To the Editor:
I read with interest the manuscript by Hinderliter and other members of the Primary Pulmonary Hypertension Study Group entitled “Effects of Long-term Infusion of Prostacyclin (Epoprostenol) on Echocardiographic Measures of Right Ventricular Structure and Function in Primary Pulmonary Hypertension.”1 I was disappointed that the authors stated conclusions with no supporting data. Previously, these authors reported that a 12-week continuous intravenous infusion of prostacyclin improved survival in patients with severe primary pulmonary hypertension.2 Hinderliter and associates now claim that a 12-week period of prostacyclin therapy also had important beneficial effects on right ventricular structure and function in this same group of patients. Unfortunately, the results were incompletely reported without appropriate statistical comparisons. The baseline echocardiographic features of treated patients and control subjects were well described. However, corresponding measurements, 12 weeks after randomization, were not reported. Furthermore, the authors reported none of the statistical comparisons between baseline and follow-up measurements. By comparing only the median differences between baseline and follow-up echocardiographic variables, the authors arrived at inappropriate and misleading conclusions. The median differences in most echocardiographic variables were so small that it is unlikely that significant “changes” in right ventricular structure or function occurred in either patient group. The authors have suggested that the effects of prostacyclin on cardiac function may have contributed to improved survival and exercise capacity. However, little attention was given to the possibility of a type II error in overlooking a significantly better baseline performance in the 6-minute walk for patients treated with prostacyclin.
The Primary Pulmonary Hypertension Study Group should be commended for their efforts in identifying a potential therapy for a group of patients with little hope for a good quality of life and long-term survival. Nonetheless, the results of their collaborative efforts must be interpreted accurately in order to clearly confirm the safety and efficacy of prostacyclin therapy.
- Copyright © 1998 by American Heart Association
Hinderliter AL, Willis PW, Barst RJ, Rich S, Rubin LJ, Badesch DB, Groves BM, McGoon MD, Tapson VF, Bourge RC, Brundage BH, Koerner SK, Langleben D, Keller CA, Murali S, Uretsky BF, Koch G, Li S, Clayton LM, Jöbsis MM, Blackburn SD, Crow JW, Long WA. Effects of long-term infusion of prostacyclin (epoprostenol) on echocardiographic measures of right ventricular structure and function in primary pulmonary hypertension. Circulation. 1997;95:1479–1486.
Barst RJ, Rubin LJ, Long WA, McGoon MD, Rich S, Badesch DB, Groves BM, Tapson VF, Bourge RC, Brundage BH, Koerner SK, Langleben D, Keller CA, Murali S, Uretsky BF, Clayton LM, Jöbsis MM, Blackburn SD, Shortino D, Crow JW. A comparison of continuous intravenous epoprostenol (prostacyclin) with conventional therapy for primary pulmonary hypertension. N Engl J Med. 1996;334:296–301.
We believe that the data published in our manuscript entitled “Effects of Long-term Infusion of Prostacyclin (Epoprostenol) on Echocardiographic Measures of Right Ventricular Structure and Function in Primary Pulmonary Hypertension”R1 were analyzed and interpreted correctly. Baseline echocardiographic data were summarized as mean values with standard errors for three purposes: (1) to fully characterize the patients enrolled in the trial; (2) to illustrate that the patients randomized to prostacyclin and conventional therapy were similar to those randomized to conventional therapy alone; and (3) to demonstrate the marked differences in echocardiographic measurements between our patients and normal control subjects. Providing corresponding descriptive statistics 12 weeks after randomization would be of little value and could be misleading, since 12-week means would not include values from some of our original sample; data loss occurred because of patient deaths, transplantation, and, in a few instances, unavailable or uninterpretable follow-up ultrasound studies. The Wilcoxon rank sum test was utilized to compare the effects of prostacyclin and conventional therapy with conventional therapy alone for changes in echocardiographic variables. The results of these analyses are reported in Table 5 of the article. Data analyzed using nonparametric statistics are appropriately summarized as median values, as shown in Table 5. We concur with Dr Day’s observation that the differences between treatment effects in the two groups, although statistically significant in some instances, were small in magnitude. Additional assessments for within-group change through the Wilcoxon signed rank test indicated significant (P<.05) worsening of the group treated with conventional therapy for all echocardiographic variables in Table 5 except right ventricular percent change in area; for the group randomized to prostacyclin and conventional therapy, no echocardiographic variable had significant worsening, and diastolic eccentricity index had significant improvement.
Dr Day correctly points out that the patients randomized to prostacyclin and conventional therapy had a better performance on the 6-minute walk than those randomized to conventional therapy alone, although this atypical random difference was not significant (P>.05). This raises the possibility that patients treated with prostacyclin were less impaired for this factor at baseline by chance, although they could by chance be more impaired for one or more other factors. This possibly confounding factor was discussed in detail in the previous report by Barst et al,R2 and the improved exercise performance and survival observed in patients treated with prostacyclin was statistically evident, independent of baseline 6-minute walk results. We did not adjust for baseline exercise capacity in our analyses of treatment effects on echocardiographic parameters since this difference between groups was unexpected and there was no a priori reason to expect that changes in echocardiographic measurements would be influenced by initial 6-minute walk results. However, in response to Dr Day’s concerns, we have examined our data with adjustment for baseline 6-minute walk through a rank ANCOVA for each echocardiographic variable. The variables with the lowest unadjusted probability values (diastolic eccentricity index and maximal tricuspid regurgitant jet velocity) remained statistically significant after this adjustment, but those with more marginal unadjusted probability values (right ventricular end-diastolic area and systolic eccentricity index) had adjusted probability values that were in the range of .05 to .15. These results are consistent with those of the unadjusted analyses, although they should be recognized as having some bias against the group treated with prostacyclin.
In summary, we agree with the observations that differences between treatment groups in the echocardiographic changes observed over the 12 weeks of therapy were small in magnitude and that the small difference between the two groups in baseline exercise capacity complicates interpretation of the data. Nonetheless, analyses reported in our manuscript are appropriate and support the interpretation that improved right heart structure and function may play a role in the clinical improvement and prolonged survival noted in patients treated with prostacyclin.
Hinderliter AL, Willis PW IV, Barst RJ, Rich S, Rubin LJ, Badesch DB, Groves BM, McGoon MD, Tapson VF, Bourge RC, Brundage BH, Koerner SK, Langleben D, Keller CA, Murali S, Uretsky BF, Koch G, Li S, Clayton LM, Jöbsis MM, Blackburn SD, Crow JW, Long W, for the Primary Pulmonary Hypertension Study Group. Effects of infusion of epoprostenol (prostacyclin) on echocardiographic measures of right heart structure and function in primary pulmonary hypertension. Circulation. 1997;95:1479–1486.
Barst RJ, Rubin LJ, Long WA, McGoon MD, Rich S, Badesch DB, Groves BM, Tapson VF, Bourge RC, Brundage BH, Koerner SK, Langleben D, Keller CA, Murali S, Uretsky BF, Clayton LM, Jöbsis MM, Blackburn SD, Shortino D, Crow JW, for the Primary Pulmonary Hypertension Study Group. A comparison of continuous intravenous epoprostenol (prostacyclin) with conventional therapy for primary pulmonary hypertension. N Engl J Med.. 1996;334:296–301.