|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Acute Myeloid Leukemia and Myelodysplastic Syndrome After Doxorubicin-Cyclophosphamide Adjuvant Therapy for Operable Breast Cancer: The National Surgical Adjuvant Breast and Bowel Project Experience
From the National Surgical Adjuvant Breast and Bowel Project Operations Center and Biostatistical Center, Pittsburgh, PA; and Bristol-Myers Squibb, Wallingford, CT. Address reprint requests to Roy Smith, MD, Medical Affairs and Medical Oversight, National Surgical Adjuvant Breast and Bowel Project, East Commons Professional Building, 4 Allegheny Center, 5th Floor, Pittsburgh, PA 15212-5234.
Purpose: We reviewed data from all adjuvant NSABP breast cancer trials that tested regimens containing both doxorubicin (A) and cyclophosphamide (C) to characterize the incidence of subsequent acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). Materials and Methods: Six complete NSABP trials have investigated AC regimens (B-15, B-16, B-18, B-22, B-23, and B-25). Six distinct AC regimens have been tested and are distinguished by differences in cyclophosphamide intensity and cumulative dose and by the presence or absence of mandated prophylactic support with growth factor and ciprofloxacin. In all regimens, A was given at 60 mg/m2 q 21 days x 4. C was given as follows: 600 mg/m2 q 21 days x 4 ("standard AC"); 1,200 mg2 q 21 days x 2; 1,200 mg/m2 q 21 days x 4; 2,400 mg/m2 q 21 days x 2; and 2,400 mg/m2 q 21 days x 4. Occurrence of AML/MDS was summarized by incidence per 1,000 patient-years at risk and by cumulative incidence. Rates were compared across regimens, by age, and by treatment with or without breast radiotherapy. Results: The incidence of AML/MDS was sharply elevated in the more intense regimens. In patients receiving two or four cycles of C at 2,400 mg/m2 with granulocyte colony-stimulating factor (G-CSF) support, cumulative incidence of AML/MDS at 5 years was 1.01% (95% confidence interval [CI], 0.63% to 1.62%), compared with 0.21% (95% CI, 0.11% to 0.41%) for patients treated with standard AC. Patients who received breast radiotherapy experienced more secondary AML/MDS than those who did not (RR = 2.38, P= .006), and the data indicated that G-CSF does may possibly also be independently correlated with increased risk. Conclusion: AC regimens employing intensified doses of cyclophosphamide requiring G-CSF support were characterized by increased rates of subsequent AML/MDS, although the incidence of AML/MDS was small relative to that of breast cancer relapse. Breast radiotherapy appeared to be associated with an increased risk of AML/MDS.
IN 1985, the National Surgical Adjuvant Breast and Bowel Project (NSABP) reported its experience with treatment-related leukemia in patients with operable breast cancer after adjuvant chemotherapy and postoperative radiation with regimens containing L-phenylalanine mustard.1 In recent years, adjuvant chemotherapy regimens for operable breast cancer have included alkylating agents other than L-phenylalanine mustard with or without anthracyclines. Other standard therapies include breast radiotherapy after lumpectomy and tamoxifen for patients with hormone-responsive breast cancer.2 Recently, there have been reports of the benefit of regional radiation therapy for axillary node–positive patients3,4 and of the potential benefit of paclitaxel administration after a doxorubicin-cyclophosphamide (AC) regimen.5 In the treatment of operable breast cancer, as in any therapeutic domain, the potential risks and benefits of therapy must be considered for a particular patient. One of the most serious risks associated with cancer therapy is the development of treatment-related acute leukemia and myelodysplastic syndrome (MDS). Leukemia has been reported after therapy with alkylating agents, topoisomerase II inhibitors, and radiation therapy for the treatment of a variety of tumor types.6–10 Recent NSABP studies have often used AC-containing regimens. Two of these studies (NSABP Protocols B-22 and B-25) evaluated dose-intensified cyclophosphamide.11,12 We have previously reported a greater than expected incidence of acute myeloid leukemia (AML) and MDS among patients treated on NSABP Protocol B-25.13 This current report presents the experience with AML and MDS across NSABP clinical studies using AC regimens. The primary issues addressed in this report are the rates of AML and/or MDS as a function of the AC regimen (standard v intensified and increased total-dose cyclophosphamide) and the use of radiation therapy. A discussion of the potential role of growth factors is also included.
The studies examined in this report were approved by local institutional review boards at the study sites and in accord with an assurance filed with and approved by the United States Department of Health and Human Services. All patients in the studies examined provided informed consent.
Patients
The studies and the relevant treatment arms are listed in Table 1  . B-15 included women with histologically positive axillary lymph nodes who were 49 years of age or younger or 50 to 59 years of age with progesterone receptor (PR)–negative tumors (< 10 fmol/mg of cytosol protein). Patients randomly assigned to arm I were treated with AC administered intravenously at doses of 60 and 600 mg/m2, respectively, every 21 days for four cycles ("standard AC"). B-16 was a study for women with histologically positive axillary lymph nodes who were 50 to 59 years of age and had PR-positive tumors or were 60 to 70 years of age regardless of tumor receptor status. Patients randomly assigned to B-16 arm II received standard AC, and tamoxifen was given (10 mg orally twice a day) for at least 5 years. B-18 was a two-arm study comparing postoperative standard AC (arm I) with preoperative standard AC (arm II). Patients who were 50 years of age or older at the time of randomization received tamoxifen for at least 5 years. B-22 was a three-arm study for women with histologically positive axillary lymph nodes. All patients received doxorubicin 60 mg/m2 intravenously every 21 days for four cycles. In arm I, patients received concurrent cyclophosphamide 600 mg/m2 for four cycles (standard AC). In arm II, patients received cyclophosphamide 1,200 mg/m2 for the first two cycles only, and in arm III, patients received cyclophosphamide 1,200 mg/m2 for four cycles. All patients 50 years of age or older received tamoxifen for at least 5 years. B-23 was a study for women with axillary node–negative breast cancer and estrogen receptor (ER)–negative tumors. Patients randomly assigned to arms III or IV received standard AC and 5 years of either tamoxifen or placebo. In Protocol B-25, a study for women with histologically positive axillary lymph nodes, all patients received doxorubicin 60 mg/m2 intravenously every 21 days for four cycles. In arm I, patients received concurrent cyclophosphamide 1,200 mg/m2 for four cycles. In arm II, patients received cyclophosphamide 2,400 mg/m2 for two cycles, and in arm III, patients received cyclophosphamide 2,400 mg/m2 for four cycles. All patients received prophylactic granulocyte colony-stimulating factor (G-CSF), 5 µg/kg starting day 2 of each cycle. G-CSF was continued until the granulocyte count was greater than 10,000/µL after day 7. The dose of G-CSF was doubled in all cycles after the occurrence of severe infection or febrile neutropenia (in this protocol, febrile neutropenia was defined as the occurrence of fever > 38.5°C or systemic infection in the presence of neutropenia requiring hospitalization). If febrile neutropenia recurred, patients also received prophylactic ciprofloxacin 500 mg orally bid. Tamoxifen was administered for 5 years to all patients 50 years of age or older.
AC therapy in these six clinical trials can be grouped into six distinct regimens (Table 2
Data Management Patients on the NSABP clinical trials listed in Table 1 are followed from the time of randomization until death. Submission of treatment data was requested on a per-cycle basis. Submission of follow-up data is required on a protocol-specific schedule: at a minimum, submissions were required every 3 months for the first year, every 6 months through year 5, and annually thereafter. In addition to treatment failures and deaths, NSABP participating institutions report all second primary cancers to the NSABP Biostatistical Center. Beginning in February 1996, reports of AML and MDS are also submitted to the National Cancer Institute using the Secondary AML/MDS Report Form. All reported blood dyscrasias undergo a centralized medical review of clinical records submitted by the participating institution (including reports of bone marrow examinations, pathology reports, laboratory values, and death records). The diagnoses of leukemia and MDS presented in this report are based on the investigators’ assessment in conjunction with headquarters medical review of the submitted pathology reports and other source documents. All analyses in this report are based on follow-up data received at the Biostatistical Center as of March 31, 2001.
Data Analysis and Statistical Methods
Incidence rates per 1,000 patient years (PYs) at risk were computed by regimen and by patient cohorts defined by age at surgery (
Comparisons of AML/MDS failure rates across regimens or other patient cohorts and calculation of relative risks (RRs) were based on stratified log-rank analyses. For example, comparison of the incidence of AML/MDS between lumpectomy patients (who received breast radiotherapy) and mastectomy patients (for whom radiotherapy was prohibited) was carried out by a log-rank analysis stratified by the six regimens of Table 2
A total of 8,563 patients with follow-up were randomized to AC regimens in Protocols B-15, B-16, B-18, B-22, B-23, and B-25. These patients have contributed a total of 61,810 PYs of follow-up before death or last follow-up. The distribution of follow-up across the six different AC regimens is shown in Table 3  . For each regimen, Table 3 lists the average patient age at surgery, the percentage of patients aged  49 years, the percentage of patients receiving lumpectomies, the number of reported cases of AML/MDS, and incidence of AML/MDS both in terms of rates per 1,000 PYs and 8-year percent cumulative incidence. The table also gives the standardized incidence rate, defined as the cohort incidence rate expressed as a multiple of the incidence rate in an age-matched Surveillance, Epidemiology, and End Results (SEER) Registry population.
Diagnoses of AML/MDS In total, 43 patients (0.50%) have presented with either AML or MDS (Appendix, available online at www.jco.org). An additional five patients have been diagnosed with other hematologic malignancies: one patient with acute lymphoblastic leukemia, one patient with chronic lymphocytic leukemia, two patients with chronic myelogenous leukemia, and one patient with essential thrombocytosis. Only diagnoses of AML or MDS are considered in the remainder of this report. Individual characteristics of the patients are summarized in the Appendix, including protocol and regimen, type of dyscrasia, cytogenetics, latency period, vital status at last follow-up, age at initial surgery, total doses of doxorubicin and cyclophosphamide, presence or absence of adjuvant radiotherapy, treatment with G-CSF, presence or absence of treatment failure or other second primary cancer before diagnosis of leukemia, and additional treatment before leukemia, if any. Of the 43 patients, 20 presented with AML, whereas 23 presented with MDS (eight of whom subsequently progressed to AML). Latency periods (time from initiation of therapy to diagnosis of AML/MDS) ranged from 1 to 125 months (median, 38 months). For 28 patients, the diagnosis of AML or MDS was the first protocol-defined event, whereas 15 patients had either a prior breast cancer recurrence or other cancer. Of these 15 patients, 13 received additional hormonal (four patients), cytotoxic (10 patients), and/or radiation (six patients) therapy before their AML or MDS. One patient was treated with surgery only, and a second patient received no further therapy. Six of the 43 patients were alive at last follow-up. AML and MDS are classified morphologically according to the widely accepted French-American-British (FAB) criteria. In this classification system, AML is categorized by the degree of differentiation along different cell lines and the extent of cell maturation.16–18 M1, M2, and M3 leukemia show predominantly granulocytic differentiation and differ from one another in the extent and nature of granulocytic maturation: M4 shows both granulocytic and monocytic differentiation, M5 shows predominantly monocytic differentiation, and M6 shows predominantly erythroblastic differentiation. M7 is associated with leukemic megakaryocytes. MDS is classified according to the degrees of disordered hematopoiesis, frequencies of transformation to acute leukemia, and prognosis.19–21 Patients with refractory anemia (RA) or refractory anemia with ringed sideroblasts (RARS) have a relatively good prognosis and show minimal dysplasia and macrocyctic changes in the blood and bone marrow. However, in patients with RARS, at least 15% of the RBC precursors are ringed sideroblasts. In patients with RA and RARS, progression to acute leukemia is infrequent. In contrast, refractory anemia with excessive blasts (RAEB) and refractory anemia with excessive blasts in transformation (RAEB-t) show a more disordered myelopoiesis and erythropoiesis than that seen in RA. In RAEB, the bone marrow shows 5% to 20% myeloblasts, and 1% to 5% myeloblasts may circulate in the blood. On the other hand, RAEB-t is characterized by 20% to 30% myeloblasts in the bone marrow, more than 5% myeloblasts in the blood, and frequently the presence of Auer rods. Chronic myelomonocytic leukemia (CMML) has normal-appearing erythrocytic precursors, 5% to 20% myeloblasts in the bone marrow, and an increased circulating monocyte count. Among patients with these subtypes of MDS, progression to leukemia is common (RAEB, 40%; RAEB-t, 75%; CMML, 30%) and their prognoses are poor.
French-American-British (FAB) subtypes were reported for 20 of the 28 patients who were either diagnosed with AML or who subsequently progressed to AML before death or last follow-up (Table 4
Unfortunately, the application of the widely accepted FAB classification to treatment-related MDS is often confounded because bone marrow samples from these patients often do not demonstrate the expected proportion of blasts or degree of dysmyelopoiesis.22,23 Nevertheless, when the criteria of the FAB classification are rigidly applied, RAEB and RAEB-t have been found to be the most common types of treatment-related MDS.23,24 Both RAEB and RAEB-t have a substantial risk of progressing to AML.25 Cytogenetic studies were available for 27 of the patients diagnosed with MDS/AML (Appendix, available online at www.jco.org). At the standard doses of doxorubicin and cyclophosphamide, three patients had simple cytogenetic abnormalities (deletions, duplications, or inversions of one or two chromosomes) involving chromosomes 7, 8, and the 11q23 locus. Among the 24 patients registered on the dose-intensified and dose-increased arms of protocols B-22 and B-25 for whom cytogenetic analysis was performed, 15 had more complex cytogenetic abnormalities involving multiple chromosomes (including 7, 8, 9, 11, 15, 23, and 24). Four of these patients also had chromosomal abnormalities involving the 11q23 locus, and two patients had a t(8;21) rearrangement.
Comparison of AML/MDS Incidence Rates Across Regimens
Association With Age and Adjuvant Radiotherapy Patients who underwent mastectomy were not to receive subsequent adjuvant radiation therapy, whereas breast irradiation was mandated in those who were treated with lumpectomy. Therefore, rates of AML/MDS were compared by method of surgery. Table 5 shows incidence of AML by regimen and method of operation. The crude incidence of AML/MDS among lumpectomy patients (1.00/1,000 PYs) was more than twice as large as that for mastectomy patients (0.49/1,000 PYs), and the trend is consistent across regimens. After stratification for regimen (A, B, C, D, E, F) and age at surgery ( 49 years,  50 years), the estimated RR for lumpectomy patients relative to mastectomy patients was 2.38 (95% CI, 1.29 to 4.40; P = .006). Similarly, age at surgery was associated with incidence of AML/MDS, with greater incidence in women aged 50 years at surgery than among younger women. After stratification for regimen and method of operation, the estimated RR was 2.37 (95% CI, 1.26 to 4.44; P = .008).
Because the administration of tamoxifen was based on age in most protocols (Table 1 ), there was concern that the association of age with incidence of AML/MDS that was noted in the previous paragraph could be confounded with tamoxifen treatment. To test this, a stratified proportional hazards model was used to relate the hazard of AML/MDS to tamoxifen usage (no, yes) and age at surgery (linear). After stratification for regimen and method of operation, there was no evidence that increased incidence of AML/MDS was related to treatment with tamoxifen (RR = 1.23; 95% CI, 0.47 to 3.25; P = .67). The magnitude of the estimated age effect was clinically significant (4% increase in risk per year of age; 95% CI, -0.4% to 9%) but was only marginally statistically significant (P = .08) because of the age-tamoxifen confound. When tamoxifen usage was dropped from the model, the estimated effect of age did not change much (5% increase in risk per year of age; 95% CI, 2% to 8%) but became strongly statistically significant (P = .001).
Treatment Failures or Second Primary Cancers Before Diagnosis of AML/MDS
Use of G-CSF in Protocol B-25 In theory, the effect of mandated G-CSF can be tested by comparing regimens C (B-22 arm III) and D (B-25 arm I), but the number of events is too small to be of use (10 in total). In addition, three of the six patients in regimen C actually did receive G-CSF, after recurrence but before the diagnosis of AML/MDS, so the comparison is unclear. However, some indirect evidence correlating use of G-CSF to subsequent presentation of AML/MDS can be obtained from Protocol B-25, where use of growth factor was mandated and daily doses of G-CSF were recorded.
Although use of G-CSF was mandated in all B-25 patients, there was nevertheless considerable variation in total G-CSF doses across patients. This resulted from differences in the number of days required to achieve granulocyte counts in excess of 10,000/µL and also from dose increases dictated by the protocol after incidents of febrile neutropenia or severe infection. A log-rank analysis was carried out in the cohort of B-25 patients to determine whether the total G-CSF dose (per kilogram, dichotomized as either Eighteen of the 22 B-25 patients diagnosed with AML/MDS received G-CSF doses in excess of the median dose administered to all B-25 patients (242 µg/kg). Controlling for treatment arm, patient age, and operation, the estimated risk of AML/MDS for patients receiving more than the median dose of G-CSF, relative to those receiving the median dose or less, was 3.58 (95% CI, 1.18 to 10.90; P = .02). Six of the 18 patients receiving more than 242 µg/kg of G-CSF experienced first events before diagnosis of AML/MDS, whereas none of the four patients receiving the median dose or less had prior events. Thus the result described above is sensitive to the inclusion or exclusion of AML/MDS after treatment failures or second primary cancers. If follow-up subsequent to first events is administratively censored, the estimated RR is reduced to 2.34 (95% CI, 0.72 to 7.55; P = .19). This computation still suggests a clinically significant RR, but the result is no longer statistically significant.
In this cross-protocol analysis, we present the incidence of AML/MDS from 8,563 patients (representing 61,810 PYs of follow-up) treated on six NSABP operable breast cancer studies using AC regimens. The chemotherapy regimens were defined according to their differences in cyclophosphamide intensity, cumulative cyclophosphamide dose, and the presence or absence of growth factor support. Forty-three patients (0.50%) developed AML/MDS. The incidence of AML/MDS associated with standard AC therapy was 0.32 cases per 1,000 PYs (95% CI, 0.16 to 0.57). The incidence of AML/MDS was sharply elevated in the more intense regimens compared with the incidence with standard AC; in patients receiving two or four cycles of cyclophosphamide at 2,400 mg/m2 with G-CSF support, the incidence rate was 1.75 cases per 1,000 PYs (95% CI, 1.04 to 2.77), corresponding to an RR of 6.16 (P < .0001). The corresponding cumulative incidence of AML/MDS at 5 years was 1.01% (95% CI, 0.63% to 1.62%) for patients receiving the more intense regimens, compared with 0.21% (95% CI, 0.11% to 0.41%) for patients treated with standard AC. These findings are consistent with previous reports indicating that exposure to certain chemotherapeutic agents is associated with genomic insult and an increased risk of AML/MDS.6,26 Fisher et al1 determined that the risk of AML was sharply elevated in breast cancer patients treated with surgery followed by melphalan-containing chemotherapy, relative to those treated with surgery alone. The cumulative risk for developing AML after receiving melphalan-containing regimens was estimated to be 1.29% at 10 years, as compared with 0.27% in patients treated with surgery alone. Several other studies indicate that the risk for developing AML/MDS among patients with early breast cancer who are treated with adjuvant chemotherapy containing standard-dose cyclophosphamide is somewhat higher than in the general population.27–31 Case-controlled studies evaluating the influence of the cumulative dose or the duration of alkylating agent therapy on patients with breast cancer or non-Hodgkin’s lymphoma indicate that in both circumstances, increased exposure heightens the risk for the development of AML/MDS.27,32 The risk of developing AML/MDS is also increased in patients receiving bone marrow–ablative chemotherapy regimens.33 Chemotherapy-induced AMLs have been reported to exhibit characteristics that differentiate them from primary leukemias.34,35 Alkylating agent–related AMLs often occur after an average interval of 5 to 7 years, are usually preceded by MDS, and are frequently described as being either M1 or M2 according to the FAB classification. Karyotypic analysis shows alterations of chromosomes 5 and 7 in 60% to 90% of cases. The most common cytogenetic findings are losses of chromosomes 5 or 7 and deletions at 7q.36 Conversely, AML/MDS associated with topoisomerase II inhibitor (doxorubicin, epipodophyllotoxin) treatment present after a much shorter interval from therapy (2 to 3 years) as either FAB M4 or M5 and are frequently associated with translocation of the long arm of chromosome 11 (11q23).36–42 RAEB and RAEB-t have been found to be the most common types of chemotherapy-induced MDS.23,24 In our data, the cytogenetic abnormalities observed were most consistent with a mixed leukemogenic effect from both cyclophosphamide and doxorubicin. This is evidenced by the presence of monosomy 5 or 7 or trisomy 8 and the presence of defects at the 11q23 locus. The presence of more complex genomic defects seemed to be more frequent among patients treated with the dose-intense regimens than in those treated with standard AC, implying the presence of a greater genomic insult. However, it is possible that this apparent difference may have resulted from the more extensive cytogenetic testing that was performed in the later protocols and, in particular, Protocol B-25. The presentation of our patients was mixed between those with either AML (nonmyelodysplastic phase) or MDS (myelodysplastic phase), and this is consistent with the presence of both a cyclophosphamide and doxorubicin genomic effect. Among patients receiving standard AC, only one of four patents had an FAB subtype that was arguably M5 (M2 v M5). In contrast, in those patients receiving the more intense regimens (B through F), the M4/M5 subtypes were frequently observed (nine of 14 subtyped cases). Putatively, this is consistent with a cyclophosphamide-induced promotion of a doxorubicin-associated leukemogenic effect.
Although an increased rate of AML/MDS is associated with the more intense regimens, interpretation is difficult because it is not clear which aspects of the dosing schedules are most associated with that increase. Differences in the six regimens are characterized in terms of increased dose-intensity of cyclophosphamide, increased total cyclophosphamide dose, and mandated use of growth factors, but because these three aspects are highly correlated, unambiguous attribution is not possible. In the data of Table 3 The data are more consistent with the hypothesis of an association between dose-intensity and incidence of AML/MDS, after controlling for cumulative dose and use of G-CSF. This follows by comparing regimen E to regimen D (compare two cycles at 2,400 mg/m2/cycle with four cycles at 1,200 mg/m2/cycle, both with G-CSF support) and regimen B to regimen A (compare two cycles at 1,200 mg/m2/cycle with four cycles at 600 mg/m2/cycle, without mandated G-CSF support). Here the estimated effect of doubling dose-intensity is large (RR = 2.21), but the result only approaches statistical significance (P = .07). Recently there has been speculation that growth factors may be leukemogenic and that the increased risk of AML/MDS associated with aggressive chemotherapy may, in part, result from growth factor use.43 Factor exposure either alone or in combination with intensified chemotherapy might increase the patient’s risk for developing AML/MDS. In Protocol B-25, a positive association was seen between total dose of G-CSF and the subsequent incidence of AML/MDS. However, even if such an association were reproducible, various interpretations are possible because the result is not based on a randomized comparison. The use of G-CSF was undoubtedly correlated with other factors that could not be adequately accounted for, including the treatment with ciprofloxacin or other antibiotics, the use of which was not documented in the research data. It is also possible that patients achieving an unusually high plasma level of doxorubicin and/or cyclophosphamide are at higher risk for AML/MDS and are simultaneously at higher risk for febrile neutropenia and severe infection. In this case, an association would be induced between use of G-CSF and subsequent incidence of AML/MDS that may have no causal basis. An additional caveat is based on the observation that six of the 18 patients receiving more than 242 mg/kg of G-CSF experienced first events before diagnosis of AML/MDS, whereas none of the four patients receiving the median dose or less had prior events. When follow-up subsequent to first events was administratively censored, the estimated RR was reduced to 2.34 (95% CI, 0.72 to 7.55; P = .19), which is a clinically significant but no longer statistically significant RR. In NSABP trials B-15, B-16, B-18, B-22, B-23, and B-25, breast irradiation was mandated for patients having a lumpectomy; patients treated with mastectomy were not to receive radiation. We therefore had the opportunity to investigate the effect of low-volume irradiation on the incidence of AML/MDS in patients receiving AC chemotherapy. We found that in patients who underwent breast radiotherapy, AML/MDS occurred at more than twice the rate than in patients who underwent no radiotherapy (RR = 2.38; P = .006 on the basis of all follow-up; RR = 2.26, P = .045 if follow-up was administratively censored after patients experienced a treatment failure or second primary cancer). Exposure to ionizing radiation has previously been associated with genomic insult and an increased risk of AML/MDS.6,26 Fisher et al1 reported an increased risk of AML in patients receiving regional radiotherapy in early NSABP trials using both melphalan-based chemotherapy and no systemic treatment. In a case-controlled study of a cohort of 82,700 women with breast cancer, Curtis et al27 detected a 2.4-fold increase in risk for AML among those women receiving regional radiotherapy relative to those receiving no radiotherapy. In an analysis of the SEER registry, Anderson and Bryant (unpublished work) similarly report an increased risk of secondary AML owing to radiotherapy in operable breast cancer. In a population-based study, Chaplain et al44 recently reported an increased risk of acute leukemia after adjuvant chemotherapy with radiotherapy. A cohort of 3,093 women with primary breast cancer was evaluated. The study was limited by the small number of patients with acute leukemia that occurred in this cohort (12 patients in total; 10 patients before breast cancer recurrence or second primary breast cancer). The risk of acute leukemia was significantly increased in patients treated with regimens that included mitoxantrone (standardized incidence ratio relative to the general population = 64.7; 95% CI, 25.9 to 133.2). The data suggested that commonly used regimens containing anthracyclines may be less leukemogenic than those containing mitoxantrone. Among women receiving regimens that contain an anthracycline, the standardized incidence ratio relative to the general population was 11.4 (95% CI, 0.3 to 63.7). To the best of our knowledge, this is the first report of elevated risk of AML/MDS in patients receiving breast (as opposed to regional) radiotherapy. As a caveat, Fisher et al1 found no incidence of increased risk in an analysis of NSABP Protocol B-06, in which patients were randomly assigned to receive mastectomy alone, lumpectomy alone, or lumpectomy followed by breast radiotherapy. Node-negative patients on Protocol B-06 received no systemic chemotherapy, whereas node-positive patients received melphalan. In addition, prolonged follow-up of B-06 patients subsequent to Fisher’s report has failed to suggest such an effect. As of March 31, 2001, eight cases of AML/MDS have been reported among B-06 patients randomized to receive mastectomy alone (two of which were preceded by contralateral events), three among those receiving lumpectomy alone and four among those randomized to receive lumpectomy followed by breast radiotherapy. Twelve of these patients received melphalan chemotherapy. The discrepancy between the B-06 results and those reported here could result from chance or, speculatively, could reflect a differential effect of the combination of doxorubicin and cyclophosphamide as compared with melphalan alone. We found a statistically significant increase in secondary AML/MDS among patients 50 years of age or older, relative to those 49 years of age or younger at surgery. An age-related increase in risk of secondary AML/MDS has previously been reported in patients treated with alkylating agents and radiotherapy for Hodgkin’s disease.45–47 Similar evidence in breast cancer patients is not as strong. Fisher et al1 noted a somewhat greater proportion of AMLs in breast cancer patients 50 years of age or older at surgery than among younger patients. Similarly, Anderson and Bryant (unpublished work) noted a significant increased risk of secondary AML in older breast cancer patients relative to younger patients on the basis of data obtained from the SEER registry.
In our data, the use of tamoxifen was confounded with patient age at surgery. In protocols B-18, B-22, and B-25, patients 50 years of age or older received tamoxifen, whereas those 49 years of age or younger did not. Similarly, entry criteria for protocol B-16 (in which patients received tamoxifen) and B-15 (in which they did not) in part were based on age (Table 1 In conclusion, our data clearly demonstrate a sharply increased risk of secondary AML/MDS in AC regimens in which the dose-intensity or cumulative dose of cyclophosphamide is increased beyond current standards. This risk must be taken into account in situations where such regimens are under consideration. It is important to emphasize that the results of NSABP Protocols B-22 and B-25 did not show any survival advantage for patients treated in this manner.11,12 It should also be kept in mind that although the risk of AML/MDS was elevated six-fold in arms II and III of Protocol B-25 relative to that for standard AC, the cumulative risk of AML/MDS (1.01% at 5 years) was still quite small in comparison to the relapse rate for node-positive breast cancer patients (approximately 35% at 5 years). Data from Protocol B-25 suggest that use of G-CSF may be associated with an increased risk of secondary AML/MDS, but for reasons discussed above, these findings should be considered exploratory. Analyses of other large databases are needed to provide confirmation or refutation of this hypothesis. Evidence correlating the use of regional radiotherapy with increased risk of secondary AML/MDS is now firmly established, and this fact should be taken into account if regional irradiation is contemplated. Our finding of an association owing to breast radiotherapy requires further confirmation.
Characteristics of Patients Who Developed Dyscrasias
This article is dedicated to Helen Louise Smith. We thank Barbara C. Good, PhD, for editorial assistance.
Supported by Public Health Service grant nos. U10-CA12027 (Treatment, Operations Center) and U10-CA69651 (Treatment, Biostatistical Center) from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Bethesda, MD.
1. Fisher B, Rockette H, Fisher ER, et al: Leukemia in breast cancer patients following adjuvant chemotherapy or postoperative radiation: The NSABP experience. J Clin Oncol 3:1640–1658, 1985 2. Fisher B, Redmond C: Systemic therapy in node-negative patients: Updated findings from NSABP clinical trials—National Surgical Adjuvant Breast and Bowel Project. J Natl Cancer Inst Monogr 11:105–116, 1992
3. Ragaz J, Jackson SM, Le N, et al: Adjuvant radiotherapy and chemotherapy in node-positive premenopausal women with breast cancer. N Engl J Med 337:956–962, 1997
4. Overgaard M, Hansen PS, Overgaard J, et al: Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive adjuvant chemotherapy: Danish Breast Cancer Cooperative Group 82b Trial. N Engl J Med 337:949–955, 1997 5. Henderson IC, Berry DA, Demetri GD, et al: Adjuvant Chemotherapy: Taxanes—The "Pro" Position. NIH Consensus Development Conference on Adjuvant Therapy for Breast Cancer, November 1–3, 2000 (monograph). Bethesda, MD, National Institutes of Health, p 79 6. Levine EG, Bloomfield CD: Leukemias and myelodysplastic syndromes secondary to drug, radiation, and environmental exposure. Semin Oncol 19:47–84, 1992[Medline]
7. Curtis Re, Boice JD Jr, Moloney WC, et al: Leukemia following chemotherapy for breast cancer. Cancer Res 50:2741–2746, 1990
8. Smith MA, Rubinstein L, Anderson JA, et al: Secondary leukemia or myelodysplastic syndrome after treatment with epipodophyllotoxins. J Clin Oncol 17:569–577, 1999
9. Obedian E, Fischer DB, Haffty BG: Second malignancies after treatment of operable breast cancer: Lumpectomy and radiation therapy versus mastectomy. J Clin Oncol 18:2406–2412, 2000 10. Hahn P, Nelson N, Baral E: Leukemia in patients with breast cancer following adjuvant chemotherapy and/or postoperative radiation therapy. Acta Oncol 33:599–602, 1994[Medline]
11. Fisher B, Anderson S, Wickerham DL, et al: Increased intensification and total dose of cyclophosphamide in a doxorubicin-cyclophosphamide regimen for the treatment of primary breast cancer: Findings from National Surgical Adjuvant Breast and Bowel Project B-22. J Clin Oncol 15:1858–1869, 1997
12. Fisher B, Anderson S, DeCillis A, et al: Further evaluation of intensified and increased total dose of cyclophosphamide for the treatment of primary breast cancer: Findings from National Surgical Adjuvant Breast and Bowel Project B-25. J Clin Oncol 17:3374–3388, 1999 13. DeCillis A, Anderson S, Bryant J, et al: Acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) on NSABP B-25: An update. Proc Am Soc Clin Oncol A16:130a, 1997 (abstr 459) 14. Korn EL, Dorey FJ: Applications of crude incidence curves. Stat Med 11:813–829, 1992[Medline] 15. Peto R, Peto J: Asymptotically efficient rank invariant test procedures. J R Stat Soc A 135, Part 2:185–207, 1972 16. Bennett JM, Catovsky D, Daniel MT, et al: Proposals for the classification of acute leukemias. Br J Haematol 33:451–458, 1976[Medline]
17. Bennett JM, Catovsky D, Daniel MT, et al: Proposed revised criteria for the classification of acute myeloid leukemia: A report of the French-American-British Cooperative Group. Ann Intern Med 103:620–625, 1985 18. Cheson BD, Cassileth PA, Head DR, et al: Report of the National Cancer Institute-sponsored workshop on definitions of diagnosis and response in acute myeloid leukemia. J Clin Oncol 8:813–819, 1990[Abstract] 19. Van Der Weide M, Sizoo W, Nauta JJ, et al: Myelodysplastic syndromes: Analysis of clinical and prognostic features in 96 patients. Eur J Haematol 41:115–122, 1988[Medline] 20. Economopoulos T, Stathakis N, Foudoulakis A, et al: Myelodysplastic syndromes: Analysis of 131 cases according to the FAB classification. Eur J Haematol 38:338–344, 1987[Medline] 21. Bennett JM, Catovsky D, Daniel MT, et al: Proposals for the classification of the myelodysplastic syndromes. Br J Haematol 51:189–199, 1982[Medline]
22. Michels SD, McKenna RW, Arthur DC, et al: Therapy-related acute myeloid leukemia and myelodysplastic syndrome: A clinical and morphologic study of 65 cases. Blood 65:1364–1372, 1985 23. Bennett JM, Moloney WC, Greene MH, et al: Acute myeloid leukemia and other myelopathic disorders following treatment with alkylating agents. Hematol Pathol 1:99–104, 1987[Medline] 24. Kantarjian HM, Keating MJ: Therapy-related leukemia and myelodysplastic syndrome. Semin Oncol 14:435–443, 1987[Medline] 25. Kouides PA, Bennett JM: Morphology and classification of myelodysplastic syndromes. Hematol Oncol Clin North Am 6:485–499, 1992[Medline]
26. Rowley JD, Golomb HM, Vardiman JW: Nonrandom chromosome abnormalities in acute leukemia and dysmyelopoietic syndromes in patients with previously treated malignant disease. Blood 58:759–767, 1981 27. Curtis RE, Boice JD Jr, Stovall M, et al: Risk of leukemia after chemotherapy and radiation treatment for breast cancer. N Engl J Med 326:1745–1751, 1992[Abstract]
28. Valagussa P, Moliterni A, Terenziani M, et al: Second malignancies following CMF-based adjuvant chemotherapy in resectable breast cancer. Ann Oncol 5:803–808, 1994
29. Tallman MS, Gray R, Bennett JM, et al: Leukemogenic potential of adjuvant chemotherapy for operable breast cancer: The Eastern Cooperative Oncology Group experience. J Clin Oncol 13:1557–1563, 1995 30. Holdener EE, Nissen-Meyer R, Bonadonna G, et al: Second malignant neoplasms in operable carcinoma of the breast: Recent results. Cancer Res 96:188–196, 1984
31. Diamandidou E, Buzdar AU, Smith TL, et al: Treatment-related leukemia in breast cancer patients treated with fluorouracil-doxorubicin-cyclophosphamide combination adjuvant chemotherapy: The University of Texas MD Anderson Cancer Center Experience. J Clin Oncol 14:2722–2730, 1996
32. Greene MH, Young RC, Merrill JM, et al: Evidence of a treatment dose response in acute nonlymphocytic leukemias which occurs after therapy on non-Hodgkin’s lymphoma. Cancer Res 43:1891–1898, 1983 33. Laughlin MJ, McGaughey DS, Crews JR, et al: Secondary myelodysplasia and acute leukemia in breast cancer patients after autologous bone marrow transplant. J Clin Oncol 16:1008–1012, 1998[Abstract] 34. Haas JF, Kittelmann B, Mehnert WH, et al: Risk of leukaemia in ovarian tumor and breast cancer patients following treatment by cyclophosphamide. Br J Cancer 55:213–218, 1987[Medline]
35. Ratain MJ, Rowley JD: Therapy-related acute myeloid leukemia secondary to inhibitors of topoisomerase II: From the bedside to the target genes. Ann Oncol 3:107–111, 1992
36. Pedersen-Bjergaard J, Pedersen M, Roulston D, et al: Different genetic pathways in leukemogenesis for patients presenting with therapy-related myelodysplasia and therapy-related acute myeloid leukemia. Blood 86:3542–3552, 1995 37. Brenner B, Carter A, Sharon R, et al: Acute leukemia following chemotherapy and radiation therapy: A report of 15 cases. Oncology 41:83–87, 1984[Medline]
38. Pedersen-Bjergaard J, Phillip P, Larsen SO, et al: Chromosome aberrations and prognostic factors in therapy-related myelodysplasia and acute nonlymphocytic leukemia. Blood 76:1083–1091, 1990 39. Pedersen-Bjergaard J, Philip P, Larsen SO, et al: Therapy-related myelodysplasia and acute myeloid leukemia: Cytogenetic characteristics of 115 consecutive cases and risk in seven cohorts of patients treated intensively for malignant diseases in the Copenhagen series. Leukemia 7:1975–1986, 1993[Medline] 40. Whitlock JA, Greer JP, Lukens JN: Epipodophyllotoxin-related leukemia: Identification of a new subset of secondary leukemia. Cancer 68:600–604, 1991[CrossRef][Medline] 41. Rowley JD: Rearrangements involving chromosome band 11q23 in acute leukemia. Semin Cancer Biol 4:377–385, 1993[Medline]
42. Andersen MK, Johansson B, Larsen SO, et al: Chromosomal abnormalities in secondary MDS and AML: Relationship to drugs and radiation with specific emphasis on the balanced rearrangements. Haematologica 83:483–488, 1998 43. Brodsky RA, Bedi A, Jones RJ: Are growth factors leukemogenic? Leukemia 10:175–177, 1996[Medline]
44. 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 18:2836–2842, 2000 45. Aisenberg AC: Acute nonlymphocytic leukemia after treatment of Hodgkin’s disease. Am J Med 75:449–454, 1983[CrossRef][Medline] 46. Van Leeuwen FE, Somers R, Taal BG, et al: Increased risk of lung cancer, non-Hodgkin’s lymphoma, and leukemia following Hodgkin’s disease. J Clin Oncol 7:1046–1058, 1989[Abstract] 47. Pedersen-Bjergaard J, Specht L, Larsen SO, et al: Risk of therapy-related leukemia and preleukemia after Hodgkin’s disease: Relation of therapy-related leukemia and preleukemia after Hodgkin’s disease. Lancet 2:83–88, 1987[Medline] 48. Early Breast Cancer Trialists’ Collaborative Group: Tamoxifen for early breast cancer: An overview of the randomized trials. Lancet 351:1451–1467, 1998[CrossRef][Medline] Submitted March 22, 2001; accepted December 23, 2002.
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
M5
AML, M1
