Sex Differences in Saphenous Vein Graft Patency: A Systematic Review and Meta-Analysis

Abstract

Background:

Saphenous vein grafts (SVG) are the most commonly used conduits in coronary artery bypass grafting (CABG). Graft failure is observed in up to 50% of SVG at 10 years after surgery. Whether a difference in SVG patency rates exists between men and women remains unclear.

Methods:

We performed a study-level meta-analysis to evaluate sex-related differences in follow-up patency rates of SVG after CABG. A systematic literature search was conducted to identify studies on CABG that reported follow-up SVG patency rates in men and women. The primary outcome was SVG patency rates by sex at follow-up.

Results:

Seventeen studies totaling 8,235 patients and 14,781 SVG grafts were included. There was no significant difference in follow-up SVG patency rates between men and women (incidence rate ratio 0.96, 95% confidence interval 0.90–1.03, p=0.24), with mean angiographic follow-up of 33.5 months (standard deviation 29.2). Leave-one-out and cumulative analysis were consistent with the main analysis.

We concluded that follow-up SVG patency rate is similar between men and women undergoing CABG.

Introduction

Coronary artery bypass grafting (CABG) is the most common adult cardiac surgery operation in the United States.^1,2 Prior studies have reported higher mortality³ and major postoperative adverse events including stroke and myocardial infarction (MI) among women.^4,5 While women who are referred for CABG are generally older and have more cardiovascular risk factors than men,⁶ technical factors and differential graft patency may contribute to the reported sex differences in CABG outcomes.⁷ Saphenous vein graft (SVG) remains the most-used graft in CABG, despite a well-described poorer patency rate than arterial grafts^8–10 and an increased susceptibility to occlusion compared to arterial grafts.¹¹ Women are less likely to receive any arterial conduits than men, and are also less likely to receive all-arterial conduits.¹² However, SVG patency in women and its correlation to clinical outcomes remains poorly described. We conducted a systematic review of the literature and meta-analysis to evaluate if there is a difference in SVG patency between men and women.

Methods

The present review was registered with the National Institute for Health Research International Registry of Systematic Reviews (PROSPERO) (CRD42020160695). The manuscript follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline.²²

A medical librarian (MD) performed a comprehensive literature search to identify both randomized trials and observational studies of patients undergoing CABG reporting follow-up SVG patency rates by sex. Searches were run initially in April 2019 and updated on July 12,2022 in the following databases: Ovid MEDLINE (ALL; 1946 to present), Scopus, and PubMed. The search strategy included the following keywords, in combination with the Boolean operators as appropriate: coronary artery bypass, saphenous vein, vascular patency, graft occlusion, women, female, and gender. The complete search strategy for Ovid MEDLINE is available in Supplementary Table 1.

After de-duplication, records were screened by 2 independent reviewers (ML and LH). Any discrepancies were adjudicated by the senior author (MG). Titles and abstracts were reviewed against pre-defined inclusion and exclusion criteria. Studies were considered for inclusion if they were written in English and reported SVG patency rates after elective CABG at long-term follow-up (≥ 1 year). Animal studies, abstracts, case reports, commentaries, editorials, expert opinions, conference presentations, experimental studies, pediatric studies (<18 years old), studies on reoperation patients only, studies of angiography performed for symptoms, studies with <10 subjects of either sex, and studies not reporting the outcomes of interest or with a follow-up <1 year were excluded. The full text of the selected articles was pulled for a second round of eligibility screening. The reference lists were also reviewed for relevant studies not captured by the original search. The quality of the included observational studies was assessed using the Newcastle-Ottawa Scale (NOS). Details of vein graft patency definition in each study are provided in Supplementary Table 2.

Two investigators (ML and LH) independently performed data extraction, and the accuracy was verified by the senior author (MG). The variables included were study characteristics (publication year, institution, country, study period, study design, sample size, mean follow-up), patient demographics (age, sex, diabetes, smoking, hypertension, chronic renal failure, chronic obstructive pulmonary disease (COPD), prior percutaneous coronary intervention (PCI), prior cardiac surgery), intraoperative characteristics (off-pump coronary artery bypass (OPCAB), cross-clamp time, method of vein harvest), in-hospital MI, need for reintervention, follow-up SVG patency rates, and follow-up mortality.

The primary outcome was follow-up SVG patency rate by sex. The secondary outcome was follow-up mortality by sex.

Assessment of follow-up graft patency and of follow-up mortality by sex was performed using incidence rate ratio (IRR) with its 95% confidence interval (CI) and was reported using generic inverse variance method. IRR was estimated through several means based on the available study data. If hazard ratio (HR) was available, it was used for IRR, otherwise, we used reported events and follow-up time or digitized Kaplan-Meier curves. Cumulative analysis and leave-one-out analysis were performed for the primary outcome to assess the robustness of the obtained estimate and its variation over the studied years and are provided in Supplementary Figure 1 and 2.

Meta-regression was performed to explore the association of baseline and intraoperative patient demographics (age, sex, diabetes, smoking, hypertension, chronic renal failure, COPD, prior PCI, prior cardiac surgery), intraoperative characteristics (OPCAB, cross-clamp time, method of vein harvest), as well as mean follow-up time, with the pooled estimate for the primary outcome.

Publication bias was visually assessed by funnel plot (using the trim and fill method) and Egger’s test and is provided in Supplementary Figure 3. Heterogeneity was reported as low (I²=0–25%), moderate (I² =26–50%), or high (I²>50%). All statistical analyses were performed using R (version 4.2.1, R Project for Statistical Computing) within RStudio. Random-effect meta-analysis was performed using “metafor” and “meta” packages.

Results

Among the 2,092 screened articles, a total of 17 studies, published between 1993 and 2021, were included in the present analysis.^13–29 The PRISMA flow diagram outlining the study selection process and the quality of included studies according to the NOS are provided in Supplementary Figure 4 and Supplementary Table 3, respectively. All studies were of high quality (NOS score 8–9). Nine studies (52.9%) were from Asia, 4 from Europe (23.4%), 3 from North America (17.6%), and 1 (5.8%) from South America. A summary of the included studies is provided in Table 1.

Table 1:

Summary of included studies

Study (year)	Study period	Type	Country	Cohort size	Women in cohort (%)
Buxton (2020)	1996–2005	RCT	Australia	112	21 (18.8)
Cataldo (1993)	1987–1990	RCT	Italy	349	30 (8.6)
Deb (2019)	2011–2013	RCT	Sweden, Canada, Israel	250	31 (12.4)
Kara (2013)	2004–2011	R	Turkey	48	17 (35.4)
Kawamura (2020)	2002–2016	R	Japan	84	16 (19.0)
Li (2018)	2012–2014	R	China	341	68 (19.9)
Lopes (2012)	2002–2003	R	USA	3014	630 (20.9)
Malinska (2017)	2008–2015	R	Poland	718	237 (33)
Sarzaeem (2010)	2006–2009	P	Iran	403	101 (25.1)
Tam (2018)	1996–2001	RCT	Canada, New Zealand	561	72 (13.4)
Tang (2021)	2017–2018	RCT	China	147	58 (39.5)
Une (2013)	2006–2009	RCT	Canada	92	12 (10.6)
Willemsen (2020)	2015–2019	RCT	Netherlands	496	64 (12.9)
Yamane (2017)	2007–2014	R	Japan	123	31 (25.2)
Yavuz (2020)	-	R	Turkey	534	148 (27.7)
Yoshida (2017)	1997–2017	R	Japan	264	74 (28.0)
Zhao (2018)	2014–2015	RCT	China	500	91 (18.2)

P= prospective cohort study, R = retrospective cohort study, RCT= randomized control trial

A total of 8,235 patients and 14,871 SVG grafts were included in the pooled analysis. The number of patients in each study ranged from 48 to 3,104 with a median sample size of 341 patients (interquartile range (IQR) 123–500). The percentage of women in these cohorts ranged from 8.6% to 39.5%, with a median of 19.9% (IQR 13.4–27.7%).

Overall, the mean age range was 57–73.1 years old, the prevalence of diabetes ranged from 0% to 54.3%, the prevalence of smoking ranged from 14.1% to 71.4%, and the prevalence of hypertension ranged from 42.2% to 91.7% in all cohorts. Only 1 of the studies reported these data by sex. OPCAB was reported by 12 studies and was utilized in 0%−97.2% of operations, with a median of 4.8% (IQR 0%−25.2%). Demographic data of the patient population in each study are summarized in Supplementary Table 4.

The mean angiographic follow-up was 33.5 months (standard deviation (SD) 29.2), with a median of 16.6 months (IQR 12–45.5); 4 studies had at least 5-year follow-up.^13,22,27,28 The mean overall SVG patency rate was 80.4% (SD 14.3, median 82.6%, IQR 79.1–90.5). The graft patency and follow-up data for individual studies are provided in Supplementary Table 5.

There was no significant difference in follow-up SVG patency between men and women (IRR 0.96, 95% CI 0.90–1.03, p=0.24, see Figure 1). The leave-one-out analysis was consistent with the main analysis (Supplementary Figure 2). Egger’s intercept test showed evidence of publication bias (p=0.01) (Supplementary Figure 3).

Figure 1:

Forest plot for follow-up patency of saphenous vein graft CI= confidence interval, IRR= incidence rate ratio

Only 2 studies reported follow-up mortality by sex, with no significant difference noted between men and women (IRR 0.84, 95% CI 0.03–26.91, p=0.92). Meta-regression did not identify any association between the tested variables and the IRR for the primary outcome (Table 2).

Table 2:

Summary of the meta-analyses for the primary and secondary outcomes

Outcomes	Number of studies	IRR [95%CI], p-value	Heterogeneity (I-square, p-value)
Follow-up SVG patency	14	0.96 [0.90; 1.03], p=0.24	0.0%, p=0.57
Follow-up mortality	2	0.84 [0.03; 26.91], p=0.92	66.4%, p=0.08

CI= confidence interval, IRR= incidence rate ratio, SVG= saphenous vein graft

Discussion

In the present systematic review and meta-analysis of 17 studies, we found that there was no significant difference in follow-up SVG patency between men and women. Meta-regression did not find any association between preoperative risk factors, intraoperative technique, and vein graft patency.

Women have well-described higher mortality after CABG than men.³ In a retrospective analysis of over 1 million patients from the Society of Thoracic Surgeons database, women were found to have a higher operative risk of death than men (odds ratio 1.26; 95% CI 1.63–1.73).³⁰ Higher rates of major postoperative adverse events including stroke and MI among women have also been described.^4,5 In a recent meta-analysis of 4 CABG randomized trials including 13,193 patients, Gaudino et al. found that at 5-year follow up, women had a significantly higher risk of major adverse cerebrovascular and cardiovascular events than men (adjusted HR 1.12, 95% CI 1.04–1.21).⁷ Differences in the technical aspects of CABG, including the type of graft used, have been proposed to explain the sex differences in outcomes. Women have both smaller conduits and smaller target coronary arteries than men, rendering successful grafting more challenging.³¹

Venous grafts are especially susceptible to atherosclerotic burden and show higher risk of occlusion after CABG than arterial grafts, with published graft failure rates as high as 40–50%.^8–11,32 Despite this evidence, the most common type of graft used in both sexes remains the SVG. Women are even more likely than men to receive SVG in lieu of arterial grafting. In a study of 57,943 patients undergoing primary elective CABG (19% women), Attia et al. found that women received less all-arterial conduits, and fewer arterial conduits at all, than did men.¹² Gaudino et al. performed a retrospective analysis of the New York State Cardiac Surgery Reporting System of all CABG patients from 2005 to 2014, and found that only 13.8% of women, compared with 21.9% of men, received multiple arterial grafts.³³

No previous meta-analysis has assessed if SVGs perform equally well in women and men, and given the preponderance of SVG use in women, it is critical to ascertain if venous graft patency contributes to the observed sex differences in post-CABG mortality and adverse events. Our study suggests that it does not.

The etiology of the sex disparity in outcomes after CABG therefore remains unclear. Women present for CABG later than men and have more cardiovascular risk factors than men at time of presentation.⁶ This higher burden of preoperative comorbidity may contribute to poorer postoperative outcomes. However, the underlying pathophysiology of coronary artery disease (CAD) in women may also play a role. While obstructive CAD predominates in men, women are over-represented among patients with ischemia secondary to non-obstructive CAD.^34,35 MI secondary to non-obstructive CAD, including coronary spasm, dissection, or emboli,³⁶ is similarly more prevalent among women.^37–39 In a retrospective study by Smilowitz et al. of patients who underwent coronary angiography for MI, non-obstructive disease was found in 18,918 patients, of whom women comprised 10.5%, a significantly higher proportion than men (3.4%, p<0.0001).³⁸ The higher prevalence of non-obstructive CAD among women challenges the accepted belief that clinical status and events are tied to CABG graft patency,⁴ and may explain how graft patency could be equivalent between sexes while clinical outcomes are not.

Our findings must be interpreted considering their context and limitations. Criteria for patency were not universally defined in each study. Also, many of the screened studies could not be included as they did not report patency by sex, and therefore our results do not represent all published SVG studies.

Conclusions

In conclusion, in this meta-analysis of 17 studies, we did not find a difference in SVG patency between women and men at a mean follow-up of 33.5 months. Given the described difference in post-CABG outcomes between men and women, further study is needed to understand the relationship between sex, graft patency, and clinical outcomes. Of note, the final stage of our literature search yielded 234 studies to be assessed for eligibility. However, 100 of these studies failed to report patency by sex and therefore were excluded (Supplementary Figure 4), and an additional 44 studies had <10 women subjects and were also excluded. This represents exclusion of valuable data on SVG patency from 144 (61.5%) of 234 studies due to a lack of inclusivity in study design, enrollment, or reporting. We encourage our colleagues to report sex-specific outcomes to answer these important questions.

Table 3:

Results of the meta-regression for the primary outcome

Variables	Beta ± SD, p-value	p-value
Mean age	−0.0086 ± 0.0189	p=0.64
Diabetes	0.0019 ± 0.0037	p=0.62
Hypertension	0.0041 ± 0.0051	p=0.42
Smoking	−0.0014 ± 0.0023	p=0.54
Chronic renal failure	0.0039 ± 0.0212	p=0.85
COPD	0.0584 ± 0.0620	p=0.35
Previous PCI	0.0149 ± 0.0270	p=0.58
OPCAB	0.0009 ± 0.0024	p=0.71
No-touch vein	0.0054 ± 0.0153	p=0.72
Endoscopic vein	0.0016 ± 0.0030	p=0.58
Mean cross-clamp time (minutes) ± SD ALL	−0.0095 ± 0.0128	p=0.46
Mean follow-up (months)	−0.0015 ± 0.0016	p=0.34

COPD= chronic obstructive pulmonary disease, OPCAB= off-pump coronary artery bypass, PCI= percutaneous coronary intervention, SD= standard deviation