MGM Journal of Medical Sciences

ORIGINAL ARTICLE
Year
: 2022  |  Volume : 9  |  Issue : 3  |  Page : 275--280

Principal component analysis of categorized parameters in periprocedural myocardial injury following percutaneous coronary intervention


Malay Acharyya1, Tanushree Mondal2,  
1 Department of Cardiology, Midnapur Medical College and Hospital, Midnapur, India
2 Department of Community Medicine, Medical College Kolkata, Kolkata, West Bengal, India

Correspondence Address:
Dr. Malay Acharyya
Department of Cardiology, Midnapur Medical College and Hospital, Midnapur 721101, West Bengal
India

Abstract

Objective: The elevation of troponin-T (Trop-T) or creatinine kinase myocardial isoform (CKMB) is very common during the percutaneous coronary intervention (PCI). A study was attempted to determine the correlation between elevated Trop-T or CKMB and the parameters of PCI by using multivariate analysis, especially principal component analysis (PCA). Materials and Methods: A prospective observational study was carried out among 100 patients who underwent PCI for stable coronary artery disease in which 31 and 37 patients were found to have elevated Trop-T and CKMB (>3 times) following PCI. The correlation was studied between Trop-T or CKMB (dependent variable) and different parameters, viz., total stent length (mm), fluoroscopy time (min), lesion strength, left ventricular (LV) function, procedural complications, type of lesions, vessels treated with drug eluting stent (DES), and major adverse cardiac events (MACE) as independent variables. Results: For Trop-T, the principal component (PC)-1 and PC-2 obtained 63.49% and 30.88% of the original variation. For PC-1 and PC-2, maximum positive loading was recorded for stent length followed by fluoroscopy time and for LV but negative loading for the type of lesion and type of stent (DES vs bare metal stent [BMS]). For CKMB, the PC-1 and PC-2 obtained 61.22% and 32.08% of the original variation. For PC-1 and PC-2, maximum positive loading was recorded for stent length and fluoroscopy time followed by vessel treated but negative loading for the type of stent and MACE, and maximum positive loading recorded for LV function but negative loading for the type of lesion. Conclusion: This study indicates which factors are most important in preventing periprocedural myocardial injury during PCI and may be a suitable tool to prevent myocardial injury and for subsequent less MACE and better patient outcomes.



How to cite this article:
Acharyya M, Mondal T. Principal component analysis of categorized parameters in periprocedural myocardial injury following percutaneous coronary intervention.MGM J Med Sci 2022;9:275-280


How to cite this URL:
Acharyya M, Mondal T. Principal component analysis of categorized parameters in periprocedural myocardial injury following percutaneous coronary intervention. MGM J Med Sci [serial online] 2022 [cited 2022 Nov 29 ];9:275-280
Available from: http://www.mgmjms.com/text.asp?2022/9/3/275/357474


Full Text



 INTRODUCTION



The elevation of the creatinine kinase myocardial isoform (CKMB) and cardiac troponin-T (Trop-T) is a indicator of myocardial injury following percutaneous coronary intervention (PCI) and is not uncommon after successful PCI. Contrast-enhanced magnetic resonance imaging studies have shown that even a mild elevation of Trop-T after PCI is associated with discrete microinfarction findings.[1],[2],[3]

Myocardial damage in patients who have successful PCI is mainly related to distal vessel thrombosis, embolization of plaque debris, platelet aggregates, and side branch occlusion.[4],[5] Periprocedural elevation of CKMB or Trop-T level occurs in 5%–50% of patients undergoing PCI,[6] whereas the adverse prognostic association of CKMB elevation is well established.[7]

In a study by Acharyya et al.,[8] it was shown that Trop-T and CKMB are suitable markers for periprocedural myocardial injury following PCI. In PCI, several parameters such as fluoroscopy time, total stent length, number of stents implanted, PTCA balloon inflation time, type of lesion, etc., were related to postprocedural Trop-T and CKMB elevation and major adverse cardiac events (MACE) in 6-month follow-up.[9]

Many reports emphasize a well-known statistical method called principal component analysis (PCA). This statistical method is used by researchers in clinical research to know the association between a dependent variable and independent variables.[10],[11],[12] Moreover, several studies have emphasized the importance of elevated Trop-T and CKMB related to the above-mentioned parameters,[6],[13],[14] but the correlation study between Trop-T or CKMB and different parameters are lacking in the eastern part of India.

The objective of the present study was to know the correlation between elevated Trop-T or CKMB (dependent variable) and the parameters used in PCI (independent variables) by using multivariate analysis through PCA.

 MATERIALS AND METHODS



Study population

A prospective observational study was carried out among 100 patients in which 31 patients were observed with elevated Trop-T (>3 times) and 37 patients were found with elevated CKMB (>3 times) during PCI. The study was carried out in the Department of Cardiology, ICVS, IPGME&R, and S.S.K.M Hospital, Kolkata, India, for 12 months.

Assessment of biomarkers

Blood samples for serum Trop-T and CKMB were drawn and analyzed as per standard protocol at the baseline and then at 12 and 24 h after the procedure as per an earlier study by Acharyya et al.[8] The upper normal value for Trop-T was <0.03 ng/mL. The upper normal value for CKMB was 24 IU/L. Furthermore, post-PCI patients were divided into three categories for both biomarkers (Trop-T and CKMB) as normal value, elevation ≤3 times the upper reference limit, and elevation >3 times the upper reference limit. In the present study, the cut-off value was used as elevation >3 times.

Assessment of parameters used in balloon angioplasty

Balloon angioplasty and stent implantation were performed according to the standard clinical practice by femoral approach. At the end of the PCI anterograde coronary flow in the target, the vessel was assessed according to the TIMI classification. The number and duration of balloon inflation were recorded, and the cumulative inflation time was computed by adding the time of each inflation. Total stent length (mm) in the case of multiple stent implantations was calculated by adding each stent length. MACE were deaths, nonfatal acute myocardial infraction, repeat PCI related to either a target lesion or a new lesion, and coronary artery bypass grafting. The presence or absence of multivessel disease, fluoroscopy time (min), lesions strength treated (<10 mm, 10–20 mm, and >20 mm), left ventricular (LV) function (≥55%: normal LV function ejection fraction; 45–54: mild LV dysfunction; 30–44: moderate LV dysfunction; and <30%: severe LV dysfunction), whether any procedural complications viz. yes or no, type of lesions (A, B, and C) as per intervention, vessel treated whether left anterior descending coronary artery (LAD)/left circumflex coronary artery (LCX)/right coronary artery (RCA)/Ramus/diagonals/multivessel, and the use of drug eluting stent (DES) as per yes or no category were recorded and gathered all data.

Ethical consideration

The study conformed to the institutional ethical guidelines, and all patients were included after obtaining informed consent. The ethical approval has been given by the Institutional Ethics Committee of the Institute of Post-Graduate Medical Education and Research, and Seth Sukhlal Karnani Memorial Hospital, Kolkata, West Bengal, India with written permission vide letter no. 1564 dated 16.03.2012 for conducting the present research work.

Statistical analyses

PCA was performed by using the PAST tool[15] to reduce the variables into a smaller number of uncorrelated predictor variables. Individual participant principal component (PC) scores were comprised of their factor loadings for different parameters and used to create PCA of the nine factors in each case (Trop-T or CKMB), which was ultimately centered and scaled. PCs were confirmed based on eigenvalues >1 as per the study of Liberda et al.[16] and an earlier study by Mondal and Acharyya.[17]

 RESULTS



Trop-T versus different parameters of PCI

The PCA revealed that Trop-T versus different parameters of balloon angioplasty loading scores, which were generated from the individual PC scores, is presented in [Figure 1].{Figure 1}

The PC-1 (63.49% of the original variation elucidated) observed a relatively higher value of positive loading for stent length followed by fluoroscopy time, a moderate value of positive loading for vessel treated followed by LV function with negative loading, a lower value of positive loading for complications and MACE with negative loading, while a least value of positive loading for the type of lesion and type of stent (DES or non-DES) [Figure 2].{Figure 2}

The PC-2 (30.88% of the variation explained) was observed a higher value of positive loading for LV function, a moderate value of positive loading for fluoroscopy time, a lower value of positive loading for stent length followed by vessel treated, while a least value of positive loading for complications and MACE but negative loading for the type of lesion and type of stent (DES or non-DES) [Figure 3].{Figure 3}

[Figure 4] describes a Scree plot in which PCs were formed in the order of the amount of variation within the plot in which PC-1 gained the most variation (63.49%) and then the second most was PC-2 (30.88%) followed by other PCs. It was noted in the present PCA that the highest eigenvalue was captured in PC-1 followed by PC-2 in Trop-T versus different parameters of PCI.{Figure 4}

CKMB versus different parameters of PCI

The PCA revealed that CKMB versus different parameters of balloon angioplasty loading scores, which were generated from the individual PC scores, is presented in [Figure 5].{Figure 5}

The PC-1 (61.22% of the original variation elucidated) observed a relatively higher value of positive loading for stent length and fluoroscopy time followed by vessel treated, a lower value of positive loading for complications and negative loading of LV function, while a least value of positive loading for the type of lesion and negative loading for the type of stent (DES vs bare metal stent [BMS]) and MACE [Figure 6].{Figure 6}

The PC-2 (32.08% of the variation explained) was observed a higher value of positive loading for LV function, a lower value of positive loading for fluoroscopy time, stent length, vessel treated, and MACE, while a least value of positive loading for complications and type of stent (DES vs BMS) but negative loading for the type of lesion [Figure 7].{Figure 7}

[Figure 8] describes a Scree plot in which PCs were formed in the order of the amount of variation within the plot in which PC-2 gained the most variation (63.49%) and then the second most was PC-2 (32.08%) followed by other PCs. It was noted in the present PCA that the highest eigenvalue was captured in PC-1 followed by PC-2 in CKMB versus different parameters of PCI.{Figure 8}

 DISCUSSION



This is an Indian study using periprocedural data comprising Trop-T and CKMB with various procedural parameters.

The present study determined the correlation between elevated Trop-T or CKMB (dependent variable) and the parameters used in PCI (independent variables) by using multivariate analysis through PCA.

Earlier Acharyya et al.[8] reported that Trop-T and CKMB are suitable biomarkers for periprocedural myocardial injury. Moreover, Prasad et al.[9] reported that several parameters such as fluoroscopy time, total stent length, number of stents implanted, PTCA balloon inflation time, type of lesion, etc., are used in balloon angioplasty and it is varied in different patients, and these parameters were closely related to postprocedural Trop-T and CKMB elevation and MACE in 6-month follow-up.[8],[9]

The present study reports for the first time that the association of parameters is associated with the PCI about the elevated Trop-T and CKMB through PCA, which are lacking in the clinical field of the eastern part of India. A past report mentioned that PCA is an essential tool in clinical research to determine different PCs for many parameters, which may reduce a set of intercorrelated variables as procedural parameters into a few dimensions and collecting of a large amount of the variability of the original variables.[18] Among several PCs, PC-1 has the largest possible variance to detect easily some components that are chosen to characterize the correlated variables.[12]

Few past studies have reported on the incidence of cardiovascular diseases through PCA,[17],[19],[20],[21] but the parameters of PCI about the elevated Trop-T and CKMB as the markers of periprocedural myocardial injury by PCA are lacking.

 CONCLUSIONS



The association of different parameters of PCI procedure with periprocedural myocardial injury evaluated through the elevated Trop-T or CKMB is an important finding. It is important to determine the most important variables among them associated with periprocedural myocardial injury through PCA. This will contribute to identifying the risk factors in patients and thus reducing future MACE. A future study is suggested with a larger sample size and multicentric approach for establishing its applicability in routine PCI procedures.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1Abdelmeguid AE, Topol EJ, Whitlow PL, Sapp SK, Ellis SG Significance of mild transient release of creatine kinase-MB fraction after percutaneous coronary interventions. Circulation 1996;94:1528-36.
2Desmet WJ, Dens J, Coussement P, Van de Werf F Does adenosine prevent myocardial micronecrosis following percutaneous coronary intervention? The ADELINE pilot trial. Adenosine limit myocardial necrosis. Heart 2002;88:293-5.
3Wang FW, Osman A, Otero J, Stouffer GA, Waxman S, Afzal A, et al. Distal myocardial protection during percutaneous coronary intervention with an intracoronary beta-blocker. Circulation 2003;107:2914-9.
4Ricciardi MJ, Wu E, Davidson CJ, Choi KM, Klocke FJ, Bonow RO, et al. Visualization of discrete microinfarction after percutaneous coronary intervention associated with mild creatine kinase-MB elevation. Circulation 2001;103:2780-3.
5Selvanayagam JB, Porto I, Channon K, Petersen SE, Francis JM, Neubauer S, et al. Troponin elevation after percutaneous coronary intervention directly represents the extent of irreversible myocardial injury: Insights from cardiovascular magnetic resonance imaging. Circulation 2005;111:1027-32.
6Balian V, Galli M, Marcassa C, Cecchin G, Child M, Barlocco F, et al. Intracoronary ST-segment shift soon after elective percutaneous coronary intervention accurately predicts periprocedural myocardial injury. Circulation 2006;114:1948-54.
7Varani E, Balducelli M, Vecchi G, Gatti C, Ricci Lucchi G, Maresta A Occurrence of non-Q wave myocardial infarction following percutaneous coronary intervention in the stent era: Systemic monitoring of the three markers of myocardial necrosis. J Interven Cardiol 2005;18: 243-8.
8Acharyya M, Haldar SK, Mukhopadhyay D Assessment of peri-procedural myocardial injury in patients of post percutaneous coronary intervention: Diagnosis through serum biomarkers as cardiac troponin T and creatine kinase myocardial isoform. Int Arch BioMed Clin Res 2018;4:19-26.
9Prasad A, Rihal CS, Singh M, Lennon R, Jaffe A, Holmes D Significance of periprocedural myonecrosis for outcomes following percutaneous coronary intervention: An analysis of pre and post-intervention troponin T levels in 5487 patients. Circ Cardiovasc Intervent 2008;1: 10-9.
10Gabriel KR, Odoroff CL Biplots in biomedical research. Stat Med 1990;9:469-85.
11Castelló A, Buijsse B, Martín M, Ruiz A, Casas AM, Baena-Cañada JM, et al; GEICAM researchers. Evaluating the applicability of data-driven dietary patterns to independent samples with a focus on measurement tools for pattern similarity. J Acad Nutr Diet 2016;116:1914-24.e6.
12Zhang Z, Castelló A Principal components analysis in clinical studies. Ann Transl Med 2017;5:351.
13Prasad A, Singh M, Lerman A, Lennon RJ, Holmes DR Jr, Rihal CS Isolated elevation in troponin T after percutaneous coronary intervention is associated with higher long-term mortality. J Am Coll Cardiol 2006;48:1765-70.
14Koskinas KC, Ndrepepa G, Räber L, Karagiannis A, Kufner S, Zanchin T, et al. Prognostic impact of periprocedural myocardial infarction in patients undergoing elective percutaneous coronary interventions. Circ Cardiovasc Intervent 2018;11:e006752.
15Hammer Ø, DAT, Ryan PD. PAST: Paleontological statistics software package for education and data analysis. Palaeontol Electron 2001;4:1-9.
16Liberda EN, Zuk AM, Tsuji LJS Complex contaminant mixtures and their associations with intima-media thickness. BMC Cardiovasc Disord 2019;19:289.
17Mondal T, Acharyya M Principal component analysis of categorized parameters used in the study of diabetes mellitus and dyslipidemia in association with carotid intima medial thickness. Int Arch Bio Med Clin Res 2020;6:GM14-18.
18Burt C Factor analysis and canonical correlations. Br J Psychol 1948;1:95-106.
19Agarwal S, Jacobs DR Jr, Vaidya D, Sibley CT, Jorgensen NW, Rotter JI, et al. Metabolic syndrome derived from principal component analysis and incident cardiovascular events: The multi ethnic study of atherosclerosis (MESA) and health, aging, and body composition (health ABC). Cardiol Res Pract 2012;2012:919425.
20Badaruddoza KR, Kaur M Principal component analysis of cardiovascular risk traits in three generations cohort among Indian Punjabi population. J Adv Res 2015;6:739-46.
21Ricciardi C, Valente AS, Edmund K, Cantoni V, Green R, Fiorillo A, et al. Linear discriminant analysis and principal component analysis to predict coronary artery disease. Health Informatics J 2020;26:2181-92.