Papers

Risk of gynaecomastia associated with cimetidine, omeprazole, and other antiulcer drugs

Boston Collaborative Drug Surveillance of Program, Boston University Medical Centre, Lexington, MA 02173-5207, USA.

Abstract

Objective : To study the risk of gynaecomastia associated with cimetidine, misoprostol, omeprazole and ranitidine.
Design : Open cohort study with nested case-control analysis.
Setting : General practices in United Kingdom that had computerised offices, 1989-92.
Subjects : 81 535 men aged 25-84 years who received at least one prescription for cimetidine, misoprostol, omeprazole, or ranitidine during the study period.
Main outcome measures : New occurrences of idiopathic gynaecomastia diagnosed by general practitioner.
Results : The relative risk of gynaecomastia for current users of cimetidine compared with non-users was 7.2 (95% confidence interval 4.5 to 11.3). Relative risks for misoprostol, omeprazole, and ranitidine were 2.0 (0.1 to 10.7), 0.6 (0.1 to 3.3), and 1.5 (0.8 to 2.6), respectively. Current users of cimetidine on a daily dose greater than equal to 1000 mg had more than 40 times the risk of developing gynaecomastia than non -users. The period of highest risk was seven to 12 months after starting cimetidine treatment. Spirono-lactone (relative risk 9.3 (3.3 to 26.1) and verapamil (9.7 (2.6 to 36.0)) were associated with a relative risk of gynaecomastia comparable to one for cimetidine.
Conclusions : Use of cimetidine, but not the three other antiulcer drugs, is associated with a substantially greater risk of gynaecomastia in men. A strong dose-response relation was present among cimetidine users.

Introduction

Gynaecomastia (enlargement of true breast tissue as opposed to adipose tissue) was a common clinical finding in case series.*RF 1-3* The differential diagnosis of gynaecomastia depends on physiological and pathological criteria, and pathological gynaecomastia can be further classified into that associated with other medical conditions and idiopathic gynaecomastia. Cimetidine has repeatedly been reported as causing gynaecomastia,*RF 4-7* and ranitidine was associated with gynaecomastia in a single case.⁸ More recently omeprazole, a proton pump inhibitor also used as an ulcer healing drug, has been associated in more than a dozen cases with the development of gynaecomastia.*RF 9-11* No epidemiological study has been published comparing the incidence of gynaecomastia among the various ulcer healing drugs. We performed a large population based cohort study which provides estimates of the absolute and relative risk of idiopathic gynaecomastia in patients who received cimetidine, misoprostol, omeprazole, or ranitidine, four drugs used primarily for treating peptic ulcer disease. The results are based on information held on British general practitioners' computers.

Methods

Over four million residents in the United Kingdom are enrolled with selected general practitioners who use office computers provided by Value Added Medical Products (VAMP) Research and have agreed to provide data for research purposes. They record medical information in a standard manner and supply it anonymously to VAMP Research on an ongoing basis. Among other items, the recorded information includes the patient's characteristics, medical diagnoses and comments arising from the patient's visit to the general practitioner, referral letters from consultants, and details of stays in hospital. Also, the general practitioners generate prescriptions directly with the computer, and these drugs are automatically transcribed into the patient's computer record. A distinctive attribute of this computerised medical information system is that the general practitioner is asked to provide the indication for each new course of medication. A modification of the Oxford Medical Information System classification is used to enter medical diagnoses, and a coded drug dictionary based on the Prescription Pricing Authority's dictionary is used for the recording of prescriptions. Two recent validation studies determined that information from manual medical records in the general practitioners' offices was recorded on the computer over 90% of the time.^12,13

Study population - For this study, 478 practices contributed information from 1 January 1989 through 15 September 1992. Men between 25 and 81 years old in 1989 and registered with one of these practices entered the study population when they were prescribed cimetidine, misoprostol, omeprazole, or ranitidine during the study period. Men with a history of gynaecomastia before they received one of these drugs during the study period were excluded, as were those with Klinefelter syndrome, testicular cancer, breast cancer, or liver disease or who had used replacement androgen therapy. A total of 81 535 men formed the final study population and were followed until either the first occurrence of gynaecomastia, any malignant neoplasm, liver disease, a diagnosis of alcoholism, use of replacement androgen therapy, transfer to another practice, death, or 15 September 1992, whichever came first (table 1).

TABLE I - Prescriptions for ulcer healing drugs in study population. Values are numbers (percentages)
--------------------------------------------------------------------------------------------------------------------------------
                                   Age distribution of patients receiving prescriptions
                                 -----------------------------------------------------------------------------------------------
                No of patients*    No of prescriptions      25-39           40-54              55-69           70-84
Drug             (n = 81 535)         (n = 585 134)     69 291 (11.8)   140 607 (24.0)    223 892 (38.3)   151 344 (25.9)
--------------------------------------------------------------------------------------------------------------------------------
Cimetidine          37 202               215 191         25 388(11.8)    51 668(24.0)      83 591(38.8)    54 544(25.3)
Misoprostol          2 663              9196 682(7.4)     2 221(24.2)     3 794(41.3)       2 499(27.2)
Omeprazole           9 972                33 503          5 026(15.0)     8 580(25.6)      12 253(36.6)     7 644(22.8)
Ranitidine          45 674               327 244         38 195(11.7)    78 138(23.9)     124 254(38.0)    86 657(26.5)
--------------------------------------------------------------------------------------------------------------------------------
*Receiving at least one prescription for any of the four ulcer healing drugs.
About 16% of the study subjects received more than one drug during the study.

Exposure Definition - Thirty days was the most common duration of an antiulcer prescription, accounting for 72% of all prescriptions; 60 days was the second most common duration, accounting for 15% of all prescriptions. Person time was aggregated across calendar time and age groups into two mutually exclusive time windows for each of the four drugs: "current use," starting from day 1 of each prescription until the earlier of day 60 or the date of the next prescription; and ``past use,'' from day 61 of each prescription until the earlier of day 120 or the date of the next prescription. The person time beginning from day 120 after each prescription for any ulcer healing drug and continuing until any subsequent prescription was combined into a ``non-use'' category. Duration of use was defined as the number of consecutive prescriptions (that is, prescriptions that followed within a maximum of 60 days of the same ulcer healing drug) among current users. Duration was categorised as ``unknown'' when there were less than six months' data in the automated prescription file. Daily dose was estimated according to the strength prescribed and the prescription's instructions as recorded on the computer.

Case ascertainment - We identified among the study population 160 subjects who had a first time diagnosis of gynaecomastia. We reviewed their computerised profiles without knowing their drug use and eliminated six with clear concurrent causes for their breast enlargement (liver disease, two; cancer, three; or trauma, one). To validate the diagnosis of gynaecomastia recorded on a computer for a random sample of half of the 160 subjects, we sent the general practitioners a questionnaire requesting details of some of the clinical features and any correspondence available related to the diagnosis of interest (referral letters, discharge letters). All patients' personal identifiers were suppressed before any material was sent to us. We excluded only one additional patient (trauma) after reviewing the information on these 80 patients.

Cohort analysis - We calculated standardised rates of gynaecomastia for each antiulcer drug, using as an internal standard the distribution of person time for the total study population over categories of calendar time and age groups. We also performed Poisson regression with the generalised linear interactive modelling (GLIM) package¹⁴ and obtained similar results.

Nested case-control analysis - A random index date between 1 January 1989 and 15 September 1992 was assigned for each of the 81 535 study members. If the random date of a study member were included in his eligible time at risk (from study entry to end of follow up), we considered the person as an eligible control.¹⁵ From the list of eligible controls, we selected a random sample of 1000 controls and compared them with the 153 cases. We used unconditional logistic regression to estimate the relative risk and 95% confidence interval for gynaecomastia within each category of exposure to ulcer healing drugs, using the non-user group as the reference group. We examined duration-response and dose-response relations among current users of cimetidine and we calculated relative risks for other selected drugs implicated in the literature as causing gynaecomastia (box).

* Drugs associated with gynaecomastia

Antibiotics:
Isoniazid
Ketoconazole
Metronidazole
Miconazole

Cardivascular drugs:
Atenolol
Captopril
Digoxin
Enalapril
Methyldopa
Nifedipine
Spironolactone
Verapamil

Psychoactive drugs
Diazepam
Tricyclic
antidepressents

Results

In all, 153 men presented with a first occurrence of idiopathic gynaecomastia. Table II presents a summary of the clinical features for the 80 patients for whom we received general practitioners' medical reports. In 12 patients the gynaecomastia did not regress, and five underwent a subareolar mastectomy. In none of the cases was a diagnosis of any other endocrine changes, such as impotence, recorded on computer.

TABLE II - Clinical features of cases of gynaecomastia*
-----------------------------------------------------------------------
Finding                                          No (%) of patients
-----------------------------------------------------------------------
Clinical presentation:
  Self reported                                        67 (84)
  Incidental finding                                   11 (14)
  Unknown                                               2 (2)
Site:
  Unilateral:                                          34 (43)
  Left                                                 12 (15)
  Right                                                12 (15)
  Unknown                                              10 (13)
  Bilateral                                            38 (47)
  Unknown                                               8 (10)
Patient referred to consultant or hospital:
  No                                                   55 (69)
  Yes                                                  21 (26)
  Unknown                                               4 (5)
Regression:
  Complete                                             37 (46)
  Partial                                              14 (18)
  None                                                 12 (15)
  Unknown                                              17 (21)
-----------------------------------------------------------------------
 *Based on 80 patients with medical records. The patient excluded after
review of medical records had a self reported presentation in the right breast
site. He was referred and the regression status was unknown.

The incidence for age and calendar time standardised among non-users and current users of the four antiulcer drugs is shown in table III. Only cimetidine users had a substantially greater risk of gynaecomastia than non -users: the relative risk associated with current use of cimetidine was 7.2 (95% confidence interval 4.5 to 11.3). The risk among users of the three other antiulcer drugs was not greatly different from the risk among non-users. These estimates of relative risk were virtually identical to the ones obtained in the nested case-control analysis.

TABLE III - Incidence and relative risk of gynaecomastia among
non-users and current users of ulcer healing drugs
------------------------------------------------------------------------------------------------------------
                               No of cases*         Incidence                       Relative risk
              Person years      (n = 153)    (per 1000 person years)(dagger)   (95% confidence interval)
------------------------------------------------------------------------------------------------------------
Non-users       55 763             24                 0.46                        1.0
Cimetidine      24 056             86                 3.29                        7.2 (4.5 to 11.3)
Misoprostol        967              1                 0.74                        2.0 (0.1 to 10.7)
Omeprazole       3 386              1                 0.25                        0.6 (0.1 to 3.3)
Ranitidine      35 133             26                 0.69                        1.5 (0.8 to 2.6)
------------------------------------------------------------------------------------------------------------
 *15 cases that occurred during past use of one of these drugs.
 (dagger)Standardised rates for age and calendar year.

Table IV presents the risk among current users of cimetidine broken according to duration of treatment, daily dose, and the joint effects between increasing duration and increasing daily dose. The period at highest risk seemed to be between the seventh and 12th month after starting treatment with cimetidine. Dose was also a major predictor of the occurrence of gynaecomastia, with most of the risk associated with greater than equal to 1000 mg daily. There was a strong synergy of effects between the duration and dose of cimetidine. When we used a common reference group (patients who received six or fewer prescriptions for cimetidine at a dose of 800 mg or less) we found relative risks of 2.0 in those who received seven or more prescription at a dose of 800 mg or less, 4.2 in those receiving six or fewer prescriptions at a dose of 1000 mg or more, and 13.1 in those receiving seven or more prescriptions at a dose of 1000 mg or more.¹⁶ The excess relative risk due to interaction between an increasing duration and an increasing dose was 7.9; owing to the small numbers in each group, this estimate of interaction presented some statistical uncertainty.¹⁷ The proportion of gynaecomastia among current users of cimetidine with both

TABLE IV - Influence of duration and daily dose of cimetidine
on risk of gynaecomastia among current users
----------------------------------------------------------------------------------------------------------------------
                                                      No of cases      No of controls         Relative risk*
                                                       (n = 153)        (n = 1000)       (95% confidence interval)
----------------------------------------------------------------------------------------------------------------------
Duration (No of prescriptions):
  1-3                                                     14                52                3.9 (2.0 to 7.7)
  4-6                                                     10                17                7.0 (2.9 to 16.6)
  7-12                                                    27                28               14.3 (7.6 to 26.9)
  >=13                                                    22                40                6.5 (3.5 to 12.2)
  Unknown                                                 13                32                6.2 (2.8 to 13.6)
Daily dose (mg):
  <=600                                                   18                64                3.4 (1.8 to 6.4)
  601-999                                                 52                100               7.7 (4.8 to 12.3)
  >=1000                                                  16                5                43.5 (14.8 to 127.3)
Duration (No of prescriptions) and daily dose (mg):
  1-6; <1000                                              20                66                4.1 (2.3 to 7.5)
  >=7;<1000                                               40                66                8.1 (4.9 to 13.4)
  1-6;>=1000                                               4                 3               18.3 (3.8 to 87.9)
  >=7;>=1000                                               9                 2               54.7 (11.2 to 267.3)
  Unknown                                                 13               32                 6.1 (2.8 to 13.4)
----------------------------------------------------------------------------------------------------------------------
 *From unconditional logistic regression models including the
categories of exposure listed, age, and calendar year by
comparison with non-users.
long term prescriptions and high daily doses, due to their synergistic effect,
was 60%.

Spironolactone (relative risk 9.3 (3.3 to 26.1) and verapamil (9.7 (2.6 to 36.0)) were the only two drugs associated with a risk of gynaecomastia comparable in magnitude with the one for cimetidine (table V).

TABLE V -  Relative risk of gynaecomastia associated with current use of
certain drugs
---------------------------------------------------------------------------------
                     No of cases   No of controls       Relative risk*
                      (n = 153)      (n = 1000)    (95% confidence interval)
---------------------------------------------------------------------------------
Digoxin                   11             18          2.7 (1.2 to 6.1)
Nifedipine                18             38          2.9 (1.6 to 5.3)
Spironolactone            11              6          9.3 (3.3 to 26.1)
Verapamil                  6              4          9.7 (2.6 to 36.0)
---------------------------------------------------------------------------------
 *From unconditional logistic regression models including the drugs listed,
age, and calendar year by comparison with non-users.

Discussion

These results indicate that use of cimetidine is associated with clinically important idiopathic gynaecomastia but that the use of misoprostol, omeprazole, and ranitidine is not. Clinical trials found that up to 0.8% of the subjects exposed to high doses of cimetidine developed gynaecomastia. ⁵ In our study population there were close to 2000 men who took 1000 mg a day or more; 15 went on to develop gynaecomastia. Despite the obvious differences between the two studies, especially with respect to surveillance of the subjects, estimates of risk for cimetidine are closely comparable, indicating that our source of information and study design had adequate validity to assess this outcome.

A recent publication presented a case series of endocrine adverse effects associated with omeprazole reported to the World Health Organisation Collaborating Centre for International Drug Monitoring.¹¹ The authors suggest that omeprazole causes impotence and gynaecomastia. Our study did not examine the risk of impotence. However, there is no indication in the present study that use of omeprazole is associated with a greatly increased risk for gynaecomastia.

The mechanism by which cimetidine causes gynaecomastia has been discussed by Galbraith and Michnovicz, who showed that giving cimetidine to normal volunteers decreases 2-hydroxylation of oestradiol resulting in an increase in the serum oestradiol concentration.¹⁸ In contrast, ranitidine did not affect 2-hydroxylation nor did omeprazole.¹⁹ Our results are in agreement with these biochemical findings.

We thank the staff at VAMP, particular Dr Alan D Dean and Ms Anne Costello, and the participating general practitioners for their excellent cooperation with this study. The Boston Collaborative Drug Surveillance Program is supported in part by grants from Bayer AG, Burroughs Wellcome, Ciba-Geigy Corporation, Hoechst AG, HoffmannLa Roche, R W Johnson Phamaceutical Research Institute, Marion Merrell Dow, Pfizer, Procter and Gamble, and Sanofi Winthrop Pharmaceuticals.

1. Carlson HE. Gynecomastia. N Engl J Med 1980;303:795-9. [Medline]
2. Nuttall FQ. Gynaecomastia as a physical finding in normal men. J Clin Endocrinol Metab 1979;48:338-40. [Abstract/Free Full Text]
3. Braunstein GD. Gynecomastia. N Engl J Med 1993;328:490-5. [Free Full Text]
4. Spence RW, Celestein LR. Gynaecomastia associated with cimetidine. Gut 1979;20:154-7. [Abstract/Free Full Text]
5. Jensen RT, Collen MJ, Pandol SJ, Allende HD, Raufman JP, Bissonnette BM, et al. Cimetidine - induced impotence and breast changes in patients with gastric hypersecretory states. N Engl J Med 1983;308:883-7. [Medline]
6. Delle Fave GE, Tamburrano G, De Magistris L. Gynaecomastia with cimetidine. Lancet 1977;i:1319.
7. Mignon M, Vallot T, Bonfils S. Gynaecomastia and histamine-2 antagonists. Lancet 1982;ii:499.
8. Tosti S, Cagnoli M. Painful gynaecomastia with ranitidine. Lancet 1982;ii:160.
9. Santucci L, Farroni F, Fiorucci S, Morelli A. Gynaecomastia during omeprazole therapy. N Engl J Med 1991;324:635. [Medline]
10. Convens C, Verhelst J, Mahler C. Painful gynaecomastia during omeprazole therapy. Lancet 1991;338:1153.
11. Lindquist M, Edwards IR. Endocrine effects of omeprazole. BMJ 1992;305:451-2.
12. Jick H, Jick SS, Derby LE. Validation of information recorded on general practitioner based computerised data resource in the United Kingdom. BMJ 1991;302:766-8.
13. Jick H, Terris BZ, Derby LE, Jick SS. Further validation of information recorded on a general practitioner based computerized data resource in the United Kingdom. Pharmacoepidemiology and Drug Safety 1992;1:347-9.
14. Payne CD, ed. Generalized linear interactive system. Release 3.77. Oxford, United Kingdom: Royal Statistical Society, 1986.
15. Walker AM, ed. Observation and inference: an introduction to the methods of epidemiology. Newton, MA: Epidemiology Resources Inc, 1991.
16. Rothman KJ, ed. Interactions between causes. In: Modern epidemiology. Boston: Little, Brown, 1986:311-26.
17. Hosmer DW, Lemeshow D. Confidence intervals for interaction. Epidemiology 1992;3:452-6. [Medline]
18. Galbraith RA, Michnovicz JJ. The effect of cimetidine on the oxidative metabolism of estradiol. N Engl J Med 1989;321:269-74. [Abstract]
19. Galbraith RA, Michnovicz JJ. Omeprazole fails to alter the cytochrome P450-dependent 2-hydroxylation of estradiol in male volunteers. Pharmacology 1993;47:8-12. [Medline]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    StumbleUpon    Technorati    What's this?

This article has been cited by other articles:

• Braunstein, G. D. (2007). Gynecomastia. NEJM 357: 1229-1237 [Full text]  
• Dubey, R. K., Jackson, E. K. (2001). Genome and Hormones: Gender Differences in Physiology: Invited Review: Cardiovascular protective effects of 17{beta}-estradiol metabolites. J. Appl. Physiol. 91: 1868-1883 [Abstract] [Full text]  
• Dubey, R. K., Jackson, E. K. (2001). Estrogen-induced cardiorenal protection: potential cellular, biochemical, and molecular mechanisms. Am. J. Physiol. Renal Physiol. 280: F365-F388 [Abstract] [Full text]  
• Meier, C. R, Jick, H. (1997). Omeprazole, H2 blockers, and polyarthralgia: case-control study. BMJ 315: 1283-1283 [Full text]