Comparison of Survival in Patients With PulmonaryHypertension Associated With Fenfluramine toPatients With Primary Pulmonary Hypertension

Stuart Rich, MD, Alicia Shillington, RN, MPH, and Vallerie McLaughlin, MD

To test whether the clinical presentation and progno-sis of fenfluramine-induced pulmonary hypertension(PH) differs from primary PH (PPH), we compared theclinical profile and outcome of 10 patients with fen-fluramine-induced PH with that of 70 patients withPPH referred to our center over the same time frameand treated identically. Patients with diet pill PH weresimilar to those with PPH with respect to hemodynam-ics. However, patients with fenfluramine-induced PHhad poorer survival: 1-year survival 50% versus 88%,and 3-year survival 17% versus 60%. �2003 byExcerpta Medica, Inc.

(Am J Cardiol 2003;92:1366–1368)

A relation between the use of diet pills and pulmo-nary hypertension (PH) was first observed in the

late 1960s when a 20-fold increase in unexplained PHwas reported in Europe after the introduction of theappetite suppressant aminorex.1 Recently, a causalrelation between the use of fenfluramine diet pills andPH in Europe and the United States was established inthe 1990s, resulting in their withdrawal from the mar-ketplace.2 Patients who present with PH after expo-sure to fenfluramines appear to be comparable topatients with primary PH (PPH) with respect to symp-toms, clinical features, and hemodynamics.2 What isunknown, however, is whether PH induced by fenflu-ramines has a different prognosis. To address this, wecompared the survival of patients with PPH (no pre-vious exposure to fenfluramines) to patients with PHexposed to fenfluramines who were referred to ourinstitution.

• • •The diagnosis of PPH was established according to

the criteria of the National Institute of Health Registryon PPH.3 Clinical data, the results of exercise testing(Naughton-Balke protocol), and cardiac catheteriza-tions that were performed for clinical assessment wereextracted from patients’ medical records. Our institutedeveloped a customized patient database to collectspecific variables in every patient referred to us andsubsequently treated with intravenous epoprostenol.Patient data analyzed in this report come from thedatabase. Use of the registry was approved by theinstitutional review board.

All patients underwent an exercise assessment us-ing a treadmill test after a Naughton-Balke protocol.After that, cardiac catheterization was performed andbaseline measurements of hemodynamic parametersobtained; response to a challenge of intravenous aden-osine to establish the magnitude of pulmonary vaso-dilator reserve was determined. Patients whose meanpulmonary artery pressure decreased to �40 mm Hgand pulmonary vascular resistance decreased by�50% were offered a clinical trial of calcium channelblockers. Patients who were unable to respond to thesedrugs were offered intravenous epoprostenol.

Epoprostenol therapy was initiated after insertionof a Hickman catheter into a subclavian or jugularvein and administered continuously with the use of aportable infusion pump. Epoprostenol was begun at adose of 2 ng/kg/min and gradually increased to amaximum tolerated dose during the initial hospitaliza-tion. Subsequently, after patient discharge, the dosewas gradually up-titrated according to patients’ symp-toms of PH and the adverse effects of epoprostenol. Inaddition to epoprostenol, all patients without contra-indications were given warfarin anticoagulation. Di-uretic drugs were freely prescribed and adjusted.Digoxin was prescribed to patients whose cardiacoutput was reduced. Patients with an arterial oxygensaturation at rest of �90% were prescribed continuousnasal oxygen and those with hypoxemia with exercisewere recommended to wear nasal oxygen during ac-tivities. Patients were followed up on an outpatientbasis and seen a minimum of every 6 months. Vitalstatus was confirmed as of July 1, 2002.

From December 1996 to July 2000, 10 patientswith PH induced with fenfluramines were referred fortreatment. To adjust for potential referral biases andthe effect of treatment on outcome, we elected tocompare these patients to those with PPH who werereferred to our center over the same time frame, andwho were treated with similar therapies.

In addition, to adjust for potential differences indisease severity, we performed a subset analysis of thepatients with fenfluramine-induced PH versus thosewith PPH who were matched for their baseline func-tional class and the level of pulmonary vascular resis-tance at cardiac catheterization.

Demographic, clinical, and hemodynamic vari-ables from the database were analyzed. The date ofinitial catheterization was used as the index date fordetermining survival, which was calculated usingKaplan-Meier estimates. Patients were censored fromthe study if they underwent lung transplantation.Baseline age, treadmill time, and hemodynamics—including right atrial pressure, pulmonary artery mean

From the Rush Heart Institute, Center for Pulmonary Heart Disease,Rush-Presbyterian, St. Luke’s Medical Center, Chicago, Illinois. Dr.Rich’s address is: Center for Pulmonary Heart Disease, Rush-Presbyte-rian, St. Luke’s Medical Center, 1725 West Harrison Street, Suite020, Chicago, Illinois 60612. E-mail: Stuart_Rich@rush.edu. Manu-script received February 17, 2003; revised manuscript received andaccepted August 8, 2003.

1366 ©2003 by Excerpta Medica, Inc. All rights reserved. 0002-9149/03/$–see front matterThe American Journal of Cardiology Vol. 92 December 1, 2003 doi:10.1016/j.amjcard.2003.08.034

pressure, and cardiac index—were compared usingStudent’s t test. Gender and baseline functional classwere compared using the chi-square test. Pulmonaryvascular resistance was compared using a Mann-Whit-ney U test because data were skewed to the right.Kaplan-Meier analysis was constructed to analyze theeffect of the underlying cause as a predictor of sur-vival.

Baseline demographics of patients with diet pill-in-duced PH were similar to those with PPH (Table 1).They had similar mean ages (47 � 10 vs 43 � 15 years)and severity of clinical symptoms based on functionalclass. Of those with diet pill-related PH, 40% were inclass III and 60% were in class IV versus 45.7% in classIII and 54.3% in class IV for those with PPH. They alsohad comparable exercise tolerance, with diet pill-relatedpatients with PH walking on the treadmill for a mean of132 � 263 versus 188 � 150 seconds for those withPPH. All patients with diet pill-induced PH were womenversus 73% of those with PPH.

Baseline hemodynamics of these 2groups are listed in Table 1. Therewere no significant differences notedwith respect to the level of right atrialpressure, pulmonary artery pressure,cardiac index, or pulmonary vascularresistance. In addition, both groups hada similar magnitude of decrease in pul-monary vascular resistance after theadenosine challenge. However, not asingle patient with diet pill-inducedPH had an acute response to adenosinethat would warrant a clinical trial withcalcium antagonists.

Because of the severity of theirillness, all patients with fenflura-mine-induced PH were treated withintravenous epoprostenol therapyand followed prospectively for 60months. When compared with pa-

tients with PPH who were also treated with epopro-stenol, patients with PPH had a markedly better 3-yearsurvival (median 49 months) versus those with fenflu-ramine-induced PH (median 14 months, p �0.05;Figure 1).

To address whether patients with diet pill-inducedPH had more advanced disease than those with spo-radic PPH, which may have accounted for their poorerprognosis, we compared patients with diet pill-in-duced PH with a subset of patients with sporadic PPHmatched for the level of pulmonary vascular resistanceand functional class. Even with this correction, wefound a dramatic difference between 5-year survivalin the 2 groups. Those with diet pill-induced PH hadan 18% 5-year survival versus 39% in those who werenot exposed to diet pills (Table 2).

• • •The pathogenic mechanisms of PH associated with

fenfluramine derivatives remain unknown. Several po-tential pathways have been suggested including alter-ations in serotonin and direct pulmonary vasoconstric-tion.4,5 Because only a few patients exposed tofenfluramines develop PH, it is likely that some typeof genetic susceptibility is present for fenfluramine toexhibit its deleterious effects. This susceptibility hasnever been specifically characterized, but it is unlikelyto come from the BMPR-2 mutation that is attributedto familial PPH.6

Brenot et al7 reported a high incidence of fenflu-ramine-induced PH with a prognosis similar to that inpatients with PPH in an era before epoprostenol wasavailable as a treatment. Epoprostenol has clearlybeen shown to improve survival substantially in pa-tients with PPH.8 The fact that patients with dietpill-induced PH have poor recovery, even whentreated with epoprostenol on presentation, suggeststhat fenfluramines induce a pathogenic mechanismthat is either resistant to intravenous epoprostenol orresults from a novel pathway for which the establishedtherapies of PH are ineffective.

The rapidly progressive nature of the death of thesepatients is alarming. Because epoprostenol remains

TABLE 1 Baseline Demographics and Hemodynamics

PH

Diet Pill(n � 10)

Primary(n � 70)

Mean age (yrs) 47 � 10 43 � 16Men 0 19 (27%)Women 10 (100%) 51 (73%)NYHA functional class at initial evaluation

III 4 (40%) 32 (46%)IV 6 (60%) 38 (54%)

Mean treadmill time at initial evaluation (min) 132 � 263 188 � 150Right atrial pressure (mm Hg) 12 � 5 14 � 7Mean pulmonary artery pressure (mm Hg) 57 � 10 59 � 12Cardiac index (L/min/min) 2 � 1 2 � 0.5Pulmonary vascular resistance (Wood U) 19 � 10 18 � 9Pulmonary vascular resistance with adenosine challenge 16 � 10 15 � 8

Values are expressed as mean � SD or number (%).NYHA � New York Heart Association.

FIGURE 1. Survival of patients with PPH (n � 70) are comparedwith patients with diet pill-induced PH (n � 10). All patientswere treated with intravenous epoprostenol. CI � confidenceinterval.

BRIEF REPORTS 1367

the most effective medical therapy for this disease,lung transplantation would be the only other option forthese patients. Given the malignant clinical course thatthese patients appear to follow, referral for lung trans-plantation may need to be undertaken as soon as thediagnosis is established.

1. Gurtner HP. Aminorex and pulmonary hypertension. Cor Vasa 1985;27:160–171.2. Abenhaim L, Moride Y, Brenot F, Rich S, Benichou J, Kurz X, HigenbottamT, Oakley C, Wouters E, Aubier M, Simonneau G, Begaud B. Appetite-suppres-sant drugs and the risk of primary pulmonary hypertension. N Engl J Med1996;335:609–616.3. Rich S, Dantzker R, Ayres S, Bergofsy EH, Brundage BH, Detre KM, FishmanAP, Goldring RM, Groves BM, Koerner SK, et al. Primary pulmonary hyper-tension: a national prospective study. Ann Intern Med 1987;107:216–223.4. Rothman RB, Ayestas MA, Dersch CM, Baumann MH. Aminorex, fenflura-mine, and chlorphentermine are serotonin transporter substrates. Implications forprimary pulmonary hypertension. Circulation 1999;100:869–875.5. Weir EK, Reeve HL, Huang JM, Michelakis E, Nelson DP, Hampl V, ArcherSL. Anorexic agents aminorex, fenfluramine, and dexfenfluramine inhibit potas-sium current in rat pulmonary vascular smooth muscle and cause pulmonaryvasoconstriction. Circulation 1996;94:2216–2220.6. Newman JH, Wheeler L, Lane KB, Loyd E, Gaddipati R, Phillips JA III, LoydJE. Mutation in the gene for bone morphogenetic protein receptor II as a cause ofprimary pulmonary hypertension in a large kindred. N Engl J Med 2001;345:319–324.7. Brenot F, Herve P, Pettiprez P, Parent F, Duroux P, Simonneau G. Primarypulmonary hypertension and fenfluramine use. Br Heart J 1993;70:537–541.8. McLaughlin VV, Shillington A, Rich S. Survival in primary pulmonaryhypertension: the impact of epoprostenol therapy. Circulation 2002;406:1477–1482.

Usefulness of B-Type Natriuretic Peptide as aNoninvasive Screening Tool for Cardiac Allograft

Pathology in Pediatric Heart Transplant Recipients

Ilene Claudius, MD, Yueh-Tze Lan, MD, Ruey-Kang Chang, MD, MPH,Glenn T. Wetzel, MD, PhD, and Juan Alejos, MD

We examined the utility of B-type natriuretic peptide(BNP) in the evaluation of pediatric orthotopic hearttransplant recipients for allograft pathology by measur-ing the serum BNP levels at the time of either screeningechocardiography and biopsy, or at the time of clinicalrejection. There was a significant difference (p<0.0001) in the BNP levels in 37 patients in the groupwith evidence of pathology compared with those with-out evidence. There was also 100% sensitivity and100% negative predictive value of BNP levels >100pg/ml for identifying graft pathology. �2003 by Ex-cerpta Medica, Inc.

(Am J Cardiol 2003;92:1368–1370)

B-type natriuretic peptide (BNP) is an endogenouscardiac neurohormone secreted from membrane

granules in the cardiac ventricles,1 which has beenshown to increase in response to ventricular stretch inadults with heart failure.2–8 A recent study of 10 adultheart transplant patients showed significantly elevated

BNP levels in the setting of clinical acute rejection.9Clinical symptoms alone are unreliable for early di-agnosing rejection,10 and pediatric patients frequentlyundergo invasive and expensive procedures to assessfor graft pathology such as rejection or transplantcoronary artery disease. Therefore, a simple and safescreening tool would be useful in this population. Thisstudy examined whether BNP can be used to rule outcardiac pathology after orthotopic heart transplanta-tion (OHT) in children.

• • •

We conducted a retrospective review of 47 consec-utive pediatric OHT patients who underwent cardiaccatheterization, endomyocardial biopsy, and echocar-diography at UCLA Medical Center between Januaryand October of 2002. These patients were 0.1 to 20years of age at the time of OHT, and follow-up timeafter OHT was 3 to 173 months (mean 34 � 41,median 20). Two patients (none with any evidence ofallograft pathology) were excluded from the studybecause no BNP levels were sent for study at the time ofcatheterization. Eight patients studied within 14 weeks oftransplantation were also excluded because we have re-cently shown that BNP levels are elevated immediatelyafter transplantation and decrease to �100 pg/ml by 14weeks.11 After these exclusions, we had 59 data pointsfrom 37 patients remaining for analysis. Our study pro-tocol was approved by the institutional review board ofthe UCLA Medical Center.

From the Division of Emergency and Transport Medicine, ChildrensHospital Los Angeles and USC Keck School of Medicine; Division ofPediatric Cardiology, UCLA Mattel Children’s Hospital; and Divisionof Cardiology, Department of Pediatrics at Harbor-UCLA Hospital andResearch and Education Institute, Los Angeles, California. Dr. Clau-dius’ address is: Childrens Hospital Los Angeles, 4650 Sunset Boule-vard, MS #113, Los Angeles, California 90027. E-mail:iclaudius@chla.usc.edu. Manuscript received May 28, 2003; revisedmanuscript received and accepted August 11, 2003.

TABLE 2 Survival of Fenfluramine Pulmonary Hypertension(PH) Versus Primary Pulmonary Hypertension (PPH) Matchedfor Severity

Time Period Ending (mo)

Survival

Diet Pill(n � 10)

Non-diet Pill(n � 20)

12 53% 89%24 35% 70%36 18% 46%48 18% 39%60 18% 39%

1368 ©2003 by Excerpta Medica, Inc. All rights reserved. 0002-9149/03/$–see front matterThe American Journal of Cardiology Vol. 92 December 1, 2003 doi:10.1016/j.amjcard.2003.08.035