Literature Search

The online PubMed, Embase, and CENTRAL databases were examined for articles of interest to the review that were published between January 1, 2000, and March 1, 2021. No language restriction was applied. The search was carried out in duplicate by 2 reviewers working independently. Terms used for the search strategy were “chronic kidney disease,” “renal disease,” “renal dysfunction,” “dialysis,” “diabetes mellitus,” “percutaneous coronary intervention,” “angioplasty,” “PCI,” and “PCA.” Details of the various combinations used are presented in Supplementary Table I. The initial set of articles was examined using the titles and abstracts to recognize studies requiring full-text analysis. The full texts of selected articles were then extracted for final assessment. These were examined by the 2 reviewers separately based on the inclusion and exclusion criteria. Disagreements were resolved by discussion. The references of the included studies were also searched for any missed articles.

TABLE I Characteristics of Included Studies

View Fullscreen

Inclusion and Exclusion Criteria

The Population, Intervention, Comparison, Outcome, and Study design guide was used to assess studies for eligibility. The following criteria were used: population, patients with DM undergoing PCI; intervention, CKD; comparison, no CKD; outcomes of mortality, myocardial infarction (MI), or revascularization rates; and study design, all types.

Studies were included irrespective of the sample size and follow-up duration. Exclusion criteria was as follows: (1) Studies on patients with DM not presenting separate outcome data for CKD and no CKD; (2) studies comparing outcomes of patients with CKD but not reporting separate data for those with DM; (3) studies on all revascularization strategies and not presenting separate data for PCI; (4) studies not describing any outcomes of interest; and (5) review articles, case reports, and unpublished data. If 2 or more studies used the same database or reported overlapping data, the study with the largest sample size was included in the review.

Data Extraction and Quality Assessment

Two reviewers extracted data independently using a prepared sheet. Data were extracted regarding author name; publication year; study type and location; stent type; sample size; patient demographic details; prior MI, hypertension, or stroke; dyslipidemia; 3-vessel disease; mean left ventricular ejection fraction; insulin use; follow-up period; and study outcomes. All extracted data were cross-checked for any errors with the parent article. The outcomes of interest for the meta-analysis were mortality (all-cause and cardiac), MI, and revascularization rates between patients with DM with and without CKD. The impact of the severity of CKD on these outcomes was also assessed.

The quality of included studies was examined by risk using a bias assessment tool for nonrandomized studies.¹⁹ Two reviewers examined each study for selection of participants, confounding variables, intervention measurements, blinding of outcome assessment, incomplete outcome data, and selective outcome reporting. Disagreements were resolved by discussion.

Statistical Analysis

Review Manager (RevMan, version 5.3; Cochrane) was used for the meta-analysis. Because of the intrinsic heterogeneity among the studies, the random-effects model was chosen for the meta-analysis. First, data on the absolute number of events of interest (mortality, MI, and revascularization) were extracted. These data were then pooled to calculate risk ratios (RRs) with 95% CIs. A separate analysis was conducted for early mortality (in-hospital and 30-day) and late mortality (until the end of the follow-up period). All-cause and cardiac mortality rates were analyzed separately. Second, multivariable-adjusted hazard ratios (HRs) of the outcome data were extracted from included studies. These data were then pooled using the generic inverse variance function of the software to calculate the pooled effect size. Few studies also reported HRs of a common composite outcome. The composite outcome was defined as combined mortality, MI, and revascularization rates. These data were also pooled in a meta-analysis. Heterogeneity was calculated using the I² statistic. I² values of 25% to 50% indicated low heterogeneity, 50% to 75% indicated medium heterogeneity, and greater than 75% indicated substantial heterogeneity. Funnel plots were used to assess publication bias. A P value of <.05 was considered statistically significant.

Search Results and Study Details

A total of 1,145 unique records were identified after the literature search (Fig. 1). On exclusion of 1,110 articles after title and abstract screening, 35 studies with their full texts were analyzed. An additional 24 articles were excluded with reasons. Finally, a total of 11 studies were included in the systematic review and meta-analysis.^{15,16,20–28} The interreviewer agreement for study selection was high (κ = 0.9).

View Figure

• Download as Powerpoint
• Download Figure

The characteristics of included studies are presented in Table I.^{15,16,20–28} One study²⁸ was a post hoc analysis of 3 multicentric randomized controlled trials (RCTs), and all others were retrospective observational studies. All studies defined CKD as less than 60 mL/min/1.73 m² of estimated glomerular filtration rate. Two studies^15,25 did not report the type of stent used in the PCI. Three studies^16,23,26 were exclusively on DES, 2 studies^20,21 were on bare-metal stents (BMS), and the remaining studies used a combination of DES and BMS. Only 1 study²³ performed propensity-score matching of the study groups. Liosis et al¹⁵ reported separate data for patients with ST-segment elevation and non–ST-segment elevation MI, which were pooled separately for the meta-analysis. Two studies^15,22 reported only early outcomes; the follow-up of remaining studies varied from 1 to 6.8 years. The risk of bias assessment is depicted in Supplementary Table II.^{15,16,20–28}

TABLE II Outcomes of studies based on severity of CKD

View Fullscreen

Meta-Analysis of the Absolute Number of Events

Pooled analysis on all-cause mortality indicated that the presence of CKD was associated with significantly increased risk of early (RR, 3.45; 95% CI, 3.07–3.87; I² = 0%; P < .001) as well as late (RR, 2.78; 95% CI, 1.92–4.02; I² = 83%; P < .001) mortality compared with patients without CKD (Fig. 2). There was no evidence of publication bias in a funnel plot (Fig. 3). The risk of cardiac mortality was also significantly higher in patients with CKD than in those without CKD (RR, 2.90; 95% CI, 1.99–4.22; I² = 29%; P < .01) (Fig. 4A). The presence of CKD also significantly increased the risk of MI in patients with DM (RR, 1.40; 95% CI, 1.06–1.85; I² = 13%; P = .02) (Fig. 4B). However, there was no difference in the risk of any revascularization between CKD and non-CKD groups (RR, 0.96; 95% CI, 0.82–1.13; I² = 25%; P = .65) (Fig. 4C).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Meta-Analysis of Adjusted Outcomes

Only 4 studies reported a multivariable-adjusted comparison of patients with and without CKD, and only 3 studies were available for meta-analysis of the following outcomes. For the composite outcome, meta-analysis indicated significantly increased combined risk of mortality, MI, and revascularization in patients with CKD compared with patients without CKD (HR, 1.70; 95% CI, 1.23–2.34; I² = 0%; P = .001) (Supplementary Fig. 1). A separate meta-analysis of these 3 outcomes indicated a significantly increased risk of mortality (HR, 2.64; 95% CI, 1.91–3.64; I² = 0%; P < .001) in the CKD group but only a nonsignificant tendency of increased MI (HR, 1.59; 95% CI, 0.99–2.54; I² = 0%; P = .05) and revascularization (HR, 1.24; 95% CI, 0.94–1.63; I² = 2%; P = .12) in the CKD group (Supplementary Fig. 1).

Outcomes Based on Severity of CKD

Because of interstudy heterogeneity and inadequate data reporting, a quantitative analysis on the impact of the severity of CKD on patient outcomes could not be conducted. However, a detailed description of outcomes reported by the studies is presented in Table II.^{15,16,20,22,23,25,27}

Notwithstanding the difference in outcomes and classifications used by studies, the majority of studies reported an increased risk of adverse outcomes with increased severity of CKD.

Limitations

The limitations of this study need to be specified. Foremost, the number of studies included in this meta-analysis was not very large. Furthermore, not all studies reported data on all outcomes, which diminished the power of individual analysis. Second, as discussed earlier, the results should be interpreted with caution, as several potential confounding factors could have influenced study outcomes. The majority of studies were retrospective in nature, with a large tendency of selection bias. Meta-analysis of only adjusted outcomes from all studies would have provided the best available evidence; however, these data were not widely reported. Also, a subgroup analysis based on the severity of CKD could not be conducted; only a qualitative analysis could be carried out. Third, the included studies were conducted over a long period spanning almost 2 decades. To include only recent evidence, only studies published in 2000 or later were searched; however, there have been several changes in PCI technology and cardiovascular drugs during this period that could have influenced outcomes. Fourth, an important difference across studies was in the different types of stents used. Although earlier studies used only BMS, more recent ones used only DES. Because of limited data, the impact of stent types on study outcomes could not be assessed. Last, the way in which antidiabetic therapy may affect outcomes in patients with CKD who undergo PCI could not be assessed. In recent years, studies have shown that sodium-glucose cotransporter-2 inhibitors used in the management of DM have cardioprotective and renoprotective functions. These drugs are known to significantly reduce the risk of worsening renal function and prevent progression to end-stage renal disease.⁴⁵ In this context, it would be interesting for future studies to explore the association of sodium-glucose cotransporter-2 inhibitors and outcomes in patients with CKD who undergo PCI. There is also a need for future studies to compare outcomes of revascularization vs nonrevascularization (where feasible) in patients with CKD with cardiovascular disease to further enhance current evidence.

Based on the current results, the authors believe that clinicians should take adequate care and prioritize treatment of patients with CKD who require PCI. These patients should also be counseled regarding the increased risk of complications and be provided with highly supervised perioperative care.