INTRODUCTION

Pulmonary Hypertension (PH), a potentially lethal condition with a prevalence of approximately 1% worldwide, is most commonly associated with Left Heart Disease (PH-LHD) [1-3]. Valve malfunction and diastolic dysfunction emerged as the prominent causes [4]. World Health Organization (WHO) classified PH-LHD as Group 2 PH which represents PH secondary to Left Ventricular (LV) systolic dysfunction (Heart Failure with reduced Ejection Fraction – HFrEF), LV diastolic dysfunction (Heart Failure with preserved Ejection Fraction – HFpEF), or Valvular Heart Disease (VHD) [5]. PH in VHD is described as an enhancement in mean Pulmonary Arterial Pressure (mPAP) ≥ 25 mmHg at rest as evaluated by right heart catheterization and a combination of precapillary-postcapillary PH [6, 7]. In VHD, persistent PH causes pulmonary vascular remodeling and reduced vascular compliance [8]. Thus, controlling VHD progression is critical, and early interventions such as valve replacement or repair help in slowing down the worsening of PH [8]. Despite important improvements in the timing of valve interventions, long-standing PH after surgery is common [9]. Thus, medical intervention is required specifically in asymptomatic patients with severe VHD or in symptomatic patients with moderate VHD [10].

The treatment options for PH secondary to VHD include Phosphodiesterase-5 inhibitors (PDE5i) such as sildenafil [6, 11].

Sildenafil relaxes the pulmonary vascular smooth muscles, and thus lowers pulmonary artery hypertension and pulmonary vascular resistance in patients with different types of PH [12, 13]. The off-label use of sildenafil in treating persistent PH after the correction of VHD has received considerable critical attention [14]. To date, limited investigations and guidelines are available to recommend the usefulness of sildenafil for group 2 PH-LHD. Based on the guidelines of 2015 European Society of Cardiology (ESC) and European Respiratory Society (ERS), there is no novel evidence to support the use of pulmonary artery hypertension therapies in PH-LHD [6]. This is partly because of the lack of researches that stratify patients with PH and target this estate [6]. The 2017 American Heart Association (AHA)/American College of Cardiology (ACC) Focused Update of the 2014 AHA/ACC Guideline for the Management of Patients with Valvular Heart Disease did not provide any recommendation on slowing down the progression of valve diseases, but only focused on medical therapy for concomitant hypertension [15]. In the United States, the use of PDE5i for group 2 and 3 PH showed an increasing trend despite guidelines recommending against this low value practice [16]. 

Several systematic and meta-analysis studies have been conducted to determine the effectiveness of sildenafil in various populations [17-20]. A recent review only evaluated the effects of preoperative sildenafil in PH patients undergoing mitral valve surgery [18] and did not compare the effects of sildenafil when given postoperatively and long term for PH secondary to VHD. The current review, therefore, aimed to establish a summary of all potential benefits of sildenafil in terms of hemodynamic parameters and patient-centered outcomes as acute and chronic treatment for PH secondary to VHD. Thus, the results of this review will inform debates about the efficacy or potential benefits of sildenafil in PH with VHD.

MATERIALS AND METHODS

Search Strategies and Study Identification

A systematic study and meta-analysis of published literature was done following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [21]. Studies were identified from inception (1946) to June 2020 through a comprehensive literature search on Web of Science, Medline (Ovid), EBSCOhost, the Cochrane Library, and PubMed, as well as Google Scholar as an additional source for published and unpublished studies. Reference lists of all selected studies were further scrutinized for any additional Randomized Controlled Trials (RCTs). Missing outcome data were traced by contacting the authors of the study. Search keywords were chosen appropriate to the Population, Intervention, Comparator and Outcome (PICO) model (Table 1). The population was set at PH with VHD. The intervention was sildenafil treatment. The comparison was with placebo, other medication or without other intervention. The outcome was determined by hemodynamic parameters or clinical performance. Thus, the search keywords used were ‘pulmonary hypertension’, ‘valvular heart disease’, ‘left-sided valve disease’, ‘sildenafil’, ‘phosphodiesterase-5’, ‘6-Minute Walk Test (6MWT)’, ‘systolic Pulmonary Artery Pressure (sPAP)’, ‘mean Pulmonary Artery Pressure (mPAP)’, ‘pulmonary hemodynamic’, ‘systemic hemodynamic’, ‘WHO functional class’, ‘adverse events’ and ‘Cardiac Output (CO)’. Boolean operators such as ‘AND’ and ‘OR’ were used to increase sensitivity and specificity of the search when needed.

Selection of Studies

We included double-blinded RCTs, non-RCTs, and retrospective or prospective studies in which sildenafil was compared to a placebo. Only studies available in English were selected. Any non-English manuscripts, conference abstracts, case reports, and animal studies were excluded from this systematic review. Two reviewers (FAG and ADY) performed the screening of titles and abstracts based on the criteria presented in Table 1. All disagreements were resolved by negotiating with a third author (MMB).  

Data Extraction

Extracted data included study design, treatment doses, duration of sildenafil administration, outcome measures, and results. For the purpose of this study, the articles were sub-grouped into acute or chronic treatment. Acute treatment was defined as preoperative or post-operative sildenafil that was stopped within 48 hours of first administration. Preoperative data was extracted at two different time points: between the preoperative and intraoperative period and between the intraoperative and post-operative period. Chronic treatment was defined as preoperative or post-operative sildenafil that was continued until the next follow-up clinic, which may be prolonged to two months or more. Chronic treatment was divided into two different data analyses, based on whether sildenafil was initiated preoperatively or postoperatively.

Risk of Bias Assessment

The bias risk in RCTs was evaluated by domains proposed by the Cochrane Handbook of Systematic Reviews [22], particularly emphasizing on allocation concealment, sequence generation, outcome assessment, blinding, and selective reporting. The bias risk for each field was rated as low, unclear, or high risk. The total bias risk was labelled as high if minimally one field was at high risk of bias. The bias risk of non-RCT researches was assessed using the Black and Downs checklist. 

Data Synthesis and Analysis

A meta-analysis was conducted to summarize two or more studies with similar outcome measures. The model of inverse variance random-effects for continuous results was utilized to form Mean Differences (MD) and 95% Confidence Intervals (CIs) for Forest plots. All results are indicated as the mean ± SD unless the contrary mentioned. Heterogeneity was assessed through the Chi-square and I2 test. The Mantel-Haenzel random-effects model was utilized for dichotomous outcomes, to form risk ratios and 95% CIs for Forest plots. All statistical analyses were done by Review Manager (RevMan) version 5.4 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen). 

RESULTS AND DISCUSSION

Research Characteristics

The electronic search yielded a total of 6,074 articles (Figure 1). Out of these, 1,982 articles were duplicated and excluded. Title and abstract screening excluded 3,869 articles due to inappropriate nature of the literature, such as reviews, editorial, conference abstracts, animal studies, molecular studies, and case reports. Full text assessment excluded 10 articles due to population conflicts and case reports. In total, nine eligible studies (eight RCTs and one non-RCT) involving 614 patients were included and considered for meta-analysis. Studies conducted at acute phase represented by four RCTs for preoperative administration of sildenafil and two RCTs involved post-operative administration. Chronic sildenafil treatment was employed in three studies (two RCTs and one non-RCT). The characteristics of all studies are presented in Table 2.  

Risk of Bias within the Studies

All eight RCTs were considered to have a low overall risk of bias. Six of the RCTs met all of specified criteria (Figure 2), while those by Ayyad (2012) and Shewale et al. (2020) [23, 24] did not report random sequence generation, allocation concealment, or blinding of personnel and outcome assessment. The non-RCT study [25] was considered as fair quality of evidence due to internal validity-confounding (selection bias). 

Abbreviations: S = sildenafil; P = placebo; NA = not available; sPAP = systolic Pulmonary Arterial Pressure; mPAP = mean Pulmonary Arterial Pressure; PVRI = Pulmonary Vascular Resistance Index; SVRI = Systemic Vascular Resistance Index, T0 = time before induction; T30 = 30 minutes after 50mg oral sildenafil or placebo; T60 = 60 minutes after 50mg oral sildenafil or placebo; RVESVI = right ventricular end systolic volume index; RVEDVI = right ventricular end diastolic volume index; Cl = cardiac index; CVP = Central Venous Pressure; CPB = cardiopulmonary bypass; PCWP = Pulmonary Capillary Wedges Pressure; MAP = Mean system in Arterial Pressure; PASP = Pulmonary Artery Systolic Pressure; EF = Ejection Fraction; PA = Pulmonary Artery; sig. = Statistically Significant; non-sig. = Statistically Non-significant.

Pulmonary Hemodynamic Parameters

Effects of Acute Preoperative Sildenafil Treatment

Meta-analysis on studies by Ayyad et al. (2012) and Shim et al. (2006) [23, 26] showed a non-significant reduction in sPAP between the preoperative and intraoperative periods by 9.23 mmHg (Figure 3a); (95% CI: -26.56 to 8.10 mmHg; P = 0.30; I2 = 0.8) after preoperative administration of 25 – 50 mg sildenafil. Both studies demonstrated significant improvements in other parameters, such as mPAP and Pulmonary Vascular Resistance Index (PVRI). However, the studies by Gandhi et al. (2014) and Shewale et al. (2020) [24, 27] did not show a significant reduction in sPAP between the intraoperative and postoperative periods, and favored the placebo group (3.72 mmHg; 95% CI: -0.43 to 7.86 mmHg; P = 0.08) after acute treatment with 25 mg sildenafil three times daily. Both studies showed similar findings in mPAP (Figure 3b); (2.15 mmHg; 95% CI: -1.35 to 6.4 mmHg; P = 0.23; I2 = 0) and favored the placebo group. No significant reduction in PVRI was reported at 30.46 dynes.sec.cm5m2 (Figure 3c); (95% CI: 111.98 to 51.06 dynes.sec.cm5m2; P = 0.46, I2 = 0) following acute sildenafil treatment (intraoperative vs. post-operative).

Effects of Acute Post-operative Sildenafil Treatment

Only one study reported a reduction in sPAP (MD -5.89 mmHg ± 17.07), mPAP (MD -4.62 mmHg ± 12.24), and PVRI (MD -60.11 dynes.sec.cm5m2 ± 500.85) following acute postoperative sildenafil treatment compared to placebo [17]. Ram et al. (2019) reported a mean sPAP of 66 mmHg in the sildenafil group, which was comparable to that of the placebo group, but reported a significant reduction in mPAP from 32 ± 7 mmHg at baseline to 26 ± 3 mmHg after 36 hours of sildenafil compared to placebo (P < 0.001). However, PVR and pulmonary arterial wedge pressure were not reported in the study [27]. 

Effects of Chronic Sildenafil Treatment

The study by Ibrahim et al. (2020) involved three different groups (Group A received sildenafil for a week, group B received sildenafil for a month, and group C received a placebo), with treatment of all groups initiated preoperatively [28]. A significant reduction was observed in sPAP following Cardiopulmonary Bypass (CPB) and was lower in groups A and B than in group C (P < 0.001). Two other studies (Bermejo et al. 2018; Chapman et al. 2009) [14, 25] reported comparable hemodynamic parameters with long-term post-operative sildenafil therapy (40–50 mg three times daily). Sildenafil improved mPAP and PVR in four patients who received sildenafil between 2 – 12 months [25]. Bermejo et al. (2018), however, showed that the use of sildenafil (oral 40 mg three times daily) for 6 months resulted in no changes in sPAP compared to the placebo group, with a mean difference of -1 mmHg (SD 24.08).

Systemic Hemodynamic Parameters

Effects of Acute Preoperative Sildenafil Treatment

No significant reduction in SVRI (Figure 4a) with low heterogeneity (I² = 0%) was demonstrated in two studies [26, 27]. Both studies reported no alterations in heart rate or mean arterial pressure after induction, after CPB weaning, and during the post-operative period. Nevertheless, the study by Shim et al. (2006) demonstrated a reduction in both parameters. In addition, central venous pressure in the sildenafil group was significantly enhanced according to baseline.  

Effects of Acute Post-operative Sildenafil Treatment

Jiang et al. (2014) and Ram et al. (2019) showed no changes in cardiac index (CI); (Figure 5a) and systemic vascular resistance (SVR); (Figure 5b) following post-operative administration of sildenafil compared to placebo, and their data showed low to moderate heterogeneity (I² = 0 and I² = 68%, respectively) (26,27). Sildenafil was given through either a nasogastric tube (27) or intravenous injection (26) at specific doses (0.5 mg/kg q8h or 20 mg q8h for 4 – 36 hours). Both studies also demonstrated that sildenafil was able to maintain CO and CI during the post-operative period.