Journal of Plastic, Reconstructive & Aesthetic Surgery 84 (2023) 134–146 Review Risk-of-bias assessment of the randomized ]] ]]]]]] ]] clinical trials and systematic reviews on surgical treatments for breast cancer- related lymphedema: A mapping review ⁎ Ali M. Al-Sakkaf a, , Xavier Bonfill b,c, Sofia Ardiles-Ruesjas b, Josefina Bendersky-Kohan b,c, Ivan Sola b,c, Jaume Masia a a Department of Plastic Surgery, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Barcelona, Spain b Biomedical Research Institute Sant Pau (IIB Sant Pau), Barcelona, Spain c Iberoamerican Cochrane Centre, Biomedical Research Institute Sant Pau (IIB Sant Pau), CIBERESP, Universitat Autònoma de Barcelona, Barcelona, Spain Received 17 August 2022; Accepted 2 May 2023 KEYWORDS Abstract Background: Breast cancer treatment is the principal cause of lymphedema in the Evidence mapping; upper extremities. Breast cancer-related lymphedema (BCRL) treatments were previously Risk-of-bias based on conservative therapy; surgical treatments are alternative options that could be highly assessment; beneficial, especially for patients who are not responsive to conservative therapy. The main Breast cancer-related aim of this study was to describe and critically assess the risk of bias of randomized clinical lymphedema trials (RCTs) and systematic reviews (SRs) on surgical treatment for BCRL. Methods: We conducted an evidence mapping review according to the methodology proposed by Global Evidence Mapping (GEM). An update was done for our previous systematic search in MEDLINE, EMBASE, CENTRAL (Cochrane), and Epistemonikos from the year 2000 onward. We assessed the risk of bias for the RCTs and SRs using the RoB-2 and ROBIS tools, respectively. Results: Two surgical RCTs and eight SRs were found among the 47 surgical studies that met the eligibility criteria. The overall risk-of-bias assessments of these studies were rated as some concerns (six outcomes) and high risk (three outcomes) for the measured outcomes among the RCTs and as a high risk of bias (five studies) and low risk (three studies) for the included SRs. Abbreviations: BCRL, Breast cancer-related lymphedema; RCTs, Randomized control trials; SRs, Systematic reviews; GEM, Global Evidence Mapping; LVA, lymphaticovenular anastomosis; VLNT, vascularized lymph node transfers ⁎ Correspondence to: Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Calle de Zaragoza 79, 3-2, 08006 Barcelona, Spain. E-mail address: sakaf_250@yahoo.com (A.M. Al-Sakkaf). https://doi.org/10.1016/j.bjps.2023.05.002 1748-6815/© 2023 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. Journal of Plastic, Reconstructive & Aesthetic Surgery 84 (2023) 134–146 Conclusions: The overall evidence in the literature on surgical treatment for BCRL is low, as there are few published RCTs and SRs, and the risk-of-bias assessment for the majority was rated as high risk of bias or with some concerns. High-quality studies are needed to improve evidence-based decision-making by surgeons and patients. © 2023 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. Contents . Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 . Eligibility criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 . Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 . Study selection and data extraction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 . Assessment of risk of bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 . Data synthesis and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 . Search results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 . Surgical studies on treatment for BCRL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 . RCTs’ characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 . SRs’ characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 . Matrix of evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 . Risk-of-bias assessment for RCTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 . Risk-of-bias assessment for SRs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 . Conflicts of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 . Ethical approval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 . Funding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 . Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 . Appendix A Supporting information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Background that could be highly beneficial, especially for patients who are not responsive to standard conservative therapy, which includes Lymphedema is defined as the abnormal collection of lym­ mostly excisional and reconstructive techniques.11–13 phatic fluid within subcutaneous structures. In advanced Both excisional and reconstructive surgical approaches countries, damage to the lymphatic system due to cancer or have been described in the treatment of BCRL. Excisional or its treatment is the most common cause of secondary lym­ nonphysiological procedures include the Charles operation phedema. In the upper extremities, breast cancer treat­ and liposuction. These strategies are most often performed ment is the principal cause of lymphedema.1 in a later stage of disease when there are no remaining According to the literature, the incidence of breast cancer- functional lymphatic vessels.14,15 Reconstructive options, related lymphedema (BCRL) depends on the type of axillary on the other hand, are physiological operations that aim to treatment; axillary lymph node dissection results in lymphe­ restore lymphatic flow to aid in lymphatic drainage from the dema in up to 53.5% of cases, and sentinel lymph node biopsy affected extremity. These include lymphaticovenular ana­ results in lymphedema in up to 15.8% of cases.2,3 Other risk stomosis (LVA) and vascularized lymph node transfers factors that aggravate the condition are adjuvant radiation, (VLNT), which currently have promising results for treating docetaxel chemotherapy, infection, iatrogenic injury, and the early stages of lymphedema.16–20 obesity. Consequently, developing lymphedema leads to a Our team previously conducted a mapping review on all chronic condition that is usually challenging to treat.4–10 treatments for BCRL, without assessing the risk of bias of the BCRL treatment options have long been based on con­ included studies, and did not focus on surgical treatment.21 servative therapy, such as compression garments/bandages and Therefore, based on our previous mapping findings and the manual lymph drainage. These conservative measures are limited knowledge of the quality of the available research, the mainly aimed at alleviating lymphedema symptoms without main aim of this study was to describe and critically assess the curative intent. Surgical treatments are alternative options risk of bias of randomized clinical trials (RCTs) and systematic 135 A.M. Al-Sakkaf, X. Bonfill, S. Ardiles-Ruesjas et al. reviews (SRs) on surgical treatment for BCRL. Other objectives inclusion and exclusion criteria. Disagreements between the are to identify gaps in knowledge, enumerate the limitations two reviewers were resolved mainly by the third reviewer. At and constraints that exist in this field, and provide re­ this step, the reasons for exclusion were recorded. commendations for future research needs. For each study, data extraction was conducted sepa­ rately by the two reviewers in a predesigned spreadsheet (AMS and AIS). The results were then compared, and in case Methods of disagreement, the third reviewer (LVC) acted as a referee to reach consensus. All extracted data were recorded in a An evidence mapping review was conducted according to data extraction sheet using Microsoft Excel. the methodology proposed by Global Evidence Mapping (GEM)22 and adhered to the Preferred Reporting Items for Assessment of risk of bias Systematic Reviews and Metanalysis (PRISMA)—Extension for Scoping Reviews.23 All methods were specified a priori in a protocol (available on request). Methodological assessment of risk of bias was independently assessed by three reviewers (AMA, SAR, and JBK). Each article was assessed blindly by two reviewers (AMA and SAR), and any Eligibility criteria disagreement in the results was resolved by the third reviewer (JBK). The risk-of-bias assessment was done only for high evi­ We updated our search strategy based on our previous mapping dence studies addressing the surgical intervention (SRs and work.21 It was built on the population, intervention, compar­ RCTs); for that reason, we did not consider the necessity of ison, outcome, and type of study (PICOT) framework to for­ assessing the risk of bias for the nonrandomized studies. mulate the eligibility criteria.24 We considered eligible patients For RCTs, the Cochrane risk-of-bias tool for randomized (older than 18 years) with BCRL. Those who had either surgical trials—version 2 (RoB-2) was used for the assessment of each or nonsurgical treatments for BCRL were initially eligible. Due outcome in the RCTs.25 The domains included in the RoB-2 to the nature of this study, we included studies with any type of are as follows: bias arising from the randomization process, comparison and those without a comparison group. All out­ bias due to deviations from intended interventions, bias due comes were eligible for this mapping review. This mapping re­ to missing outcome data, bias in measurement of the out­ view included all published studies in full text from the year come, and bias in selection of the reported result. The risk- 2000 onward, including SRs with or without metanalysis, RCTs, of-bias judgment for the RCTs was then assigned to one of quasi-experimental clinical trials, and observational studies three levels to each domain: low risk of bias, some con­ (prospective and retrospective studies), to have a broader look cerns, or high risk of bias.25 at the available evidence in this field. When several studies SRs were assessed by the ROBIS tool.26 The tool is com­ published on the same topic and by the same team were pleted in three phases: phase 1 consists of assessing the identified, we considered the most recent publication. We ex­ relevance (this was optional and not applied in this article); cluded animal studies, in vitro studies, single case reports, case phase 2 consists of identifying concerns with the review series, letters to the editor, narrative reviews, studies including process, covering four domains: study eligibility criteria, different types of edemas or mixed edema, studies including identification and selection of studies, data collection and less than 10 patients or reviews with fewer than three studies, study appraisal, and synthesis and findings; and phase 3 and studies addressing other than treatment of BCRL or ad­ consists of judging the risk of bias and assessing the overall dressing both prevention and treatment together. risk of bias in the interpretation of review findings and whether this considered limitations identified in any of the Search strategy phase 2 domains. The risk-of-bias judgment for SRs is then assigned as low risk, high risk, or unclear concern.26 The search strategy was conducted in MEDLINE (via PubMed) and EMBASE (via Ovid), Epistemonikos, and the Cochrane Library from the year 2000 onward. A search algorithm was Data synthesis and analysis designed, including a combination of controlled vocabulary, the use of MeSH descriptors, free-text term, and thesaurus term As recommended in the Cochrane Handbook for Systematic when available, adapting it accordingly for each database, with Reviews of Interventions,27 a flow chart for the whole no language restriction, and no gray literature was searched. process of study selection was elaborated based on the The last update was done on 22nd of October 2021 (the search Preferred Reporting Items for Systematic Reviews and Meta- strategy is attached as Supplementary material). Analyses (PRISMA-P diagram).28 The obtained results were presented in a narrative and visual format using tables, Study selection and data extraction figures, and a bubble plot. The bubble plot was created to illustrate the study designs in relation to their risk-of-bias The studies were retrieved by titles and abstract and were assessment; the color of the figures indicated the study uploaded to Mendeley and then managed with Rayyan QCRI design (RCTs or SRs); the size of the figure reflected the software. After removing duplicates, three reviewers (AMA, number of population or number of studies in the included AIS, and LVC) independently screened all titles and abstracts, RCTs and SRs, respectively; and their positions in the graph with each article screened by at least two reviewers. were based on their overall risk-of-bias assessment. Afterward, full-text screening was done independently by the The analysis of the selected studies was divided into two same three reviewers who confirmed eligibility based on the parts: first, a general mapping presentation of the included 136 Journal of Plastic, Reconstructive & Aesthetic Surgery 84 (2023) 134–146 surgical studies included in this review, providing a more in which the full text was missing were also excluded from detailed description of the available SRs and RCTs and, the descriptive analysis. Finally, a total of 272 studies were second, an assessment of the risk of bias of the RCTs and SRs included, of which 225 were nonsurgical studies and 47 addressing the surgical intervention, using the RoB-2 and studies addressed the surgical treatment for BCRL. Of these ROBIS tool, respectively. Due to the large amount of data surgical studies, only two RCTs and eight SRs were critically collected from the eligible studies, we focused mainly on assessed for the purpose of this mapping review. the important results that contributed to the objectives of The main reason for excluding studies was that the ar­ this article. ticles were published as conference abstracts (41). Other reasons included foreign languages (other than English and Results Spanish) (19), wrong population (18), wrong design (10), wrong objective (7), published protocol (3), case report (1), Search results editorial reply (1), literature review (1), population < 10 patients (7), reviews including < 3 studies (2), and the The flowchart of the study selection of the baseline re­ aforementioned missing full text (7). search and update is shown in Figure 1. The search after the last update yielded a total of 5663 studies. After removing 1919 duplicates, we proceeded with 3744 studies to screen Surgical studies on treatment for BCRL by title and abstract. In total, 3355 studies were excluded because they were unrelated to the review’s main topic. There were 47 studies addressing surgical intervention (39 Then, a full-text review of 389 studies was conducted. After primary studies and eight secondary studies), which in­ the resolution of discrepancies by consensus between re­ cluded 15 experimental studies (13 quasi-experimental searchers, we excluded 110 studies. Similarly, seven studies clinical trials and two RCTs), 24 observational studies (14 Figure 1 PRISMA flow diagram and selection process of studies on surgical treatments for BCRL (baseline research and update). 137 A.M. Al-Sakkaf, X. Bonfill, S. Ardiles-Ruesjas et al. Figure 2 Geographic distribution of total published studies on surgical treatments for breast cancer-related lymphedema (BCRL). prospective cohorts and 10 retrospective cohorts), and the University of Western Ontario Microsurgical Skills Ac­ eight SRs with and without meta-analysis. quisition Instrument (UWOMSA) scoring34 (see Table 1). These published studies were geographically distributed among the following countries: the Netherlands had the SRs’ characteristics highest number of publications (9), followed by the United States of America (7), France (4), Sweden (4), China (4), Among the eight SRs on surgical treatments for BCRL,35–42 three Taiwan (3), Japan (2), Brazil (2), Spain (2), Italy (2), Greece SRs performed quantitative assessment (metanalysis).36,37,41 (2), the United Kingdom (1), Belgium (1), Denmark (1), Three were conducted in the Netherlands,35,36,41 three in the Poland (1), Thailand (1), and Australia (1) (see Figure 2). United States of America,37,39,40 one in Greece42, and one in Brazil.38 There was heterogeneity in the included study designs, mainly among case series, case reports, prospective studies, RCTs’ characteristics retrospective studies, and nonrandomized trials. Two SRs did not mention the type of the included study design. The range of The first RCT, by Dionyssiou et al.,29 was conducted in included studies was from five to 17 studies. One SR included a Greece and compared the VLNT to the conservative mea­ total of 67 studies, but only 13 were described in the qualita­ surements, and the second RCT, by Van Mulken et al.,30 tive synthesis, which were addressed in our results.42 These SRs conducted in the Netherlands, compared the robotic versus addressed different surgical interventions, such as VLNT,38,41 the manual LVA. Both RCTs assessed different outcomes. LVA,3 both VLNT and LVA,35,42 or combined treatment such as Dionyssiou et al. mainly assessed three patients’ outcomes: autologous reconstruction with VLNT37,40 or lipoaspiration with upper limb volume, infection episode, and subjective VLNT.39 The only common outcome that was measured in all symptoms using the subjective analog scaling system.31 Van included SRs was limb volume, although different outcomes Mulken et al. assessed various outcomes, including four were also assessed, such as subjective symptoms, quality of patients’ outcomes: the daily use of compression garment, life, infectious episodes, complications, and discontinuation of the need for manual lymphatic drainage, the arm cir­ conservative treatments (see Table 1). cumference using the mean upper extremity lymphedema index (mean UEL index), and quality of life using a validated health questionnaire, the mean Lymphedema Functioning, Matrix of evidence Disability and Health questionnaire (Lymph-ICF).32 Further­ more, this RCT assessed two surgeons’ related outcomes: We created a matrix of evidence to show the SRs linked to the duration of surgery and quality of anastomosis using the the included primary studies and the overlaps of the pri­ Structured Assessment of Microsurgery Skills (SAMS)33 and mary studies between these SRs. Because there was 138 Table 1 Characteristics of the included SRs and RCTs on surgical treatments for BCRL. Randomized Clinical Trials Number Study ID Country Title N. of population Intervention Comparison Measured outcomes Reported results 1 Dionyssiou Greece A randomized control 36 patients Vascularized lymph Physiotherapy and 1. Upper limb volume Results favor et al.29 study of treating node transfer (VLNT), compression alone 2. Infection episodes toward secondary stage II physiotherapy and 3. Subjective symptoms intervention breast cancer-related compression lymphoedema with free lymph node transfer 2 Mulken Netherlands First-in-human robotic 20 patients Robot-assisted Manual 1. Daily use of No difference et al.,30 supermicrosurgery supermicrosurgical supermicrosurgical compression between both using a dedicated lymphaticovenous lymphaticovenous garment comparative groups microsurgical robot for anastomosis (LVA) anastomosis (LVA) 2. Manual lymphatic treating breast cancer- drainage related lymphedema: 3. Mean lymph – ICF a randomized pilot 4. Mean UEL Index trial 5. Duration of surgery 6. Quality of anastomosis Systematic Reviews Number Study ID Country Title N. of studies Intervention Metanalysis Measured outcomes Reported results 139 1 Penha Netherlands Microsurgical 10 studies (Case Lymph node No 1. Limb volume or Consist positive et al.35 techniques for the series) transplantation circumference finding with regards treatment of breast (VLNT), lymph vessels 2. Lymph flow by to limb volume cancer-related transplantation, or lymphoscintigraphy reduction and lymphedema: a derivative lymphatic 3. Symptoms relief limited systematic review surgery (LVA). 4. Discontinuation of complications post operative conservative therapy 5. Complications 2 Cornelissen Netherlands The effect of 15 studies (10 Lymphaticovenous Yes 1. Limb volume or Heterogeneous et al.36 lymphaticovenous prospective and 5 anastomosis (LVA) circumference results in volume/ anastomosis in breast retrospective 2. Quality pf life circumference cancer-related studies) reduction and Journal of Plastic, Reconstructive & Aesthetic Surgery 84 (2023) 134–146 lymphedema: A review improvement of of literature quality of life in most of the studies (continued on next page) Table 1 (continued) 3 Siotos USA Delayed breast 8 (7 case series Delayed autologous Yes Any outcomes reported The VLNT et al.37 reconstruction on and 1 case report) breast reconstruction composite of the patients with upper with or without lymph autologous breast extremity node transfer (VLNT) reconstruction lymphedema a might be the systematic review of largest contributing the literature and factor leading to pooled analysis lymphedema improvement 4 Ribeiro Brazil Lymph node 10 studies Autologous lymph No Limb volume There is a et al.38 transplantation in the node considerable management of post- transplantation (VLNT) reduction in limb mastectomy volume higher lymphedema: a than 50% systematic review with metanalysis 5 Forte USA Lipoaspiration and 5 studies (3 Lipoaspiration and No Limb volume Meaningful volume et al.39 lymph node transfer prospective venous lymph node reduction was for treatment of breast studies, 1 case transfer (VLNT) achieved in all cancer-related series and 1 case cases, patients who lymphedema: a report) underwent lymph 140 systematic review node transfer first followed by lipoaspiration appear to have the best results 6 Forte USA Lymph node transfer 6 studies Autologous breast No 1. Limb volume or Most of the studies et al.40 combined with deep reconstruction with circumference demonstrated inferior epigastric Lymph node 2. Symptoms relief subjective perforators and transfer (VLNT) symptoms A.M. Al-Sakkaf, X. Bonfill, S. Ardiles-Ruesjas et al. transvers rectus improvement, also abdominis good results with myocutaneous limb circumference procedures size and infectious episodes (continued on next page) Table 1 (continued) 7 Winter Netherlands A systematic review 17 studies (non- vascularized lymph Yes 1. Limb volume Evidence from all et al.41 and metanalysis of randomized node transfer (VLNT) 2. Quality of life studies suggest that vascularized lymph trials) 3. Infectious episodes VLNT reduce node transfer for 4. Complication volume up to 40%, breast cancer-related 5. Discontinuation of although in few lymphedema compression studies it has also garment some affect in improving quality of life, infectious episodes, the use of compression garment and complication rate 8 Gasteratos Greece Microsurgical 67 studies, but LVA, LYMPHA No 1. All patients’ Microsurgical et al.42 techniques in the only 13 studies technique and VLNT outcomes techniques can be treatment of breast addressed in the 2. Efficacy of beneficial to many cancer-related synthesis (6 microsurgical patients with BCRL lymphedema: a Prospective techniques systematic review of studies, 3 3. Complications efficacy and patient retrospective outcomes studies, 2 RCTs, 1 141 SRs and 1 narrative review) *SRs: Systematic reviews, RCTs: Randomized clinical trials, BCRL: Breast cancer-related lymphedema, LVA: Lymphatic venous anastomosis, VLNT: Vascularized lymph node transfer, LYMPHA: Lymphatic microsurgical preventive healing approach. Journal of Plastic, Reconstructive & Aesthetic Surgery 84 (2023) 134–146 A.M. Al-Sakkaf, X. Bonfill, S. Ardiles-Ruesjas et al. heterogenicity in the objectives and in the assessed surgical Risk-of-bias assessment for RCTs interventions of the SRs, we found that a total of 55 primary studies were included in all SRs, but only 19 studies over­ Based on the RoB-2, nine outcomes were assessed in the two lapped between two or more SRs. Saaristo et al.43 over­ included RCTs. lapped in five SRs: Chang et al.44 overlapped in four SRs; Lin Three outcomes were assessed in the RCT of Dionyssiou et al.18, Becker et al.17, Damstra et al.45, De Brucker et al.29 Two were rated as having a high risk of bias (limb et al.46, and Montag et al.47 overlapped in three SRs; and volume and subjective symptoms) and one had some con­ the rest overlapped in two SRs (see Table 2). cerns (infection episodes). Three RoB-2 domains, ‘the Table 2 The overall SRs on surgical treatments for BCRL: the overlaps matrix of their included studies and their overall risk of bias assessment. ¶ Gasteratos et al.,42; 67 studies were included in the study, but only 13 studies were addressed in the study synthesis. $ The colors of the systematic review reveal the risk of bias assessment, where: green is low risk of bias and red high risk of bias. *SRs: Systematic reviews, BCRL: Breast cancer-related lymphedema. The included SRs, the overlaps matrix and their risk of bias assessment. 142 Journal of Plastic, Reconstructive & Aesthetic Surgery 84 (2023) 134–146 Table 3 RoB-2 assessment (per outcome) of the randomized clinical trials on surgical treatments for BCRL. RCTs Experimental Comparator Outcome Weight D1 D2 D3 D4 D5 Overall Dionyssiou LNT + Physiotherapy Physiotherapy Limb volume 1 et al.29 Dionyssiou LNT + Physiotherapy Physiotherapy Infection episodes 1 et al.29 Dionyssiou LNT + Physiotherapy Physiotherapy Subjective 1 et al.29 symptoms Mulken Robot-assisted LVA Manual LVA Daily use of 1 et al.,30 compression garment Mulken Robot-assisted LVA Manual LVA Manual lymphatic 1 et al.,30 drainage Mulken Robot-assisted LVA Manual LVA Mean lymph - ICF 1 et al.,30 Mulken Robot-assisted LVA Manual LVA Mean UEL index 1 et al.,30 Mulken Robot-assisted LVA Manual LVA Duration of 1 et al.,30 surgery Mulken Robot-assisted LVA Manual LVA Quality of the 1 et al.,30 anastomosis ¶ Domains explanation D1: Randomization process, D2: Deviation from intended intervention, D3: Missing outcome data, D4: Measurement of the outcome, and D5: Selection of the reported results. $ The colors indication Green: Low risk, Yellow: Some concerns, Red: High risk of bias. * BCRL: Breast cancer-related lymphedema, LVA: Lymph-venous anastomosis, LNT: Lymph node transfer, Lymph ICF: Lymphedema Functioning, Disability and Health Questionnaire, and UEL Index: Upper extremity lymphedema index. RISK OF BIAS IN THE REVIEW High Low 4. Synthesis and findings Darker colours 3. Data collec�on and study appraisal indicate overall ROB ra�ng; 2. Iden�fica�on and selec�on of studies lighter colours concern 1. Study eligibility criteria judgments 0% 20% 40% 60% 80% 100% Figure 3 ROBIS risk-of-bias assessment of the systematic reviews (SRs) on surgical treatments for breast cancer-related lymphe­ dema (BCRL); overall and per-domain percentages. * Total number of the included SRs = 8. randomization process domain, deviation from intended Risk-of-bias assessment for SRs intervention domain, and measurement of the outcome domain’, had a probability of introduced bias in all RCT Based on the ROBIS tool, five SRs (5/8, 62.5%) were rated as outcomes and downgraded the rating to some concerns or high risk of bias37–40,42 and three (3/8, 37.5%) were rated as high risk of bias. low risk of bias.35,36,41 Six outcomes were assessed in the RCT of Van Mulken The five SRs rated as having a high risk of bias were et al.30 Five were rated as with some concerns (daily use of downgraded because there was a probability of introducing compressive garment, use of manual lymphatic drainage, bias in more than one domain. All rated as high risk of bias the mean lymph – ICF, duration of surgery, and quality of the in the ‘data collection and study appraisal’ domain (5/8, anastomosis) and one as high risk of bias (mean UEL index). 62.5%),37–40,42 three in the ‘study eligibility criteria’ domain Two RoB-2 domains, ‘the randomization process domain and (3/3, 37.5%),38,39,42 three in the ‘synthesis of finding’ do­ deviation from intended intervention domain’, were rated main (3/3, 37.5%),38–40 and two in the ‘identification and as having some concerns regarding the probability of in­ selection of studies’ domain (2/8, 25%).38,40 Finally, 50% of troduced bias in all RCT outcomes. In addition, “the missing all included SRs had rated the domain of ‘study eligibility outcome data” domain was rated as having a high risk of criteria’ as having a low risk of bias,35–37,41 which was the bias in the mean UEL index outcome, downgrading this best rated ROBIS domain (see Figure 3). outcome to a high risk of bias (see Table 3). 143 A.M. Al-Sakkaf, X. Bonfill, S. Ardiles-Ruesjas et al. BCRL but did not focus on the surgical treatment and did not include the risk-of-bias assessment.21 Furthermore, after updating our previous search in all the databases, our findings result in only two RCTs and eight SRs were among the 47 surgical studies that met the eligibility criteria. The overall risk-of-bias assessment of the two surgical RCTs29,30 was rated as some concerns (six outcomes) and high risk (three outcomes) of bias for the measured out­ comes among the included RCTs using the RoB-2 and high risk of bias (five studies) and low risk (three studies) for the eight included SRs35–42 using the ROBIS tool. In addition to the low-quality SRs and RCTs published in the surgical treatment for BCRL, there was a significant heterogeneity in the assessed intervention, the measured outcomes, and the included studies in the case of SRs. A study with a similar scope was an SR conducted by Chang et al.49 that addressed surgical treatment and pre­ vention for secondary lymphedema. In general, it showed that there was evidence to support some efficacy of LVA and VLNT, but their evidence was mainly based on observational studies and expert consensus. Other SRs that were involved in our results had positive findings on surgical interventions but were based mainly on case series,35,37 observational studies,36,39,42, and nonrandomized trials,41 and some did not mention which study design they included, probably not Figure 4 Bubble plot for the overall risk of bias of the sys­ including high-quality studies as we assumed.38,40 Gen­ tematic reviews and randomized clinical trials outcomes on erally, there is a lack of level 1 evidence to support the surgical treatments for BCRL. Studies descriptions: Systematic efficacy of the applied intervention. reviews (Blue circles): 1: Penha et al.35, 2: Cornelissen et al.36, Chang et al.49 assessed the risk of bias for their two in­ 3: Siotos et al.37, 4: Ribeiro et al.38, 5: Forte et al.39, 6: Forte cluded RCTs: one on surgical prevention50 and another on et al.40, 7: Winter et al.41, 8: Gasteratos et al.42. Randomized surgical treatment.29 The latter RCT was also assessed in our clinical trials (Orange circles): 1–3: Dionyssiou et al.29 (Out­ study, but in contrast, they used a different risk-of-bias comes; Limb Volume, Infection Episodes & Subjective symp­ tool; nevertheless, they reached a conclusion similar to toms, respectively), 4–9: Van Mulken et al.30 (Outcomes; Daily ours. They rated that RCT29 with a high risk of bias re­ use of compressive garment, Manual lymphatic drainage, Mean garding performance and detection biases, which is com­ lymph – ICF, Mean UEL Index, Duration of surgery & Quality of parable to our rating as high risk of bias in the deviation the anastomosis, respectively). * SRs: systematic reviews, RCTs: from the intended intervention and measurement of the randomized clinical trials, BCRL: breast cancer-related lym­ outcome in the RoB-2. Similar to our finding, Gasteratos phedema. et al.42 included in their results one similar RCT on surgical treatment,30 but they did not assess the risk of bias in their included studies. The overall risk-of-bias assessment for both RCTs and SRs There are more promising surgical treatments in the field in relation with the number of population or number of of BCRL, and demonstrating the effectiveness of these inter­ studies in the included RCTs and SRs, respectively, has been ventions has become more challenging. However, both pa­ demonstrated in a bubble plot figure for an overall visual tients and surgeons need high-quality information about presentation of the results (see Figure 4). treatment outcomes to inform decision-making.51 Surgeons now have to adopt more scientific methodologies and evi­ dence-based strategies to improve the standards of care for patients undergoing surgery.52 Based on our results, the Discussion overall evidence in the literature on surgical treatment for BCRL is low, and the limited number of well-designed RCTs in The main objective of this research was to describe and this field is an established barrier that needs to be addressed. critically assess the risk of bias of RCTs and SRs on the The constraints to conducting high-quality studies in surgical treatment for BCRL. To achieve this purpose, we surgery are attributed to the challenges related to the im­ conducted a systematic mapping review, which allowed a plementation of well-designed studies, the nature of the visual understanding of the evidence base of any treatment, interventions, and the lack of methodological experience apart from supporting the process of decision-making by among surgeons. Ergina et al.53 highlighted the difficulties facilitating information in a user-friendly format. in evaluating surgical innovations, especially in comparison Furthermore, it is the best study design to identify gaps of to pharmacological research, which usually contributes to knowledge in any research topic.48 uncertainty about the risk of biases and has led to skepti­ Our previous mapping review was conducted to provide cism about the value of surgical research. Yet, this is ap­ an overview of the current situation in the treatment for plicable by understanding the processes of evaluation in 144 Journal of Plastic, Reconstructive & Aesthetic Surgery 84 (2023) 134–146 surgery and creating alternative designs to maximize va­ Disclosure lidity and reduce the chance of bias.53 This study has some limitations. First, the search The first author Ali M. Al-Sakkaf is a PhD candidate at the strategy years were from 2000 onward because we assumed Surgery and Morphological Sciences program, Universitat that the evolution of the surgical treatments occurred in Autònoma de Barcelona, Spain. the previous two decades. Second, the mapping review usually requires additional expertize to create the visual output. Finally, in this kind of study, there is a probable risk of publication bias. Appendix A. Supporting information Among the strengths of this study, the systematic screening was performed by three independent reviewers to ensure the Supplementary data associated with this article can be reliability of the reported results. The methodological quality found in the online version at doi:10.1016/j.bjps.2023. assessment consideration was adequately done by defining the 05.002. eligibility criteria and identifying the risk of bias of the studies. The graphic presentation of the results was made to be rela­ tively easy to interpret and understand. Moreover, the findings References of this research illustrate the gaps in the literature and provide a clear picture of future needs in research in the field of surgery. 1. Ahmed RL, Prizment A, Lazovich D, Schmitz KH, Folsom AR. The shortage of strong evidence in the surgical treat­ Lymphedema, and quality of life in breast cancer survivors: the ment for BCRL makes the implications of this work in re­ Iowa Women’s Health Study. J Clin Oncol 2008;26:5689–96. search and practice significantly important and indicates 2. Granzow JW, Soderberg JM, Kaji AH, Dauphine C. Review of the need to conduct higher quality studies in this field, current surgical treatments for lymphedema. Ann Surg Oncol which can guide health policy and clinical decision-making. 2014;21(4):1195–201. 3. Gebruers N, Tjalma WAA. Clinical feasibility of axillary reverse mapping and its influence on breast cancer related lymphe­ Conclusion dema: a systematic review. Eur J Obstet Gynecol Reprod Biol 2016;200:117–22. The overall evidence in the literature on surgical treatment 4. McLaughlin SA, Wright MJ, Morris KT, et al. Prevalence of for BCRL is low because there are only two RCTs and eight lymphedema in women with breast cancer 5 years after sen­ SRs among the 47 published studies. The risk-of-bias as­ tinel lymph node biopsy or axillarydissection: objective mea­ surements. J Clin Oncol 2008;26(32):5213–9. sessment for the RCTs outcomes and most SRs were rated 5. Brahma B, Yamamoto T. Breast cancer treatment-related lymphe­ either as high risk of bias or with some concerns, and only dema (BCRL): an overview of the literature and updates in micro­ three SRs were rated as low risk of bias. surgery reconstructions. Eur J Surg Oncol 2019;45:1138–45. High-quality RCTs on different surgical interventions for 6. Jorgensen MG, Toyserkani NM, Sorensen JA. The effect of pro­ BCRL should be conducted to measure their real effective­ phylactic lymphovenous anastomosis and shunts for preventing ness, risks, and complications and to compare their benefit cancer-related lymphedema: a systematic review and meta- with other nonsurgical treatments. Moreover, better quality analysis. Microsurgery 2018;38(5):576–85. SRs on BCRL surgical treatments are needed to improve 7. Akita S, Nakamura R, Yamamoto N, et al. Early detection of evidence-based decision-making by surgeons and patients. lymphatic disorder and treatment for lymphedema following breast cancer. Plast Reconstr Surg 2016;138:192e–202e. 8. Pereira A, Koifman R, Bergmann A. Incidence and risk factors of lymphedema after breast cancer treatment: 10 years of follow- Conflicts of interest up. Breast 2017;36:67–73. 9. Gross JP, Whelan TJ, Parulekar WR, et al. Development and All remaining authors have declared no conflicts of interest. validation of a nomogram to predict lymphedema after axillary surgery and radiation therapy in women with breast cancer from the NCICCTG MA20 randomized trial. Int J Radiat Oncol Acknowledgments Biol Phys 2019;105:165–73. 10. Nguyen TT, Hoskin TL, Habermann EB, Cheville AL, Boughey JC. Breast cancer-related lymphedema risk is related to mul­ The authors thank Aliaa I. Shamardal and Luis Vasconcello- tidisciplinary treatment and not surgery alone: results from a Castillo for their participation in the baseline research large cohort study. Ann Surg Oncol 2017;24:2972–80. screening and data collection. 11. Kligman L, Wong RK, Johnston M, Laetsch NS. The treatment of lymphedema related to breast cancer: a systematic review and evidence summary. Support Care Cancer 2004;12:421–31. 12. Moseley AL, Carati CJ, Piller NB. A systematic review of common Ethical approval conservative therapies for arm lymphoedema secondary to breast cancer treatment. Ann Oncol 2007;18:639–46. Not required. 13. McNeely ML, Peddle CJ, Yurick JL, Dayes IS, Mackey JR. Conservative and dietary interventions for cancer-related lymphedema: a systematic review and meta-analysis. Cancer 2011;117:1136–48. Funding 14. Brorson H. From lymph to fat: liposuction as a treatment for complete reduction of lymphedema. Inter J Low Extr Wounds This mapping review did not receive any funding support. 2012;11(1):10–9. 145 A.M. Al-Sakkaf, X. Bonfill, S. Ardiles-Ruesjas et al. 15. Allen Jr RJ, Cheng MH. Lymphedema surgery: patient selection 36. Cornelissen A, Beugels J, Ewalds L, et al. Effect of lymphati­ and an overview of surgical techniques. J Surg Oncol covenous anastomosis in breast cancer-related lymphedema: a 2016;113(8):923–31. review of the literature. Lymphat Res Biol 2018;16(5): 16. Kung TA, Champaneria MC, Maki JH, Neligan P. Current con­ 426–34. cepts in the surgical management of lymphedema. Plast 37. Siotos C, Hassanein A, Bello R, et al. Delayed breast re­ Reconstr Surg 2017;139(4):1003e–1013ee. construction on patients with upper extremity lymphedema: a 17. Becker C, Assouad J, Riquet M, Hidden G. Postmastectomy systematic review of the literature and pooled analysis. Ann lymphedema: long-term results following microsurgical lymph Plast Surg 2018;81(6):730–5. node transplantation. Ann Surg 2006;243(3):313–4. 38. Ribeiro R, Santos-Júnior L, Barra I. Lymph node transplan­ 18. Lin CH, Ali R, Chen SC, et al. Vascularized groin lymph node tation in the management of post-mastectomy lymphedema: transfer using the wrist as a recipient site for management of a systematic review with meta-analysis. Braz J Plast Surg postmastectomy upper extremity lymphedema. Plast Reconstr 2019. Surg 2009;123(4):1265–75. 39. Forte A, Huayllani M, Boczar D, Ciudad P, Manrique O. 19. Winters H, Tielemans HJP, Hameeteman M, et al. The efficacy Lipoaspiration and lymph node transfer for treatment of breast of lymphaticovenular anastomosis in breast cancer-related cancer-related lymphedema: a systematic review. Cureus lymphedema. Breast Cancer Res Treat 2017;165(2):321–7. 2019;11(11):e6096. 20. Winters H, Tielemans H, Verhulst A, Paulus V, Slater N, Ulrich D. 40. Forte A, Cinotto G, Boczar D, et al. Lymph node transfer The long-term patency of lymphaticovenular anastomosis in breast combined with deep inferior epigastric perforators and trans­ cancer-related lymphedema. Ann Plast Surg 2019;82(2):196–200. verse rectus abdominis myocutaneous procedures: a sys­ 21. Al-Sakkaf A, Masia J, Auladell-Rispau A, et al. Evidence map­ tematic review. Gland Surg 2020;9(2):521–7. ping of the treatments for breast cancer-related lymphedema. 41. Winters H, Tielemans H, Paulus V, Hummelink S, Slater N, Plast Reconstr Surg Glob Open 2022;10(1):e4045. Ulrich D. A systematic review and meta-analysis of vascularized 22. Bragge P, Clavisi O, Turner T, Collie A, Gruen R. The global lymph node transfer for breast cancer-related lymphedema. J evidence mapping initiative: scoping research in broad topic Vasc Surg Venous Lymphat Disord 2021;10(3):786–95. areas. BMC Med Res Methodol 2011;11:92. 42. Gasteratos K, Morsi-Yeroyannis A, Vlachopoulos N, Spyropoulou 23. Tricco AC, Lillie E, Zarin W, et al. PRISMA extension for scoping GA, Del Corral G, Chaiyasate K. Microsurgical techniques in the reviews (PRISMA-ScR): checklist and explanation. Ann Intern treatment of breast cancer-related lymphedema: a systematic Med 2018;169:467–73. review of efficacy and patient outcomes. Breast Cancer 24. Melnyk BM, Fineout-Overholt E. Evidence-based practice in 2021;28(5):1002–15. nursing & healthcare: a guide to best practice. Philadelphia: 43. Saaristo AM, Niemi TS, Viitanen TP, Tervala TV, Hartiala P, Lippincott Williams & Wilkins; 2005. p. 29. Suominen EA. Microvascular breast reconstruction and lymph 25. Higgins JP, Savović J, Page MJ, Elbers RG, Sterne JA. Cochrane node transfer for postmastectomy lymphedema patients. Ann handbook for systematic reviews of interventions – chapter 8: Surg 2012;255:468–73. assessing risk of bias in randomized trial. 2019. 44. Chang DW. Lymphaticovenular bypass for lymphedema man­ 26. Whitinga P, Savovi J, Higginsa J-PT, et al. ROBIS: a new tool to agement in breast cancer patients: a prospective study. Plast assess risk of bias in systematic reviews was developed. J Clin ReconstrSurg 2010;126:752–8. Epidemiol 2016:225–34. 45. Damstra RJ, Voesten HG, van Schelven WD, van der Lei B. 27. Lefebvre C, Glanville J, Briscoe S, et al. Cochrane handbook Lymphatic venous anastomosis (LVA) for treatment of secondary for systematic reviews of interventions – chapter 4: searching arm lymphedema. A prospective study of 11 LVA procedures in for and selecting studies. 2019. 10 patients with breast cancer related lymphedema and a 28. Moher D, Shamseer L, Clarke M, et al. Preferred reporting critical review of the literature. Breast Cancer Res Treat items for systematic review and meta-analysis protocols 2009;113:199–206. (PRISMA-P) 2015 statement. Syst Rev 2015;4:1. 46. De Brucker B, Zeltzer A, Seidenstuecker K, et al. Breast cancer- 29. Dionyssiou D, Demiri E, Tsimponis A, et al. A randomized con­ related lymphedema: quality of life after lymph node transfer. trol study of treating secondary stage II breast cancer-related Plast Reconstr Surg 2016;137:1673–80. lymphoedema with free lymph node transfer. Breast Cancer 47. Montag E, Okada AY, Arruda EGP, et al. Influence of vascular­ Res Treat 2016;156(1):73–9. ized lymph node transfer (VLNT) flap positioning on the re­ 30. Van Mulken T, Schols R, Scharmga A, et al. First-in-human ro­ sponse to breast cancer-related lymphedema treatment. Rev botic supermicrosurgery using a dedicated microsurgical robot Col Bras Cir 2019;46(2019):2156. for treating breast cancer-related lymphedema: a randomized 48. Anaya M, Franco J, Ballesteros M, Sola I, Urrutia G, Bonfill X. pilot trial. Nat Commun 2020;11(1):757. Evidence mapping and quality assessment of systematic re­ 31. Klimek L, Bergmann KC, Biedermann T. al. Visual analogue views on therapeutic intervention for oral cancer. Cancer scales (VAS): measuring instruments for the documentation of Manag Res 2019;11:117–30. symptoms and therapy monitoring in cases of allergic rhinitis in 49. Chang D, Dayan J, Greene A. Surgical treatment of lymphe­ everyday health care. Allergo J Int 2017;26(1):16–24. dema: a systematic review and meta-analysis of controlled 32. Devoogdt N, Kampen MV, Geraerts I, Coremans T, Christiaens MR. trials. Results of a consensus conference. Plast Reconstr Surg Lymphoedema functioning, disability and health questionnaire 2021;147(4):975–93. (lymph-ICF): reliability and validity. Phys Ther 2011;91(6):944–57. 50. Boccardo F, Casabona F, Friedman D, et al. Surgical prevention 33. Chan WY, Niranjan N, Ramakrishnan V. Structured assessment of arm lymphedema after breast cancer treatment. Ann Surg of microsurgery skills in the clinical setting. J Plast Reconstr Oncol 2011;18(9):2500–5. Aesthet Surg 2010;63(8):1329–34. 51. Potter S, Mills N, Cawthorn S, Donovan J, Blazeby J. Time to be 34. Temple C, Ross D. A new, validated instrument to evaluate BRAVE: is educating surgeons the key to unlocking the potential competency in microsurgery: the University of Western Ontario of randomised clinical trials in surgery? A qualitative study. Microsurgical Skills Acquisition/Assessment instrument [out­ Trials 2014;15:80. comes article]. Plast Reconstr Surg 2011;127(1):215–22. 52. McLeod RS. Instilling a culture of evidence-based surgery in 35. Penha T, Ijsbrandy C, Hendrix N, et al. Microsurgical techni­ Canada. World J Surg 2007;31:1551–5. ques for the treatment of breast cancer-related lymphedema: 53. Ergina P, Cook J, Blazeby J, et al. Challenges in evaluating a systematic review. J Reconstr Microsurg 2013;29(2):99–106. surgical innovation. Lancet 2009;374:1097–104. 146