We discuss six themes that emerged for 1996: HIV-related disease, cardiovascular disease, anticoagulation for atrial fibrillation, efficacy of selective serotonin reuptake inhibitors, treatment of postmenopausal osteoporosis, and several topics in preventive medicine. Much of the material overlaps with that in other Updates in this series; when this overlap occurs, we summarize the results of studies and leave the detailed descriptions to the authors of other Updates.

HIV Disease

In 1996, the rate of death from HIV-related illness decreased dramatically in the United States. Most experts believe that this decrease reflects the success of changes in treatment. These advances will be discussed in more detail in the upcoming Update in Infectious Diseases.

Three clinical trials seem to have put to rest the use of zidovudine monotherapy [1-3]. The results of these studies indicate that multidrug therapy is now the standard of care. Four protease inhibitors are now available, and initial studies have shown that these drugs are associated with major improvements in laboratory outcomes [4].

The major problem now facing clinicians is the complexity of the newer drug combinations. Patients are taking several pills a day; some must be taken with food and others must be taken on an empty stomach. Resistance occurs when treatment is interrupted, and physicians must be mindful of drug interactions. Despite these barriers, however, protease inhibitors are probably the most important advance in HIV treatment in the past 5 years.

Predicting the rate of HIV disease progression has become more reliable with the introduction of testing for HIV viral load. In a multicenter AIDS cohort analysis, researchers followed patients until death or for as long as 11 years. They found that the initial HIV viral load was a better prognostic marker than CD4⁺ cell count [5, 6].

As a result of these and other studies, an expert panel designed a set of guidelines for antiretroviral therapy [7]. Although the "standard of care" evolves rapidly, there is consensus that patients with symptoms, asymptomatic patients with CD4⁺ cell counts less than 500 cells/mm³, and patients with high viral load titers-at least 30 000 copies/mL-should be treated. Appropriate treatment regimens include zidovudine or stavudine in combination with lamivudine, didanosine, or zalcitabine. Most experts would now add a third drug (either a protease inhibitor or a non-nucleoside reverse transcriptase inhibitor) when treating any patient. So-called triple therapy is particularly important for patients who are sick or who have very poor prognostic markers. The goal of therapy is to reduce HIV viral load titers at least threefold (optimally to undetectable levels).

Cardiovascular Disease

A much more common problem in clinical practice is cardiovascular disease. Several themes emerged in 1996, and most will be covered in detail in the upcoming Update in Cardiology.

A major question for acute myocardial infarction has been which treatment is better: thrombolysis or percutaneous transluminal coronary angioplasty (PTCA)? Most studies of these treatments have been done in academic medical centers. However, in a community-based retrospective comparison of 12 331 patients who had had acute myocardial infarction and were admitted to 19 Seattle hospitals, researchers found that thrombolysis and PTCA produced similar clinical results [8]. Length of stay was shorter in the PTCA group, but PTCA cost more (median cost, $12 600 for thrombolysis and $16 300 for PTCA). Stroke was slightly more common in the thrombolysis group than in the PTCA group (1.5% and 0.9% of patients, respectively). The main difference was that patients who underwent PTCA had more procedures, especially angiography and repeated PTCA.

Another controversy has focused on whether PTCA or coronary artery bypass grafting is superior treatment for multivessel coronary artery disease. A 5-year clinical trial of 1829 patients with this disease showed that the two procedures yielded similar outcomes; the one exception was the clearly greater survival rate in diabetic patients who had coronary artery bypass grafting. Patients who underwent PTCA had considerably more procedures during follow-up [9].

Another clinical issue is maintaining patency of diseased coronary arteries. Most survivors of acute myocardial infarction have total cholesterol levels less than 240 mg/dL, and some experts believe that decreasing these "average" levels might prevent future acute myocardial infarction. In a 5-year clinical trial, 4159 patients with "average" cholesterol levels were randomly assigned to receive either pravastatin (40 mg/d) or placebo after acute myocardial infarction [10]. In the pravastatin group, cholesterol levels decreased, and 10.2% of pravastatin recipients had a subsequent coronary event compared with 13.2% of placebo recipients (number needed to treat [NNT] to prevent one death in 5 years, 33). This study further supports the efficacy of the statin drugs after acute myocardial infarction [11].

Congestive heart failure is a common and increasing problem as more patients survive acute myocardial infarction. Estimates of the prevalence of congestive heart failure in developed countries range from 1% to 10%, and the 5-year mortality rate exceeds 50% [12]. Angiotensin-converting enzyme inhibitors and diuretics are the cornerstones of therapy. In 1996, a third drug showed promise: carvedilol, an - and ß-adrenergic receptor blocker. In a clinical trial of 1094 patients with congestive heart failure, the addition of carvedilol to angiotensin-converting enzyme inhibitors and diuretics reduced the mortality rate from 7.8% to 3.2% (NNT to prevent one death in 6 months, 22) [13].

Another traditional mainstay in the treatment of congestive heart failure has been digoxin. Although this drug has been used for 221 years, its efficacy has been poorly defined except in the setting of atrial fibrillation. However, a conclusion seems to have been reached. In a clinical trial of 7788 patients who had congestive heart failure and were in sinus rhythm, researchers found that after 3 years, the mortality rate in the digoxin and placebo groups was equal [14]. However, patients receiving digoxin had fewer hospitalizations for congestive heart failure (26.8% compared with 34.7%; NNT to prevent one hospitalization in 3 years, 13).

Diuretics Were Effective in Elderly Diabetic Patients with Hypertension

Curb JD, Pressel SL, Cutler JA, Savage PJ, Applegate WB, Black H, et al. Effect of diuretic-based antihypertensive treatment on cardiovascular disease risk in older diabetic patients with isolated systolic hypertension. Systolic Hypertension in the Elderly Program Cooperative Research Group. JAMA. 1996; 276:1886-92.

The use of thiazide diuretics to decrease blood pressure in elderly diabetic patients has been debated; the adverse effects of these drugs include hyperglycemia, hypercholesterolemia, and postural hypotension. Curb and colleagues addressed the use of diuretics in a study of systolic hypertension in the elderly.

From their large database, the researchers selected the 583 patients with diabetes and isolated systolic hypertension. Mean patient age was 71 years, and 58% of the patients were women. As part of a larger study, patients had been randomly assigned to receive chlorthalidone, 12.5 mg/d, or placebo. Atenolol and reserpine were added, if necessary, to control hypertension. Outcome measures were major coronary artery events, strokes, and all-cause mortality.

The study results are summarized in Table 1. The relative risk reduction among the treated diabetic patients was about the same as that among hypertensive patients without diabetes who used the drug. Because adverse cardiovascular outcomes were about twice as common among diabetic patients, however, the absolute benefit of treating diabetic patients was greater.

Calcium Channel Blocker Was Associated with Poor Outcomes

Borhani NO, Mercuri M, Borhani PA, Buckalew VM, Canossa-Terris M, Carr AA, et al. Final outcome results of the Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS). A randomized controlled trial. JAMA. 1996; 276:785-91.

Another active issue in hypertension is the safety of calcium channel blockers. Several observational studies have suggested the possibility of harm associated with use of short-acting dihydropyridine calcium blockers (such as nifedipine, nicardipine, and isradipine) in hypertensive patients.

This 3-year clinical trial measured progression of atherosclerosis in 883 hypertensive patients with carotid atherosclerosis. The patients were randomly assigned to receive hydrochlorothiazide, 12.5 to 25 mg/d, or isradipine, 2.5 to 5 mg/d. The major outcome was progression of carotid artery stenosis; secondary outcomes were development of other vascular symptoms.

Progression of thickness of the carotid artery wall did not differ between groups. Angina developed in 0.7% of patients in the hydrochlorothiazide group and 2.5% of patients in the isradipine group (P = 0.03). Vascular events and procedures occurred in 7.48% of hydrochlorothiazide recipients compared with 12.21% of isradipine recipients (P = 0.02).

These results support evidence from observational studies showing that short-acting calcium channel blockers may produce worse outcomes than do other therapies. The data are worrisome because calcium channel blockers have become the most commonly used drugs for hypertension.

Other Studies

A retrospective review of 5332 New Jersey Medicare claims from 1987 to 1992 revealed another unwelcome trend: underuse of ß-blockers after acute myocardial infarction among elderly persons and the potentially avoidable increase in morbidity and mortality [15]. The researchers identified 3737 patients who had had acute myocardial infarction and were clinically eligible to receive ß-blocker therapy. Only 21% received a ß-blocker, whereas 54% received a calcium channel blocker. The relative risk for death was 0.57 (95% CI, 0.67 to 0.90) for patients receiving ß-blockers compared with those receiving calcium channel blockers.

A university-based cross-sectional study of 2601 elderly hypertensive patients recruited for studies found that calcium channel blockers were used more often than other antihypertensive agents [16]. Among these patients, 23.9% were taking calcium channel blockers, 17.9% were taking diuretics, 17.5% were taking angiotensin-converting enzyme inhibitors, and 10.0% were taking ß-blockers. Of the 24% receiving combination therapy, a diuretic combined with a calcium channel blocker was the most common regimen.

Anticoagulation and Atrial Fibrillation

The accumulated evidence on anticoagulation in patients with atrial fibrillation now provides a basis for treatment guidelines. We know that treatment generally offers more benefit than harm; we are now coming to some consensus on the optimal intensity of anticoagulation.

Low-Intensity, Fixed-Dose Warfarin Plus Aspirin Was Inadequate

Adjusted-dose warfarin versus low-intensity, fixed-dose warfarin plus aspirin for high-risk patients with atrial fibrillation: stroke prevention in atrial fibrillation III randomised clinical trial. Stroke Prevention in Atrial Fibrillation Investigators. Lancet. 1996; 348:633-8.

In patients with atrial fibrillation, an international normalized ratio (INR) of 2.0 to 4.0 is generally considered optimal for preventing strokes while keeping complications to a minimum. However, it is difficult to maintain this range in elderly patients who may be taking other drugs. Therefore, these investigators tested whether aspirin plus low-intensity and fixed-dose warfarin would be effective and safe in preventing stroke.

A total of 1044 patients (39% women; mean age, 72 years) with atrial fibrillation and at least one other risk factor for thromboembolism (congestive heart failure, previous thromboembolism, hypertension, or being a woman older than 75 years of age) were randomly assigned to aspirin (325 mg/d) plus low-intensity, fixed-dose warfarin (INR of 1.2 to 1.5 used for initial dose adjustment) or standard adjusted-dose warfarin (INR, 2.0 to 3.0). The investigators hypothesized that the drugs would have equal efficacy, and they planned a 2.5-year follow-up.

The study was stopped after 1 year because the rate of stroke or systemic embolization was much greater in the group treated with aspirin and low-intensity warfarin (7.9% per year compared with 1.9% per year; P < 0.001). Major bleeding rates did not differ between groups.

Risk Factors for Stroke Were Identified

Hylek EM, Skates SJ, Sheehan MA, Singer DE. An analysis of the lowest effective intensity of prophylactic anticoagulation for patients with nonrheumatic atrial fibrillation. N Engl J Med. 1996; 335:540-6.

At about the same time the results from the Stroke Prevention in Atrial Fibrillation study were published, Hylek and colleagues reported on their attempt to identify risk factors for ischemic stroke in persons receiving warfarin for atrial fibrillation. In this case–control study, case-patients were 74 consecutive patients receiving warfarin for nonrheumatic atrial fibrillation who were admitted to a hospital between 1989 and 1994 with ischemic stroke, and controls were 222 randomly selected patients with atrial fibrillation who were enrolled in an anticoagulation clinic and had not had a stroke.

Using a reference INR of 2.0, the researchers found that as the INR decreased, the adjusted odds ratio for stroke sharply increased (Table 2). Other risks were a previous stroke, smoking, diabetes, and hypertension.

Consensus on Use of Anticoagulation for Atrial Fibrillation

As a result of these studies, the following consensus was reached on the use of anticoagulation for atrial fibrillation. In patients at high risk for stroke, a combination of aspirin and low-intensity warfarin provides inadequate protection. The lowest effective intensity of warfarin for atrial fibrillation seems to be an INR of 2.0.

Selective Serotonin Reuptake Inhibitors

Selective serotonin reuptake inhibitors (commonly referred to as SSRIs) are widely used in both primary care and psychiatry. Until now, however, little evidence was available on their efficacy relative to the older tricyclic antidepressant agents.

Fluoxetine Was Effective for Depression

Simon GE, VonKorff M, Heiligenstein JH, Revicki DA, Grothaus L, Katon L, et al. Initial antidepressant choice in primary care. Effectiveness and cost of fluoxetine vs tricyclic antidepressants. JAMA. 1996; 275:1897-902.

The use of selective serotonin reuptake inhibitors has increased dramatically since fluoxetine was introduced in 1988. This increase is probably related to 1) the ease with which these drugs can be used and 2) the lower incidence of adverse effects and low risk for death from overdoses of these drugs compared with tricyclic antidepressant agents. Selective serotonin reuptake inhibitors cost more than tricyclic antidepressant agents, however, and available information from clinical trials suggests that selective serotonin reuptake inhibitors are no more effective than the tricyclic agents.

Simon and colleagues compared the clinical, quality-of-life, and economic outcomes of a selective serotonin reuptake inhibitor (fluoxetine) and two tricyclic antidepressant agents (imipramine and desipramine) for initial treatment of depression in community-based primary care clinics. They enrolled 536 adults (72% were women) aged 18 to 90 years (mean age, 41 years) who had a depressive disorder (major depression, 67%; dysthymia, 7%; minor depression without dysthymia, 26%) from 1992 to 1994. Primary care physicians made all decisions about initial dose and subsequent management.

No differences were seen in depression rating scales or quality-of-life scores. Clinical remission rates were 49% for fluoxetine, 55% for desipramine, and 48% for imipramine (P > 0.05). Patients in the fluoxetine group were more likely to continue therapy with the antidepressant agent and to receive adequate treatment for at least 90 days (61% of the fluoxetine group compared with 48% of the desipramine group and 49% of the imipramine group). The proportion of patients who discontinued therapy because of adverse effects was lower in the fluoxetine group than in either tricyclic group (9% compared with 27% and 28%, respectively). Despite the higher cost of fluoxetine, total costs of patient care did not differ because fluoxetine recipients had fewer outpatient visits and fewer hospitalizations. Overall direct cost for 6 months was about $2000 in all groups.

Compared with the tricyclic agents, fluoxetine for initial treatment of depression in primary care results in similar relief of symptoms of depression, quality of life, and total costs of care and is associated with fewer adverse drug effects and changes in medication.

Sertraline Was Effective for Dysthymia

Thase ME, Fava M, Halbreich U, Kocsis JH, Koran L, Davidson J, et al. A placebo-controlled, randomized clinical trial comparing sertraline and imipramine for the treatment of dysthymia. Arch Gen Psychiatry. 1996; 53:777-84.

Dysthymia is a chronic mood disorder that affects 3% to 6% of adults. Dysthymic patients often report that they have felt depressed throughout life, and a specific time of onset usually cannot be identified. Symptoms of dysthymia are milder than those of major depression, but dysthymia is associated with a marked increase in use of medical services and sedatives. The efficacy of treatment has been poorly studied.

Thase and colleagues compared the safety and efficacy of sertraline, imipramine, and placebo in patients with primary dysthymia. The 12-week clinical trial enrolled 416 patients (65% were women) with long-standing dysthymia (mean duration, 30 years). The patients were randomly assigned to receive sertraline (50 mg/d), imipramine (50 mg/d), or placebo. If no clinical response occurred, the dose was increased by 50 mg weekly to a maximum of 200 mg for sertraline or 300 mg for imipramine.

Both drugs produced improvement compared with placebo. The median dose at the time of clinical response was 50 mg for sertraline and 150 mg for imipramine. The trial was completed by 84% of the sertraline group, 67% of the imipramine group, and 76% of the placebo group. The proportion of patients who discontinued therapy because of adverse effects was 6% for sertraline, 18% for imipramine, and 4% for placebo.

Summary

Evidence of the efficacy of selective serotonin reuptake inhibitors is building. For initial treatment of depression, fluoxetine and tricyclic agents have about equal costs, but fluoxetine causes fewer adverse effects. This latter issue is important for patients with major depression because they often discontinue therapy even when adverse effects are mild.

Treatment of Osteoporosis

Osteoporosis affects 30% to 50% of postmenopausal women in the United States and is associated with 1.5 million fractures annually. Alendronate represents a major advance in the treatment of osteoporosis. Slow-release sodium fluoride, when approved for clinical use, may also have a role.

Alendronate Reduced Vertebral Fractures

Liberman UA, Weiss SR, Broll J, Minne HW, Quan H, Bell NH, et al. Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. The Alendronate Phase III Osteoporosis Treatment Study Group. N Engl J Med. 1995; 333:1437-43.

Bisphosphonates inhibit bone resorption and normalize bone turnover rates. Alendronate is more potent than other drugs of this type and, unlike etidronate and pamidronate, does not impair bone mineralization. The major clinical questions concern the effectiveness of alendronate in reversing osteoporosis and the safety of the drug.

In an international clinical trial, Liberman and colleagues enrolled 994 postmenopausal women aged 45 to 80 years (mean age, 64 years; 87% were white persons). All patients had decreased bone mineral density of the lumbar spine, which was defined as density at least 2.5 SDs below the mean density in premenopausal white women; 20% had previously had vertebral fractures. The patients were randomly assigned to receive placebo or alendronate (5, 10, or 20 mg/d) for 3 years. All received 500 mg of calcium carbonate per day. Volunteers were excluded if they had a history of hip fracture; had bone loss due to an identifiable cause; had metabolic bone disease; or used estrogen, calcitonin, or fluoride.

All alendronate doses produced increases in bone density, whereas bone density decreased in the placebo group. The 10-mg dose was as effective as the 20-mg dose and was more effective than the 5-mg dose (Table 3). Alendronate was associated with a lower incidence of new vertebral fractures (3.2% compared with 6.2%) and a risk reduction of 48% (CI, 5% to 72%; NNT to prevent one vertebral fracture over 3 years, 33). Women with and those without previous fractures were protected. A trend toward fewer nonvertebral fractures was seen (relative risk, 0.79 [CI, 0.52 to 1.22]; NNT, 48). Alendronate recipients lost less height than placebo recipients (3.0 mm compared with 4.6 mm). No increased incidence of esophageal symptoms was noted. However, an international surveillance study conducted by the manufacturer of alendronate (Merck & Co.) found the incidence of esophagitis to be about 4.2 per 10 000 users [17].

These results were recently confirmed in a U.S. randomized, controlled trial involving 2027 postmenopausal women with previous vertebral fractures [18]. After 3 years, 8.0% of women in the alendronate group and 15.0% of women in the placebo group had new vertebral fractures (relative risk, 0.53 [CI, 0.41 to 0.68]). Alendronate also prevented fractures of the hip and wrist.

Daily use of alendronate (10 mg) increases bone mineral density and prevents fractures in women with postmenopausal osteoporosis. Although clinical esophagitis is unusual, clinicians should warn patients to ingest alendronate with a liberal amount of water and to avoid lying down soon after ingestion.

Fluoride Prevented Initial Vertebral Fractures

Pak CY, Sakhaee K, Adams-Huet B, Piziak V, Peterson RD, Poindexter JR. Treatment of postmenopausal osteoporosis with slow-release sodium fluoride. Final report of a randomized controlled trial. Ann Intern Med. 1995; 123:401-8.

In the early 1980s, sodium fluoride was seen as a major advance against osteoporosis because it stimulated osteoblastic proliferation and increased bone mineral density in research participants. Clinical use, however, showed that sodium fluoride actually increased the fracture rate in long bones and did not prevent vertebral fractures. Therefore, its use was abandoned around 1990. Pak and colleagues hypothesized that fluoride given cyclically and in lower doses, combined with calcium citrate, might be clinically beneficial.

The clinical trial enrolled 110 women (mean age, 67 years) who had radiologic osteopenia and had had at least one nontraumatic fracture. All patients were given calcium citrate (800 mg/d) and were then randomly assigned to receive placebo or slow-release sodium fluoride (25 mg twice daily) in four cycles (12 months receiving fluoride followed by 2 months not receiving it).

Women who received fluoride had increased bone mass in the lumbar spine and femoral neck but not in the radial shaft. After 3.5 years, women receiving fluoride had fewer fractures overall (0.064 per patient-year compared with 0.205 per patient-year in the placebo group; NNT, 7). Baseline lumbar bone mass was a major determinant of the efficacy of fluoride. Among women with mild to moderate bone loss, fluoride resulted in a 96.4% fracture-free rate compared with a rate of 67.9% with placebo. Fluoride was only marginally more effective than placebo among women with severe bone loss (fracture-free rates, 73% and 70%, respectively).

It therefore seems that cyclic administration of slow-release sodium fluoride (when approved for clinical use) may be most effective in preventing initial vertebral fractures in women with mild to moderate bone loss. Thus, sodium fluoride should be considered a preventive therapy for women who have higher risk for severe osteoporosis (Table 4).

Preventive Medicine

In 1996, four important advances occurred in the field of clinical prevention. The first advance is the discovery of the futility of obtaining vaginal Papanicolaou smears after hysterectomy for benign disease. The second is the finding that pelvic inflammatory disease can be prevented by selectively testing for chlamydial infection. The third is the discovery of the association of pulmonary hypertension with appetite suppressants. Finally, major trials reported results on the use of antioxidants to prevent important causes of death.

Papanicolaou Smears after Hysterectomy for Benign Disease Were Not Useful

Pearce KF, Haefner HK, Sarwar SF, Nolan TE. Cytopathological findings on vaginal Papanicolaou smears after hysterectomy for benign gynecologic disease. N Engl J Med. 1996; 335:1559-62.

A controversial issue is whether Papanicolaou smears should be obtained for screening purposes in women who have had hysterectomy for benign diseases. Gynecologists have generally argued in favor of screening and cite vaginal cancer as the target of screening. Others have argued that vaginal cancer is too rare for screening to be effective. Pearce and colleagues sought to determine whether periodic, routine Papanicolaou smears obtained from the vaginal apex are beneficial in such women.

The researchers retrospectively analyzed all 9610 Papanicolaou smears obtained after hysterectomy for benign disease in 5682 women with low socioeconomic status at Charity Hospital in New Orleans from 1992 to 1994. The mean patient age was 52 years, and 83% of the patients were black. These women were at relatively high risk for neoplasia.

Seventy-nine women had 104 abnormal smears (1.1% of 9610 smears). Of these smears, 52 (0.5%) were categorized as containing atypical squamous cells of undetermined significance; 44 (0.5%) showed low-grade squamous intraepithelial lesions; 6 (0.1%) showed high-grade squamous intraepithelial lesions; and 2 (0.02%) showed squamous-cell carcinoma. The positive predictive value of an abnormal Papanicolaou smear for detecting vaginal cancer in these women approached 0% (CI, 0% to 33%).

The evidence is now overwhelming: Obtaining Papanicolaou smears for screening women who have had hysterectomy for benign disease is not a sound preventive practice. Moreover, no data support the use of the bimanual examination as a screening test for ovarian cancer. This topic will be discussed in more detail in the upcoming Update in Women's Health.

Screening for Chlamydial Infection Reduced Incidence of Pelvic Inflammatory Disease

Scholes D, Stergachis A, Heidrich FE, Andrilla H, Holmes KK, Stamm WE. Prevention of pelvic inflammatory disease by screening for cervical chlamydial infection. N Engl J Med. 1996; 334:1362-6.

Growing evidence indicates that screening high-risk women can prevent pelvic inflammatory disease. Through questionnaires completed by 20 836 unmarried women in a health maintenance organization, Scholes and colleagues identified 2607 women at high risk for chlamydial infection. These researchers had previously determined risk factors for pelvic inflammatory disease [19] and developed a scoring system in which a score of 3 or higher indicated "high risk" (Table 5). The high-risk women were randomly assigned to have screening for chlamydial infection and subsequent treatment as appropriate or to receive usual care. Screening was done through pelvic examination with cervical chlamydial enzyme-linked immunosorbent assay and culture.

After 1 year, the rate of pelvic inflammatory disease was 96 per 10 000 women-years in the screened group and 216 per 10 000 women-years in the standard care group (relative risk reduction, 56%; number needed to screen annually to prevent one case of pelvic inflammatory disease, 83).

Screening women who are at high risk for chlamydial infection and treating those who have positive test results seem effective in reducing the rate of pelvic inflammatory disease, a condition associated with considerable morbidity. In addition, urinary polymerase chain reaction testing should soon be available for both Chlamydia trachomatis and Neisseria gonorrhoeae, making screening much less cumbersome for patients.

Anorectic Drugs Were Associated with Pulmonary Hypertension

Abenhaim L, Moride Y, Brenot F, Rich S, Benichou J, Kurz X, et al. Appetite-suppressant drugs and the risk of primary pulmonary hypertension. International Primary Pulmonary Hypertension Study Group. N Engl J Med. 1996; 335:609-16.

The release and marketing of dexfenfluramine have been associated with a marked increase in demand for drug therapy for obesity. In the 1960s, European researchers noted a higher incidence of pulmonary hypertension among users of appetite-suppressing drugs, but the association is difficult to prove because primary pulmonary hypertension is very rare (estimated incidence, about one case per 500 000 persons).

In a multinational case–control study, researchers at European referral centers identified 95 patients (69% women; mean age, 45 years) with primary pulmonary hypertension and 355 age- and sex-matched controls who were selected from general practices. All participants were interviewed and were asked to report use of any of 35 medications. (Pills and packaging material were shown at the interviews.) Medications included antihypertensive agents, oral contraceptives, thyroid extracts, and appetite suppressants.

Use of appetite suppressants (>90% were derivatives of fenfluramine) was associated with an increased risk for primary pulmonary hypertension (odds ratio, 6.3 [CI, 3.0 to 13.2]; NNT, approximately 94 000). The highest risk was seen with recent use (odds ratio, 10.1 [CI, 3.4 to 29.9]) and with use for more than 3 months (odds ratio, 23.1 [CI, 6.9 to 77.7]). Other factors associated with an increased risk for primary pulmonary hypertension were a positive family history, HIV infection, cirrhosis, and use of cocaine or injection drugs.

Use of appetite suppressants is associated with an increased risk for primary pulmonary hypertension. At the same time, this condition is rare, as reflected by the large NNT to "cause" a case of PPH. The use of appetite suppressants is controversial. Recent U.S. guidelines state that data are inadequate to support the routine use of these drugs [20]. The average effects of appetite suppressants are a 2- to 10-kg weight loss, which is modest in very obese patients whose health would probably benefit from extensive weight loss. Moreover, weight loss tends to reach a plateau at 6 months of therapy, and weight is commonly regained if patients continue to receive the drugs for a longer period. Guidelines recommend that a 6-month course of therapy be the maximum.

ß-Carotene Was Not Protective

Hennekens CH, Buring JE, Manson JE, Stampfer M, Rosner B, Cook NR, et al. Lack of effect of long-term supplementation with ß-carotene on the incidence of malignant neoplasms and cardiovascular disease. N Engl J Med. 1996; 334:1145-9.

Antioxidants have gained popularity as preventive therapy, and their supporters cite observational studies reporting that persons who consume more ß-carotene in their diets have less malignant and cardiovascular disease. Hennekens and colleagues tested those claims.

The Physicians' Health Study is a U.S. clinical trial in which 22 071 male physicians aged 40 to 84 years were given ß-carotene (50 mg every other day) or placebo and were followed for 12 years. After 12 years, ß-carotene did not decrease the incidence of cancer (incidence of 11.5% in the ß-carotene group compared with 11.7% in the placebo group), myocardial infarction (4.2% compared with 4.4%), or stroke (3.3% compared with 3.5%). In addition, ß-carotene had no effect on rates of death from malignant neoplasm (3.5% compared with 3.4%) or cardiovascular disease (3.0% compared with 2.8%). Hennekens and colleagues caution that although their data did not show a decrease in the incidence of disease associated with use of ß-carotene, the power of the study was insufficient to disprove the hypothesis.

A combination of ß-carotene and vitamin A had harmful effects in a randomized, placebo-controlled trial of 18 314 patients at high risk for lung cancer and cardiovascular disease [21]. After an average of 4 years of follow-up, the risk for lung cancer (relative risk, 1.17 [CI, 1.03 to 1.33]) was increased in the group assigned to ß-carotene and vitamin A.

Dr. Larson: University of Washington Medical Center, Room AA 316, Box 356330, Seattle, WA 98195-4021.

Dr. Roberts (Series Editor): York Health System, 755 South Pleasant Avenue, York, PA 17313.