Short and Long-Term Outcomes of PSARP versus LAARP and Single versus Staged Repair for Infants with High-Type Anorectal Malformations: A Systematic Review and Meta-Analysis

Background: We aimed to compare among patients with high-type anorectal malformations (ARM): (i) short- and long-term outcomes of laparoscopic-assisted anorectoplasty (LAARP) compared to classic posterior sagittal anorectoplasty (PSARP) and (ii) the results of single-stage versus staged PSARP. Methods: Using a defined search strategy, two independent investigators systematically reviewed the English literature. PRISMA guidelines were followed, and meta-analysis was performed using RevMan5.3. Results: Of 567 abstracts screened, 7 papers have been included (254 pts; 121 PSARP, 133 LAARP) in the first systematic review and meta-analysis. The length of hospitalization was shortened in LAARP versus PSARP (10.9 versus 14.4 days; p < 0.0001). PSARP and LAARP were comparable in terms of early postoperative complications (28.9% versus 24.7%; p = ns) and rectal prolapse (21.6% versus 17.5%; p = ns). At long-term follow-up, the presence of voluntary bowel movements (74.0% versus 83.5%; p = ns) and the incidence of soiling (45.5% versus 47.6%; p = ns) were similar in both PSARP and LAARP. Six papers (297 pts) were included in the second systematic review, with three comparative studies included in the meta-analysis (247 pts; 117 one-stage, 130 staged procedures). No significant difference in terms of presence of voluntary bowel movements after single-stage versus staged procedures (72.6% versus 67.3%; p = ns) has been detected. Conclusions: LAARP seems to be a safe and effective procedure, showing short- and long-term outcomes similar to PSARP. One-stage PSARP could be a safe alternative to the classic three-stage procedure, even for those infants with high-type ARM. Further and larger comparative studies would be needed to corroborate these partial existing data.

The classic PSARP approach is a three-stage procedure, which starts with a colostomy creation, followed by the anorectoplasty, with the later closure of the colostomy usually after 3 months [1,5,6].
Continence after the classic PSARP procedure ranges from 2 to 84%, depending on the type of ARM, the grade of development of the musculature, and the sacral ratio.The range of continence decreased to 0-32% when analyzing only the high-type ARM [6][7][8].
In recent years, a single-stage repair of ARMs has been proposed in order to reduce the number of surgeries and anesthesia, to reach the early restoration of the anatomy and physiology of the lower GI tract, and to avoid the colostomy creation, thus ideally ameliorating the outcomes of the classic PSARP procedure [2,[4][5][6][7]12,13].
The one-stage procedure for infants with low ARMs has shown encouraging results [14,15].However, little is known about the short-and the long-term outcomes of patients with high-type ARMs (i.e., recto-urethral, recto-vesical fistulas, and rectovaginal fistulas) [12].
Therefore, we aimed to compare the short-and long-term outcomes of male patients with high ARMs treated with LAARP or PSARP procedures.Moreover, we aimed to assess the feasibility of the one-stage procedure in the same cohort of patients, thus comparing the outcomes of the single-stage versus the staged procedures in infants with high-type ARMs.

Data Sources and Study Selection
This study was registered on the international prospective register of systematic reviews PROSPERO (registration #CRD42022359940) (National Institute for Health Research) [16].The systematic review was drafted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [17].
A systematic review of the English literature was made using a defined search strategy (Table 1).those case series with less than 10 patients, opinion articles, and experimental papers.Grey literature (i.e., reports, theses, conference proceedings, bibliographies, commercial documentations, and official documents not published commercially) were excluded.Full text manuscripts of theoretically suitable papers have been saved and autonomously considered for appropriateness by the same two authors (MEM, GLa).
All those papers (trials, cohort, and case-control) reporting at least one outcome of interest have been included, according to our PICO (Supplementary File S2).Moreover, we included in the meta-analysis all papers comparing the outcomes of LAARP versus PSARP procedures to repair high-type ARMs and/or the single versus staged repair in the same patients.In cases of overlapping cohorts of patients, the authors selected the paper with the largest number of children for each outcome.The divergence over the appropriateness of a particular paper was solved thanks to the discussion with a further investigator (GLi).

Statistical Analysis
Categorical variable rates were compared with Pearson's chi-square test or the two-tailed Fisher exact probability test.When median and range have been reported, mean ± SD were estimated, as previously reported [18].Meta-analysis was managed with RevMan 5.4 [19], with a random effects model.Risk ratio (RR) has been considered for categorical variables.Mean differences (MDs) have been selected in those continuous variables.Results were reported with 95% confidence intervals (CIs).Data are expressed as mean ± SD.I 2 values were considered to assess homogeneity and quantify the dispersion of effect sizes.Biases among those studies included were assessed with the funnel plot.Quantitative and demographic data were compared using Fisher's exact test and are expressed as number, percentage, or mean ± SD, using the RR and 95% CIs.A p < 0.05 was considered significant.

Quality Assessment
Two independent authors (DDR and VC) evaluated the risk of bias for individual papers, thanks to the methodological index for non-randomized studies (MINORS) [20].Dissimilarities between them were solved through debate and agreement with a further investigator (GLa).The score for this 12-item index ranges between 0 and 24 points.The validated "gold standard" cut-off was 19.8 points.With regards to the quality for each outcome, authors graded the quality of evidence thanks to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology [21].The quality of evidence has been ranked as high, moderate, low, and very low for all outcomes.Observational studies were assessed with a low quality of evidence.The quality of evidence has been decreased in case of risk of bias, inconsistency, indirectness imprecision, and publication issues.The MINORS index has been assumed to assess the risk of bias in comparative studies.Inconsistency has been judged according to heterogeneity, and the I 2 value was applied to evaluate heterogeneity.I 2 values of 0-40, 30-60, 50-90, and 75-100% were considered as low, moderate, substantial, and considerable heterogeneity, respectively.Imprecision was evaluated using optimal information size (OIS), which was based on 25% relative risk reduction, 0.05 of α error, and 0.20 of β error [22].

2.
Six papers (297 pts) in the systematic review and three comparative studies (247 pts; 117 one-stage, 130 staged procedures) in the meta-analysis to compare the results of the single-versus the three-stage correction of high ARMs [4][5][6][7]12,13].All studies were retrospective.

Single-Stage versus Staged Procedures
Systematic Review: Among the 297 patients included in the study, 167/297 (56.2%) patients underwent a one-stage repair of the fistula and 130 (43.8%) a staged procedure.A total of 104/297 (35.1%) patients had a recto-urethral fistula and 1/297 (0.3%) patients had a recto-vesical fistula.No data were available regarding the type of fistula for the remaining 192 patients (64.6%).
Meta-analysis: We found the same incidence in terms of presence of voluntary bowel movements after the single-stage versus staged procedures (61/84, 72.6% versus 74/110, 67.3%; p = ns, Figure 7).Only the study by Xiao et al. [4] compared the post-operative complications and the incidence of constipation and soiling after the single versus staged

Single-Stage versus Staged Procedures
Systematic Review: Among the 297 patients included in the study, 167/297 (56.2%) patients underwent a one-stage repair of the fistula and 130 (43.8%) a staged procedure.A total of 104/297 (35.1%) patients had a recto-urethral fistula and 1/297 (0.3%) patients had a recto-vesical fistula.No data were available regarding the type of fistula for the remaining 192 patients (64.6%).
Meta-analysis: We found the same incidence in terms of presence of voluntary bowel movements after the single-stage versus staged procedures (61/84, 72.6% versus 74/110, 67.3%; p = ns, Figure 7).Only the study by Xiao et al. [4] compared the post-operative complications and the incidence of constipation and soiling after the single versus staged repair: no differences were found between the two groups (p = ns) (Table 4).

Single-Stage versus Staged Procedures
Systematic Review: Among the 297 patients included in the study, 167/297 (56.2%) patients underwent a one-stage repair of the fistula and 130 (43.8%) a staged procedure.A total of 104/297 (35.1%) patients had a recto-urethral fistula and 1/297 (0.3%) patients had a recto-vesical fistula.No data were available regarding the type of fistula for the remaining 192 patients (64.6%).
Meta-analysis: We found the same incidence in terms of presence of voluntary bowel movements after the single-stage versus staged procedures (61/84, 72.6% versus 74/110, 67.3%; p = ns, Figure 7).Only the study by Xiao et al. [4] compared the post-operative complications and the incidence of constipation and soiling after the single versus staged repair: no differences were found between the two groups (p = ns) (Table 4).
The aims of surgical repair of ARMs are to divide the rectum from the urinary tract, to correctly place the rectum at the center of the muscle complex, and to achieve voluntary bowel movements and continence [1].However, irrespectively from the procedure used, continence is also dependent on a well-developed sacrum (sacral ratio > 0.6) and a good perineal musculature [6].Moreover, an early stimulation of a brain-defecatory reflex is of utmost importance, since during the early neonatal period, the neuronal networks responsible for the reflex formation develop [5][6][7]12].
The aims of surgical repair of ARMs are to divide the rectum from the urinary tract, to correctly place the rectum at the center of the muscle complex, and to achieve voluntary bowel movements and continence [1].However, irrespectively from the procedure used, continence is also dependent on a well-developed sacrum (sacral ratio > 0.6) and a good perineal musculature [6].Moreover, an early stimulation of a brain-defecatory reflex is of utmost importance, since during the early neonatal period, the neuronal networks responsible for the reflex formation develop [5][6][7]12].
The gold standard treatment to repair ARMs is the classic PSARP, firstly described by Peña in the 1980s [6,10].This approach is based on three stages, with a stoma creation at first, followed by the anorectoplasty, and finally with the stoma closure [4,6].
In the early 2000s, Georgeson introduced the use of laparoscopy for the ARM correction [2,8].The LAARP has several advantages over the classic PSARP, thanks to a better and direct visualization of the fistula, with an improved placement of the rectum at the center of the perineal muscle complex.Moreover, LAARP guarantees no or minimal dissection of the muscle fibers, thus reducing the risk of nerve injury [2,3].The less manipulation of the muscles and nerves may be responsible for the better outcomes of the LAARP over the PSARP.In fact, some authors have reported a higher incidence of the recto-anal inhibitory reflex (RIRA) after LAARP when compared to PSARP, with better outcomes in terms of fecal continence [1,2,9,11].Other authors have reported comparable functional outcomes among the two groups [3,10].Nevertheless, many studies agreed that the outcomes improve over time, especially if a post-operative bowel management program has been planned at follow-up [8,9,23].
The main complication after LAARP is the rectal mucosal prolapse, which is reported in 9-46% of patients [9].However, this complication could be avoided by performing a minimal dissection of the mesorectum and/or performing a rectopexy [2,3].
However, the majority of the studies did not differentiate among the types of ARMs, thus leading to a bias linked to the different prognosis of the different types of ARMs themselves.
In our study, we have focused on newborns with high-type ARMs (i.e., recto-urethral and recto-vesical fistulas, and rectovaginal fistulas) and we have found a shorter hospitalization after LAARP compared to the PSARP procedure, leading to a quicker recovery.Moreover, we did not find a higher incidence of rectal prolapse among the LAARP group, thus suggesting that it is not as frequent as it has been postulated.In terms of early and late post-operative outcomes, we found that the presence of voluntary bowel movements and the incidence of soiling was not different among patients undergoing LAARP or PSARP repair of ARMs.All these data seem to suggest that LAARP is a valid alternative to PSARP.
The LAARP was also used to perform a single-stage repair of the ARM, without the colostomy formation [2].The rationale for a stoma creation is to allow the correct location of the fistula by performing a late cologram, to decompress the colon, to avoid infections, and to protect the anastomosis from stool passage [4].However, colostomy seems not to be free of complications.In fact, up to 74% of patients have reported developing stoma-related complications (such as prolapse, retraction, and adhesions).Moreover, many patients with a colostomy have been lost at long-term follow-up, especially in developing countries [6,12,13].Furthermore, the presence of a colostomy requires several surgeries in order to close it [5,7,12].
Therefore, in order to avoid the problems related to staged procedures, some authors started to perform a single-stage approach in well-selected patients (i.e., those without major cardiac anomalies or sepsis) [24].The advantages of a single-stage approach were the avoidance of multiple surgeries and anesthesia, the absence of a colostomy (and colostomy-related complications), and an early restoring of the anatomy, thus allowing an early training of the perineal musculature [4][5][6][7]12,13].The early restoration of the bowel continuity has been reported to be especially important for a precocious creation of the neuronal networks responsible for the brain-defecation reflex, thus ideally leading to better outcomes in terms of continence [4][5][6][7]12,13].Nonetheless, the single-stage procedure would also reduce the burden related to the psychological familial aspect of repeated surgeries and long follow-up [12].
The main concerns related to the single-stage procedure are the unawareness of the fistula location, the risk of secondary infections due to the presence of a colon dilated with meconium, and the lack of follow-up data [6,[12][13][14].However, these aspects do not seem to emerge from our review.To avoid the problems related to the level of the fistula, some authors have suggested performing a preoperative cystoscopy in order to visualize the fistula and even to perform a trans-fistula enema [12,25]; some other authors have suggested a transperineal evacuation of the meconium in order to obtain colon decompression, thus reducing the risk of infections [6,12].
To date, few reports are available regarding the short-and long-term outcomes of the one-stage correction of infants with high ARMs.As a matter of fact, in our study, we could include only three comparative papers.Nonetheless, we found that the presence of voluntary bowel movements did not differ among patients after single-stage or staged repair.Moreover, the reported incidence of soiling and constipation was comparable between the two groups.These data partially confirm and reinforce the idea that a singlestage repair of patients with high-type ARMs could be a safe and feasible option in wellselected cases.

Limitations of the Study
There are few limitations in the present meta-analysis.All but one paper were retrospective, as already mentioned.This aspect could produce selected bias.Moreover, not one study specified a sample-size calculation.As predictable, a blinded evaluation of impartial outcomes was not feasible.Furthermore, the number of children reported were generally not considerable.Both the LAARP and single-stage technique could be related to the volume of procedures.Therefore, those centers with a high volume of patients could present improved short-and long-term results compared to low-volume centers.Additionally, none of the included papers reported the loss of patients during the follow-up.Consequently, in our study, none of the included papers obtained the gold standard cut-off on MINORS (Table 5).Looking at the GRADE assessment, our meta-analysis reached a low quality of evidence on the incidence of complications and the presence of voluntary bowel movements, comparing LAARP versus PSARP, and very low with regards the other outcomes (Table 6).Either the reduced quantity of patients or the significant heterogeneousness of the records could produce potential bias.LOS: length of Hospital stay; LAARP: laparoscopic-assisted ano-rectoplasty; PSARP: posterior sagittal ano-rectoplasty.a Bias due to possible confounding; b OIS not met GRADE Working Group grades of evidence.High quality: Further research is very unlikely to change our confidence in the estimate of effect.Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.⊗⊗ Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.⊗ Very low quality: We are very uncertain about the estimate.
Children 2024, 11, x FOR PEER REVIEW 7 of 14

Figure 2 .
Figure 2. Forest plot comparison of length of hospitalization in the LAARP versus the PSARP groups [2,10].

Figure 2 .
Figure 2. Forest plot comparison of length of hospitalization in the LAARP versus the PSARP groups [2,10].

Figure 2 .
Figure 2. Forest plot comparison of length of hospitalization in the LAARP versus the PSARP groups [2,10].

Figure 2 .
Figure 2. Forest plot comparison of length of hospitalization in the LAARP versus the PSARP groups [2,10].

Figure 2 .
Figure 2. Forest plot comparison of length of hospitalization in the LAARP versus the PSARP groups [2,10].

Figure 6 .
Figure 6.Forest plot comparison of the incidence of soiling in the LAARP versus the PSARP groups [1,8].

Figure 6 .
Figure 6.Forest plot comparison of the incidence of soiling in the LAARP versus the PSARP groups [1,8].

Figure 6 .
Figure 6.Forest plot comparison of the incidence of soiling in the LAARP versus the PSARP groups [1,8].

Figure 7 .
Figure 7. Forest plot comparison of the presence of voluntary bowel movements after the single stage versus staged procedures [4,7,13].

Figure 7 .
Figure 7. Forest plot comparison of the presence of voluntary bowel movements after the single stage versus staged procedures [4,7,13].

Table 1 .
Inclusion criteria of the systematic review.

Table 2 .
Demographic data on studies included in meta-analysis.

Table 3 .
Summary of the results of the comparison of PSARP versus LAARP.p is statistically significant when <0.05.

Table 4 .
Summary of the results of the comparison of single-stage versus staged correction of ARMs.p is statistically significant when <0.05.

Table 4 .
Summary of the results of the comparison of single-stage versus staged correction of ARMs.p is statistically significant when <0.05.

Table 5 .
[20] of bias assessment for individual studies, adapted from the methodological index for nonrandomized studies (MINORS)[20].