A Comparative Evaluation of C-Reactive Protein as a Short-Term Prognostic Marker in Severe Unstable Angina A Preliminary Study

Transcription

1 Original Article A Comparative Evaluation of C-Reactive Protein as a Short-Term Prognostic Marker in Severe Unstable Angina A Preliminary Study S Bhagat*, M Gaiha*, VK Sharma**, S Anuradha* Abstract Objective : The present study was done to evaluate the short-term prognostic significance of C-reactive protein (CRP) in unstable angina (UA) and to compare it with other known prognostic markers of UA, as there is a paucity of data from our country. Methodology : This prospective study comprised of 44 UA patients (Braunwald Class II A, B) with age <65 years, along with 40 age and sex matched healthy controls. Patients with h/o myocardial infarction (MI) in the preceding 1-month and evidence of infection, inflammation or neoplasm were excluded from the study. Complete clinical evaluation was done and presence of any prognostic variables of UA was noted (including Braunwald high-risk variables). Apart from biochemical investigations and ECG, CRP was measured at the time of admission in the study group and controls using Microwell ELISA assay. Standard treatment protocol was followed for all patients and they were closely monitored during hospital stay and subsequently for 4 weeks for occurrence of any adverse cardiac events. Results : There was a statistically significant difference in the mean level of CRP between study group (6.12 ± mg/l) and controls (1.52 ± mg/l); p = Among the 44 patients, 19 (43%) experienced any one or more outcome measures. Of the three variables (i.e. ongoing chest pain, ST depression 1 mm and CRP 4 mg/l) which showed statistical significance on univariate analysis with respect to adverse outcome measures (p = 0.001, and respectively), only CRP 4 mg/l and ST depression 1 mm showed independent prognostic significance on multiple logistic regression analysis (p = and respectively). The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and relative risk (RR) for CRP 4 mg/l were 78.9%, 96%, 93.75%, 85.74% and 6.56 (95% Cl; ) and that for ST depression 1 mm were 89.47%, 64%, 65.38%, 88.89% and 5.88% (95% Cl: ), respectively. Trop T could be done in eight patients only. Conclusion : CRP 4 mg/l by ELISA is an independent predictor of adverse cardiac outcome in severe UA in the short-term follow up and, hence, is useful for risk stratification of these patients. CRP has a higher specificity, PPV and overall RR for prediction of an outcome than ST segment depression, although it is less sensitive. INTRODUCTION Unstable angina (UA) patients constitute a large proportion of admissions in any coronary care unit and their management and risk stratification is of immense importance. It is recognized that 6-8% of all patients with UA have a non-fatal myocardial infarction (MI) or die within *Department of Medicine; **Department of Microbiology; Maulana Azad Medical College, New Delhi. Received : ; Accepted : the first year of diagnosis. 1,2 It is important to identify this high-risk subset of patients at initial admission, so that appropriate, early intervention may be directed towards preventing further life-threatening events. Various clinical and electrocardiographic parameters have been used to prognosticate patients with UA. 3 In recent years, inflammation has been recognized to underlie the plaque disruption that contributes to the occurrence of UA. 4 The role of inflammation is suggested by histological studies of unstable coronary plaques, presence of activated circulating JAPI VOL. 51 APRIL

2 leucocytes, evidence of systemic release of thromboxanes, leukotrienes and increased concentrations of acute phase reactants of inflammation like C-reactive protein (CRP), serum amyloid A etc. 5,6 Among these, CRP has shown much promise as a marker of inflammation and of prognosis in patients with UA. CRP levels have been reported to be raised in 65-90% of patients of UA compared to 13% patients of chronic stable angina. 7 Few studies have also addressed the utility of CRP as a prognostic marker in patients with UA, alone 8 or in combination with other biochemical markers like Troponin T. 9 Since Indian data on this important aspect is scarce, the present study was done to evaluate the short-term prognostic significance of CRP in UA and to compare it with other known prognostic markers (clinical and electrocardiographic) of UA. MATERIAL AND METHODS The present study was conducted prospectively over a 12 months period. All patients admitted with a diagnosis of UA were considered for enrolment into the study if they satisfied the following inclusion criteria: a. Angina at rest with one or more episodes lasting >20 minutes within the preceding 48 hours (Braunwald class III; A, B). b. Age <65 years Patients with presence of any one of the following criteria were excluded: a. An episode of acute MI within the past 1 month (to avoid raised CRP due to tissue necrosis of MI as a confounding variable). b. Raised creatinine kinase (CK) - total and MB fraction - at admission and at hours after admission and a raised lactate dehydrogenase (LDH) level at admission (to exclude both acute MI and non-q wave MI). c. Any infective or inflammatory and neoplastic condition that is known to be associated with an acute phase response, thereby causing a rise in CRP. The study group comprised of 44 patients along with 40 age and sex matched healthy controls (medical and paramedical personnel and attendants of patients). After taking a detailed history pertaining to chest pain (i.e., nature, duration, character, etc.) and for all possible risk factors including diabetes, hypertension, smoking, dyslipidemia, obesity and family history of CAD, all patients were subjected to any evidence of hypotension and left ventricular failure (LVF). In all the patients and controls the following investigations were performed: Complete hemogram with ESR including peripheral smear. Whole blood venous sugar - fasting (F) and post-prandial (PP), and glycosylated hemoglobin (HbA1c). Complete lipoprotein analysis - Total cholesterol (TC), high-density lipoprotein cholesterol (HDL), low density lipoprotein cholesterol (LDL) and triglycerides (TG). Blood urea, serum creatinine and urine examination (routine and microscopy). Additionally in the study group, the following investigations were done: 12-lead standard electrocardiogram (ECG) - patients were put on continuous ECG monitoring during the first 48 hours. ECG was done at least once daily after 48 hours till discharge and also during any intercurrent episodes of chest pain. X-ray chest, PA view Cardiac enzymes: 1. CK (total and MB fraction) at admission and once at hours. It was repeated in the event of persistent/recurrent chest pain, if indicated. 2. LDH - once, at the time of admission. 3. Troponin T - performed at admission in as many patients as possible due to limited availability, using TROPT sensitive Rapid Test kit (Roche Diagnostics Ltd). The result of the test was interpreted as follows: Positive test - Troponin - T 0.1 ng/ml Negative test - Troponin - T < 0.1 ng/ml CRP Estimation : CRP estimation was done once at the time of admission in the study group and controls. The sera obtained from aseptically collected blood samples were stored at -20 o C and tested within two to four weeks. Measurement of CRP by ELISA Assay 10 - ELISA assay based on Sandwich technique, using polystyrene microplates as the solid phase, was standardized for estimation of CRP in serum. It was reproducible and the assay arrange was wide enough to detect from physiological to raised pathological values (0.1 mg/l to 30 mg/l). Necessary reagents, enzyme substrates, unlabelled and peroxidase labelled right immunoglobulin to human CRP and microplates were obtained commercially. For calibration value graph, nine calibration samples from 10 mg/l CRP and diluted serially two-fold, were used. Treatment Protocol : Standard protocol for treatment of patients with severe unstable angina was followed. The treatment consisted of complete bed rest alongwith an intensive medical regimen including nitrates (initially intravenously, later orally), beta blockers (metoprolol started in a dose of mg twice daily), calcium channel blockers (diltiazem in a dose of 30 mg 3 times a day, if required), IV infusion heparin (80 U/kg stat followed by 18 U/kg/hour) for 5 days and aspirin 325 mg stat followed by 165 mg/day which was continued indefinitely. Measure of Outcomes : All patients were assessed after discharge by weekly visits for four weeks for occurrence of outcome measures, which included: Primary Outcome Measures Cardiac death - death due to MI, cardiac arrhythmias, 350 JAPI VOL. 51 APRIL 2003

3 cardiogenic shock or congestive heart failure; and Non-fatal MI - as diagnosed on the basis of WHO criteria. Secondary Outcome Measures Recurrent angina - defined as two or more episodes of ischemic chest pain lasting more than one minute after 24 hours of initiation of treatment. Also included were any readmissions for UA. Acute pulmonary edema Serious arrhythmias - sustained ventricular tachycardia, ventricular fibrillation, and supraventricular tachycardia with hemodynamic compromise. Intervention - CABG/PTCA Statistical Analysis : All the data collected was statistically analyzed using the following tests where applicable: analysis of variance test (ANOVA), student s t- test, Fisher s Exact test and chi-square test. The statistical significance was estimated by calculating the p value with significance assigned at p The prognostic value of CRP to predict an adverse outcome was defined in terms of sensitivity, specificity, positive predictive value, negative predictive value and relative risk, and was compared to those of other known prognostic variables. To find out the independent prognostic value of CRP, a simple stepwise multiple logistic regression analysis was done. The variables evaluated were - age, HT, DM, family history of premature CAD, past MI, previous CABG/PTCA, prior angina, prior aspirin use, Troponin T and Braunwald high risk group variables, including dynamic ST segment changes 1 mm and ongoing chest pain 20 min (not resolved). OBSERVATIONS AND RESULTS The mean age (in years) of the study group was ± 7.54 (range 36-65) and of the controls was ± 8.02 (range 37-65). The maximum number of subjects in both the groups was in the age group of years % and 47.5% respectively. The male:female ratio in the study group and controls were 1.44:1 and 1.85:1 respectively (p = 0.57). The baseline characteristics of the study group are summarized in Table 1. As can be seen, smoking was the most common risk factor present in 55% cases followed by HT in 50% cases. T wave inversion, alone or in combination with ST segment depression was the most common ECG change (70%) seen at admission in the patients. The mean level of ESR in males was 7.3 mm/h and in females was 11.6 mm/h. The mean CK levels (IU/L) - Total and MB fraction and LDH levels (IU/L) observed in the study group were ± 31.18, ± 3.96 and ± respectively. The normal reference range for the above were - CK (total) IU/L, CK (MB) 0-24 IU/L, and S. LDH IU/L respectively. Troponin T test was performed in eight patients only. Out of these only one showed a positive result and the rest seven showed a negative result. CRP Level in Present Study : The mean CRP level (by ELISA method) in mg/l in the study group was significantly higher ± (range mg/l) compared to the control group ± 0.75 (range mg/l); p = The distribution of CRP levels (mg/l) in the study subjects is depicted in Table 2. It can be seen that 95.5% of the study group had CRP levels 2 mg/l while 72.5% of the controls had CRP <2 mg/l. Outcome Measures : The outcome measures (primary Table 1 : Baseline characteristics of the study group (n=44) at admission Demographics and prior history n=44 (%) Age, years (Mean ± SD) 54.7 ± Male 26 (59%) Risk factors for CAD DM 21 (48%) HT 22 (50%) Smoking 24 (50%) Dyslipidemia 16 (36%) Family history of premature CAD 21 (48%) Obesity 8 (18%) Prior medication use Aspirin 25 (57%) Oral nitrates 24 (55%) Beta blockers 9 (18%) ECG changes Any ECG change 37 (84%) ST segment depression 1 mm 26 (59%) ST segment elevation 1 mm 0 (0%) T wave inversion 30 (70%) Length of qualifying episode of chest pain, hours (mean ± SD) 2.15 ± No of chest pain episodes at rest in last 48 hours (mean ± SD) 1.68 ± Table 2 : Distribution of CRP, mg/l (ELISA) in study group and controls CRP (mg/l) Patients, n=44(%) Controls, n=40(%) 1 0 (0%) 13 (32.5%) (4.5%) 16 (40%) (29.5%) 8 (20%) (29.5%) 3 (7.5%) 4 16 (36.4%) 0 (0%) Range of CRP Table 3 : Outcome measures (primary and secondary) in the study group Outcome measure No. of cases, n=44(%) Any outcome measure 19 (43%) Primary outcome measure Cardiac death 1 (2.3%) MI 2 (4.5%) Secondary outcome measure Recurrent angina 19 (43%) Revascularization (CABG/PTCA) 1 (2.3%) New onset serious arrhythmias/chf 1 (2.3%) JAPI VOL. 51 APRIL

4 and secondary) in the study group at 4 weeks are summarized in Table 3. Recurrent angina was the most frequent outcome, observed in 43% of study group. On analyzing the level of CRP with respect to occurrence of outcome measures (Table 4), all the statistical parameters for prediction of occurrence of any outcome measure were maximum at a CRP level of 4 mg/l. This suggests that a CRP level of 4 mg/l (by ELISA) at admission can be taken as a cut off point beyond which the likelihood of any adverse event is increased. Table 4 : Statistical parameters at various cut off levels of CRP for the prediction of occurrence of any outcome CRP (mg/l) Sensitivity Specificity Positive Negative ELISA method predictive predictive value value 1 100% 0% 43.18% 0% 2 100% 8% 45.23% 100% % 48% 53.57% 75% % 96% 93.75% 85.74% % 96% 92.30% 77.40% Importance of individual prognostic variables of unstable angina with respect to outcomes (primary and secondary) : An important aspect of this study was to compare the ability of known variables and CRP to predict any short-term adverse outcome. The univariate analysis of various prognostic variables of UA with respect to any outcome measure (primary and secondary) in the study group is described in Table 5. Only three variables showed a statistically significant positive correlation with the occurrence of any outcome measure. These included: 1. CRP level 4 mg/l (by ELISA) 2. ST depression 1 mm at admission 3. On-going chest pain 20 min (not resolved) The stepwise multiple logistic regression analysis of various prognostic variables for the occurrence of different primary and secondary outcome measures yielded the following results: 1. Among the various variables tested, CRP level (by ELISA) was the best independent predictor of any adverse outcome during short-term follow up (p = ). ST segment depression 1 mm was the next best predictor of adverse outcome (p = ). Thus CRP and ST depression were both found to be independent predictors of prognosis in UA during short term follow up of four weeks. 2. No other variable was able to reach statistical significance in this analysis. 3. The classification success percentage for CRP (ELISA) alone was 79.55%, and for ST depression 1 mm alone was 75%. On combining the two, the classification success percentage increased to 86.36%. Thus, by combining CRP 4 mg/l and ST segment depression 1 mm, the ability to predict an adverse outcome was enhanced. There was also a significant positive correlation observed between ST depression 1 mm on admission ECG and CRP 4 mg/l at admission (p = 0.005). CRP 4 mg/l versus ST depression 1 mm - The comparison of the statistical parameters determining the Table 5 : Comparative correlation of various prognostic variables of UA with respect to outcome measures Prognostic variable Occurrence of any No outcome (n=25) p value one or more outcome (n=19) ST depression 1 mm (S) Ongoing chest pain (S) Age, years (mean ± SD) 55 ± ± (NS) Past h/o MI (NS) Family h/o premature CAD (NS) Previous CABG/PTCA (NS) HT (NS) DM (NS) Prior angina (NS) Prior aspirin use (NS) Trop T* 1 out of 5 0 out of (NS) CRP 4 mg/l (ELISA) (S) * Trop T was done in eight cases only. Table 6 : Comparison of statistical parameters determining the prognostic value of CRP 4 mg/l (ELISA) and ST depression 1 mm at admission in present study Sensitivity Specificity Positive Negative Relative risk of any outcome predictive value predictive value outcome CRP 4 mg/l (ELISA) 78.9% 96% 93.75% 85.74% 6.56 (95% CI: ) ST depression % 64% 65.38% 88.89% 5.88 (95% CI: ) 352 JAPI VOL. 51 APRIL 2003

5 prognostic value of CRP 4 mg/l (ELISA) and ST depression 1 mm at admission is summarized in Table 6. The table demonstrates that the sensitivity of ST depression >1 mm at admission for prediction of an adverse outcome during short term is higher than that of CRP 4 mg/l (89.47% versus 78.9%). However, the specificity for prediction of an adverse outcome was much higher for CRP 4 mg/l than ST depression 1 mm (96% versus 64%). Also, the overall relative risk for occurrence of any outcome was higher for CRP 4 mg/l than ST depression 1 mm. DISCUSSION The mean CRP level by ELISA method in the study group was significantly higher than those of controls [6.12 ± 6.13 mg/l versus 1.52 ± 0.75 mg/l (p = 0.00)]. CRP is a nonspecific acute phase reactant, the levels of which rise with any form of tissue damage and/or inflammation. All patients with known infective or inflammatory disease states and recent MI were excluded from the study population in the present study to remove confounding factors. Further, normal levels of ESR (mean value in males was 7.3 mm/h and in females was 11.6 mm/h) at admission rules out infectious process as a cause of rise in CRP levels. Also, the normal CK levels (total and MB fraction) at admission and at 24 hours along with normal LDH levels at admission (mean ± SD of CK total, CK MB and S. LDH were ± IU/L, ± 3.96 IU/L and ± IU/L respectively) indicate that elevation of CRP in the study populations cannot be attributed to myocardial necrosis. Thus the rise in CRP observed in patients of UA in this study could possibly be attributable to the underlying plaque inflammation. These results corroborate the results of previous findings of raised CRP levels in UA patients. 6,11,12 Haverkate et al have attributed this rise in CRP levels to intrinsic inflammation and tissue damage within the arterial lesions themselves. 12 Outcome Measures, CRP Levels and Their Correlation : As shown in Table 3, 19 out of 44 patients had any one adverse outcome during the 4 week follow up - most common being recurrent angina. Of these 19 patients, 15 had CRP 4 mg/l and only four had CRP < 4 mg/l, which was statistically significant. Biasucci et al, who studied 53 patients of Braunwald class III UA, found that 69% of UA patients with CRP level 3 mg/l had an adverse cardiac event compared to only 15% of patients with CRP < 3 mg/l (p <0.001). 13 Liuzzo et al, similarly found a higher incidence of adverse cardiac events beyond a cut off level of CRP 3 mg/l (ELISA). 11 Haverkate et al identified a cut off level of CRP >3.6 mg/l (ELISA) as that indicative of a higher risk of adverse outcomes. 12 The statistical parameters of CRP 4 mg/l for prediction of an outcome in this study have been compared to other previous studies in Table 7. This table highlights the fact that higher levels of CRP are predictive of a worse outcome in UA. To evaluate the independent prognostic significance of CRP in UA, a multiple logistic regression analysis was done. Only CRP 4 mg/l and ST segment depression 1 mm were found to have strong independent prognostic value for prediction of an adverse outcome (p value and respectively). In the TIMI III B data base study, ST depression, prior angina, family history of premature CAD, prior use of heparin/aspirin and increasing age, were all identified as independent predictors of cardiac events. 14 The small number of patients included in the present study could account for these differences observed. The TIMI II A sub-study evaluated the role of CRP as a predictor of 14 days mortality in 630 UA/non-Q-wave MI patients alone or in combination with Troponin T. 9 The results of this study were very impressive. Firstly, CRP levels were much higher in the patients who died compared to those who survived, strongly suggesting a prognostic role for CRP with respect to short-term mortality. Further, among patients with a negative TROP T test, a markedly raised CRP level identified patients who remained at increased risk of death at 14 days. This suggested an additional prognostic role for CRP. Also, CRP and TROP were said to provide complementary information on patient mortality. Infact, the authors went to support the use of a combination of CRP and TROP T for a more comprehensive risk assessment in patients with UA and non-q-wave MI. In the light of these observations, we shall try to assess the data of the eight patients in whom TROP T test was done in this study (Table 8). Out of these eight patients, seven had a negative TROP T test. It is observed that four out of these Table 8 : CRP, Troponin T and outcomes in present study Patient no. Troponin T CRP Any outcome 0.1 ng/ml mg/l measure 1 Positive Negative Negative Negative Negative Negative Negative Negative Table 7 : Predictive value of CRP for occurrence of short-term adverse events in various studies Study Sensitivity Specificity Positive predictive value Negative predictive value Liuzzo et al 11 (CRP 3 mg/l) 90% 82% 90% Morrow et al 9 (CRP 15.5 mg/l) 86% 76% Present study (CRP 4 mg/l) 78.9% 96% 93.75% 85.74% JAPI VOL. 51 APRIL

6 seven patients with a negative TROP T test had an adverse cardiac outcome during the four week follow up. Though no conclusive results can be drawn due to extremely small sample size, it is evident that the CRP levels in these patients are still higher than the controls. However, studies involving much larger patient populations using both CRP levels and TROP T determination are required. To conclude, raised CRP levels are seen in patients with UA. Raised CRP level 4 mg/l by ELISA is an independent predictor of an adverse cardiac outcome in the short term and, hence, is useful in the risk stratification of these patients. CRP has a higher specificity, PPV and overall RR for prediction of an outcome than ST segment depression, although it is less sensitive. In conjunction with other markers like TROP T, CRP levels may in the future provide valuable information for identifying the high-risk group in UA patients. REFERENCES 1. Lincoff AM, Teheng JE, Califf RM, et al. Sustained suppression of ischemic complications of coronary intervention by platelet GP IIb/IIa blockade with abciximab: one year outcome in the EPILOG trial. Circulation 1999;99: Gibson CM, Goel M, Cohen DJ, et al. Six month angiographic and clinical follow up of patients prospectively randomized to receive either tirofiban or placebo during angioplasty in the RESTORE trial. J Am Coll Cardiol 1998;32: Braunwald E, Mark DB, Jones RH, et al. Unstable angina: Diagnosis and management. Clinical Practice Guideline number 10 (amended) AHCPR Publication No , Rockville, Md., AHCPR and the National Heart, Lung and Blood Institute, PHS, US Department of Health and Human Services, May van der Wal AC, Becker AE, van der Loos CM et al. Site of intimal rupture or erosion of thrombosed coronary atherosclerotic plaques is characterized by an inflammatory process irrespective of the dominant plaque morphology. Circulation 1994;89: Gogna A, Sinha RSK, Gupta B. C-reactive protein - marker for atherothrombotic events. J Assoc Physicians India 1999;47: Berk BC, Weintraub WS, Alexander RW. Elevation of C- reactive protein in active coronary artery disease. Am J Cardiol 1990; 65: Liuzzo G, Biasucci LM, Buffon A, et al. Elevated C-reactive protein at discharge and at 3 months after waning of symptoms in unstable angina is associated with recurrence of instability during 12 months follow up. J Am Coll Cardiol 1995;(Suppl 250A), Abstract. 8. Liuzzo G, Biasucci LM, Gallimore JR, Grillo RL, Rebuzzi AG, Pepys MB, Maseri A. The prognostic value of C-reactive protein and serum amyloid A protein in severe unstable angina. N Engl J Med 1994;331: Morrow DA, Rifai N, Antman EM, et al. C-reactive protein is a potent predictor of mortality independently of and in combination with troponin T in acute coronary syndromes. A TIMI II A sub study. Thrombolysis in Myocardial Infarction. J Am Coll Cardiol 1998;31: WHO Expert Committee on Biological Standardization. WHO Technical Report Series 760, Geneva, Switzerland Liuzzo G, Biasucci LM, Rebuzzi AG, et al. Plasma protein acute phase response in unstable angina is not induced by ischemic injury. Circulation 1996; 94: Haverkate F, Thompson SG, Pyke SD, et al. Production of C- reactive protein and risk of coronary events in stable and unstable angina. European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group. Lancet 1997; 349: Biasucci LM, Liuzzo G, Grillo RL, et al. Elevated levels of C- reactive protein at discharge in patients with unstable angina predict recurrent instability. Circulation 1999;99: Stone PH, Thompson B, Zaret BL, et al. Factors associated with failure of medical therapy in patients with unstable angina and non-q-wave myocardial infarction: a TIMI-III B database study. Eur Heart J 1999;20: Announcement XIVth Annual Conference of Indian Society of Organ Transplantation will be held at Hyderabad, India on August 8-10, A Pre Conference Workshop on Renal, Liver and Heart Transplantation is on August 7, For further details please write to Dr. BV Rama Raju, Organizing Secretary, ISOT , Flat No. 301, Reliance Residency, Indira Park Road, Opp. NTR Stadium, Hyderabad , AP, India. Phone No , Hospital , Res.: , , Cell No Fax : ; E_mail : drbvr@hotmail.com Sd/- BV Rama Raju 354 JAPI VOL. 51 APRIL 2003