This article is mentioned in:

Introduction

At present the incidence of contrast-induced nephropathy (CIN) has been increasing in patients undergoing coronary angioplasty, due to the increasing use of contrast media (1). CIN is usually described as an increase in serum creatinine of 0.5 mg/dl or a 25% increase from the baseline value 48 h following the imaging procedure (2). CIN has been reported to occur in ≥14.5% of unselected patients undergoing coronary angioplasty, and is considered to be the third leading cause of hospital-acquired acute renal failure (3). It is more commonly associated with adverse clinical outcomes, increased medical care costs, prolonged hospitalization, and increased in-hospitality morbidity and mortality (4). The major risk factors of CIN are reduced circulation volume, the type and volume of contrast agent, simultaneous administration of nephrotoxic agents and pre-existing renal dysfunction, particularly that due to diabetic nephropathy (5–8). Since the poor prognosis of patients with diabetic nephropathy could largely attribute to CIN, these patients may benefit greatly from preventive interventions. The precise mechanisms underlying the pathogenesis of CIN have not been well established. However, it is widely speculated that the underlying mechanism of CIN may involve an injury to the renal medulla caused by a combination of reduced blood flow, direct tubular toxicity and an osmotic effect (9). The direct tubular toxicity may be associated with reactive oxygen species (ROS), which are generated following the administration of contrast agent (10). Currently, the preventive treatments for CIN involve reducing contrast exposure, intravenous volume expansion with a saline hydration, and usage of low or iso-osmolarity contrast agent; however, these may provide incomplete prevention of CIN and thus, adjunctive pharmacotherapies in clinical practice have emerged (11). Among these, N-acetylcysteine (NAC) has been of interest since it was initially reported by Tepel et al (12). NAC as a direct scavenger of free radicals may improve blood flow via nitric oxide-mediated pathways, and it is a precursor of glutathione synthesis, providing vasodilation and antioxidant activity against CIN (13). Therefore, oral NAC therapy may be an alternative method for CIN prevention, providing safety, low cost and few side effects (14).

It has been reported that oral NAC may more effectively provide protection against CIN compared with intravenous hydration alone (15). Results of the initial study (12) of oral NAC for the prevention of CIN were encouraging, while the bioavailability of oral NAC may be low and exhibited mixed results; a few trials demonstrated the reduction of CIN incidence by oral NAC therapy (16–21), and most trials revealed no significant CIN prevention (22–34). The aim of the present study was to determine whether oral NAC therapy is beneficial for CIN prevention in clinical practice, using a meta-analysis.

Materials and methods

Search strategy and selection criteria

A comprehensive study was performed to search all published randomized controlled trials (RCT) until January 1, 2015 which concerned oral NAC treatment to prevent CIN in patients undergoing coronary angioplasty, using searching engines such as Medline (https://www.nlm.nih.gov/bsd/pmresources.html), Embase (https://www.elsevier.com/solutions/embase-biomedical-research) and Cochrane (http://uk.cochrane.org/). The search terms were as follows: N-acetylcysteine, acetylcysteine, NAC, cardiac catheterization, coronary angioplasty, coronary angiogram, percutaneous coronary intervention, contrast-induced nephropathy, contrast-induced nephrotoxicity, contrast-medium nephrotoxicity, contrast medium-induced nephropathy and contrast-induced acute kidney injury. RCTs were limited to those with human subjects. A manual search of the results was then performed for the qualifying trials. Abstracts alone or meeting proceedings were excluded. This search strategy was performed comprehensively until no new potential citations were found on review of the reference list of retrieved papers. All of the studies published in English which met the following inclusion criteria were included: Subjects underwent coronary angioplasty, randomization of oral NAC and placebo, and data regarding CIN incidence. Exclusion criteria were as follows: <18 years of age, known allergy or hypersensitivity to NAC, dialysis patients and those with ST-segment elevation myocardial infarction undergoing primary angioplasty.

Data extraction and quality assessment

Two investigators (Dr Jing-Xiu Li and Dr Nan-Nan Liu) were assigned independently to assemble the information of each study as follows: First author name, surgery type (coronary angiography or percutaneous coronary intervention), study design (RCT, prospective or not), control types (placebo or not), blinding types (double-blinding or not), NAC regimen, sample size, mean age, percentage of males, the incidence of CIN and length of hospitalization in each group. Disagreements were settled through discussion and consensus.

Risk of bias

The majority of selected trials were conducted in randomized sequence generation and allocation concealment, and the participants were divided randomly. All of them were considered to be of low bias risk.

Statistical analysis

The relative risk (RR) was estimated with 95% confidence interval (CI) for dichotomous outcomes. Heterogeneity was reported with the I2 statistic, using a fixed-effects model, and >50% of I2 was considered to be statistically significant. Begg and Egger tests were performed for presenting the publication bias, and the potential bias was analyzed with visual inspection of the Begg funnel plots in which the log RRS plotted against their standard errors. P<0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed using STATA software, version 12.0 (StataCorp LP, College Station, TX, USA) and RevMan 5.2 (The Nordic Cochrane Centre, Copenhagen, Denmark).

Results

Description of the studies

A total of 19 placebo-control RCTs were included in this study, consisting of 4,514 patients. The flow of identified studies through the selection process is shown in Fig. 1. The characteristics at baseline and design of the selected studies are shown in Tables I and II. The range of participant number was 36–2,308, including men and women. The range of total NAC dosage was 1,200–12,000 mg. The effects of oral NAC on CIN prevention were also compared.

Figure 1. Process of study selection of place-controlled, randomized trials. NAC, N-acetylcysteine.

Table I. Characteristic data of studies included in the meta-analysis.

Table I.

Characteristic data of studies included in the meta-analysis.

Authors, year	Patients (I/C)	Renal function for inclusion (mg/dl)	CIN definition (SCr)	Contrast agent	Avg. contrast volume (ml)	Hydration regimen	Cumulative NAC dose (mg)	Diabetes mellitus (%)	Refs.
Ochoa et al, 2004	80 (36/44)	>1.8 (male) >1.6 (female)	0.5 mg/dl or ≥25% above	Low osmolarity	155	0.9% saline 1,500 ml	1,000 (1 h before and 4 h after)	55	(16)
MacNeill et al, 2003	43 (21/22)	>1.5	≥25% above baseline after 72 h	Low osmolarity	103	0.45% 1 ml/kg/h	600 (bid for 4 doses)	46.50	(17)
Briguori et al, 2002	183 (92/91)	>1.2	≥25% above baseline after 48 h	Low osmolarity	140	0.45% 1 ml/kg/h	600 (bid pre/post)	37.70	(18)
Diaz-Sandoval et al, 2002	54 (25/29)	>1.4	0.5 mg/dl or ≥25% above	Low osmolarity	189	0.45% 1 ml/kg/h 2.12 h before, 12 h after	600 bid for 4 doses	38.90	(19)
Kay et al, 2003	200 (102/98)	>1.2 mg/dl	≥25% above baseline within 48 h	Low osmolarity	120	0.9% 1 ml/kg/h 12 h before, 6 h after	600 (PO bid pre/post)	37.50	(20)
Shyu et al, 2002	121 (60/61)	>2.0 mg/dl	0.5 mg/dl after 48 h	Low osmolarity	119	0.45% 1 ml/kg/h 12 h before, 12 h after	400 (PO bid pre/post)	63.60	(21)
ACT Investigators, 2011	2,308 (1,172/1,136)	>1.5	0.5 mg/dl 48.96 h	High osmolarity Low osmolarity Iso-osmolar	100	0.9% 1 ml/kg/h 6.12 h before, 6.12 h after	1,200 (bid pre/post)	68.45	(22)
Allaqaband et al, 2002	85 (45/40)	>1.6	0.5 mg/dl after 48 h	Low osmolarity	122	0.45% 1 ml/kg/h 12 h before, 12 h after	600 (bid pre/post)	48.29	(23)
Amini et al, 2009	90 (45/45)	>1.5	0.5 mg/dl or ≥25% above baseline after 48 h	Low osmolarity Iso-osmolar	118	0.9% saline 1,000 ml	600 (bid pre/post)	N/A	(24)
Baskurt et al, 2009	145 (73/72)	>1.3	0.5 mg/dl or ≥25% above baseline after 48 h	Low osmolarity	113	0.9% 1 ml/kg/h 12 h before, 12 h after	600 (bid pre/post)	30.34	(25)
Oldemeyer et al, 2003	96 (49/47)	>1.2	0.5 mg/dl or ≥25% above baseline after 48 h	Low osmolarity	134	0.45% 1 ml/kg/h 12 h before, 12 h after	1,500 (bid for 4 doses)	44.79	(26)
Durham et al, 2002	79 (41/38)	>1.7	0.5 mg/dl after 48 h	Low osmolarity	84	0.45% 1 ml/kg/h 12 h before, 12 h after	1,200 bid pre/post	48.10	(27)
Ferrario et al, 2009	200 (99/101)	>1.5	0.5 mg/dl or ≥25% above baseline within 72 h	Iso-osmolar	180	0.9% 1 ml/kg/h 12.24 h before, 24 h after	600 (bid pre/post)	25	(28)
Fung et al, 2004	91 (46/45)	>1.5	0.5 mg/dl after 48 h	Low osmolarity	135	0.9% saline 100 ml/h 12 h before, 12 h after	600 (PO, thrice pre/post)	52.74	(29)
Goldenberg et al, 2004	80 (41/39)	>1.5	0.5 mg/dl after 48 h	Low osmolarity	111	0.45% 1 ml/kg/h 12 h before, 12 h after	600 (PO, thrice pre/post	43.75	(30)
Gomes et al, 2005	156 (77/79)	>1.2	0.5 mg/dl after 48 h	Low osmolarity	102	0.9% 1 ml/kg/h 12 h before, 12 h after	600 (PO, bid pre/post	51.90	(31)
Kimmel et al, 2008	36 (19/17)	>1.2	0.5 mg/dl or ≥25% above baseline	Low osmolarity	219	0.45% 1 ml/kg/h 12 h before, 12 h after	600 (PO, bid pre/post)	30.60	(32)
Ozcan et al, 2007	176 (88/88)	>1.2	0.5 mg/dl or ≥25% above baseline after 48 h	Low osmolarity	110	0.9% 1 ml/kg/h 6 h before, 6 h after	600 (PO, bid pre/post)	46.60	(33)
Yang et al, 2014	318 (157/161)	N/A	≥25% above baseline within 72 h	Low osmolarity	124	0.9% 1.5 ml/kg/h 6 h before, 6 h after	600 (PO, bid pre/post)	25.50	(34)

[i] 1 mg/dl=88.4 µmol/l. I/C, interventions/controls; CIN, contrast-induced nephropathy; SCr, serum creatinine; NAC, N-acetylcysteine; N/A, data not available; bid, bis in die (twice a day); PO, oral.

Table II. Baseline and difference in SCr.

Table II.

Baseline and difference in SCr.

	CIN (%)		Acetylcysteine SCr (mg/dl)		Control SCr (mg/dl)
Author, year	Acetylcysteine	Control	Baseline	Second SCr	Baseline	Second SCr	Refs.
Ochoa et al, 2004	3	25	2.02±0.56	2.10±0.81	1.93±0.53	2.10±0.74	(16)
MacNeill et al, 2003	5	32	1.89±0.38	1.90±0.36	1.88±0.41	2.14±0.87	(17)
Briguori et al, 2002	6.50	11	1.54±0.4	1.48±0.36	1.5±0.4	1.53±0.45	(18)
Diaz-Sandoval et al, 2002	8	45	1.66±0.06	1.53±0.09	1.56±0.05	1.88±0.09	(19)
Kay et al, 2003	4	12	1.35	1.22	1.36	1.38	(20)
Shyu, et al, 2002	3.30	24.60	2.8±0.8	2.5±1.0	2.8±0.8	3.1±1.0	(21)
ACT Investigators, 2011	12.70	12.70	1.2±0.5	N/A	1.2±0.5	N/A	(22)
Allaqaband et al, 2002	17.70	15.30	2.2±0.73	2.22±1.00	2.03±0.79	2.03±0.48	(23)
Amini et al, 2009	11.10	14.30	1.736±0.42	2.083±0.4	1.736±0.17	2.185±0.1	(24)
Baskurt et al, 2009	10	6.90	1.39±0.24	1.47±0.38	1.3±0.20	1.38±0.34	(25)
Oldemeyer et al, 2003	8.20	6.40	1.63±0.81	N/A	1.66±0.65	N/A	(26)
Durham et al, 2002	26.30	22	2.3±0.5	N/A	2.2±0.4	N/A	(27)
Ferrario et al, 2009	8.10	5.9	N/A	N/A	N/A	N/A	(28)
Fung et al, 2004	17.40	13.30	2.27±0.54	2.45±0.65	2.37±0.61	2.40±0.70	(29)
Goldenberg et al, 2004	10	8	2.0±0.4	N/A	1.9±0.3	N/A	(30)
Gomes et al, 2005	10.40	10.10	N/A	N/A	N/A	N/A	(31)
Kimmel et al, 2008	5.30	5.90	1.51±0.23	N/A	1.65±0.65	N/A	(32)
Ozcan et al, 2007	12.50	13.60	1.4	1.42	1.4	1.46	(33)
Yang et al, 2014	4.46	3.11	N/A	N/A	N/A	N/A	(34)

[i] CIN, contrast-induced nephropathy; SCr, serum creatinine; N/A, data not available.

Quality assessment of the trials and publication bias

The selected trials in the meta-analysis were well-designed and reasonably conducted, adequately implementing randomized sequence generation and allocation concealment. The participants among them were blinded. All of the selected studies had a low risk of bias, and the details are shown in Fig. 2. Publication bias assessed by Egger's test is shown in Fig. 3.

Figure 2. Risk-of-bias analysis. (A) Risk-of-bias summary: Author's judgments about each risk-of-bias item for the included studies. (B) Risk-of-bias graph: Author's judgments concerning each risk-of-bias item across all the included studies.

Figure 3. Publication bias for OR of the incidence of contrast-induced neuropathy. SE, standard error; OR, odds ratio.

CIN incidence

The baseline characteristics revealed no significant difference between history of coexistent disease and routine prophylactic therapies. The CIN incidence was 247 patients in the oral NAC group (n=2,269) and 278 patients in the control group (n=2,245), pooling all of the 19 trials. There was no statistical significance (RR, 0.84; 95% CI, 0.65–1.10; P=0.20, Fig. 4), with no heterogeneity between trials (I2=29%, P=0.12).

Figure 4. All included studies, relative risk (fixed effect model). CI, confidence interval; M-H, Mantel Haenszel; NAC, N-acetylcysteine.

Discussion

In this meta-analysis, 19 RCTs were combined in order to evaluate the effects of oral NAC on CIN prevention in patients undergoing coronary angioplasty. The results showed that oral NAC treatment was not associated with a reduction of CIN incidence, and there was no significant heterogeneity between trials. In addition, it was found that the combined treatments of oral NAC and sodium chloride did not provide additional benefits; therefore, the role of oral NAC therapy is yet to be defined in CIN prevention (11,35,36).

It has been reported that contrast-induced nephropathy occurred in ~14.5% of unselected patients following coronary angioplasty. CIN has been considered as the third common cause of in-hospital acute renal failure after coronary angiography/intervention (37). In present studies, the commonly accepted standard for CIN is according to the absolute or relative change in plasma creatinine concentration (38). In the majority of cases, CIN is defined as an increase in baseline serum creatinine (SCr) concentration of 25% or an absolute increase of at least 44 mmol/l within 48 h (39). It is universally acknowledged that absolute increase in SCr is superior threshold than a relative increase in SCr (40–43). However, it has been shown that SCr may not be an optimal substitute marker for glomerular filtration rate (GFR), as the alteration in renal handling, filtration, secretion and resorption may exert an influence on SCr levels (44). As has been noted previously (45), tubular creatinine secretion may be decreased by contrast media itself. Thus, it may cause a transient increase in SCr concentration, independent of the reduction in GFR. Serum cystatin C has been proposed as a sensitive biomarker for the diagnosis of CIN, as cystatin C has been confirmed to reflect contrast medium-induced deterioration in kidney function in a superior manner to serum creatinine (46). A previous study showed that oral NAC did not significantly reduce the incidence of CIN on the basis of the standard disease definition; however, by the cystatin C level disease criteria it may be considered to be efficacious (47). However, at present SCr remains the cheapest and most widely accepted standard of renal function (48). Therefore, the change of absolute or relative SCr concentration remains a key parameter in the diagnosis of CIN. Intravenous saline hydration and the use of low-osmolality contrast medium has been accepted as preventive strategies for CIN (49–51).

In the present meta-analysis, 19 placebo-control RCTs were included, consisting of 4,514 patients. The baseline characteristic revealed no significant difference between history of coexistent disease and routine prophylactic therapies. Each randomize controlled trial utilized intravenous saline hydration. The CIN incidence was 247 patients in the oral NAC group (n=2,269) and 278 patients in the control group (n=2,245), pooling all of the 19 trials. There was no statistically significant difference between the oral NAC group and the control group (RR, 0.84; 95% CI, 0.65–1.10; P=0.20), with no heterogeneity between trials (I2=29%, P=0.12). The results showed that the oral NAC treatment was not associated with a reduction in CIN incidence. A previous study (52) found that intravenous saline hydration with 0.45% saline prior to and following coronary angiography and the proper use of nonionic low osmolar iodine may be renoprotective. Previously, it has been confirmed (51) that normal saline hydration (0.9%) may be more efficacious compared with half-normal saline (0.45%). It is generally accepted that the optimal volume of normal saline hydration may be determined based on body weight, and 1.0–1.5 ml/kg/h is considered to be the normal range (39). In the present meta-analysis, it was found that the combined treatments of oral NAC and sodium chloride did not provide additional benefits, and thus the role of oral NAC therapy not yet to be defined in CIN prevention.

The precise mechanism underlying the pathogenesis of CIN remains unclear. It is widely considered (53–55) that the pathogenesis of CIN may involve injury to the renal medulla caused by reduced renal blood flow and tubular toxicity through ROS, which occurs following the administration of contrast media (1,56). NAC, a thiol-containing antioxidant, has been approved for an increase in the level of plasma glutathione, which is an oxygen-free radical scavenger (13). It has been affirmed (57) that NAC is able to prevent oxidative stress at the location of renal post-ischemia. NAC has received considerable attention in recent years following research by Tepel et al (12). In the opinion of Tepel et al, the utilization of NAC in conjunction with a fixed volume (75 ml) of low-osmolar contrast medium in patients undergoing computed tomography (CT), may significantly reduce incidence of CIN. It has become increasing recognized that NAC may result in increased nitric oxide production and intensification of nitric oxide binding (58). It has been demonstrated in human testing (59) that NAC treatment may significantly improve endothelium-dependent vasodilation. In a previous study, it was found that pretreatment of vascular smooth muscle cells with NAC clearly reduced ROS formation and prevented the reduction of cell viability (60). In the present meta-analysis, the majority of the selected trials utilized a low dose of NAC (600 mg) twice daily for 48 h in conjunction with intravenous saline hydration. It is known the oral NAC may be absorbed quickly, reaching the peak plasma concentration in 45 min, and having a half-life of 2 h. Thus, pretreatment with NAC more than a few hours prior to contrast exposure or for a prolonged period afterward may not be essential to provide beneficial effects.

There were a number of limitations inherent to this study. First, the asymmetrical appearance of the funnel plot suggests that publication bias was present. Despite the broad searching databases and manually searching the conference proceedings and reference lists from the identified trials, we could not eliminate that publication bias caused overestimation of the results from the true treatment. Second, all included studies used the endpoint of CIN as the primary outcome. Typically, this has been defined as an increase in baseline serum creatinine level of 25% or an absolute increase of 44 mmol/l. It found that NAC had no effect on preventing CIN on the basis of the standard diagnostic definition, while it showed a preventive effect based on cystatin C levels. Whether a newer urinary biomarker such as cystatin C may identify kidney damage for CIN requires further research. Finally, despite earlier studies having shown the association of CIN with increased in-hospital morbidity and mortality, particularly in patients that require dialysis, insufficient trials have been designed to investigate the effect of NAC on these clinical relevant outcomes. Thus, the present study did not identify sufficient evidence for a meta-analysis to assess the effect of NAC on these relatively rare, but key outcomes.

This meta-analysis of 19 placebo-controlled RCTs indicated that oral NAC did not significantly reduce the incidence of CIN. Also, it revealed that the combination of oral NAC and sodium chloride may not provide additional benefits compared with hydration with sodium chloride alone. Up to now, trials are too inconsistent to warrant a conclusion on efficacy. Recently, it has been found that oral NAC is able to confer a preventive effect of CIN based on cystatin C. Therefore, further high quality RCTs are required to confirm the safety and investigate the effect of oral NAC on clinically relevant outcomes, such as in-hospital morbidity, mortality and cost of medical care, particularly in patients that require dialysis.

Acknowledgements

The present study was supported by the Science Fund for Distinguished Young Scholars of the Fourth Affiliated Hospital of Harbin Medical University (grant no. HYDSYJQ201504).

References