Sialendoscopy for treatment of major salivary glands diseases: a comprehensive analysis of published systematic reviews and meta-analyses

Highlights • Sialendoscopy was effective and safe in obstructive salivary glands diseases.• Although it was an effective intervention, studies showed important heterogeneity.• All reviews had critically low quality of evidence when using the AMSTAR-2 tool.• We still lack comparative observational and interventional studies in sialendoscopy.• Future reviews should follow guidelines to improve study conduction and reporting.


Introduction
Salivary gland obstruction affects approximately 1% of the general population. Common symptoms include pain and edema that worsen when eating. Sialolithiasis is responsible for 60%---70% 1 of all salivary gland obstructions. When the stone is large enough to obstruct the salivary duct, there is accumulation of saliva, with eventual chronic inflammatory response. Persistence of the obstruction is a risk factor for retrograde infections because of stagnation of saliva.
Sialendoscopy can be used to manage both lithiasic and alithiasic diseases of salivary glands, also known as Obstructive Salivary Gland Diseases (OSGDs), and for diagnosis, treatment, and assistance in surgery. Until recently, treatment of sialadenitis in symptomatic cases consisted of gland excision, with an inherent risk for adverse events. 2 Asymptomatic or oligosymptomatic cases usually have conservative management, with satisfactory preservation of gland function, but with risk of salivary duct distension and persistence of symptoms due to saliva stagnation.
In the last few decades, minimally invasive techniques have been developed to treat symptomatic lithiasic and alithiasic salivary gland obstructions. Sialendoscopy uses small semi-rigid or semi-flexible endoscopes to access salivary ducts orally and visualize its' lumen. It was introduced as an alternative to surgical removal of salivary glands, reducing morbidity related to the procedure and preserving the salivary glands. Many studies have shown sialendoscopy as a viable option for removal of stones in salivary ducts. It can also be used in children to treat juvenile recurrent parotitis, although smaller salivary ducts in children are a complicating factor. For instance, sialendoscopy has been used as the preferred method to treat both lithiasic and alithiasic OSGDs in many countries.
There are several systematic reviews and meta-analyses assessing sialendoscopy in different lithiasic and alithiasic OSGDs in adults, adolescents, and children. Most of them have shown high efficacy and safety of the procedure. However, most studies evaluate different diseases of the salivary glands, applied to varied populations. There is no single study evaluating different outcomes in all OSDGs. Moreover, there is also unknown what is the overall quality of evidence of these published reviews. In the same way, there is high clinical heterogeneity between the different published studies what can cause confusion in the interpretation of these results.
The present study evaluated the efficacy of sialendoscopy to treat different OSGDs, lithiasic or alithiasic, analyzing all published systematic reviews and meta-analyses in the field. We also assessed all studies'results and evaluated methodological quality.

Eligibility criteria and data extraction
We made a systematic search using the terms (''sialendoscopy'' OR ''sialoendoscopy'') in the Medline database in PubMed, Embase, Lilacs and Cochrane Library. Systematic reviews and meta-analyses of clinical trials or observational studies of any language and date up to April 2022 were eligible for inclusion. We included studies with both lithiasic and alithiasic OSGDs. After the search, assessment for eligibility and data extraction were made by one reviewer (L.L.M.). Data extraction included cited studies, search strategy, language, period of search, databases searched, PICO strategy (Population, Intervention, Control and Outcome), 3 use of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline, 4 reporting of publication bias, primary outcomes, type of statistical analysis, heterogeneity, and reporting of methodological quality. Study characteristics were described by their search strategy, language, period of search, databases searched, other search strategies, use of PRISMA, reporting of publication bias, number of included articles, population, intervention, outcome, reporting of quality of evidence, main results, and reporting of heterogeneity. These characteristics were summarized in Tables. Methodological quality was assessed using the AMSTAR-2 tool. 5 AMSTAR-2 is a critical appraisal tool for systematic reviews that include both randomized and non-randomized studies. It is used to assess quality of evidence taking into account critical domains in construction and reporting of systematic reviews and based on that, rates the confidence in the results of the review as low, medium or high. The tool was used by two independent investigators (L.K.S. and M.F.D.). Lack of consensus in any item was resolved by a third author (L.L.M.).
AMSTAR-2 tool defines several critical domains that should be accounted for when evaluating study quality. These are: prior establishment of review methods through protocol (Item 2), use of a comprehensive literature search strategy (Item 4), list of exclusions with justifications for exclusions (Item 7), assessment of risk of bias (Item 9), use of appropriate method in meta-analyses (Item 11), interpretation and discussion of the impact of risk of bias in the results (Item 13), and assessment of publication bias (Item 15).
Other items are considered non-critical by the authors but are also important to be assessed for. According to AMSTAR-2 authors, studies with one critical flaw are considered of low quality. Studies with more than one critical flaw are considered of critically low quality. In case there are no critical flaws, studies have moderate quality if they have more than one non-critical flaw and have high quality if they have one or no non-critical flaw.
We also summarized studies' results separately for studies that included only lithiasic or alithiasic OSGDs, as well as studies that included both lithiasic and alithiasic OSGDs in their analyses. PRISMA reporting guideline was used in manuscript preparation.

Study selection
With the presented search strategy, we identified a total of 1,260 studies. Of those, 28 were identified to be systematic reviews or meta-analyses. After exclusion of 12 duplicate studies, 16 studies were assessed by full text. Three articles were excluded because sialendoscopy was not the intervention of interest and 13 studies 1,2,6---16 were included in the final analysis. A flowchart of the inclusion of studies is presented in Fig. 1.

Study characteristics
Most of the assessed systematic reviews and meta-analyses included only observational studies, 1,2,6---11,13,15,16 with most of them being retrospective studies. All of them reported their search strategy and all but one, 9 reported search period. Most of them searched through a variety of databases, except for two, 12---15 which included only pubmed. Seven studies described the use of PRISMA guideline for reporting results. 6,8,10,12---14,16 The population evaluated in different studies significantly varied. Of the nine studies in adult populations, four studies included only lithiasic OSGDs, 1,7,13,15 one study included OSGD with underlying Sjogren syndrome, 14 one  10 and three studies included both lithiasic and alithiasic OSGDs. 6,9,16 Of the 4 studies in children and adolescents, two included only lithiasic OSGDs 9,11 and two included juvenile recurrent parotitis. 8,12 All of them included use of sialendoscopy as an intervention, although some included associated interventions, such as corticosteroid use, 14,16 saline solution, 14 and other medication treatments. 12 In two studies, the intervention was surgery combined to sialendoscopy assistance. 1,7 The studies assessed different outcomes: sialendoscopy effectiveness, 1,2,6,9,11,16 symptoms resolution, 7,8,10,13---16 safety and adverse events1, 6,11 recurrence of the disease or symptoms, 12 and salivary gland preservation. 13 Study PICO strategy is shown in Table 1. We did not report control analyses because most primary studies had  Table 2. We also evaluated what primary studies were cited in each systematic review and metaanalysis using a citation matrix (Table S1; Supplementary Material). As noted, there was a large variability in the primary studies included in the reviews, although pediatric studies were more likely to include the same primary studies in their reviews.

Overview of study results
In general, the sialendoscopy method was referred to as effective and well tolerated. Individual study results are presented in Table 3. In studies evaluating exclusively adults with only lithiasic OSGDs, success rate in stone removal was highest in Roland (2017), 1 and all other reviews identified high success rate, although Jadu (2014) 7 included studies with moderate heterogeneity of success rates. Symptom improvement and resolution was high, with a low rate of complications. In children and adolescents, although studies had a very low quality of evidence, sialendoscopy was effective with high improvement of symptoms.
In studies including both lithiasic and alithiasic OSGDs in adults, we identified a large heterogeneity of outcomes, which makes it difficult to interpret their individual results. Although success rates of sialendoscopy alone was lower in Strychowsky (2012) 6 and Atienza (2015), 2 there was a high success rate when sialendoscopy was used combined with open surgery. Donaldson (2021) 16 had a high success rate. All three studies had few adverse events.
In studies analyzing specifically alithiasic OSGDs in adult populations, Cung (2017) 10 identified a high clinical improvement in patients with radioiodine-induced sialadenitis, with moderate quality of evidence and heterogeneity (although meta-analysis was not conducted). Coca (2020) 14 identified a high clinical improvement in patients with underlying Sjogren's syndrome. In children and adolescents with juvenile recurrent parotitis, success rate (as defined by absence of recurrence) was moderate in both studies.

Assessment of methodological quality
Assessment of methodological quality using the AMSTAR-25, 17,18 tool is shown in Table 4. Since we included studies regarding different populations, we will present the results according to the research question. Although not evaluated through the AMSTAR-2 tool, it is important to note that most systematic reviews and meta-analyses included low quality primary studies, most of them being observational retrospective studies. Few studies included Randomized Controlled Trials (RCTs).
Overall compliance was 46% (range 25%---75%), and compliance in critical domains was 23% (range 0%---57%). All reviews had critically low quality of evidence. Critical domains with the lowest compliance were previous protocol registration (Item 2) and assessment of publication bias (Item 15). The only domain with no compliance was report-ing of funding in individual studies, which is a non-critical domain.
Of four studies in adult populations that included only lithiasic OSGDs, Roland (2017) 1 and Chiesa-Estomba (2020) 13 had higher methodological quality in adult populations. When regarding lithiasic OSGDs in pediatric populations, Schwarz (2017) 11 had a slightly higher methodological quality than Silva (2016). 9 In studies that included both lithiasic and alithiasic OSGDs, the three included studies had very similar methodological quality in adults. In studies with only alithiasic OSGDs in adults, Coca (2020) 14 had a higher methodological quality, although it included only OSGDs with underlying Sjogren syndrome. Ramakrishna (2014) 8 had a higher methodological quality in the investigation of juvenile recurrent parotitis.

Discussion
In the present study, we assessed the efficacy and safety of sialendoscopy to treat several OSGDs in different populations and also in different clinical scenarios. Our analysis included all systematic reviews and meta-analyses in the topic. Studies' overall results and methodological quality were assessed. According to the available evidence, sialendoscopy has shown to be an effective and safe technique to treat OSGDs. However, we found that all systematic reviews published in the topic have critically low quality of evidence, when assessed by the AMSTAR-2 checklist. The present study provides evidence-based guidance for clinical practice, considering different populations, several diseases, and outcomes.
There are several evidence-based studies published in the literature considering sialendoscopy in the treatment of OSGDs. These studies include systematic reviews with or without meta-analysis with different populations and different outcomes. These studies, sometimes, evaluate the same outcome, but with different inclusion criteria, making the comparison between them sometimes impossible and also demonstrating divergent results. Moreover, the lack of an objective analysis in terms of the quality of the generated evidence, makes it impossible to safely employ these results into clinical practice.
Most analyzed publications included primarily retrospective studies, resulting in low to moderate quality of evidence. There are important biases to consider when analyzing retrospective studies, such as recall bias in subjective outcomes and lack of randomization and close follow-up of patients. Short follow-up was also an important limitation of the primary studies, hindering our capability of identifying long-term symptoms. Study outcomes were assessed and reported differently, adding more bias and heterogeneity. Symptom improvement may be the most important outcome since it is a clinical outcome, but lack of validated standardized methods for evaluating symptoms makes it harder to define what is a relevant improvement. Thus, bias was largely present in our evaluation, which importantly limits reliability of results in studies of sialendoscopy.
Heterogeneity played an important role in our analyses. Generally, included reviews had moderate heterogeneity in their data. This could be due to small sample sizes, variability in technique and surgical equipment and inconsistent    reporting. Most meta-analyses included several case reports and case series. These study designs should not be used in meta-analyses, due to large variability of methodology. This also adds heterogeneity to the studies. There was also large heterogeneity in our review, since most studies included different populations, interventions, and outcomes. Even when included reviews evaluated the same research question, there was important heterogeneity in their results, which we attribute to variable methodological quality. The presence of many sources of heterogeneity limits the interpretability and generalizability of all included reviews. A point should be made regarding the separation of the analysis between lithiasic and alithiasic OSGDs. As noted, 3 studies included both lithiasic and alithiasic OSGDs in their analyses. We opted to include these studies to guarantee that all available data in OSGDs was covered. Moreover, since there was large overlap of primary studies included in the reviews, we cannot properly evaluate the accuracy of sialendoscopy through a simple division of lithiasic and alithiasic OSGDs. However, this does not impact the quality of evidence of presented in each population, since the critically low methodological quality was present in all studies.
Methodological quality of reviews of studies in sialendoscopy was also an important limitation. All included studies had a critically low methodological quality, according to our assessment using the AMSTAR-2 tool, making this point one of the largest weaknesses in the outcome assessment of sialendoscopy technique. Studies lacked especially critical domains, leading to poor ratings. According to AMSTAR-2, studies should be classified with critically low quality when they have more than one critical flaw. The great majority of studies did not include Items 2 (prior establishment of methods) and 15 (assessment for publication bias), considered to be critical, which already led to most of them having critically low quality. We also noted that PRISMA guideline use played an important role when assessing methodological quality. Studies that used it to report their results had generally higher quality of evidence.
Additionally, we would like to make some comments regarding our assessment of methodological quality. In Item 2, regarding prior establishment of review methods, we considered as a ''no'' when studies reported use of a protocol, but the protocol could not be found in supplementary material or in online platforms for protocol registration. Following AMSTAR-2 recommendations, studies that verified only one database received a ''no'' in Item 4, regarding search strategy. In Item 7, when authors did not report reading articles in full text, the review received a ''no''. This is a point of caution that should be a warning for all authors that intend to conduct a systematic review, with or without a meta-analysis. The better the authors describe the methodology applied at the study, the better the quality score. We also strongly recommend that authors register their projects into specific platforms, assuring a high-quality publication.
Finally, it is important to note that some studies performed meta-analyses, whereas others do not. Naturally, studies with no meta-analyses have lower ratings since they do not score in some items. Even though this is a limitation, it is important to note that, even in other items that do not include only meta-analyses, studies that made metaanalyses also had less flaws.
More primary and secondary studies should be performed using sialendoscopy, with more rigid methodologies and less predictable bias, to better establish the method as goldstandard for OSGDs treatment.

Conclusion
In this analysis of systematic reviews and meta-analyses, we found sialendoscopy to be efficacious and safe. However, the included studies showed critically low quality of evidence. We still lack randomized studies in this field, and future systematic reviews on the topic should follow current guidelines to improve conduction and reporting.

Disclosures
The preliminary results of this work were presented at the 2022 American Academy of Otolaryngology -Head and Neck Society Foundation Annual Meeting as a poster.

Financial disclosure
This research did not receive any specific funding from funding agencies in the public, commercial, or not-for-profit sectors.