Breast cancer–related postmastectomy lymphedema (BCRL) is the most common chronic condition following mastectomy, with an incidence ranging from 8% to 52% in the first 2 postoperative years.1 A study conducted by Armer et al2 reported that approximately 75% of lymphedema cases occur in the first year after breast cancer surgery. As reported in the literature, BCRL is defined as an increase in arm volume by at least 150 mL compared with preoperatively, an increase in arm circumference by at least 2 cm, or an increase in arm volume of the affected arm by at least 10% compared with the unaffected arm.3 Breast cancer–related postmastectomy lymphedema can lead to many serious consequences, such as pain with arm exercise necessitating movement restrictions, emotional disorders; increased the risk of infection in the affected arm; and in severe cases may progress to upper limb disability.1 Clearly, it can seriously impair a patient’s life. To avoid more serious and chronic irreversible sequelae, various interventions have been developed over the past few decades to treat BCRL, including complete decongestive therapy, intermittent pneumatic compression therapy, exercise, and education.2 Complete decongestive therapy is a widely used combination management for lymphedema encompassing several elements that is considered effective in treating BCRL.2 The components includes intensive lymphedema therapy, continued meticulous skin and nail care, range-of-motion exercises, manual lymphatic drainage (MLD), and application of compression garments and/or bandages.2,4 Manual lymphatic drainage is a key component of complete decongestive therapy; the specific massage technique instructions are summarized as follows: draw 5 diaphragmatic breaths and massage from under the ipsilateral ear auricle to the bottom of the neck, from inferior to the lateral aspect of the clavicle to its medial extent, from the operated side to the opposite arm’s armpit, and the operated side armpit itself.1
Manual lymphatic drainage has been shown to have multiple physiological effects including increasing the rate of lymphatic contraction, increasing protein reabsorption in the lymphatic vessels, reducing microlymphatic hypertension, and improving regional collateral skin lymphatic drainage.5,6 Despite the aforementioned apparent beneficial effects, the actual measured therapeutic effect of MLD on BCRL still remains unclear. Some studies have reported that MLD may be ineffective in the reduction of upper lymphedema of BCRL.7,8 In contrast, 2 systematic analyses in 20119 and 201610 reported that MLD successfully reduced upper limb volume. Hence, we conducted this meta-analysis to evaluate the efficacy of MLD on postmastectomy BCRL, including the reduction of upper limb lymphedema and improvement of arm symptoms and functions. If the intervention proved effective, which particular format of MLD was most effective?
Materials and Methods
PubMed, EMBASE, Web of science, and the Cochrane Library were systematically searched from the time of their establishment to January 2020 using the following search terms: “breast cancer,” “breast neoplasm,” “breast tumor,” “breast adenocarcinoma,” “breast sarcoma,” “breast carcinoma,” “mastectomy,” “axillary dissection,” “sentinel node dissection,” “lymph node excision,” “lymph node dissection,” “axillary node dissection,” “lymphadenectomy,” “manual lymphatic drainage” and “weight training,” “physiotherapy,” “rehabilitation,” “shoulder immobilization,” “stretching,” “psychosocial therapy,” “movement program,” “mobilization,” “physical therapy,” “exertion,” “musculoskeletal manipulations,” “strength training,” “physical activity” or “exercise.” The language was limited to English, and retrieval was based on the subject terms, key words, or title. References of studies with potential relevance and review articles were manually checked.
Inclusion criteria included the following: 1) randomized controlled trials (RCTs), including cross-over and cluster RCTs; 2) patients with BCRL (diagnostic criteria: an increase in arm volume by at least 150 mL compared with preoperatively, an increase in arm circumference by at least 2 cm, or an increase in arm volume of the affected arm by at least 10% compared with the unaffected arm); 3) patients ≥18 years old enrolled in studies; 4) the treatment of the experimental group was MLD; 5) presence of a control group managed with compression bandaging or other standard treatment; 6) outcome measures that included changes in upper limb volume or circumference; and 7) a minimum of 20 patients in the trial.
Publications not meeting the inclusion criteria or patients with serious complications (vital signs unstable) were excluded.
The primary efficacy outcome was lymphedema change of the operated-side upper limb from baseline to postintervention, as measured by the volume change or arm circumference change. We analyzed the data reported to determine the efficacy of MLD to treat BCRL. The symptoms and functions were evaluated by heaviness, pain, anxiety, mobility, and quality of life. We evaluated the variation of symptoms and functions by the mean change score from baseline to postintervention according to validated scales (eg, numeric rating scale, the EuroQol 5D questionnaire, visual analog scale).11
Data Extraction and Quality Assessment
Data extraction was performed by 2 authors, independently reviewing primary texts, supplementary appendices, and protocols, carefully evaluating study characteristics (publication year, first author, journal, and country of origin), patient characteristics (diagnostic criteria, age range, mean age, and sample size), intervention details (number of intervention sessions, duration and format of intervention, management method of control group), and outcome measures (as outlined above). The methodological quality of RCTs was assessed in accordance with the risk-of-bias tool described in the Cochrane Handbook for Systematic Reviews of Interventions (Cochrane Handbook 5.2).12 Random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessors, incomplete outcome data, selective reporting, and other potential risks of bias were assessed. Trials with high risk of bias of ≥1 key domain were considered to be at high risk of bias. Trials with low risk of bias for all key domains were judged as being at low risk of bias; otherwise, they were considered to be at an unclear risk of bias.
The principal emphasis of this meta-analysis was to analyze the efficacy of MLD for BCRL. The measurements were expressed as standardized mean difference (SMD) with 95% confidence intervals (CIs) for continuous outcomes, and significance was set at P < .05. We used funnel plots to detect possible publication bias. We conducted subgroup analyses to examine whether outcomes were influenced by the number of treatment sessions and duration of intervention. We also performed sensitivity analyses for primary efficacy outcomes (eg, excluding those studies that were low-quality). Analyses were performed using Review Manager 5.3 (Cochrane Collaboration, London, UK).
Study Selection and Study Characteristics
Our initial search identified 2091 studies from the 4 databases, of which 28 full-text articles were carefully screened by the authors. The PRISMA statement flowchart shows the process of literature screening, study selection, and reasons for exclusion (Figure 1). Only 8 high-quality articles met our stringent criteria to be included in our meta-analysis. Our final study group comprised 457 patients, 236 in the treatment group and 221 in the control group. All patients were female, the ages ranged from 25 to 84 years of age, and the treatment duration extended from 1 to 4 weeks. Two articles included the change of heaviness, 2 articles included the change of pain, and 2 articles included the change of mobility.
Comparing the MLD treatment group with the no-MLD treatment control group, there was no significant reduction in upper limb volume for BCRL (SMD, 0.43; 95% CI, −0.10 to 0.96; P = .11; I2 = 86%; Figure 2).
Improvement of Symptoms and Arm Function
Perceived heaviness in the arm was reduced by 18% in the study by Andersen et al,13 whereas Tambour et al7 reported the reduction of heaviness at the end of treatment in the experimental group had an average of 0.4 points, and the control group had an average of 1.5 points, P = .50; there was no significant reduction in heaviness according to Andersen et al13 and Tambour et al.7 Andersen et al13 reported that limitation of arm mobility was reduced by 22% in their treatment group. Arm mobility scores reduced to 0.5 and 0.6 in the treatment and control groups, respectively, P = .80, as reported by Tambour et al.7 There was no significant reduction in mobility in the 2 groups.7,13
Manual lymphatic drainage may not have a positive effect on reducing pain in the upper limbs. There was no significant reduction in pain according to Tambour et al7 and Bergmann et al14 (SMD, −0.09; 95% CI, −0.43 to 0.25; P = .61; Figure 3).
We conducted multiple subgroup analyses to examine the relative efficacy of MLD treatment compared with the no-MLD treatment group, for example, the treatment session (<20 or ≥20 sessions) and the duration of intervention (≤2 or >2 weeks). In the group with <20 MLD sessions,4,7,13–15 there was no significant difference between the MLD group and the no-MLD treatment group (SMD, 0.82; 95% CI, −0.17 to 1.82; P = .11; Figure 4A). In contrast, for patients in the group with ≥20 sessions,16–18 there was a significant difference compared with the no-MLD treatment group (SMD, 0.31; 95% CI, 0.03 to 0.58; P = .03; Figure 4A). In the ≤2 weeks’ duration of the MLD group,3,11,13 there was no significant difference between the MLD group and the no-MLD treatment group (SMD, 2.03; 95% CI, −0.39 to 4.44; P = .10; I2 = 96%; Figure 4B). In the >2 weeks’ duration of MLD group,7,14,16–18 there was a significant difference between the 2 groups (SMD, 0.23; 95% CI, −0.02 to 0.44; P = .03; I2 = 0%; Figure 4B).
We carried out sensitivity analyses for volume reduction of BCRL. We performed a sensitivity analysis by excluding those studies that were of low quality.4 The effect size was much lower with low heterogeneity (SMD, 0.16; 95% CI, −0.06 to 0.39; P = .15; I2 = 26%; Figure 5).
Risk of Bias for Independent Studies
From the risk-of-bias graph, the overall quality of the literature was found to be below average (Figure 6). Five of the studies described the methods of random sequence generation (selection bias) in detail and were classified as low risk. There were studies that did not describe the methods of random sequence generation (selection bias). The funnel plot would indicate that there may be publication bias (Figure 7), but the results must be viewed with caution because of the small number of publications.
Our final analysis determined that there is no statistical benefit of MLD in reducing the upper limb volume of BCRL compared with the no-MLD treatment control group managed with standard care (SMD, 0.43; P = .11). Our conclusion does not agree with some previous studies. For example, McNeely et al17 found a statistically significant benefit from the addition of MLD massage for reducing upper extremity lymphedema volume. Similarly, Shao and Zhong10 found a statistical benefit in volume reduction between adding MLD to routine treatment and routine treatment only. Despite our exhaustive literature review, we were limited to only 8 reports that met our study criteria, and the resultant limited number of patients for analysis may be one of the reasons for the lack of demonstrating treatment benefit.
In addition, the difference of inclusion and exclusion criteria may also be the reason for the different conclusions of 2 previous systematic reviews and meta-analyses.10,17 First, McNeely et al17 included all cancer-related edema and all intervention methods; we included only BCRL and MLD. Second, there were 7 studies on MLD treatment of lymphedema after breast cancer–related mastectomy. Among them, 3 studies19–21 that did not meet the inclusion criteria were excluded in our meta-analysis. Turning to Shao and Zhong,10 their study design was quite different from ours as they compared the efficacy of standard treatment alone with standard treatment combined with MLD for BCRL. Shao and Zhong10 limited their statistical analysis to only 3 studies,13,17,18 which may further account for the differences between theirs and our conclusions.
Subgroup analysis demonstrated statistical reduction of upper limb volume in BCRL patients who underwent ≥20 MLD treatment sessions or a treatment duration >2 weeks. It appears that the clinical use of MLD to treat BCRL requires multiple treatment sessions or a prolonged duration to achieve the therapeutic effect. However, because of our strict study criteria, the sample size was limited, and additional high-quality randomized trials are needed.
Recently, some studies have attested to the efficacy of adding MLD to bandage treatment15,17 and MLD to standard therapy.7,13,14 Perhaps comprehensive treatment may be more effective than MLD alone, but it is more complicated and time-consuming. Therefore, we intend to explore additional methods to enhance the effectiveness of MLD and further investigate whether there are other benefits of MLD to the heaviness, pain, and mobility of the upper extremity. In conducting this meta-analysis, some of the included studies suggested that MLD may not reduce the upper limb tightness, mobility, and arm pain compared with no MLD. However, the data of tightness and mobility were insufficient for secondary analysis. The question of whether MLD truly and demonstrably improves symptoms and function for the treatment of postmastectomy BCRL will require more randomized trials to be finally resolved.
This report has certain limitations. First, the publications included in this meta-analysis are limited in number, are relatively low-quality, and have apparent publication bias (Figure 6, Figure 7); additional high-quality RCTs are needed. Second, we used upper limb volume to evaluate upper limb lymphedema only; unfortunately, insufficient data were available to conduct secondary analysis of improvement in upper limb symptoms and function. Again, larger comprehensive clinical studies will be required to resolve these issues. Third, upper extremity volume of BCRL can increase or decrease over time, so variations in the time points at which the upper limb volume was measured could have influenced the results. Finally, even a precise definition of BCRL varies among published reports (Table); a consensus standard for this complication needs to be established to facilitate future comparative research.
Characteristics of Included Trials
|Intervention Program (Experimental Group; Control Group)
||Definition of BCRL
||Initial Arm Volume, Mean SD), mL (Experimental Group; Control Group)
||Age Range, Mean (SD), y
||BMI Range, Mean (SD), kg/m2
||Treatment Duration, wk
||No. of Sessions
||No. of Chemotherapy Sessions
||No. of Radiotherapy Sessions
|Johansson et al, 1999
||CB + MLD (n = 20); CB (n = 18)
||NA (2841); NA (2882)
|Andersen et al, 2000
||ST + MLD (n = 23); ST (n = 21)
||340 (1323); 361 (1296)
|Sitzia et al, 2002
||MLD (n = 15); SLD (n = 13)
||NA (3395); NA (3473)
|Williams et al, 2002
||MLD (n = 29); SLD (n = 31)
||NA 57.9; 59.3
|McNeely et al, 2004
||CB + MLD (n = 25); CB (n = 25)
||D-volume >150 mL
||535 (NA); 630 (NA)
|Dayes et al, 2013
||CB + MLD (n = 57); CB (n = 46)
||3422 (838); 3266 (781)
||61; 59 NA
||(32; 30) NA
|Bergmann et al, 2014
||ST + MLD (n = 28); ST (n = 29)
||D-circumference >3 cm
||62.16; 63.55 (9.06;10.98)
||30.44; 29.08 5.14;5.97)
|Tambour et al, 2018
||ST + MLD (n = 39); ST (n = 38)
||D-circumference >3 cm
||2942.9 (593.9); 2889.7(491.3)
||62.0; 60.9 (11.5;10.8)
||29.6; 29.4 (5.5;5.3)
Abbreviations: BCRL, breast cancer–related postmastectomy lymphedema; CB, compression bandaging; D-circumference, circumference difference; D-volume, volume difference; LR, the number of patients lost to follow-up; MLD, manual lymphatic drainage; PCEV, percentage change in excess limb volume; SLD, simple lymphatic drainage; ST, standard treatment.
In summary, when the number of MLD treatment sessions is ≥20 or the duration is >2 weeks, there is an effective reduction of upper limb volume; therefore, it is necessary to have sufficient number of sessions and duration of interventions for the treatment of BCRL in clinical. But the current evidence is not sufficient to evaluate the most effective format of clinical intervention in BCRL patients. More high-quality randomized clinical trials are needed to analyze the efficacy of MLD.
1. Gol MK, Aghamohamadi D. Effect of massage therapy with and without elastic bandaging on pain, edema, and shoulder dysfunction after modified radical mastectomy: a clinical trial. Int J Womens Health Reprod Sci
2. Armer JM, Ostby PL, Ginex PK, et al. ONS Guidelines (TM) for cancer-related lymphedema. Oncol Nurs Forum
3. Devoogdt N, Geraerts I, Van Kampen M, et al. Manual lymph drainage may not have a preventive effect on the development of breast cancer
–related lymphoedema in the long term: a randomised trial. J Physiother
4. Sitzia J, Sobrido L, Harlow W. Manual lymphatic drainage
compared with simple lymphatic drainage in the treatment of post-mastectomy lymphoedema. Physiotherapy
5. Leduc O, Bourgeois A, Peeters A, et al. Bandages: scintigraphic demonstration of its efficacy on colloidal protein reabsorption during muscle activity. Lymphology
6. Moseley AL, Carati CJ, Piller NB. A systematic review of common conservative therapies for arm lymphoedema secondary to breast cancer
treatment. Ann Oncol
7. Tambour M, Holt M, Speyer A, et al. Manual lymphatic drainage
adds no further volume reduction to complete decongestive therapy on breast cancer
–related lymphoedema: a multicentre, randomised, single-blind trial. Br J Cancer
8. Devoogdt N, Christiaens MR, Van Kampen M. Efficacy of manual lymphatic drainage
in preventing secondary lymphedema after breast cancer
9. McNeely ML, Peddle CJ, Yurick JL, et al. Conservative and dietary interventions for cancer-related lymphedema: a systematic review and meta-analysis
10. Shao Y, Zhong DS. Manual lymphatic drainage
for breast cancer
–related lymphoedema. Eur J Cancer Care (Engl)
11. Sorensen J, Davidsen M, Gudex C, et al. Danish EQ-5D population norms. Scand J Public Health
12. Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0
. Oxford: The Cochrane Collaboration; 2019. https://www.cochrane-handbook.org
. Accessed July 16, 2020.
13. Andersen L, Hojris I, Erlandsen M, Andersen J. Treatment of breast-cancer-related lymphedema with or without manual lymphatic drainage
: a randomized study. Acta Oncol
14. Bergmann A, da Costa Leite Ferreira MG, de Aguiar SS, et al. Physiotherapy in upper limb lymphedema after breast cancer
treatment: a randomized study. Lymphology
15. Johansson K, Albertsson M, Ingvar C, et al. Effects of compression bandaging with or without manual lymph drainage treatment in patients with postoperative arm lymphedema. Lymphology
16. Williams AF, Vadgama A, Franks PJ. A randomized controlled crossover study of manual lymphatic drainage
therapy in women with breast cancer
–related lymphoedema. Eur J Cancer Care (Engl)
17. McNeely ML, Magee DJ, Lees AW, et al. The addition of manual lymph drainage to compression therapy for breast cancer
related lymphedema: a randomized controlled trial
. Breast Cancer Res Treat
18. Dayes IS, Whelan TJ, Julian JA. Randomized trial of decongestive lymphatic therapy for the treatment of lymphedema in women with breast cancer
. J Clin Oncol
19. Didem K, Ufuk YS, Serdar S, et al. The comparison of two different physiotherapy methods in treatment of lymphedema after breast surgery. Breast Cancer Res Treat
20. Jahr S, Schoppe B, Reisshauer A. Effect of treatment with low-intensity and extremely low-frequency electrostatic fields (deep oscillation) on breast tissue and pain in patients with secondary breast lymphoedema. J Rehabil Med
21. Radakovk N, Popovic-Petrovic S, Vranjes N, et al. A comparative pilot study of the treatment of arm lymphedema by manual drainage and sequential external pneumatic compression (SEPC) after mastectomy. Arch Oncol