Prospective Prognostic Assessment of Blood Pressure Response During Exercise in Patients With Hypertrophic Cardiomyopathy
Background Previous studies revealed that an abnormal blood pressure response (ABPR) during exercise was common in young hypertrophic cardiomyopathy (HCM) patients and was associated with a family history of premature sudden cardiac death (SCD). This study was performed prospectively to assess the prognostic significance of blood pressure response during exercise in young patients with HCM.
Methods and Results Maximum symptom-limited treadmill exercise testing with continuous blood pressure monitoring was performed in 161 consecutive patients 8 to 40 years old (27±9). A normal blood pressure response, defined as an increase in the systolic pressure of at least 20 mm Hg from rest to peak exercise in the absence of a fall of >20 mm Hg from peak pressure, was seen in 101 (63%). In 60 (37%), the blood pressure response was abnormal. There was no significant difference in patients with normal blood pressure response and ABPR in terms of age, sex, follow-up, or recognized risk factors for SCD. During the follow-up period (mean, 44±20 months), SCD occurred in 12 patients: 3 (3%) in the normal blood pressure response group versus 9 (15%) in the ABPR group (P<.009). ABPR had a sensitivity of 75%, a specificity of 66%, a negative predictive value of 97%, and a positive predictive value of 15% for the prediction of SCD. There was no significant difference in the incidence of other recognized risk factors between patients with SCD and the survivors.
Conclusions A normal exercise blood pressure response identifies low-risk young patients with HCM. An ABPR identifies the high-risk cohort; the low positive predictive accuracy, however, indicates that further risk stratification is warranted.
Sudden cardiac death is a common complication of HCM, especially in adolescents and young adults, in whom it may be the first clinical manifestation of the disease in previously asymptomatic patients.1-4 Data from referral centers estimate an annual incidence of SCD of 4% to 6% in adolescence and early adulthood, decreasing to 1% to 2% in later life.4-10 A major goal in the management of patients with HCM is to accurately identify the high-risk cohort so that appropriate interventions can be applied. In adults, the finding of nonsustained VT on Holter recordings11,12 identifies the high-risk cohort with a low predictive accuracy (20% to 25%). The absence of this arrhythmia identifies the majority of low-risk patients7 with a sufficient predictive accuracy (>90%) to permit reassurance. In children, adolescents, and young adults, recurrent syncope6 and family history of premature SCD5,10 are relatively specific but fail to identify the majority of young patients who die suddenly.
Exercise-induced hypotension is a well-recognized feature of HCM. Its importance was underlined by the fact that a significant number of SCDs occur during or shortly after exercise.3,4 We have previously reported that exercise-induced hypotension occurred in one third of patients with HCM.13 It was more common in the young and was associated with a family history of SCD. The observation of an abnormal BPR during exercise in 4 of 6 patients who had experienced nonfatal sudden cardiac arrest led us to conduct this prospective study to assess the prognostic significance of BPR during exercise in young patients (<40 years old) with HCM in whom the risk of SCD is highest.
Between January 1987 and February 1994, 167 consecutive HCM patients ≤40 years old who underwent maximum treadmill exercise testing off cardiovascular medication at St George’s Hospital, London, were selected for this study. Six patients who had experienced nonfatal sudden cardiac arrest before the initial exercise test were excluded from this prospective analysis. The study then comprised 161 patients (105 men) with a mean age of 27±9 years (range, 8 to 40 years). Symptomatic evaluation at the time of exercise testing revealed that 61 (38%) experienced dyspnea (NYHA functional class II in 54 and class III in 7), 65 (40%) chest pain, and 42 (26%) one or more syncopal episodes during the preceding 5 years. Fifty-one patients (32%) had a family history of HCM, and an additional 52 (32%) had a family history of HCM and premature SCD (at <35 years old) among relatives. Thirty-one (19%) had nonsustained VT, defined as more than three consecutive ventricular beats at a rate >120 bpm during 48 hours of ambulatory ECG recordings. The diagnosis of HCM was based on the typical clinical, ECG, and hemodynamic features,14,15 with left ventricular wall thickness of ≥15 mm demonstrated on two-dimensional echocardiography (mean left ventricular wall thickness, 21.9 mm) in the absence of any other cardiac or systemic disease that could have caused hypertrophy.15,16 Twenty-five patients who did not fulfill the above echo criteria but who were obligate or proven gene carriers were included.
Treadmill Exercise Testing
The exercise test was performed as part of the initial prospective evaluation. Before the exercise test, all cardioactive medications were discontinued for at least 5 half-lives. All patients were in sinus rhythm at the time of the exercise test except one who was in atrial fibrillation. One patient had a rate-responsive dual-chamber pacemaker implanted for atrioventricular conduction disease.
Maximum symptom-limited treadmill exercise testing with continuous measurement of Vo2 was performed with the standard Bruce protocol in 150 patients and a modified Bruce protocol in 11 who could not sustain the more aggressive protocol. ECG monitoring was performed with the Marquette Max 1 system with automatic recordings. Twelve-lead ECGs were recorded at the end of each stage and every minute during the recovery period. Systolic BP was recorded by the same trained technician by auscultation of the Korotkov sounds at the left brachial artery with a mercury sphygmomanometer. The technique required continuous inflation and deflation of the cuff to maintain contact with the Korotkov sounds. Eleven patients in whom systolic BP recording could not be accurately obtained by auscultation underwent repeat exercise BP recording with direct intra-arterial measurements from the nondominant brachial artery. BP values were recorded at rest, at 1-minute intervals during exercise, and at 30-second intervals for 5 minutes during the recovery period. A cutoff value of 20 mm Hg was used to define two BP patterns.13,17 A normal BPR was defined as a gradual increase of at least 20 mm Hg in systolic BP during exercise, with a gradual decline during recovery. An abnormal BPR included hypotensive and flat responses. A hypotensive BPR was defined as either (1) an initial increase in systolic BP with a subsequent fall by peak exercise of >20 mm Hg compared with the peak BP value or (2) a continuous decrease in systolic BP throughout the exercise test of >20 mm Hg compared with baseline. A flat response was defined by a change in systolic BP during the whole exercise period of <20 mm Hg compared with the resting systolic BP.
Respiratory gases were collected by use of a face mask firmly applied to prevent escape of exhaled gas. Gas analysis was performed with established methodology with a metabolic cart (Sensor Medics Horizon MMc, Beckman Instruments) before December 1991 and a Marquette system (Marquette Electronics) with a mass spectrophotometric gas analyzer thereafter. A temperature-controlled polarographic sensor with an on-board microprocessor measured the Vo2, issuing printouts of minute ventilation, Vo2 (in mL/min), carbon dioxide production (in mL/min), and RER at 15-second intervals. Before the study, patients were guided in the techniques of exercise and respiratory gas collection. Peak Vo2 was defined as Vo2 (in mL · min−1 · kg−1) at peak exercise calculated as the mean values of the four samples taken during the last minute of exercise. Patients were encouraged to continue the exercise test until their RER was ≥1.0. The predicted maximal Vo2 was calculated according to the following formulas, taking into account sex, age, body weight, and height18,19: for men, and for women, The exercise was terminated for one of the following reasons: fatigue, severe dyspnea, significant chest pain, near syncope, or development of arrhythmias. During analysis of the exercise test, the following parameters were noted: the protocol used (Bruce or modified Bruce), the maximal workload achieved (in METs), the total exercise time (in minutes and seconds), the peak systolic BP, the systolic BP at peak exercise, the systolic BP at 1-minute intervals during recovery, the maximal Vo2 achieved (in mL · min−1 · kg−1) and percentage of the predicted Vo2, the peak heart rate (in beats per minute), and the percentage of predicted maximal heart rate for age and sex. The predicted maximal heart rate was calculated as follows: maximal heart rate=(220−age) for men20 and maximal heart rate=(210−age) for women.20
The end point in this study was the occurrence of SCD, other causes of death, or heart transplant. SCD was defined as witnessed, instantaneous collapse leading to death within 1 hour of onset of symptoms or successful resuscitation from documented ventricular fibrillation. The coding of the mode of death was performed by the first author (N.S.) blinded to knowledge of the patient’s exercise BPR and risk factor status. Subsequently, 158 patients were reevaluated at Saint George’s Hospital outpatient clinic, and follow-up data were accumulated until August 1995. Three patients were followed up elsewhere, and information was obtained from their physicians.
Data are expressed as mean±SD. Group comparisons were with the χ2 test, Fisher’s exact test, and unpaired Student’s t test where appropriate. A two-tailed probability value of <.05 was considered statistically significant.
All patients completed exercise testing without complication. The principal reason for termination of the exercise was fatigue in 101 patients (63%), shortness of breath in 31 (19%), dizziness and presyncope in 14 (9%), chest pain in 8 (5%), reaching the end of the protocol in 6 (4%), and nonsustained VT in 1. Adequate exercise as evidenced by an RER >1.0 was achieved in all but 17 patients (9 women) in whom RER was between 0.90 and 0.99. The reasons for stopping in these 17 patients were not different from those of the remaining 144 patients.
A normal BPR was recorded in 101 patients (63%), and 60 (37%) had an abnormal response. Of these, 27 had a hypotensive response and 33 a flat response (Table 1⇓). The two groups achieved a similar degree of cardiovascular stress as assessed by maximal predicted heart rate attained (93±10% in the normal BPR group versus 88±12% in the abnormal BPR group, P=NS) (Table 1⇓). Six of 101 patients (6%) with a normal BPR did not reach an RER >1, compared with 11 of 60 patients (18%) with an abnormal BPR (P<.001). The percentage predicted Vo2 achieved was higher in the normal BPR group (73±17% versus 59±18%, P<.001). There was no difference in the two groups in terms of age (28±8 versus 25±8 years), sex ratio (2.1 versus 1.6), VT on Holter (18% versus 21%), syncope (24% versus 33%), family history of HCM (34% versus 28%), family history of HCM and SCD (35% versus 28%), symptoms at the time of exercise, and echocardiographic left ventricular dimensions (left ventricular end-diastolic diameter, 44±6 versus 44±7 mm; left ventricular end-systolic diameter, 26±6 versus 27±6 mm; maximum left ventricular wall thickness, 21±8 versus 23±9 mm). Left atrial dimension was smaller in patients with a normal BPR (39±8 versus 43±8 mm; P<.002) (Table 1⇓).
The mean follow-up was 44±20 months. The predominant symptomatic treatment during the period of evaluation was β-blockers in 23 patients and verapamil or diltiazem in 17. Thirty-one patients received amiodarone: 7 for sustained palpitation, 9 for documented recurrent supraventricular tachyarrhythmias, and 15 who were considered to be at high risk of sudden death. Two patients underwent myectomy, and 2 received an ICD for nonfatal sudden cardiac arrest (see below). In addition, dual-chamber pacemakers were implanted in 8 patients: 3 for recurrent syncope associated with bradycardia and 5 for drug-refractory obstructive HCM. All these pacemakers were implanted in 1993 to 1994. The proportion of patients receiving β-blockers, calcium blockers, and amiodarone was similar in patients with an abnormal and a normal BPR (9 versus 14, 7 versus 10, and 13 versus 18, respectively).
During the follow-up, events occurred in 17 patients (Fig 1⇓).
SCD occurred in 12 patients: 3 of 101 (3%) in the normal BPR group compared with 9 of 60 (15%) in the abnormal BPR group (P<.005; odds ratio, 3). The actuarial survival curve (Kaplan-Meier) based on the BPR is given in Fig 2⇓. Abnormal BPR had a sensitivity of 75%, a specificity of 66%, a negative predictive value of 97%, and a positive predictive value of 15% for the prediction of SCD (Table 2⇓).
Among the 12 patients who experienced sudden cardiac arrest, 2 were successfully resuscitated. The characteristics of these 12 patients as well as the circumstances and possible mechanisms of the episodes are shown in Table 3⇓. The patients were young (mean age, 24±7 years; range, 15 to 37 years), and 6 had a family history of HCM and sudden death. Six patients had experienced syncope before the event, and 3 had nonsustained VT on 48-hour Holter recordings. Three of the 12 episodes of SCD occurred during or after exercise, and all of the remaining episodes occurred during moderate daily activities or at rest. Two patients had documented monomorphic VT (ECG in one, Holter in one) degenerating into VF. There was no difference in terms of preexisting established risk factors (ie, syncope, family history of HCM and sudden death, nonsustained VT on Holter recordings) between these 12 patients who experienced sudden cardiac arrest or VF and the remaining 149 patients (Table 4⇓).
Three patients died of other causes: 1 of endocarditis at 24 months, 1 of end-stage heart failure at 2 months, and 1 after mitral valve replacement at 19 months after the exercise test was performed. Two patients underwent orthotopic heart transplantation 24 and 78 months after exercise.
Exercise-related hypotension is a well-recognized feature in patients with HCM. The potential importance of such abnormalities is underlined by the fact that a significant number of SCDs occurred during or after periods of exertion.3,4 We have previously shown in a study of 129 consecutive patients (mean age, 41 years) that exercise-induced hypotension occurred in one third of patients with HCM.13 Exercise-induced hypotension was more common in young patients and was associated with a family history of SCD. Invasive assessment showed that the abnormal BPR during exercise was due to an abnormal peripheral vascular response with an exaggerated decrease in total systemic vascular resistance despite an appropriate rise in cardiac index.13,17 This was associated with reversal of the normal constrictor response seen in nonexercising vascular beds during supine dynamic leg exercise (References 13 and 1713 17 , and G.A. Haywood, MD, unpublished observations). These findings suggest that an abnormal BPR might be important in relation to the genesis of SCD, especially in young patients. Similar observations were reported by another group from Poland.21
This prospective study in the young (average age, 24 years) shows that an abnormal BPR was associated with an increased risk of SCD that was independent and not associated with other markers of increased risk. Risk stratification in HCM remains a major challenge in clinical management. The incidence of SCD is highest in the young and can be the first manifestation of the disease. In a study of 78 patients with HCM who died suddenly, Maron et al4 showed that 89% were <40 years of age. Fifty-four percent had experienced no functional limitation before death, and in 43%, SCD was the initial manifestation of the disease. Data generated from referral centers have shown that the incidence of SCD is ≈2% to 4% per year in adults and 4% to 6% per year in children and adolescents.22 Although these percentages may be biased toward more severely ill patients,23 identification of a high-risk cohort has always been a major goal in the management of patients with HCM, especially those <40 years of age. However, the mechanisms of SCD are undoubtedly not identical and may be age-related.
In adults (>25 years old), the single most useful marker of risk for SCD is the finding of episodes of nonsustained VT on Holter monitoring.11,12 Pooled data from two independent studies have shown nonsustained VT on Holter monitoring to have a specificity of 80%, a sensitivity of 69%, and a high negative predictive value (97%) for SCD during 3 years.7 This practical, noninvasive evaluation provides the basis for reassurance for the majority of adults. In the young, however, Holter arrhythmias are uncommon, and the absence of nonsustained VT does not confer safety.24 In addition, other recognized risk factors (syncope, family history of sudden death) do not identify the majority of young patients who die suddenly.
The mechanism by which an abnormal BPR on exercise may lead to SCD remains unclear. Documentation of the initiating mechanisms of SCD in HCM arise from anecdotal reports in which a patient has been fortuitously monitored during an event. Sinus tachycardia, ischemia, and atrial or ventricular tachyarrhythmias have been documented to cause SCD or hemodynamic collapse.2,7,9,25 The hemodynamic consequences of such events may be exaggerated by the concurrent existence of an abnormal vascular pressure response. This could explain the occurrence of SCD during or soon after periods of exertion in some patients.3 Moreover, the evidence suggests that patients with an abnormal BPR have a generalized abnormality of vasomotor control that is likely to be a determinant of risk at rest as well as during exercise. Abnormal exercise BP is strongly associated with an inappropriate increase in renal (G.A. Haywood, unpublished data) and forearm blood flow17 during mild supine leg exercise and with paradoxical forearm vasodilatation during the minor central volume unloading associated with subhypotensive lower-body negative pressure.26 Preliminary evidence reveals unexplained hypotension during daily life activity in >40% of a small cohort of young HCM patients during beat-to-beat ambulatory BP monitoring.27 Regardless of the initiating mechanism, these data led us to propose that altered vascular responses are an important determinant of outcome.
Limitations of the Study
Ideally, the prognostic significance of an abnormal BPR during exercise would have been assessed in untreated groups; this was not ethically possible. A similar proportion of normal and abnormal BP responders received β-blockers, calcium antagonists, and amiodarone. There is little reason to suspect that symptomatic treatment modified risk of SCD. It is important to emphasize that treatment was not altered on the basis of exercise BP findings. Whether amiodarone prevented SCD in the 15 patients who were considered to be at high risk or the 16 patients treated for arrhythmias is uncertain.28 Thirteen of these 31 patients had an abnormal BPR, and 18 had at least one of the conventional risk factors; ostensibly, they were a high-risk group. Of interest, only 1 of the 12 SCD patients was receiving amiodarone at the time of the event.
An abnormal BPR during exercise has a potential value in the clinical management of patients with HCM. Its high negative predictive accuracy in the entire population in the present study allows reassurance of young patients with a normal BPR. However, the low positive predictive value probably indicates the high heterogeneity in the cohort and underlines the multifactorial mechanisms of SCD in HCM as well as the need for additional risk stratification.
Selected Abbreviations and Acronyms
|BPR||=||blood pressure response|
|RER||=||respiratory exchange ratio|
|SCD||=||sudden cardiac death|
This study was supported by grants from the Fédération Française de Cardiologie, Paris, France (N.S.), and the British Heart Foundation (P.M.E.). We gratefully acknowledge Anne O’Donoghue, Shaughan Dickie, Carol Page, and Michele Lee Jones for their technical assistance.
- Received November 25, 1996.
- Revision received August 4, 1997.
- Accepted August 11, 1997.
- Copyright © 1997 by American Heart Association
Frank S, Braunwald E. Idiopathic hypertrophic cardiomyopathy: clinical analysis of 126 patients with emphasis on the natural history. Circulation. 1968;37:759-788.
Maron BJ, Roberts WC, McAllister HA, Rosing DR, Epstein SE. Sudden death in young athletes. Circulation. 1980;62:218-229.
Maron BJ, Roberts WC, Epstein SE. Sudden death in hypertrophic cardiomyopathy: a profile in 78 patients. Circulation. 1982;65:1388-1394.
McKenna WJ, Deanfield JE. Hypertrophic cardiomyopathy: an important cause of sudden death. Arch Dis Child. 1984;59:971-975.
McKenna WJ, Camm AJ. Sudden cardiac death in hypertrophic cardiomyopathy: assessment of patients at high risk. Circulation. 1989;80:1489-1491.
Maron BJ, Fananapazir L. Sudden cardiac death in hypertrophic cardiomyopathy. Circulation. 1992;85(suppl I):I-57-I-63.
Maron BJ, Cecchi F, McKenna WJ. Risk factors and stratification for sudden cardiac death in patients with hypertrophic cardiomyopathy. Br Heart J. 1994;72(suppl):S13-S18.
McKenna WJ, England D, Doi YL, Deanfield JE, Oakley CM, Goodwin JF. Arrhythmia in hypertrophic cardiomyopathy, I: influence on prognosis. Br Heart J. 1981;46:168-172.
Frenneaux MP, Counihan PJ, Caforio ALP, Chikamori T, McKenna WJ. Abnormal blood pressure response during exercise in hypertrophic cardiomyopathy. Circulation. 1990;82:1995-2002.
Report of the World Health Organization/International Society and Federation of Cardiology Task Force on the definition and classification of cardiomyopathies. Circulation. 1996;93:841-842.
Counihan PJ, Frenneaux MP, Webb DJ, McKenna WJ. Abnormal vascular responses to supine exercise in hypertrophic cardiomyopathy. Circulation. 1991;84:686-696.
Jones NL. Evaluation of microprocessor controlled exercise testing. J Appl Physiol. 1984;57:1312-1318.
Wasserman K, Hansen JE, Sue DY, Whipp BJ. Principles of Exercise Testing and Interpretation. Philadelphia, Pa: Lea & Febiger; 1987.
Chojnowska L, Bijak A, Woroszylska M, Ruzyllo M. Exercise blood pressure response in young patients with hypertrophic cardiomyopathy. Br Heart J. 1994;715:44. Abstract.
McKenna WJ, Goodwin JF. The natural history of hypertrophic cardiomyopathy. In: Harvey P ed. Current Problems in Cardiology. Chicago, Ill: Year Book Medical Publishers; 1981;6:5-26.
Thomson H, Lele S, Atherton J, Gaines J, Britten A, McKenna WJ, Frenneaux M. Venous control in hypertrophic cardiomyopathy: responses to exercise and lower body negative pressure. Eur Heart J. 1995;16:503. Abstract.
Prasad K, Gould M, Lu F, Elliott PM, Sadoul N, McKenna WJ. Abnormal blood pressure response during exercise predicts unstable hemodynamic status during daily activity in patients with hypertrophic cardiomyopathy. Br Heart J. 1995;73:61. Abstract.
McKenna WJ, Oakley CM, Krikler DM, Goodwin JF. Improved survival with amiodarone in patients with hypertrophic cardiomyopathy and ventricular tachycardia. Br Heart J. 1985;53:412-416.