This re-analysis of previously published data confirms that a strict therapy regime in a 3-week decongestion phase in patients with BCRL leads to a significant volume reduction of almost 10% of the initial volume.

The question, therefore, arises as to which component of CDT has led to the great decongestion effect here.

In this study, we supplemented the classical CDT with IPC, which was applied daily from the 3rd day onwards, also at weekends. However, since there was also an increase in volume, albeit small, at the weekends, the decongestant effect during the working week cannot be attributed to the IPC. This conclusion is also strengthened by the fact that the IPC was only applied from the third day onwards in the first week, i.e., there were only three IPC applications in the first working week instead of the usual five, but the volume decrease was greatest in the first working week.

Self-management and patient education as an essential component of CDT are aimed at the maintenance phase (phase II) and therefore excluded as a factor for the decongestion phase.

Compression therapy cannot be the decisive factor here, either. During the working week, all patients received the same conventional compression bandage; only at the weekend did some patients receive an alginate bandage instead of the conventional bandage. Thus, compression therapy cannot be considered a significant factor for the observed volume decrease during the working week.

The skin care, in our case with a cream containing polidocanol, is to reduce the risk of infections and skin problems but does not influence the volume and possible increases or decreases.

Exercise therapy (ET) certainly has an influence. It activates the muscle pump and thus promotes lymph flow. The improved lymph flow, in turn, can reduce the volume of the diseased limb. In our case, physiotherapy during the working week consisted of a uniform water gymnastics program in the morning and dry gymnastics in the afternoon. At the weekend, there was no structured ET, but the patients activated their muscle pump through their own measures (independent visits to the swimming pool, walks, etc., in each case with compression), thus stimulating the lymph flow.

This leaves MLD as the last component of CDT, which can be responsible for the different volume changes. With our patients, we assume that MLD has caused the observed major effects. However, this is contrasted by studies that deny precisely this decongestive effect of MLD.

In their meta-analysis, Huang et al. compared, among others, six clinical studies that investigated the decongestant effect of MLD compared to “standard therapy” [2]. The weighted mean volume difference was 75.12 mL (95% CI: − 9.43, 159.58) in favor of MLD application. The test for the overall effect yielded a Z of 1.74 with a probability of p = 0.08. The maximum probability of falsely rejecting the null hypothesis—here: both therapies have the same effect—is thus 8%, i.e., insignificant according to common parlance. Nevertheless, a clear trend in favor of MLD is recognizable.

Ezzo et al. used the same studies for their Cochrane review as Huang et al. [3]. In contrast to Huang et al., however, Ezzo et al. divided these six studies again after a more detailed analysis of the therapy regimes. In general, the results are always in favor of MLD; significant, for example, are the results for the comparison of MLD with compression bandaging versus compression bandaging alone (2 studies) for the outcomes of volume reduction in mL and percentage volume reduction as well as for the comparison of MLD with compression stocking versus IPC with compression stocking.

In their meta-analysis, Liang et al. [14] included finally eight RCTs, some of them also part of the previously mentioned meta-analyses [2, 3]. They summarized that MLD cannot significantly reduce lymphedema in patients after breast cancer surgery.

Lin et al. [15] analyzed ten RCTs and summarized that pain of BCRL patients undergoing MLD is significantly improved, while their findings did not support the use of MLD in improving volumetric of lymphedema and quality of life.

Xing et al. [16] analyzed and summarized seven systematic reviews and meta-analyses. They did not recommend the addition of MLD to CDT [sic!] or compression therapy for patients with BCRL based on the results of the present review. More well-designed and large RCTs are needed to provide a higher level of evidence to confirm the role of MLD in CDT.

A common conclusion of all these meta-analyses was that well-designed RCTs with a larger sample size are required. An overview of the studies included in these previously described meta-analyses is given in Table 4. A further study, which was hitherto not included in a meta-analysis, was made by De Vries and colleagues [8]. They compared BCRL patients with three different therapy concepts over 3 weeks with 14 therapy sessions. One group (n = 64) received CDT with placebo MLD, a second group (n = 63) received classical CDT including classical MLD, while the third group (n = 63) received CDT with fluoroscopy-guided MLD. Unfortunately, the authors did not give absolute volumes but limited themselves to percentages. The excess lymphedema volume (volume sick—volume healthy) decreased after 3 weeks of intensive treatment in each group: 5.3 percentage points of percentage excess volume (corresponding to a relative reduction of 23.3%) in the fluoroscopy-guided MLD group, 5.2% (relative reduction of 20.9%) in the traditional MLD group, and 5.4% (relative reduction of 24.8%) in the placebo MLD group. However, no clinically significant differences in volume reduction were found between the groups. This raises the question of what effect the placebo MLD had. In this group, deep massage was performed with relaxing transverse movements of the ipsilateral neck, back, shoulder, arm, and hand muscles.

Except for the study by de Vrieze et al., all other studies have small group sizes, so any trend favoring MLD is not (yet) statistically significant. De Vrieze et al. aimed to establish fluoroscopy-assisted MLD as a better alternative to classical MLD. To this end, however, they did not apply CDT without MLD in the actual control group but instead used a placebo treatment, which they assumed had no draining effect on the lymph vessels. However, the described deep massage influences the musculoskeletal system and can have similar effects to exercise therapy.

This study is a pooled re-analysis of an initially randomized controlled trial with a very special study design. This sequential design with three intervention phases of 5 days each, and interrupted and followed by three control phases of 2 days each (see Table 2), has the advantage of an identical intervention and control group, but the disadvantage that it might be difficult to compare with real randomized controlled trials. This re-analysis pooled the data of two groups, which differed in the treatment in the now control phases. By pooling these groups, the number of investigated patients increased and resulted in a significant volume reduction during the intervention phases. The number of patients exceeds the number of patients in the intervention groups of most of the clinical trials published up to now. As the initial study was done in 2008 [published 2010 as reference 9], not all original data were still available. For example, we could not find the absolute volumes of the healthy arms any more, but only the relative volume increase of the affected arms, expressed by the V-class of the LVF-classification by one of us [17].