Progressive Muscle Relaxation and Guided Imagery in Breast Cancer: A Systematic Review and Meta-analysis of Randomised Controlled Trials

Description of the condition

For a patient with breast cancer, the treatment journey is psychologically demanding.^[1] The uncertainty of survival, a sense of worthlessness to near and dear ones and treatment-related hardships compromise the mental well-being. The change of body image following the surgery and adverse effects of chemotherapy further add to the problem. Multiple hospital visits for external beam radiotherapy or isolation following brachytherapy worsens the situation.^[1-3] The survivors need long-term follow-up and complex assessments, which affect the quality of life.

They need coping strategies and a positive outlook on the scenario. Good family support and empathetic approach of physicians, nurses and staff improve the coping response of the patients to the disease. Integrative therapies such as yoga, meditation and mental relaxation are increasingly gaining importance. The exact mechanism by which they work is largely unknown. It is a general belief that these techniques help by reducing stress, restoring sympathetic and parasympathetic balance and restoring the hormonal balance. Progressive muscle relaxation (PMR) and guided imagery (GI) are two different techniques that are often advised in combination with the hope of an overall better outcome.

Description of the intervention

Why it is important to review the combined use of PMR+GI

PMR+GI is in practice for a long and have been studied independently in various condition.^[8] The combined practice of both interventions is an old practice. Their effects complement each other and give promising results in multiple clinical scenarios. Recently, a review of eight randomised control trials has been published in cancer patients, which reported a positive impact on the mental state as well as on chemotherapy-induced nausea and vomiting. However, it was not specific for breast cancer and also no attempt was made to perform the meta-analysis.^[9]

As breast cancer is the most common cancer of women worldwide, we need to understand the role of the combined practice of PMR+GI in its management.^[10] The disease is mainly limited to the female population, and therefore, the results can be specific. The meta-analysis will help in estimating the impact on different aspects of mental health and in understanding its limitations. This will give information on the present status of this intervention in breast cancer management and will also help in defining the scope of further research.

Objectives

We wanted to study the effect of the combined practice of PMR+GI on stress, anxiety, depression and mood of the patients during breast cancer management. Our secondary objective was to evaluate its effect on common chemotherapy-related complications and quality of life.

Criteria for selecting studies

We included in our study all randomised controlled trials, in which the combined practice of PMR+GI was studied in patients with breast cancer. The articles in which the study population included cases other than breast cancer were excluded from the study. We also excluded studies in which the outcome measure did not have psychological effects or chemotherapy-related adverse effects. The study protocol was registered with PROSPERO with registration number-CRD4202019622.

Search strategy

We searched the electronic databases of PubMed/Medline, CENTRAL and CINAHL (EBSCO) from April to May 2021. We also searched the ClinicalTrials.gov registry. The articles were searched irrespective of language and year of publication. We searched PubMed using the search strategy: (“Autogenic Training” [Mesh] OR “Imagery, Psychotherapy” [Mesh]) AND “Neoplasms” [Mesh]. A similar search strategy was used for searching articles in the CENTRAL database. We searched CINAHL with the search strategy “Progressive muscle relaxation OR Guided Imagery AND neoplasm” without selecting fields. Terms “Progressive muscle relaxation” and “Guided Imagery” were separately searched in the ClinicalTrials.gov registry for the completed projects. The “related articles” function of PubMed was used to broaden the search. We also searched the reference list of selected articles and some of the reviews manually. The relevant references were further searched in Google Scholar also. The whole process was conducted in accordance with the protocol of the preferred reporting item for systematic reviews and meta-analysis.

Data extraction

A pre-designed pro forma was used to extract the data from eligible studies. The authors’ MS and AB extracted the data and discussed the discrepancies. We collected information about the place and year of study, participants characteristics (number of breast cancer patients), nature of the intervention, measurement tools, the outcome and a summary of the results.

Methodological quality assessment

Authors MS and AB assessed the risk of bias independently, and any disagreement was resolved by the third assessor MG. The bias was assessed for random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting and other sources of bias. According to criteria given in the Cochrane handbook, the studies were graded as having low risk of bias, high risk of bias or unclear risk of bias. Risk of bias summary was prepared.

Data synthesis and statistical analysis

We extracted the data on mean and standard deviation from the studies. The extracted data were continuous and belonged to different scales. The summary statistics were prepared using a standardised mean difference (SMD) approach. They were entered into the Review Manager 5.4, and forest plots were constructed. We used the fixed effect model for the construction of forest plots if the value of I² was less than 50%, and for others, the random effect model was used. The heterogeneity was assessed using I² and t² statistics. I2 values of 25%, 50% and 75% were defined as low, moderate and high heterogeneity, respectively. P < 0.05 was considered statistically significant.

Measurement of treatment effect

For repeated measures value, we took the last recorded value for calculation. For studies with more than 2 intervention groups, we included the data of the arms with PMR+GI and controls only. We carried out analysis on intention to treat basis and included all the participants randomised to each group for analysis. We divided the studies into radiotherapy and chemotherapy subgroups based on the treatment stage. This was done for subgroup analysis.

Description of studies

We identified a total of 935 records [Figure 1]. We removed the duplicates and excluded the studies that were not related to cancer or were not randomised controlled trials. After this exclusion, we were left with 49 full-text articles. Out of these, in 35 studies, the combined practice of PMR+GI was not performed, and in two studies (Eremin et al., 2009; Walker et al., 1999), the outcome measures were different.^[11,12] In four studies (De Paolis et al., 2019, León-Pizarro et al., 2007, Charalambous et al., 2015, and Shahriari et al., 2017), the population was heterogeneous and included cases other than breast cancer.^[13-16] In one study (Elias et al., 2015), the intervention had spirituality and PMR+GI.^[17]

Close

Figure 1:

The study flow diagram.

Export to PPT

We excluded all these studies from further analysis and were left with seven studies for qualitative synthesis. We failed to retrieve data from two studies (Molassiotis, 2000; Molassiotis et al., 2002), and finally, five studies were selected for quantitative synthesis.^[18,19] All included studies were in English literature.

Characteristics of included studies [Table 1]

The included studies were from the United Kingdom, Taiwan, South Korea and Brazil. Three studies were conducted on patients receiving chemotherapy and two on patients receiving radiotherapy.^{[20,21,22,23,24]}

Training and practice of PMR+GI were not exactly the same in all the studies. In some of the studies, a single session training was given at the hospital, and then, instructional CDs were given for practice at home (Bridge et al., 1988; Chen et al., 2015). In the other three studies, either multiple sessions of training were organised in the beginning or they were evenly spread over a long duration. Instructional CDs were given to patients for practice at home in all studies.

In one study (Nunes et al., 2007), patients were advised to practice the PMR+GI twice daily at home. In others, they were recommended for the same once daily. Walker, 1999, used patients’ diary for checking the compliance to instructions. Methods for checking the same were not elaborated in other studies.

A variety of scales was used in different studies to assess pain, mood, stress, anxiety, depression, nausea and vomiting and quality of life. The day on which these scales were recorded was also different.

Assessment of risk bias

Assessment of the risk bias in included studies was done as per the guidelines given in the Cochrane handbook of systematic reviews and meta-analysis.^[25]

Allocation

Random sequence generation was used in all studies. Allocation concealment was described in only one study (Walker et al., 1999).

Blinding

The intervention was non-pharmacological in nature, therefore, blinding of the participants was difficult. However, this was not described in two studies (Yoo et al., 2005; Nunes et al., 2007), and we categorised them as having an unclear risk of bias. Blinding of the outcome assessment was mentioned in the study of Bridge et al., 1988.

Incomplete outcome data

We could not find data on attrition in two studies (Walker, 1999; Yoo et al., 2005), and they were assigned a high risk of bias in this category.

Selective reporting bias

We evaluated each of the studies for the possible risk of selective reporting bias. We agreed that this was high in the study by Nunes et al., 2007.

Other sources of bias

We assessed other sources of bias. We checked for care provider experience and a standardised intervention protocol. Based on common consensus, we assigned the study of Bridge et al., 1988, as having lesser risk of bias.

The data were entered into Review Manager 5.4, and a methodological quality summary and graph were prepared [Figure 2].

Close

Figure 2:

Quality assessment of included studies. (a) Methodological quality summary. (b) Methodological quality graph.

Export to PPT

Effects of intervention

Stress

Five studies, including a total of 344 patients, reported the effect on stress. Three studies measured stress during chemotherapy and two during radiotherapy. The time of score measurement ranged from the 7^th day to 15^th week. The heterogeneity was minimal, and a fixed effect model was used for analysis. The stress was significantly lower in intervention group (SMD, –0.38; 95% CI, [–0.59, –0.16]; I² = 8%, P < 0.001). However, this reduction was not significant in radiotherapy subgroup [Figure 3a].

Close

Figure 3:

Forest plot of comparison showing effect on: (a) Stress, (b) anxiety, (c) depression, (d) mood, (e) HRQol, (f) nausea and vomiting, (g) pain, (h) fatigue.

Export to PPT

Anxiety

Five studies with 344 patients reported the effect on anxiety. Three studies reported anxiety during chemotherapy and two during radiotherapy. We used a fixed effect model for analysis as the heterogeneity was small. The period range for the measurement of scores was from the 7^th day to 18^th weeks. In one study (Nunes et al., 2007), the anxiety in the intervention group was more than that in the control group. Overall, PMR+GI was found effective in reducing the anxiety (SMD, –0.30; 95% CI, [–0.51, –0.09]; I² = 24%; P = 0.006). However, a significant reduction in anxiety was not seen in radiotherapy subgroup [Figure 3b].

Depression

Five studies reported the effect on depression. A total of 363 patients were evaluated. The period range for score measurement was from the 7^th day to 18 weeks. There was a significant heterogeneity in the reported outcome, and therefore, random effect model was used for analysis. PMR+GI reduced depression but this was not statistically significant (SMD, −0.37; 95% CI, [–0.81, 0.07]; I² = 76%; P = 0.1) [Figure 3c]. Significant improvement in depression was not observed in either the radiotherapy or the chemotherapy subgroups.

Mood

Only one study (Walker et al., 1999) reported mood on a total of 96 patients. Mood rating score was assessed at the 18^th week in patients receiving chemotherapy. This study showed improvement in the mood of the patients. The effect was not significant (SMD, 0.13; 95% CI, [–0.27, 0.53]; P = 0.51) [Figure 3d].

Quality of life

Two studies reported quality of life on a total of 294 patients. The studies were on chemotherapy patients, and the effect was measured at the 6^th week and 6^th months. There was no heterogeneity, and therefore, a fixed effect model was used for analysis. The quality of life was significantly better in patients with PMR+GI (SMD, 0.60; 95% CI, [0.28, 0.92]; I2 = 0%; P = 0.0002) [Figure 3e].

Nausea

Two studies reported the effect on nausea on a total of 125 patients. Both the studies were on patients receiving chemotherapy. The time range of reporting was at the 7^th day and at the 18^th weeks. There was a significant heterogeneity in the reported outcome; therefore, random effect model was used for analysis. Nausea was less in patients receiving intervention, but this was not statistically significant (SMD, –0.50; 95% CI, [–1.07, 0.06]; I² = 79%; P = 0.08) [Figure 3f].

Pain

Only one study (Chen et al., 2015) reported an effect on pain. The result was studied on the 7^th day, on 65 patients receiving chemotherapy. Pain on the subjective scale was lesser in groups receiving the intervention, but the difference was not statistically significant (SMD, –0.35; 95% CI, [–0.84, 0.14]; P = 0.16) [Figure 3g].

Fatigue

Only one study (Chen et al., 2015) reported an effect on fatigue. The effect was studied on the 7^th day on 65 patients receiving chemotherapy. Fatigue was less in the group receiving intervention, but this was not statistically significant (SMD, 0.14; 95% CI, [–0.35, 0.63]; P = 0.57) [Figure 3h].

Subgroup analysis

There were two studies in the radiotherapy subgroup, and they reported the effect on stress, anxiety and depression. No significant improvement was observed on subgroup analysis following the intervention [Figure 3a-c].

Implications for practice

Based on the small number of randomised controlled trials, it appears that the combined practice of PMR+GI is a valuable intervention. This improves the psychological health of the patient, especially during the period of chemotherapy. In the long term, patients experience a better quality of life. However, the effect on chemotherapy-induced nausea, pain and fatigue may not be significant. This intervention may be of limited use during radiotherapy.

Implications for research

Breast cancer negatively impacts the psychology of the patient. This can be reduced by PMR+GI, especially during chemotherapy. More high-quality randomised controlled trials are needed for better evidence. Effect on chemotherapy-related side effects needs further evaluation. Furthermore, the role of this intervention during the follow-up period should be evaluated. The research methodology should include novel measures for checking patient compliance. The research should also report the difficulties in the technical implementation of the program. More well-planned studies are required to explore its impact on tumour pathology and immunity status.