External application of liver compresses to reduce fatigue in patients with metastatic cancer undergoing radiation therapy, a randomized clinical trial

This study was approved by the Ethical Committee of the Charité Universitätsmedizin Berlin, Germany (Reference number EA1/078/17) and the pilot trial was registered in the Cochrane Central Register of Controlled Trials (DRKS00012999).

For logistical reasons the decision was made to conduct this trial among patients with metastatic cancer undergoing radiation therapy being on an inpatient basis within the Department of Radiation Oncology, Charité Universitätsmedizin Berlin.

The criteria for patient eligibility were age ≥ 18 years, with at least minor fatigue according to the general fatigue subscale of the multidimensional fatigue inventory (MFI-20) [13], estimated life expectancy > 3 months, Karnofsky performance score ≥ 60%, indication for inpatient palliative whole brain radiation therapy (WBRT) or palliative analgesic RT of bone metastasis. Patients were required to be inpatients due to the foreseen logistic hurdles in preparing and applying liver compresses in an outpatient setting. Patient written consent was mandatory. Patients were ineligible when severe psychiatric disorders or allergies prohibited participation and in cases with medical conditions potentially causing fatigue such as severe hypothyroidism, sleep apnea, insomnia, anemia with hemoglobin levels < 8 g/dL, cachexia with a body mass index < 18.5, chronic kidney failure or acute depression. Patients with liver metastasis were also ineligible because in this case warming compresses were contraindicated, as were patients participating in other clinical trials. Patients receiving both WBRT and bone RT were also ineligible.

Patients in both groups were treated by external beam RT. All patients underwent a dedicated planning computed tomography scan. For WBRT patients received 10 fractions of 3 Gy over 2 weeks using a three-dimensional conformal RT (3D-CRT) approach, wearing a thermoplastic mask for immobilization. Patients with bone metastasis received either 10 fractions of 3 Gy over 2 weeks or a slightly intensified regimen in the case of soft tissue tumor mass with up to 12–13 fractions of 3 Gy using 3D-CRT or an intensity modulated radiation therapy (IMRT) or volumetric modulated arc treatment (VMAT) approach. For WBRT the brain was defined as gross tumor volume (GTV) surrounded by a 5 mm margin to establish the planning target volume (PTV). For the treatment of bone metastasis, the involved bones/bone regions were defined as GTV and a 5–10 mm margin added to establish the PTV. RT was performed using a 6-MV linear accelerator with multileaf collimators (Varian, USA).

During the inpatient course of treatment patients in both groups were offered palliative care in the form of visits from a dedicated multi-professional palliative care team of physicians, nurses, psycho-oncologists, physiotherapists and social workers, depending on the requirements of the respective patient.

External application of yarrow liver compresses was performed once daily over the course of two weeks of RT, the minimum number of applications was set to seven. Applications were preferably performed after lunch or in the evening. The liver compresses were prepared and applied by trained members of our scientific team in cooperation with the staff on the ward. For preparation of the compresses a dedicated room on the in-patient ward outside the patient room was used. One tablespoon of yarrow tea with cut blossoms, stems and leaves of common yarrow was infused with 500 ml boiling water and brewed for 10 min. The infusion was then strained and cooled for 5 min at room temperature to reach the required application temperature of 60 °C. A muslin cloth (made of cotton) was folded to the size of 20 × 30 cm, rolled up firmly (into a small roll) and placed on a rough cotton cloth (such as a tea towel). Both together were rolled up tightly, twisting the rough cotton cloth so that the two ends formed handles. This roll was then placed in a bowl. The necessary materials where brought to the patient’s room where the patient was then comfortably positioned on the bed lying on an outer wrap. The wound-up roll in the bowl was then doused with the 60 °C hot yarrow infusion and wrung out firmly, wearing gloves to protect the hands from the heat. The muslin cloth was removed, carefully placed on the patient’s right upper abdomen and lower costal arch. In case the patient experienced the muslin cotton as too hot, it was fanned to make the heat bearable. It was then covered with a cotton compress cloth which had been previously folded to the size of 20 × 30 cm. The outer wrap underneath the patient’s back was then folded tightly over the compress from both sides. A hot-water bottle was placed on top and could be removed at the patient’s convenience. The patient was then covered with a blanket and allowed to rest for 20–25 min. Care was taken to leave the patient alone in the room without external disturbance during the rest period. After 20–25 min the moist inner muslin cloth was removed while the compress cloth and outer wrap stayed in place, including the hot-water bottle and the patient left to rest for another 30 min (Fig. 1).

Fatigue was assessed at baseline, after one week and at the end of the treatment according to the German version of the MFI-20 [13]. The MFI-20 is a 20-item psychometric inventory designed to evaluate five dimensions of fatigue: general fatigue, physical fatigue, reduced motivation, reduced activity, and mental fatigue [14]. The primary trial endpoint was the general fatigue subscale of the MFI-20 at the end of treatment. A mean difference of two points compared to baseline was considered clinically relevant [15].

Secondary endpoints included psychological distress as measured according to the German version of the distress thermometer of the National Comprehensive Cancer Network (NCCN) [16], patient’s quality of life was assessed according to the European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 questionnaire [17]. Moreover, a qualitative analysis of patients of the intervention group was performed using questions regarding their condition on self-established visual analogue scales (VAS). The VAS covered tension, restlessness, fatigue, lack of drive, exhaustion, experience of warmth (cold hands/feet) and pain. All questionnaires were administered at baseline, after one week of treatment and at the end of treatment. As a sub-project, physiological data was assessed using heart rate variability and body temperature, which will be reported separately. Liver compress-related symptoms were assessed after each application and the time of onset and resolution as well as the intensity (mild vs. intermediate vs. severe) was noted.

The trial was designed as a randomized, non-blinded prospective clinical pilot trial. A pilot phase was planned with subsequent sample size recalculation. Block randomization was performed with stratification according to the RT site brain vs. bone.

The power calculation for the pilot phase was based on a 3-point reduction on the general fatigue subscale of the MFI-20 in the intervention group vs. control group at the end of treatment compared to baseline. With a common standard deviation of 3 points, alpha 0.05 and a power of 80% an initial sample size of 34 patients (17 per group) was required for an unpaired t-test. Accounting for 10% drop-outs 38 patients should initially be included. Due to the unblinded non-confirmatory design were p-values are regarded as exploratory, no alpha-adjustment was planned. The sample size calculation was performed using G*Power Version 3.1.3.

Analysis of the pilot trial was performed using analysis of covariance (ANCOVA) with end and baseline differences on the general fatigue subscale as dependent and group variables, RT site brain vs. bone (strata) as independent variables and baseline general fatigue as covariate variable. A p value < 0.05 will be regarded as significant. P-values were not corrected for multiple testing. The effect size partial eta squared (η²) was presented to show the difference between groups as well as omega squared (ω²) for correcting biased from a small sample size [18].

After the pilot trial had experienced severe accrual problems, due to many patients being ineligible and several other patients withdrawing from the trial after inclusion (Fig. 1), the pilot trial was halted in September 2019. Given the mean difference of two points on the general fatigue subscale (considered clinically relevant) in favor of the intervention group already found, the trial team decided to stop the pilot trial and moved ahead to analyze the results. All statistical tests were performed using Stata IC15 (StataCorp, 2017, College Station, TX, USA).

Between September 2017 and August 2019 a total of 79 preselected patients were screened, of whom 39 were randomized, and after exclusion of drop outs a total of 24 patients (12 per group) were available for analysis (Fig. 2). The baseline patient characteristics are summarized in Table 1, there were no significant differences between the two pilot trial groups. Patients in the intervention group received a mean number of 10.5 applications of yarrow liver compresses (range 7–14).

For the primary endpoint, the general fatigue subscale of the MFI-20, we observed a mean improvement over baseline at the end of treatment of 2 points in favor of the intervention group (p = 0.13, effect size (ω²) = 0.062) (Table 2 and Fig. 3). All other MFI-20 subscales tended in the same direction towards improvement of fatigue in patients of the intervention group. For the MFI-20 subscale reduced motivation the mean improvement at the end of therapy was 2.1 points the difference being statistically significant between the pilot trial groups (p = 0.035, effect size (ω²) = 0.158) (Fig. 3). Mean changes from baseline (t0) at one week (t1) and at the end of treatment (t2) for all MFI-20 subscales are provided in the Additional file 1: Fig. 4.

Differences in psychological distress also tended to favour the intervention group (Table 2). Moreover, the VAS results showing significant improvement in the symptoms fatigue, tension and lack of drive in patients from the intervention group when compared to the control group also suggested a positive effect of the intervention (Table 2). There was no difference in the global health status/quality of life between the two groups (Table 2).

There were no significant differences between patients treated with WBRT or bone metastasis with respect to reduction of fatigue.

In five patients from the intervention group (42%) local redness of the skin was observed at least once after the application of the compress which was not perceived as uncomfortable by the patient and completely returned to normal without intervention. One patient developed mild nausea and in another patient a single episode of temporary localized crampy pain in the liver region was observed after application of the compress. Treatments were continued in both patients and neither of the two symptoms occurred again. Major compress-related toxicities were not observed.