INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: hypertension; pulmonary; resuscitation

Pulmonary arterial hypertension (PAH) results from progressive obliteration of the pulmonary vascular bed. The rise in pulmonary vascular resistance causes an increased workload of the right heart that eventually results in right ventricular failure (1). In fact, the most common cause of death in PAH is progressive right heart failure (2, 3). However, sudden and unexpected death may occur (2). Moreover, invasive hemodynamics are part of the diagnostic and therapeutic workup of PAH and complications may occur during the procedures (4, 5). For these reasons, cardiopulmonary resuscitation (CPR) may be performed on patients with PAH as a result of multiple circumstances leading to acute circulatory or respiratory failure.

There is little information in the medical literature on how often CPR is attempted on patients with PAH and on the outcome of CPR in this group of patients. From discussions with colleagues, we found that some physicians are very reluctant to perform CPR on patients with severe PAH because they have found their attempts to be uniformly futile, whereas others believe that vigorous CPR attempts are warranted, at least in selected patients.

To obtain data on the frequency and outcome of CPR in patients with severe PAH, we conducted an international survey in Europe and in the United States, collecting data from large pulmonary hypertension centers which had expertise as well as high volumes of patients with pulmonary hypertension. In this report, we report the results of this study.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Referral centers specialized in the evaluation and treatment of PAH in Europe and in the United States received two questionnaires. The first sheet asked for general information (the number of patients with pulmonary hypertension who were treated at the center between 1997 and 2000, the number of patients who died, and in how many patients CPR had been initiated). The second sheet was designed to obtain individual information on the patients in whom CPR had been instituted (age, sex, diagnosis, medication, significant intercurrent illnesses, setting of CPR, cause of cardiorespiratory arrest, initial electrocardiogram [ECG], interval between collapse and CPR, outcome of CPR, autopsy results when available, and data from right heart catheterization within 3 mo before CPR when available). Furthermore, the participants were asked to provide a brief description of the course of the resuscitation attempts and any special intervention used.

This was a retrospective analysis and the patients were not identified as individuals in our analysis. We did not obtain approval of the institutional review boards from all participating centers. The ethics committee of Hannover Medical School approved the study protocol and supported our decision not to submit the study to other review boards.

Statistical Analysis

All results are given as mean ± SD. Two-sided unpaired Student's t test was used to compare hemodynamic variables between survivors and nonsurvivors of CPR. A p value below 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Seventeen centers (10 European centers and 7 American centers) participated in this study. In the 4-yr period between 1997 and 2000, approximately 3,130 patients with pulmonary hypertension were treated in these centers. An exact number could not be given because two centers could only estimate the total number of patients. Of the 3,130 patients who were treated at the participating centers, 513 (16%) had circulatory and respiratory arrest within the study period. In 132 cases (26%), CPR was attempted. The demographic data of these patients are shown in Table 1. The outcome of the patients in whom CPR was attempted is shown in Figure 1. Overall, only 8 of 132 patients (6%) were long-term survivors without neurologic deficit. Two of these patients died 300 and 468 d, respectively, after CPR; the remaining six patients are currently alive at the time of this writing for a median period of 425 d (range, 100 to 700 d). Autopsy was performed in 47 of 124 patients (38%) and confirmed the diagnosis of pulmonary vascular disease in each case.

View larger version (29K): [in this window] [in a new window]   Figure 1.   Outcome of CPR in patients with PAH.

The most common causes of death were progressive right heart failure (n = 64; 49%), respiratory failure (n = 24; 18%), and sudden (unexpected) death (n = 23; 17%). Other, rare causes of death included vasovagal reactions, hemoptysis, septic complications, gastrointestinal bleeding episodes, and seizures. In 11 cases (8%), cardiorespiratory arrest was associated with catheter interventions (including two patients who died within 24 h after atrial septostomy). Four additional deaths were related to infectious complications of central venous catheters used for administration of continuous intravenous prostacyclin therapy.

In 71 cases (54%), cardiorespiratory arrest was associated with a significant intercurrent illness. The single most common intercurrent illness was a respiratory tract infection in 24 patients (18%). Other concomitant diseases were viral enteritis, inguinal hernia, thrombophlebitis, and gastroenteritis. Only 5 of 132 CPR attempts (4%) took place outside the hospital; 127 patients (96%) were hospitalized. Most patients (n = 83; 63%) were already in an intensive care unit (ICU) when cardiorespiratory failure occurred. Eight CPR attempts (6%) were performed in the emergency room, 7 (5%) in the cardiac catheterization laboratory, 22 (17%) on a medical or surgical ward, and 7 (5%) in other in-hospital locations.

Corresponding with these data, the interval between collapse and the onset of CPR was less than 1 min in 95 cases (73%), between 1 and 5 min in 30 cases (23%), and more than 5 min in only 6 cases (4%). The time interval between circulatory arrest and onset of CPR was undetermined in one patient.

The initial ECGs at the time of CPR showed bradycardia in 58 cases (45%), electromechanical dissociation in 37 cases (28%), asystole in 19 cases (15%), ventricular fibrillation in 10 cases (8%), and other rhythms in 6 cases (4%). The initial ECG rhythm was unknown in two cases. Data from right heart catheterization within 3 mo before CPR were available for 80 patients (61%). The hemodynamic variables confirmed the presence of severe PAH in these patients (Table 2).

As previously indicated, only eight of 132 patients (6%) were long-term survivors of CPR. The demographic characteristics, circumstances of CPR, and hemodynamic variables of these patients are listed in Table 3. In all eight patients, CPR was instituted almost immediately but this was also the case in the majority of nonsurvivors. In three survivors, cardiorespiratory arrest occurred during cardiac catheterization. The first patient (Patient 3) had right heart failure during heart catheterization and recovered after prolonged resuscitation that included the intravenous administration of a 50-µg iloprost bolus. The second patient (Patient 5) had pericardial tamponade during cardiac catheterization and was successfully resuscitated after pericardiocentesis. The third patient (Patient 8) had a vasovagal reaction with the development of third-degree atrioventricular (AV) block during catheterization and required brief CPR without medications or intubation. Patient 1 was admitted to the ICU because of hemodynamic instability in association with a respiratory tract infection and became pulseless after coughing. Chest compression was performed for 1 min and the patient recovered after receiving inhaled nitric oxide and an intravenous bolus of 50 µg iloprost. This patient died 468 d later from sepsis and right heart failure. Patient 2 required CPR for 5 min for third-degree heart block due to digitalis toxicity. Patient 4 collapsed before entering the emergency room, where resuscitation was started immediately. Prolonged resuscitation attempts were initially unsuccessful, although the patient subsequently recovered after receiving 50 µg iloprost intravenously. This patient died 300 d later from right heart failure. Patient 6 had a vasovagal syncope after going to the bathroom and required 15 min of resuscitation before stabilization. Patient 7 required brief CPR for respiratory failure caused by an epileptic seizure.

There was no significant difference between the survivors and nonsurvivors with respect to sex, age, underlying disease, or initial rhythm (data not shown). The time interval between collapse and onset of CPR was also no discriminating factor because CPR was started immediately in most of the survivors as well as the nonsurvivors. Except for Patient 4, all long-term survivors had identifiable and correctable causes of circulatory arrest. Right heart failure, which was the most common cause of death in the study population, was indicated as the cause of death in only one of the long-term survivors. Surprisingly, there were also no significant differences in the hemodynamic variables obtained during right heart catheterization in the survivors compared with the nonsurvivors; however, there was a trend toward lower right atrial pressures in the long-term survivors (Table 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study shows that CPR is attempted in approximately 25% of patients who subsequently die from progression of their PAH. Unfortunately, survival after CPR in this group of patients is quite poor. In this survey of 132 patients with PAH in whom CPR was instituted, only eight patients (6%) survived for more than 90 d. Seven of these patients had correctable causes of cardiopulmonary arrest, such as vasovagal reactions, digitalis toxicity, or pericardial tamponade. Thus, the single most important factor that seems to influence success of CPR is the presence of a defined reason for cardiopulmonary arrest and its prompt correction.

Reported data on survival of out-of-hospital resuscitation in patients with cardiac arrest from other causes range from 4 to 34% (6, 7). Survival rates of inpatient cardiac arrest are reported to be as high as 37% (8). The most important factors determining the outcome of cardiac arrest include the rhythm associated with arrest, the interval until CPR is initiated, and the underlying clinical condition of the patient (11). In fact, patients with chronic medical conditions who undergo CPR usually have a bad outcome (12).

There may be several reasons for the poor results of CPR in patients with PAH. In most studies investigating CPR in patients with cardiovascular disease, the time interval between collapse and onset of resuscitation was one of the most important determining factors of survival (11). In the present investigation, this time gap was less than 1 min in 73% and less than 5 min in 96% of the cases. Therefore, delayed institution of CPR was not an important factor contributing to the poor outcome of CPR in patients with PAH.

The majority of patients (96%) were hospitalized when cardiorespiratory arrest occurred, and 63% of these patients had already been admitted to an ICU, suggesting that their hemodynamic condition had already been judged as unstable. All participating centers were university hospitals specialized in the treatment of pulmonary hypertension. It is therefore highly unlikely that the quality of the resuscitation measures was inadequate. The patients in this study had a mean pulmonary artery pressure of 61 mm Hg and a mean cardiac index of 1.7 L/min/m², indicating severe pulmonary vascular disease. Interestingly, these hemodynamic variables were similar in the survivors compared with the nonsurvivors except for a trend toward higher right atrial pressures in nonsurvivors that almost reached statistical significance. Thus, the severity of pulmonary hemodynamics did not seem to be an important factor that influenced the outcome of CPR.

As in previous studies in patients with pulmonary hypertension, the most common cause of death was right heart failure (49% in the present study). The initial ECGs obtained after collapse included bradycardia, electromechanical dissociation (pulseless electrical activity), and asystole, rhythms that are known to be associated with a bad prognosis (11). In addition, the presence of ventricular fibrillation, which is usually an indicator of a favorable outcome after CPR (11), was an ominous sign in patients with pulmonary hypertension; none of the 10 patients with ventricular fibrillation survived despite immediate attempts at defibrillation. In contrast, in a recent study, the survival was 74% for unselected patients with ventricular fibrillation who received defibrillation within 3 min after collapse (13).

In approximately 50% of the patients in this study, cardiorespiratory arrest resulted from progression of their underlying disease, whereas an intercurrent illness contributed to death in the remaining patients. These intercurrent diseases were often minor abnormalities such as simple respiratory tract infections or gastrointestinal infections, which underscores the notion that patients with pulmonary hypertension are very "fragile" with little or no compensatory reserve.

The poor results of CPR in patients with PAH may be explained by the underlying hemodynamic condition. The mean pulmonary vascular resistance in our study population was 1,694 dynes · s · cm⁵, which is more than eight times above the upper normal limit of 200 dynes · s · cm⁵. Under these conditions, it is extremely difficult to achieve effective pulmonary blood flow and left ventricular filling with chest compression. Although this study was not designed to prove this hypothesis, we did measure end-tidal carbon dioxide in two patients with primary pulmonary hypertension during CPR and found end-expiratory carbon dioxide levels of 5 and 7 mm Hg, respectively (data not shown). These findings suggest that pulmonary blood flow was virtually absent in these two patients during CPR (14).

Rest and exercise right ventricular ischemia resulting from reduced coronary driving pressure and hypertrophied right ventricular myocardium has been shown in a majority of patients with PAH (15). In this context of poor right ventricular coronary perfusion, cardiopulmonary arrest, whatever the cause, may further compromise right myocardial viability.

Based on these pathophysiologic considerations, measures to improve the results of CPR in patients with pulmonary hypertension need to focus on lowering pulmonary vascular resistance. In this context, it is noteworthy that three of the successful CPR attempts included the intravenous bolus administration of iloprost, a prostacyclin analogue (16). Although these are just preliminary observations, aggressive administration of prostanoids may improve the results of CPR in patients with PAH. On the other hand, in some cases, cardiac arrest is due to vagal reaction and in this context systemic hypotension could be aggravated by acute administration of prostanoids and improved by the use of atropine and adrenergic drugs.

Some limitations may have influenced the results of the present study. PAH is a rare disorder and single-center experiences are not sufficient to obtain valid results on CPR attempts in this group of patients. Our survey collected the experience of 17 referral centers in Europe and America that treated a total of 3,130 patients with pulmonary hypertension between 1997 and 2000. Based on these numbers, this survey should provide a realistic overview on the results of CPR in PAH. However, the study was retrospective in design and we therefore cannot exclude a recall bias. The total number of patients who had circulatory arrest, 513, out of a total of 3,130 patients with severe pulmonary hypertension during a 4-yr observation period appears low compared with mortality rates that have been published in the literature (2). However, 74% of the patients who died were on chronic treatment with prostacyclin, indicating that the patients had been treated in centers with expertise in the field of pulmonary vascular disease, which might have contributed to the low overall mortality rate.

In conclusion, CPR is unsuccessful in most patients with PAH. The best chance for survival is in patients who have a defined reason for their cardiorespiratory arrest that can be rapidly corrected.

Further contributors were as follows: Belgium: Robert Naeije, M.D., Erasme Hospital, Brussels. Germany: Ralph Wiedemann, M.D., H. Ardeshir Ghofrani, M.D., W. Seeger, M.D., Justus-Liebig-University, Giessen; Edda Spiekerkoetter, M.D., Hannover Medical School. Italy: Alessandra Manes, M.D., University of Bologna. Poland: Pawel Kuca, M.D., Institute of Tuberculosis and Lung Diseases, Warsaw. Switzerland: Rudolf Speich, M.D., University Hospital, Zurich. United Kingdom: Natalie Doughty, R.N., Papworth Hospital, Cambridge. USA: Amy Yoney, R.N., Columbia-Presbyterian University, New York; Maged Chane, M.D., Glenna Traiger, R.N., Wendy Hill, University of Southern California; Loretta Berger, Rush-Presbyterian-St. Luke's Hospital, Chicago; Elisabeth S. Klings, M.D., Boston University; Ashraf Uzzaman, M.D., University of Michigan; Tracey Ryan, R.N., University of Pittsburgh.