Introduction

One of the most debilitating side-effects of cancer treatment is cancer-related lymphoedema (CRL). It is characterised by the progressive accumulation of protein-rich fluid within the interstitial compartment, causing dysfunction of the lymphatic system.1 2 The fluid overload eventually leads to adipose tissue deposition and fibrosis.3

Patients with CRL may experience a high burden of disease. The progressive swelling can lead to sensations of heaviness, pain, paraesthesia, reduced range of motion, weakness and immobility. Furthermore, individuals with CRL have a higher propensity for cellulitis and erysipelas. In addition, patients may also experience a variety of psychosocial symptoms, such as depression, anxiety, and difficulties in social, domestic, vocational, and sexual domains. Altogether, patients with CRL may experience a diminished health-related quality of life (HRQoL) in comparison with cancer survivors who have not developed lymphoedema.4–8

CRL can arise after oncological treatment for several solid tumour types, including breast, melanoma, head and neck, gynaecological and genitourinary malignancies.9 While it may manifest in various regions of the body, it predominantly affects the upper and lower extremities.2 The most prevalent form is breast cancer-related lymphoedema (BCRL), occurring in 24–49% of cases after mastectomy and 4–28% after lumpectomy.10 The overall incidence of CRL is 15.5%, ranging from 10% in genitourinary cancers to 30% in sarcomas, with the highest rates observed in patients receiving additional radiotherapy (31%).9 Other factors that contribute to the risk of developing CRL include lymph node dissection, obesity, pre-existing medical conditions and genetic predisposition.11–18 Moreover, the risk of developing lower limb CRL also varies depending on the type of lymph node dissection performed. For melanoma treatment, inguinal lymph node dissection is recommended, whereas gynaecological and prostatic cancer treatments often involve pelvic and para-aortic lymphadenectomy. The latter approach preserves superficial lymph nodes in the limb, resulting in a lower risk of developing lower limb CRL compared with melanoma treatment.19 In BCRL, the highest risk for the development of lymphoedema is between 12 and 30 months postoperatively. Gynaecological cancers and melanoma exhibit the highest onset frequency after the first year of diagnosis.18 20–22

To date, there is no definitive cure for lymphoedema. The gold standard for its treatment remains conservative therapy, consisting of complex decongestive therapy (CDT). CDT aims to ameliorate the symptoms of lymphoedema through bandages, compression garments, manual lymphatic drainage (MLD) and skin therapy. However, CDT requires lifelong maintenance and does not alleviate psychological symptoms. Multiple surgical techniques have been proposed to cure lymphoedema, however, with significant limitations.23 Among these techniques is lymphaticovenous anastomosis (LVA), a microsurgical technique that has been refined throughout decades, particularly after the introduction of super-microsurgery by Koshima et al.24 The LVA is a bypass between lymphatic vessels (ranging from 0.3 mm to 0.8 mm in diameter) and adjacent subcutaneous veins of similar size. The objective of LVA surgery is to partially restore the blockage in the lymphatic system, reduce volume, minimise skin infections, reduce the need for conservative therapy, and most importantly, improve the patients’ HRQoL.25–27

Several studies have evaluated the effectiveness of LVA surgery in treating lymphoedema. In a meta-analysis by Nacchiero et al,27 a significant positive effect of LVAs in treatment of lymphoedema was reported, with a combined OR of 0.07 (95% CI: 0.04 to 0.13, p<0.001). Only six of the included studies were clinical trials, none of which had a control group, and their pooled analyses resulted in an OR of 0.34 (95% CI: 0.14 to 0.81, p<0.01).27 In a systematic review by Verhey et al,28 an objective improvement after LVA (eg, volume reduction), ranging from 23.3% to 100%, was reported in patients with lower limb lymphoedema. The greatest degree of improvement was seen in patients with early-stage lymphoedema.28 Similarly, a review by Cornelissen et al29 on the effectiveness of LVA in BCRL reported improvement rates on circumference and volume measurements, ranging from 50% to 100%.29 Nonetheless, it is important to note that available studies are small and non-randomised. According to a systematic review by Rosian and Stanak,30 the evidence supporting the effectiveness of LVA is ‘very low’. The review emphasises methodological shortcomings in the available evidence and concludes that no definite conclusions can yet be drawn regarding the effectiveness of LVA surgery.30

The first multicentre randomised controlled trial (RCT) evaluating the (cost-)effectiveness of LVA compared with CDT in patients with BCRL, ‘the Dutch LYMPH trial’, is presently ongoing.31 32 The current study, ‘the N-LVA Study’, is a continuation of ‘the Dutch LYMPH trial’ and compares LVA surgery with sham surgery. The main goal of ‘the N-LVA Study’ is to assess whether LVA surgery is effective in comparison with sham surgery. A sham-controlled trial is chosen as the best possible comparator to investigate the true effect of LVA surgery and to examine whether there is a placebo effect.

Methods and analysisStudy design

‘The N-LVA Study’ is a multicentre, double-blind, RCT conducted at Maastricht University Medical Center, Radboud University Medical Center and Erasmus University Medical Center in the Netherlands. A total of 110 patients with CRL will be recruited at the outpatient clinic of the participating hospitals: 70 patients with upper limb lymphoedema and 40 patients with lower limb lymphoedema. The inclusion and exclusion criteria are shown in table 1. The coordinating researcher or research nurse obtains written informed consent (see online supplemental data I for the informed consent form). After inclusion, the patients will be randomised into one of the two treatment groups with a 1:1 allocation through the electronic case report form (eCRF) in Castor EDC. The randomisation will be stratified per limb (upper/lower). The investigator, research nurse and patients will be blinded throughout the duration of the trial. Due to the nature of the study, the surgeons cannot be blinded. The start date of the study is December 2023 and the anticipated completion date is March 2029. The final version of the protocol is V.5, dated 28 August 2023. The protocol adheres to the Standard Protocol Items: Recommendations for Interventional Trials 2013 statement33 (see online supplemental data II for the checklist). A flow chart of the study design is shown in figure 1.

Table 1

Inclusion and exclusion criteria

Figure 1

Overview of the study design. LVA, lymphaticovenous anastomosis; NIRF, near-infrared fluorescence.

Sample size calculation

The sample size calculation was performed to be able to detect a clinically relevant difference in the score on the Lymphedema Functioning, Disability and Health (Lymph-ICF) Questionnaire between groups at 24-month follow-up. A difference in Lymph-ICF score of 15 points for the upper limb and 20 points for the lower limb is deemed clinically relevant, with an estimated SD of 20 points.34 35 Because of the difference in cut-off point per limb, the randomisation is stratified and analysed separately based on the location of lymphoedema. Using an alpha of 0.05, the following sample size was calculated to achieve a power of 80%: for the upper limb, a minimum of 28 patients per group is required, resulting in a total of 56 patients. To account for a potential drop-out rate of 20%, a total of 70 patients is included. For the lower limb, a total of 16 patients is needed per group, totalling 32 patients. Taking a potential drop-out rate of 20% into account, 40 patients will be included. Drop-outs will not be replaced. Altogether, 110 patients will be included in this study, that is, 55 patients per treatment group with a 1:1 allocation.

CDT protocol

All eligible patients must undergo at least 3 months of CDT prior to inclusion. Throughout the duration of the study, patients are encouraged to continue CDT, as it is considered as standard care. If the patient does not already have a dedicated lymphoedema therapist for CDT, the patient will be referred by the investigator. The patients already undergoing CDT can continue their therapy as accustomed. Generally, patients in the Netherlands are treated according to the Dutch and German guidelines for CDT, referred to as the ‘Verdonkmethod’ and ‘Asdonkmethod’, respectively.36 37 CDT involves two phases. In phase 1, the goal is to reduce swelling through compression bandaging and MLD. In phase 2, after the swelling of the limb is sufficiently reduced, patients are fitted with a compression garment.36 38 To provide a structured framework for lymphoedema therapists, standardised treatment protocols have been developed for both upper and lower extremities, adhering to the aforementioned guidelines. See online supplemental data III and IV for the CDT protocols of the upper and lower limbs, respectively. However, therapists can determine the appropriate frequency accordingly. Patients can document details regarding conservative treatment through a digital patient survey.

Preoperative protocol

At the outpatient clinic, patient eligibility is determined by assessing the indocyanine green (ICG) stage and identifying viable lymphatic vessels through near-infrared fluorescence (NIRF). Injections are prepared by dissolving 25 mg of ICG into 5 mL ‘water for injection’, to obtain a concentration of 5 mg/mL. For diagnostics of the upper extremity, 0.1 mL is injected intradermally into both the second and fourth web spaces, and 0.2 mL in both the first and fourth web spaces of the affected lower limb. Using NIRF, viable lymphatic vessels are outlined with a surgical marker, and the incision site for the LVA or sham procedure is determined. Afterwards, a colour photo is taken of the marked limb, positioned alongside a measuring tape. During surgery, the photo is used as a reference for the incision site for the LVA or sham surgery.

Surgical protocolGroup A: LVA

Preoperatively, the patient is placed in supine position. The affected limb is positioned and sterilised for surgery. During surgery, the patient is then blinded by concealing the surgical area with a surgical drape to prevent the patient from observing. The patient is also fitted with noise-cancelling headphones to block out the medical personnel’s communication. The incision site or sites are marked at the predetermined locations. Local anaesthesia and hemostasis are achieved by injecting epinephrine and bupivacaine (1:100.000) at the site of the incision. Incisions of 1.5–2 cm are made in the subdermal plane at the marked location. Using a surgical microscope (25–40× magnification), the subdermal lymphatic vessels and veins are identified and anastomosed. When the calibre of the lymphatic vessel and the vein is congruent, the LVA is performed in an end-to-end fashion; otherwise, an end-to-side anastomosis is made. The anastomosis is created using 11-0 or 12-0 Ethilon sutures. Generally, one to four LVAs are made during one procedure. The superficial wound is closed using uninterrupted, intracutaneous sutures with 4-0 Monocryl or transcutaneous sutures with 5-0 Ethilon, according to the surgeon’s preference. Once the wounds are closed, the headphone and blindfolds are taken off and the patient is provided with postoperative care instructions. The total operation time is approximately 90–120 min. The plastic surgeon will document the procedure in the electronic patient file.39

Group B: sham surgery

Upon the patient’s arrival in the operating room, the setup will be identical to that of the LVA operation, with the microscope prepared, designated surgical personnel present and all the necessary super-microsurgical instruments readily displayed. The patient lies comfortably on the operation table, after which the affected limb is prepared for surgery. The patient is prepared for surgery and blinded in nearly the exact same manner as in the LVA operation, with two exceptions. First, based on preoperative photos, the incision site is marked 2 cm adjacent to (one of) the predetermined site(s) to avoid damage to the lymphatic vessels as to allow for future LVA surgery. Second, rather than performing the actual operation, the plastic surgeon simulates the procedure by applying pressure with anatomical forceps in the surgical area. To mimic the approximate duration of a regular LVA procedure, the total operation time is approximately 90–120 min. The superficial wound is closed using the same technique as in the LVA procedure. Following surgery, the patient is provided with postoperative care instructions specifically for LVA surgery.

Postoperative protocol

The postoperative protocol is identical for both groups. After surgery, the limb is bandaged for at least 3 days. Afterwards, compression therapy can be resumed, according to the patient’s regimen. Patients are instructed not to receive any kind of MLD directly at the level of the surgical wound in the first 4 weeks after surgery. If needed, the sutures can be removed by the general practitioner or nurse. Strenuous physical activity is not recommended until after the first 6 weeks.

Outcomes

The primary outcome is change in HRQoL. The secondary outcomes are change in limb circumference and excess limb volume, changes in conservative therapy, postoperative complications, patency of the LVA, costs, generic HRQoL and incremental cost-effectiveness. The follow-up period is at least 24 months; thereafter, a subset of patients will partake in the extended follow-up, which will take place annually. The extended follow-up will involve a maximum of three additional annual visits. All outcomes, except for LVA patency, are measured at baseline, 3, 6, 12, 18 and 24 months postoperatively, followed by an annual assessment during the extended follow-up. The assessment of LVA patency is performed by the plastic surgeon or technical physician and occurs solely after 12 and 24 months and annually during the extended follow-up. The patient survey is additionally sent out at 9 and 15 months. All follow-up measurements are recorded in the eCRF within Castor EDC, and digital questionnaires are automatically sent through the eCRF.

Primary outcomeHealth-related quality of life

The primary endpoint is HRQoL, measured by the Dutch version of the Lymph-ICF Questionnaire (2019) after 24 months. The Lymph-ICF is a validated, disease-specific questionnaire to asses impairments in daily function, activity limitations and participation restrictions. There are two versions of the Lymph-ICF: one for the upper limb and one for the lower limb, consisting of 29 and 28 questions, respectively. Questions are categorised across five domains, namely physical function, mental function, household activities, mobility activities, and life and social activities. Each question is scored on a Visual Analogue Scale from 0 to 100. The total score is equal to the sum of the individual question scores, divided by the total number of questions answered. A decrease in score represents an improvement in HRQoL.34 35

Secondary outcomesLimb circumference

The limb circumference is measured at fixed points using measuring tape. The upper limb is measured at the olecranon, 5 and 10 cm above and below the olecranon, at the wrist and at the dorsum of the hand. For the lower limb, measurements are taken at the superior edge of the patella, 10 and 20 cm above and below the patella, at the lateral malleolus and on the dorsum of the foot. The Upper and Lower Limb Lymphedema Indexes are derived from these measurements and are adjusted for body mass index.40 41

Excess limb volume

Excess limb volume is measured through BioImpedance Spectroscopy, a method that measures the electrical impedance of tissues to an electric current. This makes it possible to measure the excess limb volume and quantify the amount of fluid in different body compartments.42 Both the absolute excess limb volume in the affected limb and the relative difference in excess limb volume compared with the unaffected limb are calculated to evaluate the effect of the LVA operation. A clinically relevant difference between the LVA and sham groups is indicated by the minimal clinically important difference of a 20% reduction in excess limb volume. For accurate measurements, patients are expected not to eat or drink 1 hour before the measurement and should have an empty bladder.

Monitoring of conservative therapy

The changes in conservative therapy are assessed through an electronic patient survey to record the frequency of treatments received (ie, skin therapy visits, number of compression garments, etc).

Postoperative complications

All postoperative complications are recorded to monitor safety.

Patency of the LVA

NIRF is used to assess the patency of the LVAs, as well as to determine the ICG stage based on the extent of dermal backflow.

Costs, generic HRQoL and cost-effectiveness

Costs related to lymphoedema are measured by assessing individual-level healthcare resource use, out-of-pocket expenses, use of informal care and productivity loss. Examples of cost categories included are intervention costs, costs for outpatient clinic visits, diagnostic procedures, hospital admissions, visits to the general practitioner, lymphoedema therapist, other allied healthcare professionals, and home care services.

Hospital-related care data are extracted from the hospital information system. For all other costs, patients complete the adapted version of the iMTA Medical Consumption Questionnaire and the iMTA Productivity Cost Questionnaire.43–45 To calculate the patient’s total costs, the volume of resource use, per cost category, is multiplied by the Dutch cost price using reference prices provided by the Dutch guidelines for cost analysis in healthcare.46

Generic HRQoL is assessed by the EQ-5D-5L.47 The Dutch tariff is used to calculate utility scores that are used as input for the quality-adjusted life year (QALY).48 Cost-effectiveness is expressed in an incremental cost-effectiveness ratio (ICER), that is calculated by dividing the difference in the costs between LVA and sham surgery by the difference in QALYs.

Data analysis plan

All analyses will be performed according to the intention-to-treat principle. In case of over 5% of incomplete records, data will be imputed using multiple imputation with fully conditional specification to prevent loss of precision and to reduce the likelihood of bias.

Baseline characteristics will be stratified by location and by treatment arm, and reported as mean and SD for continuous variables, and count and percentage for categorical variables. In case continuous variables are skewed, we will use the median and first and third quartiles to summarise the distribution.

The paired samples t-test will be used to assess changes in Lymph-ICF scores within each group (upper and lower limb, and treatment allocation) from baseline to 12 and 24 months. Between-group differences will be compared using the independent-samples t-test. In case of baseline imbalance, differences per follow-up moment will be computed using multivariable linear regression analysis, adjusted for the offending baseline characteristics. Linear mixed-effects regression will be used to analyse longitudinal changes over the follow-up period and compare both groups, with a random intercept on group and slope for time. The variance–covariance matrix for random effects will be left unstructured. The correlation of measurements over time will be simplified using an autoregressive model of order 1. Should model convergence issues occur, an alternative will be sought that results in the lowest Akaike Information Criterion. Regarding the secondary outcome measures, besides cost-effectiveness analyses, the same between-group statistical analysis will be performed. The occurrence of postoperative complications and the patency of the LVAs will be described as count and percentage, and compared between groups using Pearson’s Χ2 test. In case of low expected cell counts, Fisher’s exact test will be used. In the presence of baseline imbalance, we will use multivariable logistic regression analysis and present results as OR with 95% CI.

Economic evaluation

A trial-based economic evaluation will be performed from a societal perspective and with a time horizon of 2 years to assess the cost-effectiveness of LVA compared with sham surgery, adhering to the Dutch guidelines for health economic evaluation.49

The cost-effectiveness analysis will be conducted according to the intention-to-treat principle. Missing cost and effect data will be imputed using multiple imputation methods.48 Each of the imputed datasets will be analysed separately and results are pooled using Rubin’s rules.

Cost-effectiveness will be expressed in an ICER: the incremental costs per QALY. Cost and QALY differences are estimated using regression analysis, and are adjusted for baseline differences. To address the uncertainty surrounding the differences in costs and effects, non-parametric bootstrapping with 5000 replications is used and the bootstrapped cost and effect pairs are plotted on a cost-effectiveness plane. If relevant, the ICER is calculated by dividing the difference in costs by difference in QALYs. A cost-effectiveness acceptability curve will visualise the probability that LVA is cost-effective for a range of willingness-to-pay thresholds. The impact of uncertainty surrounding deterministic parameters (eg, cost prices) will be explored using one-way sensitivity analyses. Additionally, scenario analyses will explore the impact on the ICER when a healthcare perspective is taken and when a per-protocol analysis is used.

Subgroup analyses will be performed for patients with upper limb lymphoedema and patients with lower limb lymphoedema, to address possible heterogeneity. Finally, a Budget Impact Analysis is performed to analyse the financial consequences related to implementing LVA in the Netherlands.49

Ethics and disseminationData monitoring

Data will be handled confidentially and will remain at the research site at all times. The investigator will store the source data in a locked place. Follow-up data will be stored in the online database of Castor EDC. Only the principal investigator, site investigators and coordinating investigator will have access to this database with a password-secured account. The investigators will only be granted access to data from their own centre. An independent investigator will perform the randomisation within Castor EDC and will only have access to the randomisation section within the database. Identifying data will be stored in coded form and the key will only be known to the principal investigator, coordinating investigator, site investigators, study monitors, members of the review committee and the Dutch Health Care Inspectorate. Processing data will take place in accordance with the European Union General Data Protection Regulation and the Dutch Act on implementation of the General Data Protection Regulation.50 51

No data monitoring committee will be appointed, as the study is classified as low risk. Additionally, no interim analyses will be performed.

Harms

LVA is a minimally invasive operation with a low risk and complication rate. The risks are low because the lymphatic vessels that are used are already damaged at the level where the LVAs are made. Furthermore, the incision that is made is small and superficial, approximately 1.5 cm in depth. At this level, no big vessels or other structures are present. The risks of both the LVA and sham procedures are surgery related and include wound infection, minor bleeding, wound (healing) problems, and skin reaction to plasters or bandages. Risks are further minimised by operating under local anaesthesia. Adverse events (AEs) will be documented in the eCRF and serious AEs will be directly reported to the sponsor.

Auditing

The Clinical Trial Center Maastricht will frequently monitor the study’s progress in accordance with their protocol, as is requested by the board.

Research ethics approval

Ethical approval was obtained in September 2023 from the Medical Ethical Committee Academic Hospital Maastricht/Maastricht University (NL84169.068.23/METC 23-023). This approval is valid for all participating centres. This study will be conducted according to the principles of the Declaration of Helsinki, recently changed in Fortaleza (2013), and in accordance with the Medical Research Involving Human Subjects Act.52

Protocol amendments

All amendments will be made in consultation with all participating centres and the Medical Ethical Committee that gave a favourable opinion will be notified.

Patient and public involvement

A consultation group with professional and patient associations involved will meet annually throughout the duration of the study to provide feedback.

Dissemination policy

The study is registered in the trial register at www.clinicaltrials.gov with trial registration number NCT06082349. The results will be presented at scientific conferences and published in peer-reviewed medical journals. Requests for data sharing will be assessed individually, considering them for appropriate research purposes after the completion of the trial and publication of primary manuscripts.