Portal de Periódicos da CAPES

Secondary Leukemia After Adjuvant Chemotherapy for Breast Cancer

2001; Lippincott Williams & Wilkins; Volume: 19; Issue: 4 Linguagem: Inglês

10.1200/jco.2001.19.4.1231

1527-7755

Richard Ghalie, Donald E. Goodkin, Claire Bonithon‐Kopp, C Milan, Gilles Chaplain, C Sgro, Paule‐Marie Carli,

Chronic Lymphocytic Leukemia Research

Article Tools SPECIAL DEPARTMENTS Article Tools OPTIONS & TOOLS Export Citation Track Citation Add To Favorites Rights & Permissions COMPANION ARTICLES No companion articles ARTICLE CITATION DOI: 10.1200/JCO.2001.19.4.1231 Journal of Clinical Oncology - published online before print September 21, 2016 PMID: 11181691 Secondary Leukemia After Adjuvant Chemotherapy for Breast Cancer Richard G. GhaliexRichard G. GhalieSearch for articles by this author , Donald E. GoodkinxDonald E. GoodkinSearch for articles by this author C. Bonithon-KoppxC. Bonithon-KoppSearch for articles by this author , C. MilanxC. MilanSearch for articles by this author , G. ChaplainxG. ChaplainSearch for articles by this author , C. SgroxC. SgroSearch for articles by this author , P.M. CarlixP.M. CarliSearch for articles by this author Show More Immunex Corp., Seattle, WACentre d’Epidémiologie de Population de l’Université de BourgogneRegistre de Cancers Gynécologiques de Côte d’OrCentre Régional de Pharmacovigilance Centre Hospitalier et UniversitaireRegistre des Hémopathies Malignes de Côte d’Or, Dijon, France https://doi.org/10.1200/JCO.2001.19.4.1231 First Page Full Text PDF Figures and Tables © 2001 by American Society of Clinical OncologyjcoJ Clin OncolJournal of Clinical OncologyJCO0732-183X1527-7755American Society of Clinical OncologyResponse15022001In Reply:We thank Drs Ghalie and Goodkin for their interest in our study.1 It is reassuring to know that biopharmaceutical companies feel concerned about possible long-term adverse effects of their products. Studies like ours are justified by the scarcity of published controlled data on the leukemogenic potential of cytotoxic drugs which are widely and increasingly used in the treatment of breast cancer. Because of the high number of women diagnosed with breast cancer worldwide, drug monitoring is of crucial importance for public health. The limitations of our observational study based on cancer registries are clear. However, in our opinion, they do not cast doubt on our major finding, that is, an increased risk of acute leukemia in patients who received a combination of radiotherapy and chemotherapy with mitoxantrone.We agree with some issues raised by Ghalie and Goodkin. Because of the small number of leukemia cases, the statistical power of our study was low and risk estimates were quite imprecise. Furthermore, in within-cohort analysis, we were unable to take into account some potential confounders, such as the intensity of radiation therapy. However, in order to explain the association we found between leukemia and the type of chemotherapy (relative risk = 0.095), the intensity of the association between the confounding factor and leukemia would have had to be of the same order. This seems quite improbable.Other comments are based on an erroneous presentation of our study and seem intended to put the reader off track. Our study was based on two cancer registries, not on “various local databases.” The high quality of their data is regularly recognized by the L’Institut National de la Santé et de la Recherche Médicale (INSERM) and the National Committee of Registries. Because of their specificity, both registries serve as references in France and Europe in the field of hematologic malignancies and gynecologic cancers. Although our study cannot strictly be considered as a cohort study, its design makes it closer to a cohort study than to a retrospective study. The patients’ exposure to radiotherapy and chemotherapy was indeed regularly and actively checked and was systematically recorded before the occurrence of acute leukemia (generally in the few months after the diagnosis of breast cancer). Consequently, differential misclassification of exposure seems quite unlikely. Ghalie and Goodkin suggest that a higher proportion of patients with a leukemia diagnosis could have been treated outside the Côte d’Or area from 1982 to 1989 than from 1990 to 1996. We presume this would imply that some cases of acute leukemia may not have been registered in the first period. We have no reason to think this was the case. High-level diagnostic and therapeutic resources have existed in the Burgundy region since 1967 and are concentrated in the Côte d’Or area. Thus, the probability of a woman residing in the Côte d’Or area being treated outside is much lower than the probability of a woman residing in the surrounding area being treated in the Côte d’Or area. Furthermore, because of the demographic and sociologic characteristics of our population (stable and rural population), access to medical resources from other regions (especially from Paris) was more likely from 1990 to 1996 than from 1982 to 1989, while remaining very low. Our analysis by period aimed to show the large changes in breast cancer care during the study period. As noted by Ghalie and Goodkin, there was an increasing use of mitoxantrone from 1982 to 1989 and 1990 to 1996. Mitoxantrone did indeed obtain French marketing authorization for the treatment of breast cancer (particularly advanced metastatic breast cancer) in 1985. Among the entire study population (all patients with resected nonmetastatic breast cancer), its use increased from 2% in 1982 to 1989 to 26% after 1990. Among patients with adjuvant chemotherapy, the proportion of women who received mitoxantrone increased from 8% to 66%. Thus, our data suggest a large drift from official recommendations in the use of mitoxantrone since 1990. During the study period, there were relatively few changes in the use of radiotherapy alone (from 60.3% to 53.7%). On the other hand, there was an increasing use of a combination of radiotherapy and chemotherapy, but not in the proportions indicated by Ghalie and Goodkin: the percentage of patients who received radiotherapy and chemotherapy as adjuvant treatment increased from 23.5% in 1982 to 1989 to 38% in 1990 to 1996. Very few women received chemotherapy without radiotherapy (1.9% from 1982 to 1989 v 1.3% since 1990). We had thus clearly stated in our article that it was impossible to determine whether the excess risk of leukemia was due to chemotherapy per se or to the interaction between radiotherapy and chemotherapy, especially chemotherapy including mitoxantrone. Very large cohort studies would be necessary to detect such interaction. As noted by Ghalie and Goodkin, there have been no published reports of acute leukemia secondary to mitoxantrone monotherapy. However, we know at least one case of acute promyelocytic leukemia reported in a patient who received a cumulative mitoxantrone dose of 50 mg/m2 as single-agent treatment of multiple sclerosis.2We agree that the results of the Cox regression should be taken with caution because of the small number of leukemia cases (10 cases, not eight as indicated by Ghalie and Goodkin). In such conditions, the estimate of the risk ratio of leukemia is somewhat unstable and imprecise (relative risk for anthracycline v mitoxantrone use at cumulative dose ≥ 13 mg/m2 = 0.095; 95% confidence interval, 0.010 to 0.90). However, it is unlikely that the estimate of the risk ratio was severely biased by the inclusion of confounding factors in the regression model. In univariate analysis, the risk ratio was 0.115 (95% confidence interval, 0.014 to 0.95; P = .014), which is very close to the adjusted risk ratio. Furthermore, it is noteworthy that comparison of the standardized incidence ratios associated with mitoxantrone and anthracycline use gives a crude estimation of the risk ratio (relative risk = 0.098) that is also very close to that found in the Cox regression. The fact that two methods with different underlying assumptions provide similar estimates emphasizes the consistency of our results.Both statistical significance and confidence intervals must be taken into account in the interpretation of results. For example, the lack of statistical significance for the standardized incidence ratio (SIR) associated with anthracycline use (SIR = 11.4; 95% CI, 0.3 to 63.7; P = .18) did not lead us to conclude that these drugs were safe. The large confidence interval did not allow a leukemogenic effect of the association of radiotherapy plus anthracyclines to be excluded. Contrary to Ghalie and Goodkin’s statement, our confidence intervals were computed by using tabulated values of exact 95% confidence limits for estimating a Poisson-distributed variable.3 Using a nonexplicated method, Ghalie and Goodkin found confidence intervals of 8.0 to 2,869 instead of 25.9 to 133 for the mitoxantrone incidence ratio (SIR = 64.7) and questioned the clinical relevance of this result. We will let the readers be the judge of the clinical relevance of a leukemia risk multiplied at least by eight relative to the general population.The leukemogenic potential of mitoxantrone is not in doubt and has been confirmed in a recent report.4 However, we agree with Ghalie and Goodkin that larger studies are needed to quantify more precisely the leukemia risk associated with mitoxantrone and other topoisomerase II inhibitors.1. Chaplain G, Milan C, Sgro C, et al: Increased risk of acute leukemia after adjuvant chemotherapy for breast cancer: A population-based study. J Clin Oncol 18:: 2836,2000-2842, Link, Google Scholar2. Vicari AM, Ciceri F, Folli F, et al: Acute promyelocytic leukemia following mitoxantrone as single agent for the treatment of multiple sclerosis. Leukemia 12:: 441,1998-442, Crossref, Medline, Google Scholar3. Breslow NE, Day NE: Rates and rate standardization, in Statistical Methods in Cancer Research, Vol II: The Design and Analysis of Cohort Studies. Lyon, France, International Agency for Research on Cancer, 1987, IARC Scientific Publication No. 82 Google Scholar4. Saso R, Kulkarni S, Mitchell P, et al: Secondary myelodysplastic syndrome/acute myeloid leukaemia following mitoxantrone-based therapy for breast carcinoma. Br J Cancer 83:: 91,2000-94, Crossref, Medline, Google Scholar