Description

Context of the problem: Improved knowledge on the diagnosis and treatment of cancer has resulted in an increased number of cancer survivors. Consequently, more patients are confronted, often life-long, with the complications of their cancer treatment. These chronic conditions impact daily functioning and quality of life (QoL), complicate the return to work and undermine psycho-social well-being. One of the most frequent and disabling complications is lymphedema, a condition undermining QoL to an extent exceeding even a 10-year age increase.

The damage of lymphedema covers many aspects of daily life: it changes body image and complicates daily activities as basic as clothing, it increases patient-related health expenditures for treatment and for coping with the disease, reduces patient autonomy in daily functioning and can even cause sensory or motor dysfunctions. So far, no definitive cure exists for lymphedema: patients are confined to life-long preventive measures, compression therapy and regular manual lymph drainage. Surgical repair may be an option in selected cases, but evidence on surgical techniques is limited and their long-term effects are still unknown. Moreover, even after successful repair, conservative treatment is still needed.

Most studies on lymphedema are limited to upper limb lymphedema (ULL), mainly after breast cancer treatment. Only a few publications focus on lower limb lymphedema (LLL), despite its detrimental impact on QoL. This could be due to the difficulties to diagnose and treat LLL in comparison to ULL: LLL often occurs bilaterally or at the level of the groins or genital region, hence impeding diagnosis through comparison. Perimetric measurement of large cone-like oedematous legs is cumbersome. Manual lymph drainage and application of compressive garments is more strenuous for legs than for arms, and the typical bilateral occurrence and large volumes of LLL inflate the cost of treatment garments. However limited, some groups did evaluate the incidence of LLL after pelvic treatment, but maintained the criteria for advanced stage lymphedema as cut off, thereby underestimating the prior phases in which patients are confronted with functional impairment due to LLL, without apparent swelling.

The observational cohort in this study is aimed at inventorying the incidence, timing and risk factors of early as well as advanced stage lymphedema. Time and cost-investment of patients for prevention and/or treatment of LLL will be analyzed. The impact of LLL on QoL and on sexual well-being will be calculated.

A sub-cohort of this trial is of interventional design: in patients, included in the observational cohort and developing early stage LLL, standard of care treatment will be started, including skin care, active exercises and manual lymph drainage. In the experimental group, compressive garments class II will be added and compliance as well as tolerance of these garments will be analyzed.

Side-kicks of this trial are the standardization of perimetric measurements of the legs, validation of a self-report LLL screening tool and testing of the sensitivity and specificity of ICG fluoroscopy to detect early stage LLL.

Sample size calculation:

The sample size is calculated for the interventional sub-cohort, to have at least 80% power to detect a difference between the compression group and the control group for progression of lymphedema. An alpha has been set equal to 0.05.

Sample size calculation is based on evaluation at year 1. No longitudinal analysis is foreseen as primary analysis, but this could be included as a secondary analysis using a GEE analysis.

To detect a difference, in the interventional sub-cohort 109 subjects are required per group (2*109=218 subjects in total for the two groups) to have at least 80% power. If 66% of the patients in the observational cohort will meet the inclusion criteria for the interventional randomized trial, at least 330 (=218/0.66) patients should be included in the observational cohort. To correct for drop out we planned to include 400 patients. This number is feasible within the planned recruitment period of 18-24 months. If the number of subjects would be reached before the end of the planned recruitment period, recruitment will be stopped at 400 patients. If the number is not attained, the recruitment period will be prolonged. If the number of 218 subjects is reached in the interventional sub-cohort, randomization will be stopped and analysis starts. However, the observational cohort will continue accrual until 400 subjects are reached.

We assume a difference in true transition rates (T1 to T3) in both study arms, with 0.4 in the interventional group versus 0.75 in the control group. The recruitment of 109 patients in each treatment arm would allow the assessment of benefit of the intervention with 80% power, based on the 95% two-sided confidence interval approach for the differences in two independent proportions.