**3. Results**

The search strategy generated 1317 articles (Figure 1). After the exclusion of 584 articles, 733 articles were identified as eligible records. However, 723 articles were excluded because they did not fulfil all eligibility criteria. Therefore, 10 articles resulted as eligible, but 7 were excluded because they were prospective [18–21] or retrospective [22–24] studies. In the end, three RCTs were included [25–27].

**Figure 1.** Flow chart of the study selection. RCT - randomized clinical trial.

The three RCTs included were an Open, Randomized, Multicentre Trial [25], a Randomized Double-Blind Clinical Trial [26] and an Open-Label Phase II Randomized Trial [27]. They were published between 2012 and 2016, involving a total of 76 patients, of which 44 patients received topical dexamethasone, 14 patients received topical clobetasol and 18 patients received topical budesonide. The studies were conducted in Israel/Germany [25], Brazil [26] and the USA [27]. The mean age of the participants varied from 43.8 to 55 years and the sex ratio was female dominant.

Oral GVHD diagnosis was done on different parameters across the included studies: WHO toxicity oral/gastrointestinal, modified oral mucosal rating scale (mOMRS), Oral Mucositis Assessment Scale (OMAS), National Institute of Health (NIH) oral cavity severity score, mucosal score and oral symptoms score. Oral lesions involved in GVHD were erythema, atrophy, ulcer, lichen, hyperkeratosis, pseudomembrane, edema and mucocele, appearing as a mucus cyst on the soft palate, on the labial

and buccal mucosa. Clinical response to these agents were 61% for WHO toxicity oral/gastrointestinal, 50–61% for mOMRS, 69% for OMAS and 50% for NIH oral cavity response. Side effects reported were cheilitis, esophagitis, fungal infections, taste alteration, burning sensations and oral cavity pain. Additional data can be found in Table 1.

Figure 2 represents the overall response between the included studies. Red rectangles represent the proportion of patients that presented a response. Because of the heterogeneity of the studies, we considered for Elad et al. [25] the mOMRS any response, for Noce et al. [26] the symptomatic response and for Treister et al. [27] the overall response described by the authors. It appears that the best overall response is present for budesonide with no difference between the four arms, followed by clobetasol and then by dexamethasone. The limitation of the current study is mainly represented by the fact that the overall response was derived in two of the studies from other parameters. Moreover, both budesonide and clobetasol were used in only one study each, while two assessed dexamethasone.

**Figure 2.** The overall response between the arms of the three studies included.

Overall, the three included RCTs were considered at "risk of bias" because of the lack of blinding of study participants, blinding of the outcome data and other bias. Figure 3 shows the analysis for the risk of bias for RCTs.



105

NR, no response;

hematopoietic cell transplantation; mOMRS, modified oral mucosal rating scale; NIH, National Institute of Health; OMAS, Oral Mucositis Assessment Scale; PR, partial response;PD,progressivedisease.

**Figure 3.** Risk of bias for the RCTs.

## **4. Discussion**

Up to this moment, in the current literature, there are two systematic reviews trying to assess the benefits of using topical agents in oral GVHD [10,28]. Albuquerque et al. observed that there are a limited number of RCTs, and, therefore, the evidence sustaining the use of topical agents for the inflammatory lesions in oral GVHD is low [10]. The same authors have stated that there is still a need for quality RCTs to assess the efficacy of these agents in GVHD [10]. The paper of Elsaadany et al. reported moderate evidence for the efficacy of topical agents for oral GVHD, showing minimal side effects of clobetasol followed by budesonide [28]. Our systematic review had a homogenous selection of randomized clinical trials allowing a calculation of the Cochrane risk of bias and of the overall response, and, therefore, giving a clear recommendation for a better efficacy of budesonide, compared to clobetasol and then dexamethasone.

Another powerful parameter we have evaluated from the RCTs included in our review is the safety profile of the therapy. In the study of Elad et al. [25], the use of topical budesonide (3 mg/10 mL) showed that this corticosteroid had a satisfactory safety profile. Topical budesonide mouthwashes improved the oral condition when it was applied for 5–10 min, 2/3 times per day. Regarding the response in all treatment arms, Elad mentioned that it was the same, in any length of exposure to treatment and in any frequency. Safety analysis was performed at a dosing schedule of budesonide 3 mg, three times per day, for a period of 10 min, representing the most important exposure to the drug [25].

In the study of Noce et al. [26], a significant improvement in the symptoms appeared comparing to the baseline after the use of corticosteroids, but with a significantly greater response in the topical clobetasol group compared to dexamethasone. The authors stated that the limitations of their research were the low number of subjects and other confounding variables. They indicate that these variables in further studies should be taken into account, with a larger sample size and stratification of subjects [26].

Treister et al. [27] observed in their study that the patients with dexamethasone obtained a response of 58%. The overall global response rates were reported to reach up to 81% including responses such as much better. In total, 96% of the patients reported the dexamethasone rinses as being well tolerated and the taste being "very pleasant" or "tolerable". According to these results, the authors concluded that intensive topical therapy with this agent is efficient for managing oral chronic GVHD and should be used as a first line therapy [27].

Other topical therapies studied in the literature include triamcinolone, fluocinonide, betamethasone, tacrolimus and prednisolone, fluocinonide, halobetasol prepared as a gel or cream and

topical platelet-rich gel, with various results. Because of their lack of availability in all countries we have used in our inclusion criteria only the studies mentioning the most common and used steroids: budesonide, clobetasol and dexamethasone [11].

Our research tried to do a systematic review and a meta-analysis. The second objective was not able to be fulfilled because of the heterogeneity of the articles included. The low number of RCTs included in this review represents a major limitation in concluding on the efficacy of topical corticosteroids in oral GVHD and establishing a therapy protocol. On the other hand, several weaknesses were observed within the studies included. First, a variation in assessing oral GVHD parameters before/after topical agents was observed. The adjustments of the oral GVHD parameters is required. Secondly, the sample sizes were too small; a larger sample size should be used in future studies. Third, all RCTs included only chronic GVHD patients, excluding the acute alarming manifestations. This research was made evaluating only three databases and included only articles written in English, leading to a possible exclusion of other important data.
