Risk of Pulmonary Hypertension in patients exposed to derivatives of Fenfluramine in appetite suppressants
Appraised by: Debra Bynum, M.D.
Date: July 18, 1997
Clinical Scenario: 55 year old female who has severely limiting osteoarthritis of both knees and in need of bilateral knee replacement. She weighs 425 lbs and cannot undergo surgery until she loses weight. "But Doctor" she asks, "how can I lose weight if I can’t walk". Dietary changes have failed and she is interested in a prescription for the new "fat pill"...
Clinical Bottom Lines:
1. Of the 95 total patients identified with Primary Pulmonary Hypertension (PPH), 31.6% had a history of appetite suppressant use; Only 7.3% of the 355 controls (no PPH) had had exposure to such agents for an overall Odds Ratio of 6.3 with any anorexic-drug use.
2. Risk appeared to be proportional to duration of use of these agents with 19% of cases compared to 1.4% of controls having an exposure time of over 3 months (OR 23.1).
3. The majority of appetite suppressants involved Fenfluramine, Dexfenfluramine, or some combination including these agents (only 2 of the 30 case exposures and 3 of the 26 control exposures did not involve these agents).
4. Depite differences between case and control groups, analysis of potential confounding variables revealed no significant impact on the results.
The Evidence: Case-Control Study involving 95 patients identified with PPH and 355 controls matched for sex, age and location.
_____________________________________________________________________________________
| Case Patients (n=95) | Controls (n=355) | Adjusted Odds ratio | |
| Definite use of appetite suppresants | 30 (31.6%) | 26 (7.3%) | 6.3 (3.0-13.2, 95%) |
| Duration of Use | |||
| <3 months | 7 (7.4%) | 12 (3.4%) | 1.8 (0.5-5.7, 95% CI) |
| >3 months | 18 (19.0%) | 5 (1.4%) | 23.1 (6.9-77.7, 95% CI) |
| Indeterminate | 5 (5.3%) | 9 (2.5%) | 2.6 (0.5-12.6, 95% CI) |
Comments:
1. The Case-Control design is necessary for a study involving an outcome that is harmful and rare, but unfortunately is retrospective and open for recall bias, surveillance bias, and interviewer bias as well as problems with potential confounding evidence (despite claims by this study that all of these potential problems could be dismissed).
2. Unable to determine causality (increased odds that someone with PPH had exposure to anorexic agents may not necessarily mean that someone taking anorexic agents is more likely to develop PPH).
3. Despite claims that analysis of multiple variables thought to be risk factors for PPH demonstrated no significant impact on the OR, there is significant potential for confounding evidence. The cases and controls were significantly different in regards to BMI, weight loss behavior, systemic hypertension, use of cocaine or IVDA, and smoking with a higher incidence for each variable in the case group!
Source: Abenhaim L, et al. Appetite-Suppressant Drugs and the Risk of PPH. NEJM.1996;335:609-16.