Breast cancer–related lymphedema (BCRL) is one of the major postoperative complications suffered by breast cancer survivors.1 It is defined as the abnormal accumulation of fluid in the interstitial space of the affected arm, breast, and chest wall.2 Breast cancer treatments, especially axillary lymph node dissection (ALND) and radiotherapy, are blamed for the occurrence of BCRL for impairing the lymphatic system of the axilla.3 Additionally, inappropriate use of upper extremity and high body mass index (BMI) may contribute to the risk of developing BCRL.4 The incidence of BCRL varies widely between 2.5% and 42.9% in different studies, depending on the patient characteristics, definition, and measurements.3,5–8 Without timely and appropriate intervention, BCRL can be a progressive condition. A study that followed up 964 patients after ALND for 10 years, in which BCRL was defined by 200-mL difference between affected and unaffected limb volume, found that the cumulative incidence was 13.5% at 2 years of follow-up, 30.2% at 5 years, and 41.1% at 10 years.5 The incidence of BCRL increased significantly over the postoperative period.

Breast cancer–related lymphedema severity also increased over time. The International Society of Lymphology classified lymphedema as stages 0 to III.9 Stage 0 refers to a latent or subclinical condition where swelling is not evident, but subjective symptoms occur. Stage I represents an early and mild accumulation of protein-rich fluid in the arm, hand, trunk, or breast that can reverse with limb elevation or compression. It may present with pits, but the skin is typically soft with no dermal fibrosis. Stages II to III develop excess subcutaneous fat, fibrosis, and lymphatic elephantiasis, which are difficult to intervene. Specific diagnostic criteria were used to assess the severity of BCRL further, regarding the difference in limb volume,9 limb circumference,10 and L-Dex value.11 With the progress of the BCRL stage, survivors might suffer from more symptoms,12 which led to functional, psychological, and social morbidity and a severe negative impact on patients’ quality of life.13,14 Preliminary research findings suggested that early diagnosis and intervention in the subclinical phase and mild stage might prevent the development of BCRL and minimize symptoms.1,15,16

A variety of interventions, including manual lymph drainage,17 physical therapy,18 compressive garments,19 upper extremity exercise,20 and yoga,21 have been applied to the subclinical and mild stage of BCRL. A randomized controlled study involving 45 breast cancer survivors demonstrated that light compression sleeves combined with physical activity applied early after surgery might be a safe and efficient option to prevent postsurgical arm swelling and development of BCRL.19 In another prospective observational study, patients diagnosed with subclinical BCRL by bioimpedance spectroscopy received a series of interventions including short-term physical therapy, compression garments, and education, and the incidence of BCRL reduced from 36.4% to 4.4% over a 20-month follow-up.11

As the irreversible and progressive nature of this complication, patients will lifelong struggle with BCRL.22 Interventions focusing on self-management held great promise for slowing the progression of BCRL.23 The-Optimal-Lymph-Flow (TOLF) program is a patient-centered nurse-led self-management program developed by Fu et al.24,25 It helps patients understand what BCRL is and how important to prevent it and empower patients to take safe actions. In 2014, 140 patients after breast cancer surgery were stratified into the ALND group (n = 81) and the sentinel lymph node biopsy group (n = 59) and received the preventive intervention of TOLF program. Findings showed that no patients in the sentinel lymph node biopsy and ALND group exceeded ≥10% limb volume increase. TOLF program is effective in risk reduction of lymphedema.24 However, there is a paucity of evidence to apply these self-management strategies to the subclinical and mild stage of BCRL. It is unclear if TOLF program targeting daily behaviors, proactive exercises, and optimal BMI can effectively reduce the risk or halt the progression of the subclinical and mild stage of BCRL.

To address this critical clinical need, we conducted a quasi-experimental study to evaluate the effectiveness of TOLF program in reducing the risk of the subclinical and mild stage of BCRL in Chinese breast cancer survivors over a 12-month intervention period. This study was guided by a conceptual framework called the Model of Self-care for Lymphedema Symptom Management.25 In this model, physiological treatment factors, such as surgery and radiotherapy, which directly lead to BCRL symptoms, are difficult to be modified.5,7 However, compromised lymphatic drainage and BMI can be modified through self-care strategies to reduce the risk of BCRL.4,25,26 By targeting compromised lymphatic drainage, shoulder mobility exercises, muscle tightening–breathing, muscle tightening–pumping exercises, and large muscle exercises are provided by TOLF program to promote lymph flow and lessen lymphedema symptom burden and help reverse the mild BCRL and prevent further development of the subclinical BCRL.24,25 Meanwhile, general instructions are provided to encourage nutrition-balanced, portion-appropriate diet, adequate hydration, and sleep to strive for maintaining optimal BMI and minimizing the risk of lymphedema.24,25

Methods Study Design

A longitudinal, quasi-experimental (single group), pre-post design with repeated measures was used to test the effect of TOLF program on the subclinical and mild stage of BCRL among Chinese breast cancer survivors.

Participants

We screened the medical records of breast cancer patients who had surgery from April 2017 to June 2018 at a large metropolitan tertiary hospital in northern China. The researcher contacted postoperative breast cancer patients by telephone and made an appointment to conduct limb circumference measurements during the next routine doctor appointment. The inclusion criteria for screening participants were as follows: (a) female, (b) aged 18 years or older, (c) initially diagnosed with unilateral breast cancer and finished surgery, and (d) completed ALND. The exclusion criteria for screening participants were as follows: (a) undergoing radiation and/or chemotherapy, (b) with tumor metastasis or recurrence, (c) other conditions that could lead to edema (eg, renal disease, malnutrition, or congestive heart failure), and (d) history of surgery or trauma on the affected axilla or arm.

Patients were screened again by the results of limb circumference measurements. Only patients with limb circumference difference of 1 to 3 cm (subclinical or mild stage of lymphedema) were invited to participate in this intervention study. Among 430 patients who received limb circumference measurements, 179 of them had a limb difference of 1 to 3 cm and were invited to participate in our study, and finally, 41 patients signed the written informed consent form and accepted intervention. Reasons provided by patients who did not enroll in the study included significant travel distance, stress from cancer treatment, or busy with work.

Interventions

The-Optimal-Lymph-Flow program is a patient-centered educational and behavioral program focused on daily self-management strategies, including basic knowledge about BCRL, exercises to promote lymph flow and drainage (muscle tightening–breathing, muscle tightening–pumping, and large muscle exercises), arm precautions to improve limb functional status, and instructions to optimize BMI.25 Our research team has translated TOLF program handbook into Chinese after authorization. Table 1 presents the main self-management strategies of TOLF program.

Table 1 - The-Optimal-Lymph-Flow Program Strategies and Actions Frequency and Situations Basic knowledge about lymphedema What is lymph system?
What is breast cancer–related lymphedema?
Why does lymphedema exist after breast cancer treatment?
What time will the lymphedema exist?
What are the common symptoms of lymphedema? Promoting lymph flow Muscle tightening–breathing and Muscle tightening–pumping At least 3 times a day before meals or as much as the patient wants throughout the day
Air travel: before take-off and after landing
Sedentary lifestyle: at least every 4 h Large muscle exercises (walking, swimming, dancing, yoga, jogging, tai chi) At least 15–30 min daily
Air travel: get up and walk around every 1 h during flight
Sedentary lifestyle: get up and walk at least every 1 h Arm precautions Skin care (keep skin clean and moist)
Avoid skin damage
Timely treatment of mild skin damage
Avoid lifting heavy objects by affected limb
Avoid repetitive and excessive activities
Avoid high-temperature environment Always Instructions to optimize body mass index Nutrition-balanced (ie, more vegetables and fruits, feeling 75% full for each meal) Each meal daily Stay hydrated Drink 1500–1700 mL (6–8 glasses) water daily Large muscle exercises (walking, swimming, dancing, yoga, jogging, tai chi) At least 15–30 min daily 3 times a week Get enough sleep At least 7-8 h of sleep per night

After signing the written informed consent, the patient completed the self-report questionnaires, including demographic and clinical information, lymphedema-related behaviors, and the Breast Cancer and Lymphedema Symptom Experience Index. The researcher then carried out a 30-minute face-to-face individual education session with the patient, including self-management strategies in TOLF program handbook, demonstration, and feedback of muscle tightening–breathing and muscle tightening–pumping exercises. The-Optimal-Lymph-Flow program handbook was given to the patient after this session.

Another 4 face-to-face sessions were carried out at 1, 3, 6, and 12 months after the recruitment. During each session, the researcher checked the patient’s lymphedema-related behaviors, collected the Breast Cancer and Lymphedema Symptom Experience Index questionnaires, and measured the patient’s limb circumference. Based on the patient’s condition, the researcher reviewed TOLF program with the patient to ensure that each component was done correctly. Besides, at least 15 telephone contacts were carried out during the intervention period (once a week in the first month and once a month during the following 11 months). During the telephone contacts, the researcher checked the adherence of self-management strategies and answered patients’ questions about daily precautions.

Data Collection DEMOGRAPHIC AND CLINICAL INFORMATION

We applied a self-designed questionnaire to collect demographic and clinical information regarding breast cancer and lymphedema. The demographic information included age, height, weight, education, marital status, and employment status. The clinical information included the dominant arm, surgical site, tumor staging, pathological type, type of surgery, neoadjuvant chemotherapy, chemotherapy, radiotherapy, and hormonal therapy. Patients completed this questionnaire at the recruitment, and the researcher would retrieve and check the information from electronic medical records.

LYMPHEDEMA-RELATED BEHAVIORS

This checklist was used to collect information on participants’ lymphedema-related behaviors.27 The checklist comprises 12 behaviors, including 5 protective behaviors (large muscle exercises, lifting affected limb, muscle tightening–breathing and muscle tightening–pumping exercises, skin care, progressive resistance training) and 7 risk behaviors (air travel, lift heavy objects with affected limb, trim cuticle around nails, carry a shoulder bag over the affected limb, take hot baths or saunas, hold a baby with the affected limb, smoking). Each item is rated on a Likert-type scale from 0 to 3 by frequency of the behavior. The protective behaviors use normal scoring method and ranged from 0 to 15, whereas the risk behaviors use reverse-scoring method and ranged from 0 to 21. The total score is the summation of all behaviors, ranging from 0 to 36. For the three scores, higher scores indicate better behaviors. This checklist was completed 5 times by patients during the whole intervention: at the recruitment and at the 4 face-to-face sessions.

THE BREAST CANCER AND LYMPHEDEMA SYMPTOM EXPERIENCE INDEX

This scale was used to collect information on participants’ lymphedema-related symptom experience.28 It is a self-report instrument consisting of 2 parts. Part I is used to assess the occurrence of lymphedema symptoms, comprising 24 items. By choosing either yes or no to indicate the presence or absence, each symptom can be considered as a categorical variable. Each item is also rated on a Likert-type scale from 0 to 4 by symptom severity. The total score of symptom occurrence is the summation of all item scores ranging from 0 to 96. Higher scores indicate more severe symptom occurrence. Part II is used to assess the distress from the lymphedema symptoms, comprising 6 dimensions: functional, social, sleep, sexuality, emotional and psychological, and self-perception. The total score of symptom distress is the summation of all symptom distress item scores ranging from 0 to 128. Higher scores indicate more severe symptom distress. The total score of symptom experience is the summation of symptom occurrence score and symptom distress score, and higher scores indicate worse symptom experience. The English version was translated and tested in the Chinese population with the Cronbach’s α ranging from .930 to .960 and the test-retest reliability ranging from 0.572 to 0.705.29 This scale was also completed 5 times by patients during the whole intervention: at the recruitment and at the 4 face-to-face sessions.

LIMB CIRCUMFERENCE MEASUREMENTS

We used a well-established protocol for limb circumference measurements to measure both limbs: at hand proximal to the metacarpals, the wrist, and every 4 cm from the wrist to the axilla.2 Circumference differences of 1 to 2 cm between limbs were defined as subclinical lymphedema, differences of 2 to 3 cm were defined as mild, differences of 3 to 5 cm were defined as moderate, and differences of more than 5 cm were defined as severe.10,30 Limb circumference measurements were performed with a flexible, nonelastic tape according to the protocol by a single researcher to avoid errors between the evaluators.

Ethical Aspects of the Study

The research project was approved by a biomedical ethics committee (IRB00001052-15073). All participants in this study provided informed written consent before being included.

Data Analysis

The analysis was performed with IBM SPSS Statistics for Windows, version 21.0 (IBM Corp, Armonk, New York). Frequency distributions and descriptive statistics were calculated to describe the demographic and clinical characteristics of the participants. Scores of lymphedema-related behaviors were presented as means with SD. Scores of lymphedema-related symptom experience and limb circumference differences were presented as medians with quartiles. Generalized estimating equations were used to estimate the effects of the intervention on lymphedema-related behaviors, lymphedema-related symptom experience, and limb circumference differences. α Level was set at P < .05 for all the statistical tests.

Results Demographic and Clinical Characteristics

Among the 41 patients, 1 patient did not complete the 12-month follow-up. Forty-one patients were involved in the analysis. The demographic and clinical characteristics are shown in Table 2.

Table 2 - Demographic and Clinical Characteristics (n = 41) Variables n (%) Variables n (%) Age, mean ± SD, y 58.0 ± 12.48 Pathological type Body mass index, mean ± SD, kg/m2 25.17 ± 3.25  Noninfiltrative carcinoma 1 (2.4) Education  Infiltrative nonspecific carcinoma 37 (90.2)  Primary school 3 (7.3)  Infiltrative specific carcinoma 1 (2.4)  Middle school 7 (17.1)  Data deficient 2 (4.9)  Senior high school/secondary school 13 (31.7) Type of surgery  Junior college 4 (9.8)  Breast conserving surgery 6 (14.6)  College degree or above 14 (34.1)  Partial mastectomy 1 (2.4) Marital status  Total mastectomy 5 (12.2)  Married 40 (97.6)  Modified radical mastectomy 29 (70.7)  Single/divorced/separated 1 (2.4) Neoadjuvant chemotherapy Employment status  No 34 (82.9)  Unemployed 11 (26.8)  Yes 7 (17.1)  Employed 30 (73.2) Chemotherapy Surgery on dominant arm  No 5 (12.2)  No 17 (41.5)  Yes 36 (87.8)  Yes 24 (58.5) Radiotherapy Surgical site  No 16 (39)  Left side 16 (39.0)  Yes 25 (61)  Right side 26 (63.4) Hormonal therapy Staging (Union for International Cancer Control)  No 13 (31.7)  I 5 (12.2)  Yes 28 (68.3)  II 21 (51.2) Lymphedema status  III 13 (31.7)  Mild lymphedema 25 (61.0)  Data deficient 2 (4.9)  Subclinical lymphedema 16 (39.0)
Lymphedema-Related Behaviors

Generalized estimating equations were used to analyze the change of lymphedema-related behaviors among the 5 test points (Table 3 and Table 4). Table 3 presents the comparison results of scores of lymphedema-related risk behaviors, lymphedema-related protective behaviors, and lymphedema-related all behaviors at any 2 different test points. Not all the comparison results are significant, but there is a continuous improvement trend of behaviors in the first 3 months, which is well kept until the end of the follow-up.

Table 3 - Scores of Lymphedema-Related Behaviors and Effect Test of Generalized Estimation Equation (Mean ± SD) Baseline
(N1 = 41) 1-mo Follow-up
(N2 = 41) 3-mo Follow-up
(N3 = 41) 6-mo Follow-up
(N4 = 41) 12-mo Follow-up
(N5 = 40) Wald χ 2 P Lymphedema-related risk behaviors 19.85 ± 1.30a–d 20.46 ± 0.67c,e 20.56 ± 0.50c,e 20.68 ± 0.47a,b,e 20.68 ± 0.53e 27.258 <.001 Lymphedema-related protective behaviors 3.46 ± 2.63a–d 7.83 ± 2.14b,e 8.24 ± 1.76a,e 8.17 ± 1.66e 8.13 ± 2.22e 251.140 <.001 Lymphedema-related all behaviors 23.32 ± 2.83a–d 28.29 ± 2.12b,c,e 28.80 ± 1.69a,e 28.85 ± 1.54a,e 28.80 ± 2.16e 216.958 <.001

aP < .05 versus 1-month follow-up by ANOVA-GEE.

bP < .05 versus 3-month follow-up by ANOVA-GEE.

cP < .05 versus 6-month follow-up by ANOVA-GEE.

dP < .05 versus 12-month follow-up by ANOVA-GEE.

eP < .05 versus baseline by analysis of variance generalized estimation equation (ANOVA-GEE).

Table 4 - Parameter Estimation Results of Generalized Estimation Equation of Scores of Lymphedema-Related Behaviors N Parameter β SE 95% CI Wald χ 2 df P Lower Limit Upper Limit Lymphedema-related risk behaviors 41 Baseline 0a 41 1 mo 0.610 0.1613 0.294 0.926 14.284 1 <.001 41 3 mo 0.707 0.2188 0.278 1.136 10.449 1 .001 41 6 mo 0.829 0.2151 0.408 1.251 14.858 1 <.001 40 12 mo 0.825 0.2169 0.400 1.250 14.462 1 <.001 Lymphedema-related protective behaviors 41 Baseline 0a 41 1 mo 4.366 0.3002 3.777 4.954 211.475 1 <.001 41 3 mo 4.780 0.3133 4.166 5.395 232.790 1 <.001 41 6 mo 4.707 0.3313 4.058 5.357 201.905 1 <.001 40 12 mo 4.626 0.4153 3.812 5.440 124.066 1 <.001 Lymphedema-related all behaviors 41 Baseline 0a 41 1 mo 4.976 0.3475 4.295 5.657 205.030 1 <.001 41 3 mo 5.488 0.4024 4.699 6.277 185.955 1 <.001 41 6 mo 5.537 0.4112 4.731 6.342 181.316 1 <.001 40 12 mo 5.456 0.4801 4.515 6.397 129.157 1 <.001

aNuisance parameter set as 0.

Table 4 presents the β values of lymphedema-related risk behaviors, lymphedema-related protective behaviors, and lymphedema-related all behaviors at each follow-up point. The β values of lymphedema-related risk behaviors at each follow-up point were .610 (P < .001), .707 (P = .001), .829 (P < .001), and .825 (P < .001), which indicated that lymphedema-related risk behavior score was significantly improved at 4 postintervention test points compared with baseline. The β values of lymphedema-related protective behaviors at each follow-up point were 4.366 (P < .001), 4.780 (P < .001), 4.707 (P < .001), and 4.626 (P < .001), which indicated that lymphedema-related protective behavior score was significantly improved at 4 postintervention test points compared with baseline. The β values of lymphedema-related all behaviors at each follow-up point were 4.976 (P < .001), 5.488 (P < .001), 5.537 (P < .001), and 5.456 (P < .001), which indicated that lymphedema-related all behavior score was significantly improved at 4 postintervention test points compared with baseline.

Lymphedema-Related Symptom Experience

Generalized estimating equations were used to analyze the change of lymphedema-related symptom experience among the 5 test points (Table 5 and Table 6). Table 5 presents the comparison results of the scores of lymphedema-related symptom experience, the numbers of lymphedema-related symptoms, the scores of lymphedema-related symptom occurrence, and the scores of lymphedema-related symptom distress at any 2 different test points. A continuous improvement trend showed up within 6 months and was well kept until the end of the intervention. Among the 6 dimensions of lymphedema-related symptom distress, there were statistically significant differences in the scores of functional, social, emotional, and psychological and self-perception among the 5 test points (P < .05), whereas no significant difference was found in sleep and sexuality among different test points (P > .05).

Table 5 - Scores of Lymphedema-Related Symptom Experience and Effect Test of Generalized Estimation Equation [M(Q1, Q3)] Baseline
(N1 = 41) 1-mo Follow-up
(N2 = 41) 3-mo Follow-up
(N3 = 41) 6-mo Follow-up
(N4 = 41) 12-mo Follow-up
(N5 = 40) Wald χ 2 P Symptom experience 25.00 (13.00, 40.00) a–d 20.00 (11.50, 37.50)c–e 21.00 (9.50, 34.00)c,e 20.00 (10.00, 28.00)a,b,e 20.50 (10.00, 26.00)a,e 42.963 <.001 No. of symptoms 7.00 (5.50, 10.00)a–d 7.00 (5.00, 9.00)c–e 6.00 (4.50, 9.00)c–e 6.00 (4.00, 8.50)a,b,e 7.00 (4.00, 8.00)a,b,e 34.909 <.001 Symptom occurrence 11.00 (7.50, 14.00)a–d 10.00 (5.50, 12.50)c–e 9.00 (5.50, 12.00)c,e 10.00 (5.00, 11.00)a,b,e 9.50 (5.00, 11.75)a,e 24.784 <.001 Symptom distress 15.00 (7.00, 26.00)a–d 12.00 (5.50, 22.50)c–e 10.00 (4.00, 22.00)c,e 11.00 (4.50, 17.00)a,b,e 10.50 (4.25, 15.00)a,e 23.411 <.001 Functional 8.00 (3.00, 12.50)a–d 7.00 (3.00, 12.00)c,e 7.00 (2.50, 12.00)e 6.00 (2.50, 10.00)a,e 6.50 (2.25, 10.00)e 11.512 <.001 Social 1.00 (0.00, 2.00)b–d 1.00 (0.00, 2.00)b–d 0.00 (0.00, 1.50)a–e 0.00 (0.00, 1.00)a,e 0.00 (0.00, 1.00)a,b,e 22.168 <.001 Sleep 0.00 (0.00, 0.00) 0.00 (0.00, 0.00) 0.00 (0.00, 0.00) 0.00 (0.00, 0.00) 0.00 (0.00, 0.00) 5.740 .219 Sexuality 0.00 (0.00, 1.00) 0.00 (0.00, 1.00) 0.00 (0.00, 1.00) 0.00 (0.00, 0.00) 0.00 (0.00, 0.00) 1.171 .883 Emotional and psychological 3.00 (0.50, 6.00)a–d 2.00 (0.00, 5.00)e 2.00 (0.00, 6.00)e 1.00 (0.00, 5.50)e 1.00 (0.00, 4.00)e 15.657 .004 Self-perception 1.00 (0.00, 3.00) a–d