The Role of Insulin-Like Growth Factor-1 and Pregnancy-Associated Plasma Protein-A in Diagnosis of Acute Coronary Syndrome and Its Related Morbidities.

INTRODUCTION: Pregnancy-associated plasma protein-A (PAPP-A) is a metalloproteinase that plays a role in atherosclerotic plaque destabilization. In recent studies, insulin-like growth factor-1 (IGF-1) has been introduced as a mediator of atherosclerosis. PAPP-A and IGF-1 level may be important diagnostic indicators of acute coronary syndrome (ACS).
OBJECTIVE: The present study tried to assess the diagnostic role of IGF-1 and PAPP-A biomarkers in ACS spectrum.
METHODS: The serum level of IGF-1, PAPP-A and troponin I was determined in 121 consecutive patients with ACS. Relationships were assessed by t-test, ANOVA and the non-parametric equivalent. Accuracy of biomarkers was measured by the area under the ROC curve (AUC) and optimal cut-off points to diagnose STEMI and NSTEMI using Youden index.
RESULTS: In patients with acute ST segment elevation myocardial infarction (STEMI), all of these three biomarkers were significantly higher than those in patients with unstable angina (P= 0.028 for IGF-1, P<0.001 for PAPP-A and Troponin-I). Mean level of IGF-1 in patients with renal failure was significantly higher than that in patients without renal failure (137.9±35.1 vs 105.1±46.9, P=0.003), but PAPP-A and serum Troponin-I level had no significant difference in renal failure groups (P>0.05). ROC curve analysis showed that after Troponin-I, PAPP-A was a good discriminator between patients with STEMI and patients with unstable angina (AUC=0.79). Optimum cut-off value for PAPP-A was found to be 89.2 ng/ml, with sensitivity and specificity of 66.7% and 83.8%, respectively.
CONCLUSION: PAPP-A can be a novel biomarker for both identification of patients with STEMI and risk stratification in patients with ACS.

Introduction

Insulin-like growth factor-1 (IGF-1) and pregnancy-associated plasma protein-A (PAPP-A) are two biomarkers that have been recently considered in the management of patients with acute coronary syndrome (ACS).

IGF-1 is a functional encoded protein produced by IGF-1 gene. IGF-1 has been shown to oppose endothelial dysfunction by interacting with endothelial binding sites leading to nitric oxide production (1, 2). Also, IGF-1 has some anti-inflammatory and vasodilation properties that result in coronary flow preservation (3). Moreover, the atherogenicity effect of IGF-1 has been suggested due to its potential role to induce vascular smooth muscle cell (VSMC) proliferation leading to arterial obstructive lesions (4, 5). However, there are still conflicting results concerning IGF-1 and coronary artery diseases. Some studies suggested that a decreased level of IGF-1 in patient with ST segment elevation myocardial infarction (STEMI) can be used as a diagnostic marker of myocardial necrosis (6, 7). Conversely, some randomized trials reported diagnostic importance of an increased level of IGF-1 in the setting of stent restenosis and lowering IGF-1 levels with somatostatin analogues such as angiopeptin (8, 9).

PAPP-A is a human protein encoded by PAPP-A gene (10). This agent is a zinc binding metalloproteinase which splits IGF binding proteins (IGFBPs) especially IGFBP-4, therefore with increasing IGFs allowing their actions (11). Circulating PAPP-A has recently been suggested as an independent predictor of next and late adverse cardiovascular events. Moreover, increased activity of PAPP-A has been reported in ACS, hence it is a promising biomarker for risk stratification in these patients (12–15). Clinically, an increase in serum level of PAPP-A has been shown in patients with ACS that may be an indicator for coronary plaques complexity and vulnerability (16). In fact, higher levels of PAPP-A were found in cells and extracellular matrix of the plaques that showed rupture or erosion compared to stable plaques (12). Despite evidence on the potential role of IGF-1 and PAPP-A as a precursor for stimulating or preserving coronary flow and cardiac function, whether the plasma levels of these two biomarkers provide diagnostic information in patients with coronary lesions has remained unclear. The present study aimed to assess the diagnostic role of IGF-1 and PAPP-A biomarkers for a wide spectrum of coronary artery diseases and their correlation with conventional cardiovascular risk factors.

Methods

Study design and participants

This cross-sectional study was conducted from 2009 until 2010 in Imam Khomeini Hospital of Naghade, affiliated with Urmia University of Medical Science, West Azerbaijan Province. The study was approved by the Institutional Ethics Committee and written informed consent was obtained from the study population.

Study population

All patients diagnosed as ACS based on the American College of Cardiology guideline definitions by considering troponin levels, ECG abnormalities and clinical context (17); hospitalized in cardiac and critical care unit, admitted within one hour after the onset of chest pain were eligible. ACS was diagnosed based on the latest guideline of American College of Cardiology released on 2014 (especially by serum level of troponin I) (17–19). The exclusion criteria were progressive and serious systemic disorders such as liver diseases, overt heart failure, history of major surgeries, pregnancy, known infectious or inflammatory disorders, or malignancies. Sampling was conducted prospectively and consecutively. A total of 145 patients were admitted during the study and 18 were excluded due to the exclusion criteria and 6 were excluded due to incomplete lab data. Finally, 121 patients were included for the analysis.

Definitions

ACS is a spectrum of conditions from unstable angina (UA) to STEMI. UA is a condition in which biomarkers like troponin do not increase. Inversely, myocardial infarction leads to an increase in troponin. According to the American college of cardiology, STEMI is a condition defined with ST segment elevation plus a rise in troponin I levels above > 0.5 ng/mL and NSTEMI is described with positive troponin levels without ST segment elevation (20, 21). Patients were categorized with eGFR lower than 90 mL/min/1.73 m2 as chronic kidney disease (CKD) (22). Left ventricular systolic dysfunction (LVSDF) is qualitatively categorized to severe (<30%), moderate (30–40%), mild (40–50%) and preserved (>50%) stages (23).

Data gathering

Baseline characteristics including demographics, medical history, and troponin indices were collected from hospital records and approved by two cardiologists using a pre-designed checklist. Venous blood samples were taken from each patient within three hours after the onset of chest pain for routine tests and for measuring serum level of IGF-1 and PAPP-A. Serum level of IGF-1 was determined by the DEMEDITEC IGF-1 600 ELISA Kit (Enzyme-Linked Immunosorbent Assay). Standard range was defined according to the manufacturer’s protocol 10 – 600 ng/ml. The DEMEDITEC PAPP-A US ELISA Kit was also used for PAPP-A assay based on the sandwich principle. The assay ranged 0–450 ng/ml. To minimize inter-observer variability, all measurements were followed by a single researcher.

Data analysis

Results are presented as mean ± standard deviation (SD) and median [Interquartile range (IQR)] for quantitative variables and summarized by absolute frequencies and percentages for categorical variables. Quantitative variables are also compared with independent t-test for two groups and one-way ANOVA for more than two groups. We used Mann-Whitney U and Kruskal-Wallis H test for variables with non-normal distribution. Post hoc analysis and correction of error type-I in multiple comparisons were conducted based-on Bonferroni method. Spearman correlation test was applied to determine the correlation between the level of IGF-1, PAPP-A biomarkers and Troponin-I. We used Shapiro-Wilk test and graphical approaches for assessing normality.

Accuracy of biomarkers was measured by the area under the ROC curve (AUC). ROC analysis was used to choose the optimal cut-off points to diagnose STEMI and NSTEMI using highest values of the Youden index and calculated sensitivity and specificity with 95% confidence interval (CI) for optimal cut-off points. Statistical significance was determined as a p value of ≤ 0.05. All statistical analyses were performed using SPSS software version 22.0 (Armonk, NY: IBM Corp) and STATA version 14 (Stata Corp LLC).

Results

As shown in table 1, among 121 participants, their mean age was 69.09±13.00 years (range: 38 to 95 years) and the majority of them (68%) were male. Of total patients, 25.4% were smokers and more than 20% had no risk factor of hypertension, hyperlipidemia or diabetes mellitus. Almost 56% of the patients with ACS had unstable angina, and severe left ventricle systolic dysfunction (LVSDF) was detected in 17.4% (Table 1).

Table 1:

Baseline characteristics of study population

Variable	Frequency (%)
Gender
Female	38 (31.4)
Male	83 (68.6)
Risk factors
Hypertension	32 (24.6)
Hyperlipidemia	26 (20)
Smoker	33 (25.4)
Diabetes mellitus	39 (30)
Type of acute coronary syndrome
Unstable angina	68 (56.2)
NSTEMI	25 (20.7)
Extensive anterior MI	10 (8.3)
Inferior MI	12 (9.9)
Anteroseptal MI	3 (2.5)
Anterolateral MI	2 (1.7)
LV systolic dysfunction
Preserved EF	1 (0.8)
Mild LVSDF	62 (51.2)
Moderate LVSDF	37 (30.6)
Severe LVSDF	21 (17.4)

MI: Myocardial infarction; NSTEMI: Non-ST segment elevation myocardial infarction; LVEF: Left ventricular ejection fraction

IGF-1 levels and demographic characteristics

No difference was observed in the mean IGF-1 level (ng/ml) between men and women (115.0±48.3 vs 102.6±42.2, p=0.173). Also, no difference was observed in serum level of IGF-1 between hypertensive and normotensive groups (112.3±40.2 vs 110.7±49.0, p=0.866), the groups with and without hyperlipidemia (107.1±46.0 vs 112.2±47.0, p=0.618), smoker and non-smoker groups (117.3±48.9 vs 108.8 ±45.8, p=0.373), and also diabetics and non-diabetics (117.1±43.2 vs 108.3±48.2, p=0.330).

PAPP-A levels and Troponin I, and demographic characteristics

PAPP-A level was significantly higher in male than in female patients [median (IQR): 70.0 (150.0) vs 19.0 (77.0)]. Also, troponin I was higher in male than in female patients and this difference was marginally significant (p=0.054). PAPP-A levels and Troponin-I in patients with or without any risk factor was not significantly different (p>0.05) (Table 2).

Table 2:

Distribution and association of IGF-1, PAPP-A and Troponin-I with sex, risk factors, LVSDF, renal failure, type of ACS and type of STEMI

Variable	Insulin-like growth factor-1	p	Pregnancy-associated plasma protein-A	p	Troponin-I	p
	Mean (SD)		Mean (SD)		Mean (SD)
Sex		0.173		0.002		0.054
Male	115.0 (48.3)		134.2 (185.3)		0.88 (2.1)
Female	102.6 (42.2)		48.3 (68.6)		0.35 (1.2)
Hypertension		0.866		0.875		0.939
Yes	112.3 (40.2)		100.2 (183.6)		0.76 (1.7)
No	110.7 (49.0)		109.7 (155.8)		0.69 (1.9)
Hyperlipidemia		0.618		0.830		0.712
Yes	107.1 (46.0)		75.4 (84.7)		1.1 (3.0)
No	112.2 (47.0)		115.9 (177.8)		0.61 (1.4)
Smoker		0.373		0.123
Yes	117.3 (48.9)		134.0 (174.4)		1.1 (2.8)	0.219
No	108.8 (45.8)		97.1 (158.2)		0.57 (1.3)
Diabetes mellitus		0.330		0.706		0.296
Yes	117.1 (43.2)		110.5 (190.5)		0.58 (1.5)
No	108.3 (48.2)		105.6 (149.2)		0.78 (2.0)
Renal failure		0.003		0.294		0.351
Yes (n=22)	137.9 (35.1)		142.4 (214.9)		0.95 (1.8)
No (n=99)	105.1 (46.9)		99.4 (149.2)		0.66 (1.9)
LVSDF		0.307		0.792		0.093
Mild LVSDF	107.0 (42.5)		95.6 (156.0)		0.43 (1.0)
Moderate LVSDF	111.1 (49.7)		119.6 (180.5)		1.1 (2.8)
Severe LVSDF	125.2 (52.5)		121.6 (158.9)		0.82 (1.7)
Type of ACS		0.028		<0.001		<0.001
STEMI (n=27)	132.2 (58.2)		184.8 (171.5)		1.9 (3.2)
NSTEMI (n=25)	109.3 (44.8)		118.6 (125.1)		0.91 (1.6)
UA (n=68)	104.1 (40.1)		72.8 (163.7)		0.15 (0.44)
Type of STEMI		0.890		0.488		0.475
Inferior (n=12)	137.3 (64.5)		144.6 (117.5)		1.1 (1.6)
Antero Septal (n=3)	146.0 (56.1)		298.3 (192.7)		3.0 (2.7)
Extensive Anterior (n=10)	126.7 (60.9)		220.3 (219.4)		3.0 (4.7)
Antero Lateral (n=2)	108.2 (12.1)		79.0 (111.7)		0.15 (0.1)

SD: Standard deviation; LVSDF: left ventricle systolic dysfunction; ACS: Acute coronary syndromes; STEMI: ST segment elevation myocardial infarction; NSTEMI: Non-ST-segment elevation myocardial infarction; UA: Unstable angina

Biomarkers and LVSDF

There were no significant differences in serum level of IGF-1 and PAPP-A, and outcome in terms of different stages of LVSDF (P=0.307 and p=0.792 respectively). But Troponin-I was lower in mild LVSDF than moderate and severe LVSDF, and this difference was marginally significant (p=0.093) (Table 2).

Correlation between IGF-1, PAPP-A and Troponin-I

Our study showed direct significant correlation between troponin I and two parameters of IGF-1 (r=0.29, p=0.001) and PAPP-A (r=0.40, p<0.001). Moreover, the correlation between PAPP-A and IGF-1 indicators was direct and significant (r=0.29, p=0.001).

Comorbidities and level of study biomarkers

Twenty-two patients (18.2%) suffered renal failure. Mean level of IGF-1 in patients with renal failure was significantly higher than those without renal failure (137.9±35.1 vs 105.1±46.9, p=0.003). No difference was found in the level of PAPP-A and serum troponin I level between the patients with and without renal failure (p>0.05) (Table 2).

Type of ACS and biomarkers

All assessed biomarkers were significantly associated with the type of ACS (p<0.05). Mean serum level of IGF-1 was 132.1, 109.3 and 104.0 for STEMI, NSTEMI and UA, respectively (p=0.028). In post hoc analysis, a significant difference was observed just between UA and STEMI (p= 0.023) and the difference between NSTEMI and STEMI was not significant (p=0.221) (Table 2).

The relationship of PAPP-A and troponin I with type of ACS was also significant (p<0.001). Post hoc analysis showed PAPP-A was significantly lower in UA compared to STEMI [Median (IQR): 12.8 (73.3) vs 116.0 (220.0); p<0.001] and NSTEMI [Median (IQR): 12.8 (73.3) vs 108.0 (128.0); p=0.003]. Also, post hoc analysis showed Troponin-I was significantly lower in UA compared to STEMI [Median (IQR): 0.01 (0.02) vs 0.56 (2.6); p<0.001] and NSTEMI [Median (IQR): 0.01 (0.02) vs 0.18 (1.1); p<0.001]. But differences of PAPP-A and troponin I level between STEMI and NSTEMI was not significant (p=0.155 and p=0.113, respectively) (Table 2).

Accuracy of biomarkers to predict STEMI and NSTEMI

In the ROC curve, biomarker levels were plotted for their ability to predict STEMI comparing patients with UA. AUC (95% CI) for IGF-1, PAPP-A and Troponin-I was 0.61 (0.48 to 0.74), 0.79 (0.69 to 0.89), and 0.89 (0.82 to 0.96), respectively. So, IGF-1 level was not able to predict STEMI comparing patients with UA (Figure 1). The optimum cut-off value for PAPP-A was found to be 89.2 ng/ml with 67% sensitivity and 84% specificity. The optimum cut-off value for Troponin-I was found to be 0.086 with 85% sensitivity and 87% specificity (Table 3). Also based-on ROC curve, biomarker levels were plotted for their ability to predict NSTEMI comparing patients with UA. AUC (95% CI) for IGF-1, PAPP-A and Troponin-I was 0.50 (0.37 to 0.64), 0.68 (0.55 to 0.82), and 0.86 (0.78 to 0.94), respectively. So, IGF-1 level was not able to predict NSTEMI comparing patients with UA and also, PAPP-A had low accuracy for this prediction. The optimum cut-off value for Troponin-I was found to be 0.038 with 88% sensitivity and 81% specificity (Table 3).

Figure 1:

Receiver operator characteristic curve of PAPP-A, IGD-1 and Troponin-I for predict STEMI

Table 3:

Optimal cut-off point and accuracy with 95% confidence interval (CI) of PAPP-A and Troponin-I for diagnosis of STEMI and NSTEMI

Biomarker	Classify group	Optimal cut-off point	Sensitivity (95% CI)	Specificity (95% CI)
PAPP-A	STIME-UA	89.2	66.7 (46.0, 83.5)	83.8 (72.9, 91.6)
	NSTIME-UA	95.6	60.0 (38.7, 78.9)	85.3 (74.6, 92.7)
Troponin-I	STIME-UA	0.086	85.2 (66.3, 95.8)	86.8 (76.4, 93.8)
	NSTIME-UA	0.038	87.5 (67.6, 97.3)	80.9 (69.5, 89.4)

Discussion

Although evidence shows the relationship between IGF-1 concentration and different components of metabolic syndrome and cardiovascular risk profile such as body mass index, waist circumference and insulin resistance, the value of this biomarker for predicting cardiovascular events as well as for assessing the type and severity of cardiac ischemic events remains challenging (24–26). As shown in our study, IGF-1 levels were not associated with cardiovascular risk factors. Our results showed serum level of IGF-1 was significantly higher in patients with STEMI than that in patients with UA. Post hoc analysis also revealed a significant difference just between UA and STEMI. But, ROC curve analysis for IGF-1 was not significant for predicting STEMI.

It has been previously shown that IGF-1 can enhance ischemic eventuality, and reduce ischemia/reperfusion damage (27). Furthermore, expression of IGF-1 in animal models could protect cardiovascular system from cardiomyocyte death after infarction, and attenuate ventricular dilation, wall stress, and cardiac hypertrophy (28, 29). The controversy between IGF-1 level and cardiac-related events in different studies can be attributed to factors such as probable effect of genetic predisposition. For example, different genetic pathways and upregulation of IGF-1 gene in our population can produce active form of IGF-1. On the other hand, the pathway responsible for activating IGF-1 and its effects on cardiac preconditioning and plaque stability may be suppressed by some factors that should be further assessed.

Regarding IGF-1 and its relationship with ACS and its prognosis, the result of the study by Sekuri et al. showed that significantly decreased levels of IGF-1 in STEMI group of ACS may be used as a diagnostic marker for myocardial necrosis, but since no relationship existed between IGF-1 level and cardiovascular events in 90 days, this parameter cannot be suggested as a negative prognostic factor (6).

Yalcin et al. recently revealed that high IGF-1 levels may identify the patients who are high-risk for stent thrombosis (30).

Interestingly, in patients with renal failure, IGF-1 was significantly higher than that in patients without renal failure. Regarding the effects of IGF-1 on ischemic events and plaque vulnerability, it was higher just in patients with STEMI than that in patients with unstable angina. Several studies propose that IGF-1 is involved in the development of atherosclerosis, and some of its consequences (31, 32).

In the current study, level of IGF-1 was significantly higher in patients with renal failure. Interaction between IGF and kidney disease has been discussed in two recent studies which confirm our results. Bach LA and his colleges showed perturbed regulation of the IGF system in a number of kidney diseases. Enhanced IGF activity was noted in early diabetic nephropathy and polycystic kidneys, but IGF resistance was implicated in chronic kidney failure (33). In a cohort study by Teppala et al., they found higher IGF-1 levels to be positively associated with CKD, and suggested assessment of serum IGF-1 may be valuable for risk stratification of CKD (27).

As presented in our study, it appears that measuring the level of IGF-1 could have a major role in predicting adverse events in patients with ACS especially renal failure. According to our results, the occurrence or uncovering of renal failure as a one of main ACS morbidities can be predicted by high serum levels of IGF-1. Similarly, in Teppala et al. study, higher serum IGF-1 levels were positively associated with CKD after adjusting for age, sex, race/ethnicity, education levels, smoking, alcohol intake, body mass index, diabetes, hypertension and serum cholesterol (27). Thus, it appears that increasing the level of IGF-1 can be helpful to assess tendency to renal failure especially in patients with coronary artery disease. In our study, there were no significant relationship between PAPP-A level and any of cardiac risk factors. Also, relationship of this biomarker with renal failure was not significant.

In two recent studies conducted by Konev et al. and Hjortebjerg et al. it was reported that IGFBP-4 fragments from PAPP-A were suggested as a novel biomarker for cardiac risk assessment (33, 34).

The relationship between PAPP-A and coronary angiographic features in patients with ACS was examined in a study by Shehata et al. Higher PAPP-A levels in patients with ACS were associated with unfavorable coronary anatomy and complex angiographic plaque features (35). We were not able to assess angiographic features of study group. Literature review about PAPP-A and ACS or cardiac events shows similar results. In this context, Bonaca et al. conducted a study of usefulness of PAPP-A for risk assessment in NSTEMI. PAPP-A was independently associated with recurrent cardiovascular events in these patients. So, they suggest PAPP-A as a candidate prognostic marker in patients with ACS (36).

Our study revealed a significant association between serum level of PAPP-A and type of ACS and also post hoc analysis showed a significant difference between UA and STEMI. In a similar study by Gururajan et al., PAPP-A had a pivotal role in the pathogenesis of atherosclerosis. PAPP-A level was higher in patients diagnosed as ACS in comparison with the controls. PAPP-A was able to differentiate ischemic and non-ischemic patients. AUC was 0.904, 95% CI (0.874–0.929) with 85% specificity and 90% sensitivity (P < 0.0001). The acquired cut-off value was 0.55 μg/mL, and above this value, PAPP-A was regarded to be positive (37). In line with our study, Laterza et al. reported cut-off value of 0.22 mIU/L, with a sensitivity of 66.7% and a specificity of 51.1% for PAPP-A to be a predictor of adverse events at 30 days, but it was inferior to cardiac Troponin-I (38). Detailed clarification of the pathophysiologic role of PAPP-A in detecting STEMI (myocardial necrosis) could be promising for the innovation of specific biomarker-directed therapies.

Limitations

Our study has limitations that warrant consideration. These limitations originate mainly from the inclusion of the in-hospital period without long-term follow-up, failure to consider angiographic data, and lack of a healthy population as the control group. Further studies are recommended for long-term follow-up of patients, comparing IGF-1 and PAPP-A levels between patients with ACS and a control group, analysis of angiographic data including involved coronary artery, Gensini score, syntax score, need to revascularization and finally determining correlation of these parameters with IGF-1 and PAPP-A levels.

Conclusions

In summary, it appears measuring serum level of PAPP-A could predict STEMI in patients with ACS. Serum level of IGF-1 is not able to predict STEMI in cases admitted for ACS. Regarding higher IGF-1 levels in these patients with CKD, this biomarker may be valuable for cardiac risk stratification of CKD, but further studies are necessary to clarify this point.