Family history of breast or ovarian cancer modifies the risk of secondary leukemia after breast cancer: Results from a population-based study
Version of Record online: 31 OCT 2007
Copyright © 2007 Wiley-Liss, Inc.
International Journal of Cancer
Volume 122, Issue 5, pages 1114–1117, 1 March 2008
How to Cite
Verkooijen, H. M., Fioretta, G., Rapiti, E., Vlastos, G., Neyroud-Caspar, I., Chappuis, P. O. and Bouchardy, C. (2008), Family history of breast or ovarian cancer modifies the risk of secondary leukemia after breast cancer: Results from a population-based study. Int. J. Cancer, 122: 1114–1117. doi: 10.1002/ijc.23212
- Issue online: 24 DEC 2007
- Version of Record online: 31 OCT 2007
- Manuscript Accepted: 3 SEP 2007
- Manuscript Received: 26 JUN 2007
- Swiss National Science Foundation (PROSPER Grant). Grant Number: 3233-069350
- Swiss Cancer League
- breast cancer;
- secondary cancer;
- family history;
We evaluated the impact of a family history of breast/ovarian cancer on the risk of secondary leukemia following breast cancer. At the Geneva cancer registry, we identified 4,397 patients diagnosed with invasive breast cancer between 1990 and 2004. Patients were followed up for leukemia until the end of 2005. Family history was categorized as positive in patients with ≥1 first- or second-degree relative with breast/ovarian cancer. We compared leukemia rates in patients with positive and negative family histories with those expected in the general population, generating standardized incidence ratios (SIRs). With Cox regression analysis, we calculated adjusted risks of secondary leukemia in patients with familial risks compared to those without it. Breast cancer patients had a significantly increased risk of secondary acute leukemia (SIR 3.2, 95% CI: 1.2–6.9) but not of chronic leukemia (SIR 1.6, 95% CI: 0.6–3.5). Among patients with a positive family history (n = 1.125, 25.6%), the SIRs were 5.7 (95% CI: 1.2–16.6) for acute and 5.2 (95% CI: 1.4–13.3) for chronic leukemia. Among breast cancer patients, family history was independently associated with leukemia [adjusted hazard ratio (HRadj) of 3.2, 95% CI: 1.1–9.2, among patient with vs. without family history]. The effect of family history was stronger for chronic leukemia (HRadj: 11.6, 95% CI 1.3–104.7) than for acute leukemia (HRadj 1.6, 95% CI: 0.4–6.6). Breast cancer patients with a family history of breast/ovarian have an increased risk of secondary leukemia, both compared to the general population as well as to breast cancer patients without family histories. This excess risk is largely due to the increased risk of secondary chronic leukemia. © 2007 Wiley-Liss, Inc.
The risk of secondary leukemia following treatment for breast cancer has been well established.1, 2, 3, 4 Modifiers of this risk include increasing age and exposure to radiotherapy as well as certain types of chemotherapy.5, 6, 7 Most studies reported only increased risks of acute myeloid leukemia following breast cancer, but a recent large, population-based study has shown that breast cancer patients are at increased risk of developing acute lymphoblastic leukemia and chronic myeloid leukemia as well.1
A family history of leukemia or other hematologic malignancies is not considered as a major risk factor for leukemia,8, 9, 10 except for chronic lymphocytic leukemia.11 There have been some reports of breast cancer and leukemia clustering in families, in particular, in the context of some hereditary cancer syndromes like ataxia-telangiectasia and Li-Fraumeni syndrome.10, 12 In addition, Rauscher et al. reported that adults, whose sisters had been diagnosed with breast cancer, were at 1.8–3.3 times increased risk of developing acute leukemia.13
Although the excess absolute risk of secondary leukemia following breast cancer is rather low—around 9 extra cases per 100,000 person years—and the individual risk of developing leukemia following breast cancer is decreasing,1 the overall the burden of this serious complication may actually rise. The increasing incidence of breast cancer in combination with improving survival rates leads to increasing numbers of women at risk of leukemia following breast cancer. The aging of the population may also contribute to increasing numbers of leukemia following breast cancer, since some studies have indicated that the risk of leukemia following breast cancer is highest among women who developed breast cancer at an older age.2 Finally, the number of patients managed with breast conserving treatment, including lumpectomy and radiotherapy, is still on the rise, leading to increasing number of patients exposed to ionizing radiation and therefore at increased risk of leukemia.
For this reason, it would be useful to identify additional risk factors for leukemia after breast cancer, in order to be able to better predict which women will develop this often highly fatal disease. In this study, we evaluated whether the risk of secondary leukemia after breast cancer is modified by family history of breast/ovarian cancer.
We used information from the population-based Geneva Cancer Registry, which records all incident cancers occurring in the population of the Geneva canton (∼420,000 inhabitants) since 1970. It collects information from various sources and is considered accurate, as attested by its very low percentage (<2%) of cases recorded from death certificates only.14 All hospitals, pathology laboratories and private practitioners in the canton are requested to report all cancer cases. Trained tumor registrars systematically abstract data from medical and laboratory records. Physicians regularly receive enquiry forms to complete missing clinical and therapeutic data.
Recorded data include sociodemographic information, tumor characteristics coded according to the International Classification of Diseases for Oncology15 and treatment given during the first 6 months after diagnosis. The Registry staff regularly assess survival, taking as reference date the date of confirmation of diagnosis or the date of hospitalization (if it preceded the diagnosis and was related to the disease). In addition to passive follow-up (standard examination of death certificates and hospital records), active follow-up is performed yearly using the files of the Cantonal Population Office (office in charge of the registration of the resident population). Cause of death is taken from clinical files.
In the current study, we included resident patients diagnosed with invasive breast cancer between 1990 and 2004. We limited our study to this period, since for these years information on family history of cancer was available. Family history of breast and/or ovarian cancer was obtained from the Familial Breast Cancer Registry. This unit of the Geneva Cancer Registry was set up in 1999, by extending its data set to the detailed family history of cancer for all women diagnosed with invasive breast cancer in the Geneva population.16 The level of familial risk was classified into 3 categories according to the number of relatives diagnosed with breast or ovarian cancer, their age at diagnosis and their degree of kinship.17 The low familial risk category included breast cancer patients with no first- or second-degree relatives with breast or ovarian cancer (i.e., sporadic cases). The high familial risk category included patients who reported 1 of the following family histories: (i) ≥1 first-degree relative with breast or ovarian cancer ≤50 years; (ii) ≥2 first-degree relatives with breast or ovarian cancer at any age; (iii) ≥3 cases of breast or ovarian cancer among first- or second-degree relatives. Patients with other types of family history were classified into the moderate familial risk category.
Other variables of interest were age, stage, family history of breast cancer and type of adjuvant treatment. Stage was coded using the pathologic tumor node metastasis (pTNM) classification system or, when not available, the clinical cTNM classification.18 Adjuvant therapy was categorized as (i) no adjuvant therapy, (ii) radiotherapy, (iii) chemotherapy and (iv) radiotherapy and chemotherapy. Family history of breast and/or ovarian cancer was obtained from the Familial Breast Cancer Registry. This unit of the Geneva Cancer Registry was set up in 1999, by extending its data set to the detailed family history of cancer for all women diagnosed with invasive breast cancer in the Geneva population.
Follow-up of breast cancer patients for secondary leukemia occurrence started 6 months after breast cancer diagnosis. Follow-up ended on date of leukemia occurrence, date of death, emigration or December 31, 2005. We only considered microscopically proven leukemia.
The relative risk of secondary leukemia was expressed as the ratio of observed to expected number of cases, i.e. standardized incidence ratio (SIR).19 To calculate the expected number of cases, we multiplied the age- and calendar period specific incidence rates of leukemia in the background population by correspondingly stratified person-years at risk in our breast cancer cohort and summing all the products. With Cox regression analysis we evaluated whether differences in leukemia risk between women of high versus low familial risk persisted after adjustment for age, disease stage and types of adjuvant treatment.
Between 1990 and 2004, 4,861 women were diagnosed with invasive breast cancer in Geneva, Switzerland. We excluded patients diagnosed with in situ or invasive cancer before or within 6 months after breast cancer diagnosis (n = 445). In addition, we excluded 19 autopsy cases, leaving us with 4,397 patients with primary invasive breast cancer. After a median follow-up of 61 months (mean, 72 months), 14 patients (0.32%) developed secondary leukemia. The time interval between breast cancer diagnosis and date of leukemia diagnosis varied from 8 months to 7.7 years.
A moderate familial risk was reported for 885 (20%) patients, and 240 (5.5%) patients were classified as having a high familial risk. For 405 (9.2%) patients, there was no information on presence or absence of breast or ovarian cancer in family members. Patients with a moderate or high familial risk were on average 2 years younger at breast cancer diagnosis than those at low familial risk (58–59 years vs. 61 years respectively; Table I). Stage distribution was similar for women at low, moderate and high familial risk: for all 3 groups, more than 80% of cancers were stage I or stage II at diagnosis. However, women at moderate or high familial risk were more likely to have received some type of adjuvant treatment: 80% of women at low familial risk received adjuvant treatment versus 87% of women at moderate or high familial risk (p < 0.001). Women at unknown familial risk were older, had more often advanced stage disease, less often adjuvant treatment and more often missing information on stage and adjuvant treatment.
|Low||Moderate||High||Unknown||p value (chi square)|
|N||2,867 (65.2)1||885 (20.1)||240 (5.5)||405 (9.2)|
|Mean age (test with ANOVA)||60.7 [24–100]2||58.1 [25–94]||58.6 [24–97]||69.6 [29–101]||<0.0001|
|I||1110 (39)||340 (38)||93 (39)||109 (27)||<0.0001|
|II||1226 (43)||403 (46)||111 (46)||133 (33)|
|III||264 (9)||85 (10)||24 (10)||38 (9)|
|IV||159 (6)||33 (4)||7 (3)||53 (13)|
|Unknown||108 (4)||24 (3)||5 (2)||72 (18)|
|None||577 (20)||115 (13)||30 (13)||229 (57)||<0.0001|
|Chemo only||209 (7)||72 (8)||18 (8)||38 (9)|
|RT only||1088 (38)||347 (39)||102 (43)||32 (8)|
|RT and chemo||885 (31)||327 (37)||83 (35)||30 (7)|
|Unknown||108 (4)||24 (3)||7 (3)||76 (19)|
Among the 14 patients who developed leukemia after breast cancer, 6 belonged to low risk families, 7 had an increased familial risk (5 moderate and 2 high familial risks) and for 1 patient the family history was missing. There were no major differences in terms of age at breast cancer diagnosis, age at leukemia diagnosis, stage at diagnosis and type of adjuvant breast cancer treatment between women at low versus increased familial risk (Table II). Women at low familial risk had on average a significantly shorter time interval between breast cancer diagnosis and onset of secondary leukemia than those with increased familial risk.
|Low (N = 6)||Increased (N = 7)||p value|
|Mean age at breast cancer diagnosis (years)||67||66||0.861|
|Mean age at leukemia diagnosis (years)||70||71||0.812|
|Delay between breast cancer and leukemia (months)||27||60||0.007|
|Stage at diagnosis||0.796*|
|I||3 (50)1||3 (43)|
|II||3 (50)||4 (57)|
|III/IV||0 (0)||0 (0)|
|None||1 (17)||1 (20)|
|Only radiotherapy||3 (50)||5 (60)|
|Radio- and chemotherapy||2 (33)||1 (20)|
|Only chemotherapy||0 (0)||0 (0)|
Compared to the general population, breast cancer patients in general were at increased risk of developing leukemia (SIR 2.3, 95% CI 1.2–3.8). This excess risk was significant only for acute leukemia (SIR 3.2, 95% CI 1.2–6.9) but not for chronic leukemia (SIR 1.6, 95% CI 0.6–3.5). After stratification by familial risk, the SIR of leukemia was higher for patients with an increased familial risk compared to those without (Table III). Breast cancer patients without a family history of breast and/or ovarian cancer did not have an increased risk of secondary leukemia compared to the general population (SIR 1.4, 95% CI 0.5–3.1), but patients with a moderate or high familial risk had a more than 5-fold increased risk of leukemia (SIR 5.4, 95% CI 2.2–11.1). After further stratification into acute and chronic leukemia, the risk of acute leukemia was almost 3-fold, but not significantly increased for women without familial risk (SIR 2.8, 95% CI 0.9–6.9). Breast cancer patients at increased familial risk had a significantly increased risk of acute leukemia compared to the general population (SIR 5.7, 95% 1.2–16.6). The risk of chronic leukemia was not increased for breast cancer patients at low familial risk (SIR 0.4, 95% CI 0.01–2.2), but significantly increased for breast cancer patients at increased familial risk (SIR 5.2, 95% CI 1.4–13.3).
|Familial risk||All leukemia||Chronic leukemia||Acute leukemia|
|Low||6||4.3||1.4 (0.5–3.1)1||1||2.5||0.4 (0.01–2.2)||5||1.8||2.8 (0.9–6.9)|
|Increased||7||1.3||5.4 (2.2–11.1)||4||0.77||5.2 (1.4–13.3)||3||0.53||5.7 (1.2–16.6)|
|Unknown||1||0.68||1.5 (0.04–8.13)||1||0.43||2.4 (0.1–13.1)||0||0.25||NA|
|All||14||6.2||2.3 (1.2–3.8)||6||3.7||1.6 (0.6–3.5)||8||2.5||3.2 (1.2–6.9)|
Within the group of breast cancer patients, a family history of breast/ovarian cancer was associated with a significantly higher risk of developing chronic leukemia [hazard ratio (HR) 11.0, 95% CI 1.2–98.5], but not of acute leukemia (HR 1.6, 95% CI 0.4–6.5). These risks did not change after adjustment for age, stage and use of adjuvant chemotherapy and/or radiotherapy (multiadjusted HRs 11.6, 95% CI 1.3–104.7 for chronic leukemia and 1.6, 95% CI 0.4–6.6 for acute leukemia). Also, in subgroup analysis including patients treated with radiotherapy only, or patients treated with radiotherapy with or without chemotherapy, the excess risk of patients at increased familial risk remained elevated (Table IV).
|Familial risk||All leukemia||Chronic leukemia||Acute leukemia|
|Unadjusted HR||Adjusted HR||Unadjusted HR||Adjusted HR||Unadjusted HR||Adjusted HR|
|All patients (4,397)|
|Low||1 (ref)||1 (ref)||1 (ref)||1 (ref)||1 (ref)||1 (ref)|
|Increased||3.1 (1.0–9.2)||3.2 (1.1–9.5)||11.0 (1.2–98.5)||11.6 (1.3–104.7)||1.6 (0.4–6.5)||1.6 (0.4–6.6)|
|Patients treated with radiotherapy only (n = 1,537)|
|Low||1 (ref)||1 (ref)||1 (ref)||1 (ref)||1 (ref)||1 (ref)|
|Increased||4.4 (1.1–18.5)||4.4 (1.0–18.4)||5.5 (0.5–60.7)||5.5 (0.5–60.3)||3.9 (0.5–60.7)||3.9 (0.6–23.1)|
|Patients treated with radiotherapy, with or without chemotherapy (n = 2,832)|
|Low||1 (ref)||1 (ref)||1 (ref)||1 (ref)||1 (ref)||1 (ref)|
|Increased||2.9 (0.9–9.5)||3.0 (0.9–10.0)||7.5 (0.8–72.0)||7.9 (0.8–75.8)||1.8 (0.8–72.0)||1.9 (0.4–8.3)|
In consistence with previous studies, we observed an increased risk of leukemia in women treated for breast cancer.1, 2, 6 Usually, these increased risks of leukemia are being attributed to exposure to radiotherapy and certain chemotherapeutical agents.1, 3, 6 Our results suggest that familial risk factors may modify the risk of secondary leukemia after breast cancer as well. For the first time we show that breast cancer patients with first or second degree relatives with breast and/or ovarian cancer have higher risks of developing secondary leukemia, not only compared to the general population, but also compared to breast cancer patients without familial risk factors. This excess risk is chiefly due to an excess risk of chronic leukemia.
There are several potential explanations for our findings. First of all, women at increased familial risk may have received different adjuvant treatment regimens. In fact, in an earlier study we demonstrated that young breast cancer patients were more likely to receive some kind of adjuvant systemic therapy if they are at high familial risk.20 Also in the present study, women at increased familial risk were more likely to have received some form of adjuvant treatment than women without familial risk. Nevertheless, the design of this study allowed us to adjust for this potentially confounding factor, and we found that the risk estimates of leukemia following breast cancer were not modified by adjusting for use of adjuvant treatment. Nevertheless, we cannot exclude that the radiotherapy and chemotherapy regimens of women of high versus low familial risk may have differed in terms of cumulative doses and intensity, conferring different risks of leukemia.
Women at increased familial risk may have been receiving more intensive follow-up after their breast cancer than women without affected family members. Although we find this rather unlikely, we cannot exclude that more intensive follow-up of patients at increased familial risk may have increased the likelihood of being diagnosed with chronic leukemia.
Shared genes may also explain the increased risk of leukemia in breast cancer patients with a positive family history. Breast cancer and leukemia cluster in families with rare hereditary cancer syndromes12 and adult acute leukemia patients are more likely to have a family history of breast cancer as compared to healthy controls.13 In addition, a study by Rudd et al. provided evidence that inherited predisposition to chronic lymphocytic leukemia was partly mediated though low-penetrance polymorphisms in the ARM-BRCA2-Chek2 DNA damage-response axis. Mutations/variations in these same genes are also associated with increased risk of breast cancer.21
Finally, it has been hypothesized that the presence of a family history of breast cancer modified the effect of certain leukemogenic risk factors on the risk of developing leukemia.22 Rausscher et al. showed that exposure to smoking, solvents, aromatic hydrocarbon and diagnostic radiation exams doubled the risk of acute leukemia in the presence of a family history of breast cancer and often tripled it in case of an affected sibling. Extrapolating these findings to our study, we could theorize that exposure to radiotherapy and chemotherapy has a stronger impact in the presence of a positive family history of breast/ovarian cancer. Unfortunately, our study sample was too limited to test for interaction between familial risk factors and adjuvant treatment.
We recognize that our study has several shortcomings. First of all, the number of leukemia cases was rather limited, impairing detailed analyses and tests for interaction. Nevertheless, even with small numbers, rather high and statistically significant risk estimates were obtained, suggesting that the associations are real and important. As mentioned earlier, the Geneva Cancer Registry has no automated detailed information on types and doses of chemotherapy and radiotherapy administered, preventing us from evaluating dose–response relationships. Since the Familial Breast Cancer Registry only started in 1990, the follow-up of our population is rather limited. However, even with relatively short follow-up we found significantly increased leukemia risks for women at increased familial risk.
Therefore, we can conclude that the risk of leukemia following breast cancer is higher in women with a family history of breast/ovarian cancer. This is particularly true for chronic leukemia. Future research is needed to unravel the underlying reasons for this association. If our findings were to be confirmed in larger studies, we may have to rethink the optimal treatment and follow-up strategies for breast cancer patients with an increased familial risk.
Dr. H.M. Verkooijen was financially supported by PROSPER Grant No. 3233-069350 from the Swiss National Science Foundation. The Geneva Familial Breast Cancer Registry was set up with financial support of the Swiss Cancer League.
- 7Therapy-related acute lymphoblastic leukaemia with MLL rearrangements following DNA topoisomerase II inhibitors, an increasing problem: report on two new cases and review of the literature since 1992. Br J Haematol 2001; 114: 539–43., , , , , .
- 10Cancer epidemiology and prevention. New York: Oxford University Press, 2006., .
- 14Switzerland, Geneva. In: ParkinDM,WhelanSL,FerlayJ,RaymondL,YoungJ, eds. Cancer incidence in five continents, Vol. 7. Lyon: International Agency for Research on Cancer, 1997. 666–9..
- 15World Health Organization (WHO), eds. ICD-O International classification of diseases for oncology, 1st ed. Geneva: World Health Organization, 1976.
- 16Occurrence and impact of genetic factors in breast cancer among the female population in Geneva: creation of the Geneva familial breast cancer registry. Bull Suisse Cancer 2002; 2650: 165–8., , , , , , , .
- 18SobinLH, WittekindCh, eds. TNM classification of malignant tumours,6th ed. New York: UICC, 2002.
- 19Person-Years (PYRS). A Fortran program for cohort study analysis. Lyon: International Agency for Research on Cancer (IARC), 1989. Report No. 89/006., , .