RESEARCH REPORT PLATFORM PRESENTATION

LYMPHEDEMA THERAPY IN BREAST CANCER PATIENTS WITH LYMPHEDEMA; A SYSTEMATIC REVIEW. Kärki A¹, Anttila H², Rautakorpi U²; ¹Satakunta University of Applied Sciences, Pori, Finalnd. ²National Research and Development Centre for Welfare and Health (STAKES)/Finnish Office for Health Technology Assessment (Finohta), Helsinki, Finland

PURPOSE: To review systematically the effects of lymphedema therapy. RELEVANCE: Breast cancer treatments may disturb the lymphatic drainage of the upper limb causing lymphedema. Lymphedema therapy methods are various, and national guidelines are needed to support the use of the effective methods. PARTICIPANTS: Operated breast cancer patients with upper limb lymphedema. METHODS: Medline, Cinahl, DARE, HTA database, PEDro and the Cochrane Library were searched for systematic reviews, guidelines or HTA reports. We updated the latest review by a literature search from January 2004 to May 2006 for controlled trials with or without randomization. ANALYSIS: Two authors selected the trials, three performed the final selection and two extracted the data. Three authors assessed the methodological quality for studies. Disagreements were solved by consensus. We used a list developed by van Tulder (2003), 11 equally weighted criteria (yes=1, no=0, don‘t know=0), resulting 0-11 points. From studies that had been assessed in the earlier high quality reviews (Pecino 2004, Kärki 2001) we adopted the quality assessments. We also aimed to pool the data of the trials. Synthesis: We were not able to pool any studies because of clinical heterogeneity and lacking data. We made a qualitative synthesis taking into account the methodological quality of the studies, number of studies, type of design (RCR, CT), type of outcome measure, statistical significance and consistency of the findings. We applied a rating with five levels of evidence by van Tulder (2000): 1. Strong evidence: generally consistent findings in multiple (≥2) high quality RCTs. 2. Moderate evidence: generally consistent findings in one high quality RCT and multiple (>2) low quality RCTs. 3. Limited evidence: one RCT or generally consistent findings in multiple RCTs or CTs. 4. Conflicting evidence: inconsistent findings in multiple RCTs or CTs. 5. No evidence: no RCTs found or just one CT. RESULTS: We identified 10 systematic reviews and 18 trials (12 randomized). One high quality RCT compared manual lymph drainage (MLD) with compression garments (CG) to CG alone, and reported similar reduction of lymphedema in both groups after 4 weeks, the greatest reduction in the first 2 weeks. Other studies were moderate to low quality, compared MLD vs. simplified MLD (3 studies), MLD vs. compression sleeve (CS) (2 studies), MLD vs. compression pump (3 studies), CS vs. no treatment (1 study), CG vs. CS (1 study), resisted training vs. no training (1 study), and 3 studies on devices (mechanical pressure device, lower-level laser and ultrasound compared to self massage and CS, placebo, and compression pump, respectively). CG or CS was used as an add-on treatment in 14 studies. These studies reported no differences between groups, but a 4-60% reduction of lymphedema was found in all groups. MLD or compression pump alone, lower-level laser or ultrasound had no effects on lymphedema. CONCLUSIONS: CG are effective in reducing lymphedema. MLD seems to have no additional effects when combined with CG (2=Moderate evidence). Evidence of the effectiveness of other lymphedema therapy methods was insufficient (5=No evidence). IMPLICATIONS: CG or CS should be introduced to daily use immediately in case of post-treatment lymphedema in breast cancer patients. KEYWORDS: breast cancer, lymphedema, lymph therapy, systematic review. FUNDING ACKNOWLEDGEMENTS: The research project was funded by National Research and Development Centre for Welfare and Health (STAKES)/Finnish Office for Health Technology Assessment (Finohta). CONTACT: anne.karki@samk.fi

ETHICS COMMITTEE: Ethical committee in National Research and Development Centre for Welfare and Health (STAKES)have approved our study in 2006.