Following institutional review board approval, we conducted a preliminary trial (http://ClinicalTrials.gov
identifier: NCT01003951) to assess the safety and efficacy of acupuncture in patients with breast cancer-related lymphedema (BCRL). Patients were identified and screened and enrolled in the study following informed consent and their oncologists' approval. Inclusion criteria were: women aged ≥18 years with unilateral lymphedema, defined as ≥2 cm in circumference difference between affected and unaffected arms resulting from surgery and/or radiation therapy for breast cancer,6
and a clinical diagnosis of lymphedema for at least 6 months and no more than 5 years. The 6-month postsurgery time frame allowed for any surgically related nonlymphedema swelling to subside, and a cap of 5 years, which we believed to capture the broadest range of cases, has been used as a time in previous studies.6,39
Women with previous acupuncture treatment for lymphedema or currently using diuretics were ineligible. Additional exclusion criteria amended to the original protocol included metastatic cancer, history of autoimmune or fibroproliferative disorders, history of primary lymphedema or bone marrow transplant, or current treatment with corticosteroids or myelosuppressive or stimulatory drugs. Patients receiving maintenance treatment for lymphedema such as exercise, massage, or compression garments were included, as these interventions are standard practice at Memorial Sloan-Kettering and because chronic lymphedema typically persists or recurs despite such interventions.
Each patient received acupuncture treatment for lymphedema twice weekly for 30 minutes over 4 consecutive weeks (see Acupuncture Treatment Technique section). No further acupuncture treatments were given or measurements taken after the 4-week treatment period, but monthly follow-up calls were conducted for 6 months following completion of treatment to document any side effects or complications and to determine patients' self-reported lymphedema status.
Our trained research assistants asked participants the following questions, which were aimed to identify any adverse events potentially from acupuncture: 1) Have you experienced any reactions or adverse events since the last time we talked? 2) Are you currently receiving treatment for your lymphedema: exercise, arm wrapping, manual lymph drainage, pneumatic compression, use of compression garments, other? 3) How is your lymphedema since the last time we talked? Same? Better? Worse?
A Simon's 2-stage minimax design was used to discriminate between a 5% response rate and a response rate of 20%, which would merit further study. The design yields an 80% probability of a positive result if the response rate is ≥20%, and a 95% probability of a negative result if the true response rate is ≤5%. Acupuncture was considered promising if at least 3 of 27 study participants achieved a reduction of ≥30% in comparative arm circumference. This goal was determined on the basis of our clinical experience in the absence of any consistent definition found in the clinical literature.
Our pilot study met its goal after 4 responses occurred in 9 treated patients, and the data were published.21
We continued the trial to evaluate a total of 33 patients for whom data are reported here.
Acupuncture Treatment Technique
Traditional Chinese medicine (TCM) acupuncture treatment was performed by licensed Memorial Sloan-Kettering Cancer Center (MSKCC) Integrative Medicine Service staff acupuncturists with at least 5 years' clinical experience in treating patients with cancer. Alcohol swabs were applied prior to insertion of sterile single-use filiform needles (32-36 gauge; 30-40 mm in length, Tai Chi brand, made in China and distributed by Lhasa OMS, Weymouth, MA) that penetrate 5-10 mm into the skin. A total of 14 needles were inserted: 4 in both affected and unaffected limbs, 2 in acupuncture points on both legs, and 2 in unilateral points on the torso. Selected acupuncture points () were stimulated manually by gentle rotation of the needles with lift and thrust. The acupuncturists did not intentionally seek to achieve a de qi sensation.
Acupuncture points used to treat lymphedema in 30-minute sessions twice weekly for 4 consecutive weeks.
Specific acupuncture points used in this study were determined on the basis of historical context, the published literature, and the consensus of our experienced group of MSKCC staff acupuncturists.18–20,34,37
Many of these acupoints are used to treat pain, weakness, and motor impairment; others are traditionally used to drain “dampness,” a TCM concept similar to edema.
Two-point circumference measurements of the affected and unaffected arms were performed before and after each treatment session using our previously published methods.6,21
This is the most common approach used to diagnose upper-extremity lymphedema, and sequential measurements over time that include pretreatment measures are considered to be optimal clinically.40,41
Our ipsilateral and contralateral measurements at both baseline and weekly during the course of treatment also controlled for the inherent biases that occur with straight arm circumference measures. These include weight gain or loss and arm dominance and other predispositions.6
Data also indicated that 2-point measures with a 5% circumference change threshold have high sensitivity and specificity (80% and 71%, respectively), close correlation to limb volume calculations using water displacement, and compare favorably with the sensitivity and specificity of multiple-point measurements with high test-retest and interrater reliability.42–44
In addition, circumference measurements provide reliable calculations for differences >0.4 cm.45
Further, water displacement techniques have practical difficulties such as space issues, spillage, and the need for sterile technique for each measurement.
Measurements were performed by trained research assistants 10 cm above (upper arm) and 5 cm below (lower arm) the olecranon process using nonstretch tape measures. We used our published technique21
to calculate the extent of BCRL by determining the difference in circumference of the affected and unaffected arms at the site with the largest baseline difference. Percent change in lymphedema was defined as:
This formula enabled the determination of response to treatment and adjusted the outcomes relative to baseline arm circumference differences. The mean change in the extent of these differences was also analyzed as a continuous variable using the Student t test for paired data.