Survival and Predictors of Survival in Patients With Congestive Heart Failure

Transcription

1 398 Current Perspectives Survival and Predictors of Survival in Patients With Congestive Heart Failure Due to Chagas' Cardioyopathy Charles Mady, MD; Rita Helena Antonelli Cardoso, MSc; Antonio Carlos Pereira Barretto, MD; Protasio Leos da Luz, MD, FACC; Giovanni Bellotti, MD; Fulvio Pileggi, MD, FACC Downloaded fro by on Noveber 3, 218 Background The fundaental deterinant of the natural history of Chagas' disease is cardiac involveent. Methods and Results We studied 1 ale patients with congestive heart failure due to Chagas' disease to estiate the survival distribution function and to evaluate age, functional class (FC), axial oxygen consuption (Vo2ax ), and ejection fraction (EF) as predictors of survival. Statistical evaluation was perfored through univariate (Student's t test and X2 test) and ultivariate analyses (Cox's regression odel). Overall survival was 66% at 1 year, 56% at 3 years, and 8% at 5 years. Ages were not statistically different (P=.9811) between survivor (.3±8.7) and nonsurvivor (.3±9.) groups. The ejection fraction(s) were statistically different (P=.1) between survival (3.6±9.9) and nonsurvival ince the original description of Chagas' disease,1 there has been a great developent in its understanding through research in several fields. The fundaental deterinant of the natural history of the disease is cardiac involveent, and its several evolutive fors have been well studied fro the orphological, etiopathogenic, and clinical points of view.2 7 However, studies regarding functional alterations are rare, and when perfored, they were generally restricted to patients with advanced congestive heart failure (CHF); the ilder fors deserved less attention.6 13 The scarce inforation available about the functional status of these patients was not found to relate to survival. Because new therapeutic odalities are being proposed for the treatent of advanced Chagas' cardioyopathy, such as cardioyoplasty1 and heart transplantation,15 it becoes iportant to identify and quantify cardiac dysfunctions, relating the to evolution and survival of such patients. The ai of this work was to study patients with CHF due to Chagas' cardioyopathy, ranging fro ild to severe fors, in order to analyze survival and soe predictors of survival in individuals within this evolutive range of the disease. Received February 16, 199; revision accepted July 26, 199. Fro the Heart Institute, University of S6o Paulo Medical School, Sao Paulo, Brazil. Reprint requests to Charles Mady, MD, Cardiopatias Gerais, Instituto do Coragao, Av Dr Eneas C. Aguiar,, 53--Sao Paulo, SP, Brazil. ) 199 Aerican Heart Association, Inc. ( ) groups, as was Vo2ax (P=.1) (21.±.7 and 15.±.9, respectively). Most of the surviving patients were in FC II and ost of the nonsurvivors were in FC IV (P=.1). Vo2ax (P=.1) and EF (P=.8) are highly associated with survival tie in the ultivariate analysis, but FC (P=.578) is less iportant. Age (P=.9811) did not influence survival. Conclusions We conclude that 5% of the patients with heart failure due to Chagas' disease die in 7 onths and that Vo2ax and EF are iportant indices of survival in this group.(circulation. 199;9: ) Key Words * Chagas' disease * cardioyopathy * ejection fraction * heart failure Methods We studied 1 ale patients, with ages ranging between 18 and 65 years (.3±9.). Fro these, 31 (29.8%) were in New York Heart Association (NYHA) functional class (FC) II, 1 (39.%) in FC III, and 32 (3.8%) in FC IV. Follow-up ties varied between 1 and 77 onths (3.1±2.5, edian of 2 onths). All patients were treated and followed up in our hospital, and they received digitalis, diuretics, and vasodilator drugs whenever necessary and in the absence of side effects. Patients were not subitted to cardioyoplasty and heart transplantation until the end of the study. The diagnosis of Chagas' disease was based on epideiological aspects, clinical history, physical findings copatible with cardioyopathy, serological reactions (Machado-Guerreiro and iunofluorescence), and typical ECG abnoralities. The latter included 3 electrically inactive areas, 61 right bundle branch blocks, 9 left anterosuperior division branch blocks, 2 posteroinferior division branch blocks, 2 anteroedial division branch blocks, 6 left ventricular hypertrophies, 2 ventricular extrasystoles, 22 supraventricular extrasystoles, 5 first-degree atrioventricular (AV) blocks, 1 with second-degree AV block Mobitz type I, 2 left bundle branch blocks, and 9 atrial fibrillations. Patients with high-degree AV block were excluded. The proposition was to study these patients in relation to survival and to evaluate age, FC, axial oxygen consuption (Vo2ax), and ejection fraction (EF) as predictors of survival. The EF was obtained by Aloka SSD 72 and SSd 87 echocardiographic equipent and was calculated by the forula EF=EDV-ESV/EDV, in which EDV is the end-diastolic volue and ESV is the end-systolic volue.16 The volues were calculated by the forula of Pobo in accordance with the Aerican Society of Echocardiography,16 and the ean of five consecutive beats was used. The evaluation of the Vo2ax was perfored using a Beckan coputerized gas analyzer, odel MMC, which has a polarographic sensor (OM-11) and an infrared syste

2 Mady et al Survival in Patients With Chagas' Cardioyopathy 399 TABLE 1. Age, Ejection Fractions, Vo2ax, and NHYA Functional Class Data on Total Group and Survivor and Nonsurvivor Subgroups Variable Total Group (n=1) Survivors (n=5) Nonsurvivors (n=5) P Age.3±9..3±8.7.3± Ejection fractions* 37.± ± ±8.1.1 Vo2ax 18.1± ±.7 15.±.9.1 NYHA classt (29.8) 28 (51.9) 3 (6.) III 1 (39.) 22 (.7) 19 (38.).1 IV 32 (3.8) (7.) 28 (56.) NYHA indicates New York Heart Association. Values are listed as ean±sd or frequency and percentage. *Significant difference fro eans for survivors and nonsurvivors. tsignificant difference fro proportions for survivors and nonsurvivors. Downloaded fro by on Noveber 3, 218 (LB-2) for evaluation of the oxygen and carbon dioxide expired fractions. Respiratory variables were obtained under conditions of standard teperature, pressure, and huidity (StPD), using apropriate correction factors. The group was subitted to a test of axiu exercise on a Quinton odel 18-5 treadill with variable speed (ph) and inclination (%), using the odified Naughton protocol. Values of Vo2ax in L * kg.* in` were obtained. All patients agreed to take part in the study after presentation of the ethod and explanation of the purpose of the exaination. Statistical Analysis The relation of the variables age, FC, EF, and Vo2ax to ortality was assessed using both univariate and ultivariate analyses. For univariate analysis, subgroups were defined as survivors and nonsurvivors. The quantitative variables were exained by Student's t test and the qualitative variables by X 2 test to deterine differences between the subgroups. To identify the variables that are independently predictive of survival, ultivariate analysis using Cox's regression odel17 was perfored. Finally, patients were classified in different subsets according to the variables selected by Cox's analysis. Kaplan- Meier survival curves'8 were constructed to approxiate the life expectancy of each group, and the long rank test was used to assess the significance of differences in survival rates aong the various groups. All statistics coputations were provided by SAS (Statistical Analysis Syste).19 Results The age, EF, Vo2ax, and FC of the coplete group and survivor and nonsurvivor subgroups are shown in Table 1. At the end of the study, there were 5 survivors and 5 nonsurvivors. All deaths were cardiac (6% sudden death and 36% due to pup failure). The ages in the survivor group varied between 2 and 58 years (.3+±8.7) and in the nonsurvivor group between 18 and 65 years (.3±9.). There were no statistical differences between the two subgroups (P=.9811). The EF in the survival group ranged between 17% and 63% (3.6±9.9%); it was saller in the nonsurvival group and varied between 1% and 62% (3.6±8.1%, P=.1). Vo2ax in the survival group ranged between 11.1 and 3.5 L* kg` * in-1 (21.±.7) and in the nonsurvival group between 7.2 and 28.2 L* kg` in' (15.±.9), a value that is significantly saller than aong survivors (P=.1). Most of the survivors were in FC II and ost of the nonsurvivors were in FC IV (P=.1). The EF, Vo2ax, and FC were subsequently included in a ultivariate analysis that used Cox's ultiple regression odel. In the independent associations between these variables and ortality (Table 2), we noticed that Vo2ax (P=.1) and EF (P=.8) are highly associated with survival tie. FC (P=.578) was of arginal statistical significance. The Kaplan-Meier survival curve of all patients is in Fig 1. Each circle represents an individual event (death). The nubers under the graphic indicate the nuber of patients traced at that tie. The 7% confidence liit is indicated by the vertical bar around the point estiate. Overall survival was 66% at 1 year, 56% at 3 years, and 8% at 5 years. The probability of survival in relation to each individual prognostic variable was estiated by the Kaplan- Meier ethod. For this purpose, EF values were classified as ildly reduced (>5%), oderately reduced (31% to 5%), or severely reduced (c3%). Likewise, Vo2ax values were categorized as being noral (>2 L* kg`. in-t), oderately reduced (1 to 2 L), or severely reduced (<1 L* kg`. in-1). The P values represent difference (log rank test) aong the three groups. Fig 2 represents the Kaplan-Meier survival curves of patients stratified according to NYHA class. Patients in FC II had predicted survival probability of 97% at 1 and 3 years. Patients in FC III had predicted survival probabilities of 73% and 58% at 1 and 3 years, respectively. Patients in FC IV, therefore, had a considerably lower predicted survival: 38% at 1 year and 16% at 3 years. With increasing values of FC, there was a progressive increase in ortality (P=.1). TABLE 2. Independent Associations Between Variables and Mortality Variable Regression Coefficient x2 for Independence P Ejection fraction(s) Vo2ax New York Heart Association class

3 31 Circulation Vol 9, No 6 Deceber _ cn IL :- : J C6 I. -J 9 a. Downloaded fro by on Noveber 3, I l I T --I TIME (years) FIG 1. Kaplan-Meier survival curve of all patients. Each circle represents an individual event (death). Nubers indicate the reaining patients still being followed up at the tie, with reference to the tie scale of the x-axis; 7% confidence liit is indicated by vertical bars around the point estiate. Fig 3 represents the Kaplan-Meier survival curves of patients stratified according to EF. There was a statistically significant difference aong the three subgroups (P=.1). Predicted survival was 1% at 1 and 3 years in patients with EF >5%. Only 31% of patients with EF 53% survived at 1 year and 16% at 3 years copared with 78% and 7% at 1 and 3 years, respectively, of those with EF between 31% and 5%. Fig represents the Kaplan-Meier survival curves of patients stratified according to Vo2ax. There was also a highly significant statistical difference aong the three subgroups (P=.1). Survival was significantly better in patients with Vo2ax >2 L kg * All patients in who Vo2ax was below 1 L* kg`1 in-1 had died Co 3.i T a T 1 M E (years) > FIG 3. Kaplan-Meier survival curves of patients stratified according to ejection fraction. P value represents difference (log rank) aong the three groups. Other coents as in Fig 1. before 1 year. Patients presenting Vo2ax between 1 and 2 L* kg-'. in` had a predicted survival of 59% and % at 1 and 3 years, respectively. The corresponding values for patients with Vo2ax >2 were 89% and 86%. Discussion Although the poor prognosis of patients with CHF due to Chagas' cardioyopathy has long been appreciated, inforation concerning survival and predictors of survival in this disease is liited. Survival curves fro recent studies following patients with severe CHF due to several etiologies indicate 1 year survival rates of % to 7%.2-22 However, these data ay not reflect the natural history of ilder CHF. In addition, ost studies have copared the prognosis of patients with CHF due to several etiologies, soe of -_ t- CD -ct a >2 1-2 TIME (years) T M E (years) 7 8 IH III W FIG 2. Kaplan-Meier survival curves of patients stratified according to New York Heart Association functional class. P value represents difference (log rank test) aong the three groups. Other coents as in Fig 1. C > FIG. Kaplan-Meier survival curves of patients stratified according to axial oxygen consuption. P value represents difference (log rank) aong the three groups. Other coents as in Fig 1.

4 Mady et al Survival in Patients With Chagas' Cardioyopathy 311 Downloaded fro by on Noveber 3, 218 the showing different or siilar survival rates according to the underlying disease.23'2 In our group, there were only patients with Chagas' disease. We studied only ale patients because there are differences in evolution according to sex in other etiologies, and the axial functional capacity is greater in ale patients in relation to feale patients.2225 Soe articles have suggested that elderly patients have poorer prognosis.25 We did not find differences between our two groups in relation to ages. CHF ranged fro ild to severe, with a siilar distribution of patients aong these categories. This is an iportant point, because in several studies there is no clear relation between syptos and ortality except in the ost extree cases. In others, the severity of syptos is an iportant prognostic variable.21,23-25 In our patients, we found overall survival rates of 66% at 1 year, 56% at 3 years, and 8% at 5 years (Fig 1). However, when we divided the patients into subgroups according to FC, we noticed a clear relation between syptos and ortality, showing an iportant prognostic significance (Fig 2). Patients in FC II had the sae predicted survival probability of 97% at 1 year and 3 years, deonstrating the good prognosis of this subgroup. Patients in FC III had predicted survival probabilities of 73% in 1 year and 58% in 3 years. Patients in FC IV had a considerably lower predicted survival rates, 38% at 1 year and 16% at 3 years (Fig 2). Therefore, there is a clear difference in evolution of groups with this kind of cardioyopathy according to the FC, fro ild to severe. This point has considerable therapeutic iplications. Other studies were not able to show this difference, probably because they do not have a single etiology.26 Soe investigators have shown that axial oxygen uptake during exercise is related to long-ter prognosis, whereas others have noted no significant difference in exercise capacity between groups of survivors and nonsurvivors with CfF Many studies showed that the predictive value of the exercise test was related to the arkedly reduced survival of patients with the ost severe exercise liitation, with Vo2ax <1 LI kg`- in-l. When such patients were excluded, exercise capacity provided little prognostic inforation, according to these authors.29 All of our patients with Vo2ax of <1 L. kg`. in-1 died within 1 year of evolution. Other studies showed a death rate of 77% during the sae period.27 Otherwise, the survival rate of patients with Vo2ax between 1 and 2 L* kg` - in-1 is 59% in 1 year and 6% in 3 years. Above 2 L * kg`. in-, it iproves to 89% in 1 year and 86% in 3 years. In conclusion, the dearcation between survivors and nonsurvivors is very distinct in our group. Therefore, Vo2ax is a strong predictor of survival in patients with heart failure due to Chagas' disease, even in individuals with ild dysfunctions. These data also have potential clinical relevance. The value of EF also is a significant predictor of outcoe,3 but in soe studies no relation between resting left ventricular EF of survivors and nonsurvivors was found.3'31 In contrast, other investigators have found that greater degrees of depression of left ventricular EF are associated with a poorer prognosis.32'33 In our group, a highly significant statistical difference between the EF of the survival and nonsurvival groups was found, that is, those with greatly reduced EF have the worst prognosis; even at 1-year follow-up their survival was only 3%. Conclusions We can estiate the survival distribution function of patients with cardioyopathy due to Chagas' disease, and the FC, EF, and Vo2ax have been shown to be effective predictors of survival. It is notewhorty that EF and Vo2ax are stronger predictors of survival than FC. Perhaps subjective factors intrinsic to the FC deterination ay account for this difference. The central inforation fro this study is relevant to long-ter treatent of cardioyopathy due to Chagas' disease such as heart transplantation and cardioyoplasty. References 1. Chagas C. Nova tripanosoiase huana. Me Inst Oswaldo Cruz. 199; Andrade ZA. Patogenia da doenca de Chagas: Novos aspectos. Arq Bras Cardiol. 1982;38: Laranja FS, Dias E, N6brega G, Miranda A. Chagas' disease: a clinical, epideiological, and pathologic study. Circulation. 1956; 1: Arteaga-Fernandez E, Pereira Barretto AC, Mady C, Ianni BM, Bellotti G, Pileggi F. eletrocardiograa e pacientes co rea96es sorol6gicas positivas para doenca de Chagas. Estudo de 6 casos. Arq Bras Cardiol. 1985;: Mady C, Pereira Barretto AC, Ianni BM, Lopes EA, Pileggi F. Right ventricular endoyocardial biopsy in undeterined for of Chagas' disease. Angiology. 198;35: Pereira Barretto AC, Mady C, Arteaga-Fernfindez E, Stolf N, Lopes EA, Higuchi ML, Bellotti G, Pileggi F. Right ventricular endoyocardial biopsy in chronic Chagas' disease. A Heart J. 1986;111: Mady C, Pereira Barretto AC, Moffa PJ, Ianni BM, Arteaga- Fernandez E, Bellotti G, Pileggi F. vetocardiograa na fora indeterinada da doenca de Chagas. Arq Bras Cardiol. 1985;: Arteaga-Fernindez E, Pereira Barretto AC, Mady C, Martinelli M, Dauar D, lanni BM, Pileggi F. Correlagfio entre as alterac6es eletrovetocardiograficas e o desepenho ventricular esquerdo na iocardiopatia chagfisica cr6nica. Arq Bras Cardiol. 1983;1: Del Nero E Jr, Diaent J, Papaleo Netto M, Rangel FA, Serro-Azul LG, Tranchesi J, Pileggi F. Paraetros sist6licos fonoecanogrdficos do ventriculo esquerdo na fora pre-clinica da doenga de Chagas cr6nica. Rev Hosp Clin Fac Med S Paulo. 197;29(suppl): Friedann AA, Arelin E, Nelken JR, Zerbini CAF, Coibra MA, Serro-Azul LG. Estudo ecocardiografico do desepenho ventricular e fase pre-clfnica da doenca de Chagas. Rev Hosp Clin Fac Med S Paulo. 198;35: Mady C, Decourt LV. A fora indeterinada da doenca de Chagas. Arq Bras Cardiol. 1981;36: Editorial. 12. Mady C, Moraes AV, Galiano N, D6court LV. Estudo heodinaico na fora indeterinada da doenga de Chagas. Arq Bras Cardiol. 1982;38: Mady C, Yazbek P Jr, Pereira-Barretto AC, Saraiva JF, Vianna CB, Serro-Azul LG, Bellotti G, Pileggi F. Estudo da capacidade funcional axia pela ergoespiroetria e pacientes portadores da doenga de Chagas. Arq Bras Cardiol. 1986;7: Bocchi EA, Bellotti G, Uip D, Kalil J, Higuchi ML, Fiorelli A, Stolf N, Jatene A, Pileggi F. Long-ter follow-up after transplantation in Chagas' disease. Transplant Proc. 1993;25: Bocchi EA, Moreira LF, Bellotti G, Barretto ACP, Azul LGS, Stolf N, Jatene A, Pileggi F. Heodynaic study during upright isotonic exercise before and six onths after cardioyoplasty for idiopathic dilated cardioyopathy or Chagas' disease. A J Cardiol. 1991;67: Sahn DJ, De Maria A, Kisslo J, Weyann A. Recoendations regarding quantification in M-ode echocardiography: results of a

5 312 Circulation Vol 9, No 6 Deceber 199 survey of echocardiography easureents. Circulation. 1978;58: Cox DR. Regression odels and life-tables. J R Stat Soc 1972;3: Lee ET. Statistical Methods for Survival Data Analysis. Belont, Calif: Lifetie Learning Publications; 198: SAS Institute Inc. SUGI Suppleental Library User's Guide, Version 5. Cary, NC: SAS Institute Inc; 1986: Franciosa JA, Wilen M, Ziesche S, Cohn JN. Survival in en with severe chronic left ventricular failure due to either coronary heart disease or idiopathic dilated cardioyopathy. A J Cardiol. 1983; 51: Wilson JR, Schwartz JS, St John Sutton M, Ferraro N, Horowitz LN, Reichek N, Josephson ME. Prognosis in severe heart failure: relation to heodynaic easureents and ventricular ectopic activity. JA Coll Cardiol. 1983;2: Fuster V, Gersh BJ, Giuliani ER, Tajik AJ, Brandenburg RO, Frye RL. The natural history of idiopathic dilated cardioyopathy. A J Cardiol. 1981;7: Massie BM, Ports T, Chatterjee K, Parley W, Ostland J, O'Young J, Haugho F. Long-ter vasodilator therapy for heart failure: clinical response and its relationship to heodynaic easureents. Circulation. 1981;63: Massie BM, Conway M. Survival of patients with congestive heart failure: past, present, and future prospects. Circulation. 1987;75: 5(suppl IV):IV-11. Abstract. 25. Johnson RA, Palacios I. Dilated cardioyopathies of the adult. N Engl J Med. 1982;37: Califf RM, Bounous P, Harrell FE, McCants B, Lee KL, McKinnis RA, Rosati RA. The prognosis in the presence of coronary artery disease. In: Braunwald E, Mock MB, Watson JT, eds. Congestive Heart Failure. New York: Grune & Stratton; 1982: Szlachcic J, Massie BM, Kraer BL, Topic N, Tubau J. Correlates and prognostic iplication of exercise capacity in chronic congestive heart failure. A J Cardiol. 1985;55: Franciosa JA. Why patients with heart failure die: heodynaic and functional deterinants of survival. Circulation. 1987;75(suppl IV): IV-2. Abstract. 29. Franciosa JA. Exercise testing in chronic congestive heart failure. A J Cardiol. 198;53: Nelson GR, Cohn PF, Gorlin R. Prognosis in edically treated coronary artery disease: influence of ejection fraction copared to other paraeters. Circulation. 1975;52: Unverferth DV, Magorien RD, Moeschberger ML, Baker PB, Fetters JK, Leier CV. Factors influencing the one-year ortality of dilated cardioyopathy. A J Cardiol. 198;5: Burggraf GW, Parker JO. Prognosis in coronary artery disease: angiographic, heodynaic, and clinical factors. Circulation. 1975; 51: Schwarz F, Mall G, Zebe H, Schitzer E, Manthey J, Scheurlen H, Kubler W. Deterinants of survival in patients with congestive cardioyopathy: quantitative orphologic findings and left ventricular heodynaics. Circulation. 198;7: Downloaded fro by on Noveber 3, 218