchful-waiting group (8.4% in PIVOT vs. 19.5% in SPCG-4). It should be clarified whether adherence to the randomized treatment assignment (which was not negligible in PIVOT) or, much more likely, a considerable attribution bias can explain such significant differences in cancerspecific mortality. Bernardo Rocco, M.D. Andrea Conti, M.D. Elisa De Lorenzis, M.D. Ospedale Maggiore Policlinico di Milano Milan, Italy [email protected]
2150
No potential conflict of interest relevant to this letter was reported. 1. Bill-Axelson A, Holmberg L, Garmo H, et al. Radical prosta-
tectomy or watchful waiting in early prostate cancer. N Engl J Med 2014;370:932-42. 2. Wilt TJ, Brawer MK, Jones KM, et al. Radical prostatectomy versus observation for localized prostate cancer. N Engl J Med 2012;367:203-13. [Erratum, N Engl J Med 2012;367:582.] 3. Bill-Axelson A, Holmberg L, Filén F, et al. Radical prostatectomy versus watchful waiting in localized prostate cancer: the Scandinavian Prostate Cancer Group-4 randomized trial. J Natl Cancer Inst 2008;100:1144-54. DOI: 10.1056/NEJMc1403862
To the Editor: Bill-Axelson et al. report longterm follow-up results showing that the benefit of prostatectomy was greater than that of watchful waiting with respect to death from prostate cancer. “Watchful waiting” in this study was defined as doing nothing for cancer until the development of overt metastasis or the occurrence of any symptom related to cancer. However, in 2003, Klotz1 proposed “active surveillance,”2,3 and that concept has been widely adopted recently. As compared with watchful waiting, active surveillance consists of more aggressive policies such as initial selection of low-risk patients according to a specific criterion, strict and systematic monitoring, and early transfer to aggressive intervention even in patients who do not have symptoms.4 Therefore, in the current state of the art, high-risk patients should not be randomly assigned to a passive watchful-waiting group. Clinicians who want to reduce the rate of death from prostate cancer and avoid complications of surgery such as urinary incontinence and erectile dysfunction must be aware that the results of this study cannot be simply applied to daily clinical practice, since the treatment strategy
n engl j med 370;22 nejm.org may 29, 2014
The New England Journal of Medicine Downloaded from nejm.org on September 26, 2015. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
