Critique 099: Alcohol and survival after postmenopausal diagnosis of breast cancer – 20 December 2012

Vrieling A, Buck K, Heinz J, Obi N, Benner A, Flesch-Janys D, Chang-Claude J.  Pre-diagnostic alcohol consumption and postmenopausal breast cancer survival: a prospective patient cohort study.  Breast Cancer Res Treat 2012;136:195–207.  DOI 10.1007/s10549-012-2230-2

Authors’ Abstract

Study results on the association of alcohol consumption with breast cancer survival are inconsistent, partly due to the use of different survival outcomes.  We assessed the association of pre-diagnostic alcohol consumption with survival and recurrence in a prospective cohort study in Germany including 2,522 postmenopausal breast cancer patients aged 50–74 years.  Patients were diagnosed between 2001 and 2005 and vital status, causes of death, and recurrences were verified through the end of 2009.  Cox proportional hazards models were stratified by age at diagnosis and study center and adjusted for relevant prognostic factors.

Alcohol consumption was non-linearly associated with increased breast cancer-specific mortality [e.g., ≥ 12 vs.< 0.5 g/day: hazard ratio (HR) = 1.74, 95 % confidence interval (CI): 1.13, 2.67]. Results were independent of estrogen receptor status.  A non-significantly decreased risk of mortality due to other causes was found (≥ 12 vs. < 0.5 g/day: HR = 0.67, 95 % CI: 0.35, 1.29).  Alcohol consumption was not associated with overall mortality (≥ 12 vs. < 0.5 g/day: HR = 1.28, 95 % CI: 0.90, 1.81) and breast cancer recurrence (≥ 12 vs. < 0.5 g/day: HR = 1.08, 95 % CI: 0.73, 1.58).

In conclusion, our findings show that consumption of alcohol before diagnosis is non-linearly associated with increased breast cancer specific mortality but may be associated with decreased risk of mortality due to other causes.

Forum Comments

Background:  Data from previous research are inconsistent as to survival after the diagnosis of breast cancer according to alcohol consumption.  Many studies have described the pre-diagnostic consumption of alcohol in relation to the development of breast cancer, and some have reported survival from the disease.   A paper from Reding et al1 from the Fred Hutchinson Cancer Research Center in Seattle, WA, reported that women who consumed alcohol before a diagnosis of breast cancer had improved survival. 

A number of studies have also related survival related to alcohol consumption following the diagnosis of breast cancer.  Barnett et al2 reported on more than 4,000 women with invasive breast cancer who had taken part in the Studies of Epidemiology and Risk Factors in Cancer Heredity (SEARCH) project, and showed a lower risk of death among women “currently consuming” 7 or more drinks/week versus < 7 drinks/week.  Another previous large study of 3,088 early-stage breast cancer survivors found that neither light nor moderate drinking following a diagnosis of breast cancer was associated with recurrence of breast cancer, but moderate intake lowered overall mortality3. 

As for recurrence of breast cancer, Li et al4 found an increase in the development of a primary cancer in the contralateral breast among women who were smokers and consumed 7 or more drinks per week4.  On the other hand, no relation of alcohol consumption with overall mortality among breast cancer survivors was seen in several other past studies.  In the Nurses’ Health Study, for example, moderate alcohol intake after a diagnosis of breast cancer was not associated with overall mortality among 1,982 women observed for 13.1 years5.  In that study, there was a tendency for poorer survival among current smokers and obese women, although interactions with alcohol were not reported5.  In another prospective cohort study of 1,453 patients with breast cancer observed for 12.6 years in Italy, no association was observed between overall alcohol drinking or wine drinking within 1 year after diagnosis and risk of overall death and breast cancer death6,7.

Kwan et al8 reported in 2010 on a prospective study of 1,897 women with early-stage breast cancer from the Kaiser Permanente Study; alcohol intake assessed at two years after diagnosis was related to subsequent recurrence of breast cancer and breast-cancer death during follow up averaging 7.4 years, although no effects were seen on all-cause mortality.  The authors concluded that consumption of 3 to 4 or more drinks per week increased the risk of recurrent disease and breast-cancer mortality. 

In a study similar to the present one, Harris et al9 investigated whether alcohol intake pre-diagnosis of breast cancer was associated with survival among 3,146 women diagnosed with invasive breast cancer in the Swedish Mammography Cohort.  No significant association was observed between alcohol intake and breast cancer-specific survival.  A significant inverse association was observed between alcohol and non-breast cancer deaths. Those who consumed 3.4-9.9 g per day of alcohol had a 33% lower risk of death compared with non-drinkers (95% CI 0.50-0.90;p(trend)=0.04).  These authors conclude that alcohol intake up to approximately one small drink per day does not negatively impact breast cancer-specific survival and a half drink per day is associated with a decreased risk of mortality from other causes.

Overall comments on present study:   Overall, this is a reasonably large, well-done study with excellent ascertainment of outcomes (mortality from breast cancer, total mortality, recurrence of breast cancer).  There were 316 deaths in the cohort (74.4% due to breast cancer) over a mean follow-up period of 5.5 years.  A limitation is that that there were no data on alcohol consumption after the diagnosis of breast cancer (which could either increase or decrease the risk of all outcomes). 

Most of the comparisons presented in the paper are between the non-drinkers/light drinkers and the highest category of intake, but no description is given of the typical amount of alcohol consumed by women within this group, or no indication of how many heavy drinkers were included.  The authors state that “Results for breast cancer-specific mortality were in the same direction but no longer significant after exclusion of ex-drinkers from the lowest category of alcohol consumption (data not shown).”  You cannot be sure how this would have affected their results, but including ex-drinkers in the lowest group could have either increased or decreased their survival.  While the authors do not emphasize the finding, in almost all analyses the lowest risk of death was among the moderate drinkers reporting ≥ 6 to < 12.0 drinks per week.

Specific comments from reviewers:  Stated reviewer Waterhouse, “I think this study shows not the need for larger studies, as they mention in their conclusion, but the need for more information about the alcohol consumption, i.e., the pattern of consumption or type of alcohol consumed.  Their peculiar spike in breast cancer mortality risk for low alcohol consumption, with lower risk at moderate consumption, suggests that perhaps the moderate consumers were daily drinkers, while the low consumers were weekly drinkers.  However, it is surprising that it is protective even in this population of breast cancer diagnosed patients — who presumably have a very high risk of breast cancer mortality, a risk that should overwhelm mortality due to cardiovascular diseases, for instance.   While the data are of high quality and specific to diagnosed patients, I don’t see terribly ‘new’ findings here.”

Reviewer Skovenborg commented: “There are some strange aspects of the study results.  First, the association of alcohol consumption and breast cancer mortality was independent of estrogen receptor status.  In a very large majority of studies, alcohol consumption is associated with estrogen positive tumours only.10  Further, it is unusual that there was a non-linear association of alcohol consumption and breast cancer specific mortality.  Finally, the mean amount of alcohol consumed was 5.87 g/day = about ½ of a typical drink, with a large range: 0.51 – 289 g/day.  This suggests that serious confounding may be at play in the group of women consuming ≥ 12 grams of alcohol per day.”

Potential for index event bias:  There was also concern by Forum reviewers that the results of these analyses may suffer from what is known in epidemiology as index event bias or collider bias.  Since the exposure being evaluated (alcohol) affects the development of breast cancer, it is likely that the subjects in this study differed according to their previous alcohol intake (being either a drinker or a non-drinker prior to developing breast cancer).  This is an epidemiologic problem that has increasingly become recognized, and was discussed well recently by Dahabreh and Kent.11  These authors point out problems when evaluating a risk factor (e.g., alcohol) among subjects selected for follow up of a particular disease in which not only the occurrence of the disease is affected by the risk factor but the course of the disease (including subsequent mortality) may also be affected by the same risk factor.

In our Forum review of an earlier paper by Kwan (Critique 018, at www.bu.edu/alcohol-forum) we discussed problems in comparing outcomes between women with breast cancer who had been drinkers with those who had not been drinkers before the diagnosis.   The present study had only alcohol consumption data prior to the diagnosis, but still the subsequent course of women with the disease may be different from those who previously consumed alcohol and those who did not.  Thus, this makes it very difficult to compare the subsequent course (including mortality) of subjects for whom the exposure being evaluated (e.g., alcohol consumption) may relate to the development of the disease, be it cardiovascular disease or breast cancer.

References from Forum Review

  1. Reding KW et al, Effect of prediagnostic alcohol consumption on survival after breast cancer in young women.  Cancer Epidemiol Biomarkers Prev 2008;17:1988-1996.
  2. Barnett GC et al, Risk factors for the incidence of breast cancer: Do they affect survival from the disease?  J Clin Oncol 2008;26:3310-3316. 
  3. Flatt SW, Thomson CA, Gold EB, et al: Low to moderate alcohol intake is not associated with increased mortality after breast cancer. Cancer Epidemiol Biomarkers Prev 2010;19:681-688.
  4. Li CI, Daling JR, Porter PL, et al: Relationship between potentially modifiable lifestyle factors and risk of second primary contralateral breast cancer among women diagnosed with estrogen receptorpositive invasive breast cancer. J Clin Oncol 2009;27:5312-5318.
  5. Holmes MD, Stampfer MJ, Colditz GA, et al: Dietary factors and the survival of women with breast carcinoma. Cancer 1999;86:826-835.
  6. Dal Maso L, Zucchetto A, Talamini R, et al: Effect of obesity and other lifestyle factors on mortality in women with breast cancer. Int J Cancer 2008;123:2188-2194.
  7. Franceschi S, Dal Maso L, Zucchetto A, et al: Alcohol consumption and survival after breast cancer. Cancer Epidemiol Biomarkers Prev 2009;l8:1011-1012.
  8. Kwan ML, Kushi LH, Weltzien E, Tam EK, Castillo A, Sweeney C, Caan BJ.  Alcohol Consumption and Breast Cancer Recurrence and Survival Among Women With Early-Stage Breast Cancer:The Life After Cancer Epidemiology Study.  J Clin Oncol 2010;28 (published ahead of print, 10.1200/JCO.2010.29.2730)
  9. Harris HR, Bergkvist L, Wolk A.  Alcohol intake and mortality among women with invasive breast cancer.  Br J Cancer 2012;106:592-595. doi: 10.1038/bjc.2011.561.
  10. Suzuki R et al. Alcohol and postmenopausal breast cancer risk defined by estrogen and progesterone receptor status: a prospective cohort study. J Natl Cancer Inst 2005;97:1601-1608.
  11. Dahabreh IJ, Kent DM.  Commentary: Index event bias as an explanation for the paradoxes of recurrence risk research.  JAMA 2011;305, reprint No. 8.

Forum Summary

This paper assessed the association of pre-diagnostic alcohol consumption with disease recurrence and survival in a prospective cohort study in Germany based on 2,522 postmenopausal breast cancer patients, aged 50–74 years.  The authors report that alcohol consumption was non-linearly associated with increased breast cancer-specific mortality, but not with breast cancer recurrence.  Results were independent of estrogen receptor status.  The authors also report a non-significantly decreased risk of mortality due to other causes.

Forum reviewers considered this to be a well-done analysis that was limited, however, by the lack of data on drinking patterns prior to breast cancer diagnosis and because no data were available on drinking after the diagnosis (the latter limits the applicability of its results to women who have already developed breast cancer).  The comparisons were primarily between non-drinkers/light drinkers and women in the highest alcohol category; the authors state that when ex-drinkers were removed from the referent category, their results regarding breast cancer-specific mortality were no longer statistically significant.   The authors do not describe the potential effects on their main results that could result from what is known as “index event bias,” which may occur when an exposure (e.g., alcohol)  relates to the development of a disease, breast cancer in this study, and may also relate to sequelae (e.g., recurrence or death) from the same disease. 

Forum reviewers conclude that current scientific data show that alcohol consumption may increase the risk of developing breast cancer.  However, they believe that the effects on subsequent survival after such a diagnosis remain unclear.

                                                          *                       *                      *

Comments on this critique were provided by the following members of the International Scientific Forum on Alcohol Research

Andrew L. Waterhouse, PhD, Marvin Sands Professor, Department of Viticulture and Enology, University of California, Davis; Davis, CA, USA. 

Yuqing Zhang, MD, DSc, Epidemiology, Boston University School of Medicine, Boston, MA, USA

Erik Skovenborg, MD, Scandinavian Medical Alcohol Board, Practitioner, Aarhus, Denmark

Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA

R. Curtis Ellison, MD, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA

In addition to the above, included in this review were contributions from the critique of an earlier paper on this topic (Kwan et al) that were provided by the following Forum members:

Lynn Gretkowski, MD, Obstetrics/Gynecology, Mountainview, CA, Stanford University, Stanford, CA, USA

Luc Djoussé, MD, DSc, Dept. of Medicine, Division of Aging, Brigham & Women’s Hospital and Harvard Medical School, Boston, MA, USA