Midodrine in Liver Cirrhosis With Ascites: A Systematic Review and Meta-Analysis

Ascites is the most common complication of liver cirrhosis. Midodrine is a vasoconstrictor that improves splanchnic and systemic hemodynamics, reduces ascites, and improves clinical outcomes. Here, we aimed to examine the role of midodrine in cirrhosis-related ascites. Scopus, Embase, PubMed, and PubMed Central databases were searched for relevant randomized controlled trials comparing midodrine with other interventions in patients with cirrhotic ascites on November 25, 2020, using appropriate keywords like “midodrine”, “ascitic cirrhosis”, “peritoneal paracentesis” and suitable Boolean operators. Odds ratio (OR) and mean difference (MD) were used to analyze pool data as appropriate with a 95% confident interval (CI). A total of 14 studies were included in our analysis including 1199 patients. The addition of midodrine resulted in statistically significant improvement in mean arterial pressure (MAP) (MD, 3.95 mmHg; 95% CI, 1.53-6.36) and MELD (Model for End-Stage Liver Disease) score (MD, -1.27; 95% CI, -2.49 to -0.04) compared to standard medical treatment (SMT). There was also a significant improvement in plasma renin activity and plasma aldosterone concentration. However, there was no significant improvement in mortality or serum creatinine compared to SMT. In addition, there was no statistically significant improvement in MAP, plasma renin activity, plasma aldosterone concentration, MELD score, overall mortality, and paracentesis-induced circulatory dysfunction comparing midodrine with albumin. Midodrine alone leads to significant improvement in various clinical parameters in patients with cirrhotic ascites compared to standard medical care. At the same time, it was found to be non-inferior to albumin. Therefore, further well-designed studies need to be carried out on midodrine in addition to albumin for optimal clinical benefits among patients with ascites due to cirrhosis.


Introduction And Background
Ascites is one of the most common and serious complications of liver cirrhosis [1]. Ascites is managed with diuretics and sodium restriction. Ascites that does not reduce or that occurs shortly after therapeutic paracentesis despite sodium restriction and diuretic treatment is called refractory ascites [2]. Therapeutic paracentesis, combined with the expansion of plasma volume using albumin, is an effective and safe procedure with fewer risks than diuretic therapy in such cases [1]. Albumin, however, is expensive and, may have some risk of disease transmission; its use is thus controversial in some countries [1,3,4]. Peripheral arterial vasodilation has been hypothesized to be the critical factor in the pathogenesis of functional renal abnormalities in patients with cirrhosis [5]. Vasoconstrictor administration may decrease arteriolar vasodilation caused by paracentesis and prevent complications associated with a decrease in the effective arterial blood volume. Midodrine, an alpha-1 agonist directly acting on peripheral alpha-receptors, is a vasoconstrictor and is available as a cheap oral formulation. It has been commonly used to treat orthostatic hypotension and multiple secondary hypotensive disorders [6][7][8]. Recently, a single-dose administration of midodrine has been shown to substantially improve the systemic and renal hemodynamics of ascites in nonazotemic cirrhotic patients [7]. However, clinical trials evaluating Midodrine have provided inconclusive findings in patients with liver cirrhosis-related ascites, irrespective of the refractory status of the ascites [9].
We aimed to conduct a systematic review and meta-analysis to assess the effectiveness of midodrine in reducing mortality, improving response rates in patients with ascitic cirrhosis undergoing peritoneal paracentesis/drainage, assessment of MELD (Model For End-Stage Liver Disease), plasma renin, aldosterone, and creatinine.

Study Quality
The quality of individual articles was evaluated using the Cochrane ROB (Risk of Bias) 2.0 for RCTs [13]. The risk of bias was assessed ( Figure 1). Two of the authors independently assessed the design of each study, and the number of patients in outcomes including short-term mortality, paracentesis-induced circulatory dysfunction, serum creatinine, plasma renin, plasma aldosterone, and MELD scores. Third-person (among authors) resolved the disagreement.

FIGURE 1: Risk of Bias assessment of included RCTs
Included studies are reference nos. [1,3,4,[6][7][8][14][15][16][17][18][19][20][21][22] Data Analysis Data were analyzed using RevMan v5.4 (https://training.cochrane.org/). Odds ratio (OR) was used for outcomes like short-term mortality and paracentesis-induced circulatory dysfunction (PICD). Heterogeneity was measured by the I² test among the included studies. For data synthesis, a qualitative approach was planned. The handling of data and combining results of the studies was done using OR and using the random or fixed effect model based on heterogeneities. We analyzed the mean difference among the two groups for mean arterial pressure, MELD scores, plasma renin, plasma aldosterone, and serum creatinine level.

Sensitivity Analysis
Subgroup analysis was done within the respective outcomes contrasting albumin-based control and other treatments as a control.

Publication Bias
Publication bias of the included studies was assessed and presented using Funnel plots.

Results
We identified a total of 865 studies through a thorough database search. A total of 318 duplicates were removed, and we screened the title and abstracts of 547 studies. After excluding 497 studies, we assessed the full text of 50 studies, and 30 studies were excluded for definite reasons ( Figure 2). Therefore, 15 remaining studies were included in our qualitative analysis.

Qualitative Analysis
We included 15 studies in our qualitative analysis presented in Table 1. Basic details of included studies are shown in the supplementary file. Narrative summary of included studies is shown in

Mean Arterial Pressure (MAP)
A total of twelve studies reported MAP outcomes, mostly around one week of treatment. The addition of midodrine to standard medical treatment (SMT) showed a mean MAP of 2.56 mmHg higher in the midodrine group (MD, 3.95 mmHg; 95% CI, 1.53-6.36; p=0.001) compared to SMT. Midodrine when compared to albumin did not reach significant differences level in terms of MAP (MD -0.40, 95% CI -2.37 to 1.57; n= 164; I2 = 0%) (Figure 3).

FIGURE 3: Forest plots comparing MAP between midodrine and Placebo/SMT, and midodrine and albumin
The square box across the horizontal lines represents the mean difference (MD) value for the individual study, the horizontal line represents 95% confidence interval (CI), and the diamond represents the pooled MD with its CI.

FIGURE 4: Forest plot comparing mean MELD score between midodrine and placebo/SMT. (Only one study compared midodrine with albumin for MELD score)
The square box across the horizontal lines represents the mean difference (MD) value for the individual study, the horizontal line represents 95% confidence interval (CI), and the diamond represents the pooled MD with its CI.

FIGURE 5: Forest plots comparing mean PRA between midodrine and placebo/SMT, and midodrine and albumin
A square box across the horizontal lines represents the mean difference (MD) value for the individual study, the horizontal line represents 95% confidence interval (CI), and the diamond represents the pooled MD with its CI.

FIGURE 6: Forest plot comparing mean PAC among midodrine and other treatments in case of ascites due to cirrhosis
The square box across the horizontal lines represents the mean difference (MD) value for the individual study, the horizontal line represents 95% confidence interval (CI), and the diamond represents the pooled MD with its CI.

Short-Term Mortality
A total of eight studies reported mortality outcomes. There were no significant differences in short-term mortality (within three months, though it was reported heterogeneously across studies noted in footnotes) when midodrine use was compared to SMT/placebo or albumin (OR, 0.52; 95% CI, 0.13 to 2.01; P=0.34 and OR, 2.05; 95% CI, 0.38 to 11.04; P=0.40 respectively) ( Figure 7).

FIGURE 7: Forest plots showing mortality comparing midodrine to SMT/Placebo and albumin
The square box across the horizontal lines represents the Odds Ratio (OR) value for the individual study, the horizontal line represents 95% confidence interval (CI), and the diamond represents the pooled OR with its CI.

Serum Creatinine
A total of ten studies reported serum creatinine value during the study period, mostly around one week of treatment. Midodrine use was not statistically significant in lowering serum creatinine compared to SMT/placebo; however, it was nearing statistical significance (MD, -0.06; 95% CI, -0.14 to 0.03; P=0.19). On the contrary, midodrine use leads to a statistically significant reduction in serum creatinine compared to albumin (MD, -0.09; 95% CI, -0.16 to -0.02; P=0.01) (Figure 8).

FIGURE 8: Forest plots comparing mean serum creatinine between midodrine and placebo/SMT, and midodrine and albumin
The square box across the horizontal lines represents the mean difference (MD) value for the individual study, the horizontal line represents 95% confidence interval (CI), and the diamond represents the pooled MD with its CI.

FIGURE 9: Forest plot showing PICD comparing midodrine with other treatments in case of ascites due to cirrhosis
The square box across the horizontal lines represents the Odds Ratio (OR) value for the individual study, the horizontal line represents 95% confidence interval (CI), and the diamond represents the pooled OR with its CI.

Publication Bias
Publication bias of the included studies was assessed and presented in Funnel plots. Significant publication bias was present as suggested by an asymmetry of the plot for outcomes evaluated (Figures 10-11).

Discussion
Cirrhotic ascites is usually associated with hypotension due to vasodilation mediated by low effective circulatory volume. Diuretics in such cases can further worsen renal perfusion and decrease renal sodium excretion. Midodrine is an oral vasopressor that blocks vasodilation and increases blood pressure, potentially leading to improved renal perfusion and decreased ascites [20,21,23]. This possibly leads to mortality and morbidity benefits. In this meta-analysis, we focused on the role of midodrine in combination with drugs like rifaximin, octreotide, and clonidine in cirrhotic ascites. Different studies included rifaximin, octreotide, clonidine, albumin, terlipressin, hydroxyethyl starch (HES), a combination of alternative diuretics like torsemide, amiloride, furosemide, and spironolactone, repeated large-volume paracentesis as standard medical treatment (SMT). As expected, we found significant improvement in blood pressure in patients receiving midodrine compared to standard medical treatment as a potential effect of alpha-1 mediated vasoconstriction. Midodrine use was statistically significant in lowering serum creatinine compared to albumin, however, reduction in creatinine did not reach the level of significance while compared with SMT/placebo. This is likely due to the effect of midodrine, which has been found to improve renal hemodynamics, and glomerular filtration rate (GFR) and promote sodium excretion in patients with cirrhosis [5,18,20]. In our analysis, we found midodrine to decrease plasma renin and aldosterone concentration compared to standard medical treatment alone. This is significant because this explains the beneficial effect of midodrine in paracentesis-induced circulatory dysfunction and the apparent lack of difference observed between patients treated with albumin and midodrine regarding the occurrence of PICD. Midodrine was found to improve urine output and cause weight loss in multiple studies [8,17,20]. However, the patients in these studies received concomitant diuretic therapy, which also leads to these changes, and the benefit cannot be solely credited to midodrine. We also found a significant reduction in MELD scores comparing patients treated with midodrine to standard medical treatment. A reduction in MELD scores is a possible prognostic factor for patients with cirrhosis and ascites. However, a previous study suggested reversible deterioration of MELD score with midodrine, octreotide, and albumin treatment for one month in refractory ascites [24]. This might be due to the co-administration of octreotide and midodrine for one month. Our analysis of MELD scores included studies in which patients received midodrine alone and for prolonged periods.
Our analysis found no difference in PICD between patients receiving albumin and midodrine while analyzing the results of four trials that reported on PICD [1,3,14,15]. PICD was defined as increased plasma renin by 50% from baseline at day six in studies [1,14]. Therapeutic paracentesis leads to depletion of intracellular volume, thereby activating the renin-angiotensin-aldosterone system and increasing renin levels. Expansion of plasma volume with albumin decreases the risks of paracentesis-induced circulatory dysfunction in various studies. However, we found no difference in PICD between patients treated with albumin and midodrine [1,16]. This finding was similar to the previous meta-analysis done by Guo et al. [9]. However, we did not find a significant difference in short-term mortality between midodrine and SMT, midodrine and albumin. Our findings are similar to the previous meta-analysis done by Guo et al., who found no improvement in mortality at one month [9]. Sola et al. reported renal impairment, hepatic encephalopathy, gastrointestinal bleeding, hyponatremia, and sepsis as some of the adverse effects of midodrine compared to placebo [22].
Our meta-analysis is the most comprehensive meta-analysis to date, including a total of 14 studies, and the second meta-analysis to evaluate the effect of midodrine in cirrhotic ascites. We have compared multiple outcomes regarding the use of midodrine in cirrhotic ascites to albumin and standard medical treatment. Terlipressin and albumin are treatments for refractory ascites, but both require intravenous access and are expensive. Our findings of midodrine being non-inferior to albumin regarding the occurrence of PICD and decrement in plasma renin and aldosterone are significant because midodrine is available in cheap oral formulation making it much easier to use.
Our study has several limitations. The endpoints for assessment of our outcomes were variable ranging from day four, day 10, one month to three months [8,[18][19][20][21]. In some of the studies, patients received concomitant adjuvant treatment like octreotide [8,15], and rifaximin [6]. Another significant limitation was the wide variation in the dosage and duration of midodrine ranging from three days to months, which caused heterogeneity in the reported results. Finally, there were inherent limitations in included studies like small sample size, lack of proper randomization, short duration of midodrine treatment, etc.

Conclusions
Midodrine alone leads to statistically significant improvement in various clinical parameters in patients with cirrhotic ascites compared to standard medical care. At the same time, it appears to be non-inferior to albumin. We report that the addition of midodrine to SMT for diuretic-resistant cirrhotic ascites would be beneficial. The results from our study call for further well-designed studies evaluating the combination of midodrine and albumin for optimal clinical benefits.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.