Background
Lymphedema is defined as persistent tissue swelling caused by the blockage or absence of lymph drainage [
1]. Lymphedema is a major concern for patients undergoing axillary lymph-node dissection for the treatment of breast cancer. The incidence of lymphedema at 12 months after breast surgery ranges from 12% to 26% [
2,
3]. Lymphedema may result in cosmetic deformity, loss of function, physical discomfort, recurrent episodes of erysipelas ,and psychological distress [
4,
5]. Thus, an effective treatment for lymphedema is necessary.
Previous surgical techniques for the treatment of lymphedema aimed to reduce limb volume using a debulking resection approach. With the advent of microsurgery, use of multiple lymphatic-venous anastomoses has become the most common surgical treatment [
6]. However, convincing evidence of the success of lymphatic-venous anastomoses has not been demonstrated. Thus, most patients with lymphedema choose non-surgical treatments, such as the use of elastic stockings, especially in early stages of lymphedema [
7].
Complex decongestive physiotherapy (CDP) is likely to reduce upper limb lymphedema in patients with breast cancer. Evidence of the efficacy of other physiotherapy methods is limited [
8‐
10]. Compression bandaging, manual lymphatic drainage (MLD), physical exercise to maintain lymphatic flow, and skin care are combined in CDP [
11,
12]. In MLD, specialized rhythmic pumping techniques are used to massage the affected area and enhance the lymph flow. Gentle skin massage is thought to cause superficial lymphatic contraction, thereby increasing lymph drainage [
13].Vodder originally suggested the use of range-of-motion exercises to relieve various types of chronic edema, such as sinus congestion and catarrh [
14], and the use of MLD has become a common treatment for lymphedema worldwide, especially in European hospitals and clinics.
To date, several studies have been published investigating the effects of MLD in preventing and treating lymphedema after breast-cancer surgery [
15‐
18]. However, these studies have been inconclusive, probably because of small sample sizes. Therefore, we conducted a systematic literature review and meta-analysis of randomized controlled trials (RCTs) to evaluate the effectiveness of MLD in the prevention and treatment of breast-cancer-related lymphedema.
Methods
Selection criteria
We reviewed RCTs or quasi-RCTs from the literature that evaluated the outcome of MLD in preventing and treating breast-cancer-related lymphedema. For inclusion in our study, the trials were required to describe: 1) the inclusion and exclusion criteria used for patient selection, 2) the MLD technique used, 3) the compression strategy used, 4) the definition of lymphedema, and 5) the evaluation of lymphedema severity. We excluded trials that met as least one of the following criteria: 1) patients had not received axillary lymph-node dissection (such as in studies in which only sentinel node sampling was used), 2) the clinical outcomes had not been clearly stated, or 3) duplicate reporting of patient cohorts had occurred.
Search strategy and study selection
Studies were identified by keyword searches of the following electronic databases: PubMed, EMBASE, CINAHL, PEDro (Physiotherapy Evidence Database), SCOPUS, Cochrane Central Register of Controlled Trials, and the ClinicalTrials.gov registry (
http://clinicaltrials.gov/). The following terms and Boolean operator were used in MeSH and free-text searches: ‘manual lymph drainage’, ‘breast cancer OR neoplasm’, ‘lymphoedema OR lymphedema’. The ‘related articles’ facility in PubMed was used to broaden the search. No language restrictions were applied. The final search was performed in December 2012. We attempted to identify additional studies by searching the reference sections of any relevant papers and contacting known experts in the field.
Data extraction and methodological quality appraisal
Two authors (K-WT and T-WH) independently extracted details of the RCTs pertaining to the participants, inclusion and exclusion criteria, manual lymph-drainage techniques used, arm lymphedema parameters, and complications. The individually recorded decisions of the two reviewers were compared, and any disagreements were resolved based on the evaluation of a third reviewer (S-HT).
The two authors independently appraised the methodological quality of each study based on: 1) adequacy of the randomization, 2) allocation concealment, 3) blinding, 4) duration of follow-up, 5) number of drop-outs, and 6) performance of an intention-to-treat (ITT) analysis.
Outcomes assessments
The efficacy of MLD was evaluated by the incidence of lymphedema and the reduction in the volume of the patient’s arm at 1, 3, 6, 9 and 12 months after MLD treatment. The arm volume was assessed by submerging each arm in a container filled with water, and measuring the volume (ml) displaced [
19]. The absolute edema volume was defined as the difference in volume between the arm with lymphedema and the contralateral arm [
18]. The following various definitions for lymphedema were used in the studies analyzed: a difference in volume of greater than 10% between the affected arm and contralateral arm [
17,
18,
20]; an increase of 200 ml or more in the volume of the affected arm compared with the pre-surgery volume of the same arm [
16]; and an increase of 20 mm or more in the circumference of the affected arm compared with the pre-surgery circumference of the same arm [
16,
21].
Statistical analysis
Statistical analysis was conducted using Review Manager software (version 5.1;Cochrane Collaboration, Oxford, UK). The meta-analysis was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines [
22]. When necessary, standard deviations (SDs) were estimated based on the reported confidence interval (CI) limits, standard error, or range values [
23]. The effect sizes of dichotomous outcomes were calculated as risk ratios (RR), and the mean difference was calculated for continuous outcomes. The precision of an effect size was calculated as the 95% CI. A pooled estimate of the RR was calculated using the DerSimonian and Laird random-effects model [
24]. This provided relatively wide CIs and an appropriate estimate of the average treatment effect for trials that were statistically heterogeneous, resulting in a conservative statistical claim. The data were pooled only for studies that exhibited adequate clinical and methodological similarity. Statistical heterogeneity was assessed using the
I2 test, with
I2 quantifying the proportion of the total outcome variability that was attributable to variability among the studies.
Discussion
A physical treatment program combining MLD, skin care, exercise, compression bandaging, and sleeve or stocking compression is recognized as providing optimal lymphedema management [
29]. Three systematic reviews concluded that combined physical therapy provides effective treatment for lymphedema [
30‐
32]. However, the effectiveness of the individual components of such programs has not been clearly established. The relatively high cost of MLD compared with compression bandaging warrants assessment of the efficacy of these individual components. The results of our systematic review and meta-analysis did not show a significant benefit for MLD in reducing lymphedema volume. Although individual studies reported advantages associated with MLD, methodological inconsistencies between the studies confounded our attempts to conduct an overall comparison of the effects of MLD across the studies.
The published reports of the effectiveness of MLD are conflicting. One prospective study of 682 individual cases in a single lymphology unit evaluated various treatments for lymphedema. The results indicated that the risk of failure for lymphedema therapy after intensive decongestive physiotherapy was primarily associated with younger age, higher weight, and higher body mass index. By contrast, elastic sleeve and multilayer bandaging treatments were associated with a reduced risk of treatment failure, whereas the use of MLD as an adjunct to those therapeutic components was not [
33]. One retrospective study of 208 patients with lymphedema receiving palliative care showed clinical improvement in the intensity of pain and dyspnea in most patients after MLD treatment [
34]. The advantage of the RCT design is that allocation bias is minimized, resulting in a balance between the known and unknown confounding variables in the assignment of treatments. Systematic review and meta-analysis of the clinical outcomes of therapy, as reported in the summaries of the RCT results to date, may help identify the effects that are common to these trials. Such research more clearly distinguishes the effects of MLD in preventing and managing lymphedema.
Our meta-analysis examined the results of six studies that assessed the effects of MLD in patients with post-mastectomy lymphedema, compared with compression therapy [
15,
17,
18,
20,
25,
27]. Compression bandaging has been shown to be effective in managing lymphedema. Badger
et al. conducted an RCT to compare compression bandaging for 18 days followed by a compression garment (treatment group) versus the compression garment only (comparison group). These authors reported a significantly greater reduction in limb volume at 24 weeks in the treatment group compared with the comparison group [
4]. The studies that we reviewed had investigated several types of compression therapy. McNeely
et al. found that the figure-of-eight method was more effective in maintaining the correct bandage position, and was also more comfortable for the patient, compared with the spiral-bandaging method [
25]. McNeely
et al. replaced the bandages 5 times/week over the 4-week treatment period, whereas Johansson
et al. replaced the compression bandage every 2 days over a 3-week period [
18].
Sequential intermittent pneumatic compression is another nonsurgical treatment for lymphedema [
35]. Szolnoky
et al. investigated whether a combination of SPC treatments and MLD improved the outcome of CPD treatment for women with secondary lymphedema [
28]. Thus, in the studies we investigated, there was a high level of heterogeneity regarding the variables measured to represent the reduction in lymphedema volume.
We included two studies in our analysis that compared MLD with SLD in the treatment of breast-cancer-related lymphedema [
20,
27]. Although MLD and SLD involve the same principles, SLD is a less complex technique that uses simplified hand movements in a set sequence. SLD can also be applied by the patient or a caregiver without requiring specialized training [
27]. The results of both studies showed that MLD significantly reduced excess limb volume compared with SLD.
Of the ten RCT studies that we reviewed in our meta-analysis, only two investigated the effects of MLD for preventing lymphedema after breast-cancer surgery [
16,
21]. Devoogdt
et al. evaluated the effect of MLD used in combination with exercise therapy and instructional guidelines for lymphedema prevention in 160 patients with breast cancer and unilateral axillary lymph-node dissection, who were stratified by body mass index and axillary irradiation [
16]. Patients received exercise therapy plus MLD or exercise therapy only for 6 months; the results showed no significant difference in the prevention of lymphedema between the two groups [
16]. By contrast, Torres Lacomba
et al. used MLD, scar-tissue massage, and progressive active and action-assisted shoulder exercises postoperatively in patients who had undergone breast-cancer surgery, whereas their control group received only instructional guidelines for lymphedema prevention Torres Lacomba
et al. found a significant difference in secondary lymphedema between the groups at 1 year post-surgery [
21]. However, the individual contribution of MLD to the prevention of secondary lymphedema was unclear.
Variability in clinical factors and non-uniform reporting of clinical parameters contributed to the heterogeneity between the studies that we reviewed. First, the technique, duration, and frequency of MLD differed across the studies, and one study did not report the technical details of their MLD method [
15]. Second, the experience of the physiotherapist and the characteristics of the individual patient can affect clinical outcomes. For example, patients in the study by Sitzia
et al. were older than those in the other trials that we reviewed [
27]. Third, the compression and exercise strategies also differed greatly between the studies that we reviewed (Table
1); for example, the control group in the study by Torres Lacomba
et al. received only educational instructions [
21]. Fourth, the methods used for evaluating the reduction in arm volume were also different between the studies, rendering our assessment vulnerable to measurement bias.
The strengths of our review include our comprehensive search for relevant studies, the systematic and explicit application of eligibility criteria, the careful consideration of study quality, and our rigorous analytical approach. However, our review was limited by the methodological quality of the original studies (Table
2). First, several trials were small, and one study recruited only 12 patients in each treatment group [
17], diminishing the statistical power of their analysis. Second, only half of the studies included in our analysis reported adequate randomization in the study-group allocation [
16,
21,
25‐
27]. Third, in seven studies, the assessment staff were not blinded to the outcomes [
15,
17,
18,
20,
26‐
28]. Furthermore, most of the investigators analyzed their data according to the per-protocol principle, which may have biased their evaluations of the effect of MLD.
An ongoing study of 58 patients with post-mastectomy lymphedema is evaluating the effectiveness of MLD as an adjunct to standard treatment for reducing the volume of the affected arm and the consequent effects on patient quality of life and physical limitations [ClinicalTrials.gov identifier NCT01152099] [
36]. We await the results to determine whether this will provide more evidence for clinical practice.
Competing interests
The authors have no conflicts of interest or financial ties to disclose.
Authors’ contributions
K-WT and T-WH devised the study. K-WT, T-WH, and S-HT extracted the data. K-WT, T-WH, C-CL, and C-HB analyzed and interpreted the data. K-WT and T-WH wrote the first draft. All authors contributed to subsequent versions, and approved the final article. K-WT is the corresponding author.