Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Use of a prospective surveillance model to prevent breast cancer treatment-related lymphedema: a single-center experience

Use of a prospective surveillance model to prevent breast cancer treatment-related lymphedema: a... Breast Cancer Res Treat (2016) 160:269–276 DOI 10.1007/s10549-016-3993-7 CLINICAL TRIAL Use of a prospective surveillance model to prevent breast cancer treatment-related lymphedema: a single-center experience 1 2 3 3 • • • • Eun Joo Yang Soyeon Ahn Eun-Kyu Kim Eunyoung Kang 2 1 4 • • Youngmi Park Jae-Young Lim Sung-Won Kim Received: 9 September 2016 / Accepted: 19 September 2016 / Published online: 24 September 2016 The Author(s) 2016. This article is published with open access at Springerlink.com Abstract LE during follow-up. Of these, 126 (19.7 %) were in the Purpose Breast cancer patients undergoing axillary lymph surveillance group and 77 (24.3 %) in the HC group. The node dissection (ALND) are at risk of lymphedema (LE). overall 5-year cumulative incidence of LE (greater than Successful management of LE relies on early diagnosis stage 3) was 25 (95 % CI 15.4–34.6) (6.4 %) in the SLYM using sensitive modalities. In the current study, we group and 48 (95 % CI, 15.4–34.6) (15.1 %) in the HC explored the effectiveness of a surveillance program for group. In the SLYM group, poor compliance had a sig- lymphedema management (SLYM) compared to standard nificant impact on LE incidence (OR = 2.98, P = 0.002). care. Low level of self-monitoring and insight scores were sig- Methods Breast cancer patients who underwent ALND in nificantly related to LE incidence (OR = 1.31, P = 0.025) Seoul National University Bundang Hospital from January after adjusting for age, body mass index, the type of sur- 2008 to December 2015 were included in this prospective gery chosen, radiation therapy, and chemotherapy. With a study. The SLYM commenced in May 2011. The LE cut-off of 29.5 days from operation to the first visit to the outcomes of patients treated prior to initiation of the SLYM LE clinic, the sensitivity was 60 % and the specificity 61 % were compared with those of patients after SLYM in terms of predicting a LE event. implementation. Conclusions Surveillance improves LE prevention com- Results A total of 707 patients were included, 390 in the pared to clinical evaluation. The first visit to the LE clinic SLYM group and 317 in the historical control (HC) group. should be made within 1 month after surgery. In the first A total of 203 patients (28.7 %) had episodes of all-stage year, visits should be made at intervals of less than 3 months. Keywords Lymphedema  Surveillance  Breast cancer Electronic supplementary material The online version of this article (doi:10.1007/s10549-016-3993-7) contains supplementary Compliance material, which is available to authorized users. & Eun Joo Yang graceloves@gmail.com Introduction Department of Rehabilitation Medicine, Seoul National Breast cancer survivors are at increased risk for the University College of Medicine, Seoul National University development of breast cancer-related lymphedema Bundang Hospital, 300 Gumi-dong Bundang-gu, Seongnam-si, Gyeonggi-do 463-707, Korea (BCRL), a chronic, debilitating, and disfiguring condition that is progressive and requires lifelong self-management Medical Research Collaborating Center, Seoul National University Bundang Hospital, Seongnam-si, Korea of symptoms. Lymphedema (LE) is caused by a disruption of the lymphatic system that, in the initial stages, leads to Department of Surgery, Seoul National University Bundang Hospital, Seoul National University College of Medicine, fluid accumulation in the interstitial tissue spaces, and Seongnam-si, Korea eventually manifests clinically as swelling of the arm, breast, shoulder, neck, or torso [1]. Early assessment and Daerim St. Mary’s Hospital, Seoul, Korea 123 270 Breast Cancer Res Treat (2016) 160:269–276 intervention may be important to correct subtle subclinical undergone breast cancer surgery with axillary lymph node LE that, if left untreated, may progress to chronic and dissection (ALND). This was thus a single-center obser- severe LE. Previous studies suggested that regular vational study. Clinical characteristics, demographic surveillance of upper-body morbidities such as LE should information, treatment details, and LE status upon follow- be integrated into the routine postoperative care of women up were all recorded in our clinical database warehouse with breast cancer, as early diagnosis potentially con- system. A surveillance program for LE management tributes to more effective management, and prevention of (SLYM) commenced in May 2011. We sought to identify progression of troublesome conditions [2]. patients with early-stage LE and institute management at Detection and management of early-stage LE may pre- that time. vent progression to chronic disabling disease [3] and may The 767 patients were divided into three groups: one enable cost-effective conservative intervention. Fu et al. [4] historical comparison (HC) group treated prior to imple- found that patient education on the early signs and symp- mentation of the surveillance program (thus, from 2008 toms of upper-body morbidity, in particular disease pro- through 2010; HC group, n = 317); those who were gression, was important. Bioimpedance spectroscopy (BIS) screened from 2010 to 2011 (group B, n = 60); and a assesses changes in extracellular fluid levels and can current group (treated from 2011 through 2015) who were identify such changes in limbs prior to clinical presentation participating in the surveillance program (SLYM group, (thus before the condition becomes non-pitting [fibrotic] n = 390). [5]). A short trial showed that compression garments The primary aim of our study was to compare the event- effectively treated subclinical LE [6]. A systematic review free survival rate between the HC and SLYM groups. This revealed that compression garments and bandages reduced was defined as survival without advanced LE. LE was the volume of cancer-related LE. defined by reference to the guidelines of the International A prospective surveillance model may be useful to Society of Lymphology (ISL) consensus document. These detect BCRL at an early stage, when the opportunities to guidelines feature a staging system based on the amount of reduce risk or slow progression are optimal. A surveillance swelling and the condition of the skin and tissues, and can program would allow healthcare providers to detect BCRL be used to identify disease progression and severity and the symptoms early, affording better opportunities to prevent potential for successful treatment. We defined the event of progression to the subclinical stage and to institute con- interest as advanced LE (greater than stage 3). The sec- tinuous care plans from the inpatient to the outpatient ondary aim was to measure the incidence of LE of any settings [6]. Few rigorous comparative research studies stage during follow-up. have been performed on patients with BCRL, compro- Three physicians defined LE stages by reference to the mising the development of evidence-based assessment of, ISL criteria. Inter-rater agreement between physicians A (a and treatments for, hundreds of thousands of women who psychiatrist, EJY), B (a surgeon, SWK), and C (a surgeon, have, or are at risk for the development of, BCRL. EYK) was analyzed by calculation of kappa coefficients. Therefore, we hypothesized that a surveillance program We used simple randomization to select representative featuring the use of extracellular water (ECW) ratio to samples from the pre-surveillance (May 2010 to April detect subclinical LE might be effective to prevent the 2011) (n = 45) and post-surveillance (May 2012 to April development of LE of stages 2 and 3 after surgery. 2013) (n = 50) periods. Kappa coefficients were 0.87 and The purpose of our study was thus to evaluate the effi- 0.93 in the pre- and post-surveillance period, respectively. cacy of a surveillance program including ECW ratio ICCs were 0.84 and 0.92 in the pre- and post-surveillance measurement in terms of detection of subclinical LE; we period, respectively. Overall agreements (%) were 81.2 and explored whether such detection might prevent the devel- 83.7 % in the pre- and post-surveillance period, opment of advanced LE (stages 2 and 3) after surgery. respectively. Surveillance protocol Patients and methods The SLYM program was implemented in May 2011 to Study design identify high-risk LE patients who would benefit from comprehensive surveillance by a transdisciplinary team, We accessed the database of the Seoul National University Bundang Hospital and extracted records made from Jan- with an emphasis on early detection and prevention of LE. A care plan was initiated immediately after surgery for all uary 1, 2008, through to December 31, 2015. We collected patients who underwent ALND to identify patients at high data on women aged 19–99 years who were newly diag- nosed with stage 1–3 unilateral breast cancer and who had risk of LE. 123 Breast Cancer Res Treat (2016) 160:269–276 271 To detect subclinical LE, we used a reliable and valid up days, were calculated. The cut-off intervals were instrument, the breast cancer and lymphedema symptom 3 months (good compliance B 3 months; poor compliance experience index [7, 8] and a multi-frequency bioelectrical [3 months). impedance analyzer (BIA) (Inbody S10 Biospace, Bio- To assess health-related empowerment, patients were space Co. Ltd., Korea; Model JMW140), according to the asked to complete the Health Education Impact Question- manufacturer’s instructions. The BIA estimates bodily naire (HeiQ) at their first visit. The HeiQ is a well-vali- composition by comparing conductivity differences dated, widely used instrument developed in Australia, between various tissues; these reflect the biological char- containing 40 questions exploring eight different domains acteristics of the tissues. Electrodes are placed at eight (health-directed behavior, positive and active life engage- precisely defined tactile points prior to multi-segmental ment, emotional well-being, self-monitoring and insight, frequency analysis. A total of 30 impedance measurements constructive attitudes and approaches, skill and technique are obtained at six different frequencies (1, 5, 50, 250, 500, acquisition, social integration and support, and health ser- and 1000 kHz) from the following five locations: the right vice navigation) [14]. We analyzed the self-monitoring and and left arms, the trunk, and the right and left legs. Each insight dimension scores (ranges 1.0–4.0) and categorized single-frequency bioimpedance ratio is expressed as a them into low (1.0–2.0) and high level (3.0–4.0). Z /Z ratio, yielding LE index values unaffected limb affected limb [1[9]. Historical control group (HC group) The inclusion criterion for compressive intervention was a diagnosis of subclinical LE. The diagnostic criteria In the HC group, LE was diagnosed when patients com- included subjective LE symptoms and an ECF ratio[1.066 plained of subjective symptoms, or by clinicians of our at the dominant arm, or a ratio[1.106 at the non-dominant multidisciplinary team. Patients who complained and those arm. The garment provided was a 20–30 mm-Hg com- with swelling on the operative site (noted by clinicians) pression sleeve (with a gauntlet) fitted by a trained nurse. If were referred to the LE clinic for further evaluation and the volume increased by [3 %, both a compression gar- management. If LE was confirmed, CDT was applied. ment and education on manual lymphatic massage were prescribed for 4 weeks. Statistical considerations If the clinical LE stage is greater than two, complete decongestive therapy (CDT) performed by specially The actuarial rates of irreversible LE were calculated using trained LE therapists, is considered to be the international the Kaplan–Meier method. All statistics were calculated standard of care [10, 11]. Such therapy seeks to move from the date of surgery. To isolate the effect of the SLYM program, our calculation of the high risk-free survival rate lymphatic fluid to an area from which it may drain, thus reducing swelling. The therapy features manual lymph in the SLYM group included only survival without drainage, progressive active and action-assisted exercises, advanced (over stage 3) LE. Cox’s proportional hazard and compression therapy [12]. model was used to compare the high risk-free survival rates If LE recurred, we checked the status thereof using the between groups after adjusting for age, body mass index, same tools. Patients with stage 0 or 1 subclinical LE the number of cycles of chemotherapy given, and the commenced progressive strengthening exercises; they were radiotherapy field. Only variables that were significant instructed in modified progressive weight-lifting exercises upon univariate analysis were included in multivariate [13]. Patients participated in a supervised program run in analysis. Comparisons between groups were performed the hospital and continued the program at home. with the aid of the log-rank test. P values \0.05 were Strengthening exercises commenced with the aid of a deemed to be statistically significant. All statistical tests thera-band. After checking shoulder strength using a were two-sided and were performed with the aid of SPSS manual muscle strength test, an appropriate thera-band was (version 17.0) software (SPSS Inc., Chicago, IL). prescribed, as was the number of exercise repetitions. This number was progressively increased if no symptoms were evident after three sessions. If the fluid volume decreased Results after intensive treatment, the patients returned to 3-month surveillance until 1 year after surgery, and were then Patient characteristics screened at 6-month intervals. Compliance with the surveillance program was catego- Patient and treatment characteristics are summarized in rized as good or poor by reference to the intervals between Table 1. The median age at the time of surgery was visits to the LE clinic. The times from the day of operation 47.9 years (range, 23–89 years). Breast-conserving surgery to the first visit to the LE clinic, and those between follow- (BCS) was performed on 138 (35.4 %) and mastectomy on 123 272 Breast Cancer Res Treat (2016) 160:269–276 Table 1 Patient and treatment HC group SLYM group, characteristics (N = 317), n (%) (N = 390), n (%) Age at diagnosis (years) 48.6 ± 11.7 (25–82) 47.6 ± 10.7 (23–89) BMI 23.3 ± 9.8 24.3 ± 9.5 Dominant side 132 (41.6 %) 207 (53.0 %) Histopathologic stage I(%) 92 (29.0 %) 109 (28.0 %) II (%) 168 (53.0 %) 215 (55.0 %) III (%) 57 (18.0 %) 66 (17.0 %) Breast surgery BCS 28 (8.8 %) 138 (35.4 %) Mastectomy 289 (91.1 %) 252 (64.6 %) Radiotherapy Not done 41 (12.9 %) 47 (12.0 %) Breast only 174 (54.9 %) 211 (54.0 %) Breast and SCRT (%) 102 (32.2 %) 140 (36.0 %) Chemotherapy Doxetaxel 223 (70.3 %) 280 (71.9 %) Cycle of doxetaxel 5.1 ± 0.3 6.1 ± 0.2 HC historical control, SLYM surveillance program for lymphedema management, BCS breast-conserving surgery, SCRT supraclavicular radiation therapy 252 (64.6 %) patients in the SLYM group, and 28 (8.8 %) (15.1 %) in the HC group (Fig. 1). The median interval and 289 (91.9 %) patients in the HC group. Chemotherapy from surgery to initial swelling in patients with advanced (doxetaxel) was administered to 280 patients (71.9 %) in LE was 25.4 months (95 % CI 18.6–32.1) in the SLYM the SLYM group and 223 (70.3 %) in the HC group. group and 20.7 months (17.0–24.5 months) in the HC Radiotherapy was prescribed for 351 patients (90.0 %) in group. the SLYM group, of whom 211 (54.0 %) received breast or The overall 5-year cumulative incidence of LE (any chest-wall irradiation alone, and for 276 in the HC group stages) was 126 (95 % CI 106.8–145.2 (32.3 %) in the (87.1 %), of whom 174 (54.9 %) received breast or chest- SLYM group and 145 (95 % CI 127.6–162.4) (45.7 %) in the HC group. The median interval from surgery to initial wall irradiation alone. swelling in patients with any stage of LE was 16.9 months Incidence and time course of LE in the SLYM group and 18.4 months in the HC group. Of the 77 affected patients, LE first occurred within 1 year of Initially, 203 patients (28.7 %) were found to have devel- diagnosis in 97 and 80 % of patients, and within 3 years in oped LE of any stage during follow-up. Of these, 126 89 and 62 %, in the SLYM and HC group, respectively. (19.7 %) were in the SLYM group and 77 (24.3 %) in the HC group. Of these patients, 131 (18.5 %) in whom LE had Risk factors for LE resolved or improved at the next follow-up (thus after 6 months) were defined as having reversible LE. Of these Upon multivariate analysis, the following treatment-related patients, 101 (33.6 %) were in the SLYM group and 30 factors were significantly correlated with an increased risk (9.5 %) in the HC group. The initial stage of the first of LE: chemotherapy with docetaxel (hazard ratio (HR) swelling episode was two in 179 patients (25.3 % of all 4.98; P = 0.001) and radiotherapy on breast with SCRT patients; 116 in the SLYM group and 63 in the HC group) (HR 1.20; P = 0.045) (Table 2). The HR of the patho- and three in 24 patients (3.4 %) (Ten in the SLYM group logical stage was 1.52, and lacked significance. and 14 in the HC group). A total of 48 (6.8 %) patients who were initially of stage 2 progressed to stage 3 (15 in the Risk factors for LE in the SLYM group SLYM group and 33 in the HC group). The overall 5-year cumulative incidence of advanced Table 3 shows the results of multivariate logistic regres- LE (greater than stage 3) was 25 (95 % CI 15.4–34.6 sion modeling. Model 1 computed coefficients for a LE (6.4 %) in the SLYM group and 48 (95 % CI 15.4–34.6) event using demographic characteristics such as age, body 123 Breast Cancer Res Treat (2016) 160:269–276 273 Fig. 1 Kaplan–Meier with log- Kaplan-Meier survival estimates rank analysis for the irreversible lymphedema in the surveillance group compared the historical control group 0 20 40 60 Months Historical Surveillance group N at risk 12 mo 24 mo 36 mo 48 mo 60 mo Surveillance 144 94 57 49 1 Probability 97 93 89 80 80 (%) Historical 119 85 47 18 2 Probability 80 66 62 58 58 (%) Table 2 Multivariate analysis of risk factors associated with lym- Model 3 incorporated compliance with the surveillance phedema (N = 707) program in addition to demographic and treatment char- Clinical characteristics HR 95 % CI P value acteristics. Chemotherapy and radiation therapy remained significantly associated with LE events (OR = 5.55, Age (C60 years) 0.03 0.01–0.05 0.014 P = 0.041; OR = 2.01, P = 0.042, respectively). Poor BMI (C25 kg/m ) 1.60 0.69–2.75 0.255 compliance had a significant impact on LE events Dominant side 1.75 0.45–8.63 0.432 (OR = 2.98, P = 0.002). The low level of self-monitoring Histopathologic stage (CII) 1.52 0.73–3.11 0.321 and insight were significantly related to LE events Type of surgery (mastectomy) 1.17 0.42–3.29 0.766 (OR = 1.31, P = 0.025) (Fig. 2). Radiation therapy (breast with SCRT) 2.01 1.05–3.03 0.045 We explored the sensitivity and specificity of a cut-off Chemotherapy (taxel) 4.98 1.93–12.87 0.001 value (days) from operation to the first visit to the LE Surveillance protocol 0.31 0.17–0.56 \0.001 clinic in terms of predicting a LE event. With a cut-off of HR hazard ratio, CI confidence interval, SCRT supraclavicular radi- 29.5 days, the sensitivity was 60 % and the specificity ation therapy 61 %. However, the area under the ROC curve was Cox proportional hazards model 0.6417, thus affording poor diagnostic utility. However, 29.5 days can be used as a reference value (Supplemen- tary Table). mass index (BMI), and the dominant side. None of age, BMI, or the operative site was significantly associated with a LE event. Model 2 incorporated cancer and treatment character- Discussion istics, such as histopathological stage, type of breast sur- gery, radiotherapy, and chemotherapy (docetaxel), in The frequency of advanced LE after breast cancer surgery addition to demographic characteristics. Chemotherapy and was reduced on the introduction of a LE surveillance radiotherapy on breast with SCRT were significantly rela- program. The program afforded ten percentage points of ted to the development of LE events (OR = 5.56, the total reduction in advanced LE among women invited P = 0.032; OR = 2.02, P = 0.243, respectively). to participate. The first visit to the LE clinic should be 0.00 0.25 0.50 0.75 1.00 274 Breast Cancer Res Treat (2016) 160:269–276 Table 3 Multiple logistic regression analysis of patients with breast cancer in surveillance group (N = 390) Variables Model 1 Model 2 Model 3 Age 1.01 (0.98–1.05) 1.01 (0.98–1.05) 1.02 (0.98–1.05) BMI 1.02 (0.97–1.06) 1.02 (0.97–1.05) 1.03 (0.98–1.05) Dominant side 1.05 (0.96–1.16) 1.06 (0.96–1.16) 1.06 (0.97–1.15) Type of surgery (mastectomy) 1.70 (0.86–6.55) 1.67 (0.88–6.11) * * Radiation therapy (breast with SCRT) 2.02 (1.06–3.11) 2.01 (1.05–3.10) * * Chemotherapy (taxel) 5.56 (2.00–9.11) 5.55 (1.99–9.01) Poor compliance (interval of follow-up [3 months) 3.16 (1.36–6.89) Low grade of self-monitoring and insight (score B2) 1.31 (1.03–3.24) Values are odds ratio (95 % CI) Model 1: Age, BMI, and dominant side Model 2: Model 1 ? histopathologic stage, type of surgery, radiation therapy, chemotherapy Model 3: Model 2 ? compliance, self-monitoring, and insight SCRT supraclavicular radiation therapy * P \ 0.05 and is more sensitive and specific than conventional methods [16]. BIS allows earlier recognition of subclinical LE, before the disease is clinically evident upon tape measurement of water displacement. Soran et al. [17] reported that periodic monitoring of women at high risk of LE, using BIS, facilitated early detection and timely intervention. Currently, the lack of data on appropriate assessment frequency and the assessor skills required compro- mise analyses of the cost-effectiveness of prospective 0.00 0.25 0.50 0.75 1.00 surveillance models for breast cancer survivors [18]. One 1-Specificity previous study [17] recorded measurements preoperatively, Model 1 ROC area=0.5538 Model 2 ROC area=0.5906 at 3–6 months after surgery, and then annually for 5 years. Model 3 ROC area=0.6807 Reference Other studies assessed patients five times: preoperatively Fig. 2 ROC curve of lymphedema prediction in surveillance group and every 3 months for the following year [6, 19]. Clearly, such variations affect costs. The incremental benefits made within 1 month after surgery. In the first year, visits afforded by surveillance at 3- versus 4- or 6-month inter- should be made at intervals of less than 3 months. vals warrants study [18]. We explored the sensitivity and Early LE detection is important, and can prevent specificity of time from surgery to the onset of surveillance development of irreversible disease. Surveillance can pre- in terms of predicting a LE event. Surveillance should vent progression of subclinical LE [6]. No evidence-based commence within 1 month of surgery; this should be method for early detection of subclinical LE is yet avail- considered when planning clinical interventions and the able. Relevant clinical trials, using various surveillance follow-up schedule. A previous study also suggested that protocols, are in progress [15]. However, few rigorous the first postoperative assessment visit should take place comparative studies on patients with BCRL have appeared, within 1 month after surgery [20]. compromising the development of evidence-based assess- Of the patient-, treatment-, and disease-related factors, ment and treatment for women who have LE or are at risk chemotherapy with doxetaxel was associated with an of LE development [15]. We found that the hazard ratio for increased risk of LE. Neither age nor BMI correlated with LE development was three times higher in the control than development of a LE event. An earlier study found that the surveillance group. these factors were, in fact, important [21], but another We used BIS to measure extracellular fluid volumes; study found no such correlations [22]. Radiation therapy of BIS records the responses to an applied electrical current, the breast, and SCRT, were significantly associated with Sensitivity 0.00 0.25 0.50 0.75 1.00 Breast Cancer Res Treat (2016) 160:269–276 275 LE development. Regional irradiation is considered to be a self-monitoring. The take-home message is that the LE significant risk factor for LE [23]. Irradiation of the breast surveillance program achieved an absolute reduction of ten and SCRT increased the risk of LE compared to that percentage points in the frequency of advanced LE devel- associated with breast irradiation alone [24]. The question oping after breast cancer surgery. of whether chemotherapy is a risk factor for the develop- Acknowledgments This research was supported by a grant of the ment of lymphedema remains controversial. Patients Korea Health Technology R&D Project through the Korea Health receiving docetaxel-based chemotherapy were at an Industry Development Institute (KHIDI), funded by the Ministry of increased risk of developing LE [21]. Health & Welfare, Republic of Korea (Grant No. L-2010-52-1). Adherence to self-management regimens must be Compliance with ethical standards improved. Adherence to the surveillance program was checked by recording the intervals between visits to the in- Conflict of interest All the other authors declare that they have no hospital program. The self-monitoring and insight scales conflict of interest. capture the ability of an individual to monitor a medical Ethical approval All procedures performed in studies involving condition, triggering physical responses that create insight human participants were in accordance with the ethical standards of and appropriate self-management [25]. We found that the institutional and/or national research committee and with the 1964 adherence and self-monitoring were significantly associ- Helsinki declaration and its later amendments or comparable ethical ated with LE development. Visits should be at less than standards. 5-month intervals. After adjusting for cancer and treatment Informed consent Informed consent was not obtained from all factors, compliance with the surveillance program individual participants included in the study. remained significant. Cancer patients require health-related empowerment to manage the challenges of cancer and to Open Access This article is distributed under the terms of the control their lives [26, 27]. Especially in patients at high Creative Commons Attribution 4.0 International License (http://crea tivecommons.org/licenses/by/4.0/), which permits unrestricted use, risk of LE, fear or denial operates against disease preven- distribution, and reproduction in any medium, provided you give tion and management. All surveillance programs must seek appropriate credit to the original author(s) and the source, provide a to empower patients; this is becoming increasingly link to the Creative Commons license, and indicate if changes were important [28]. made. Caution must be exercised before seeking to generalize our results. Although our choice of a matched cohort may minimize the risk of systematically overestimating the effects of the surveillance program, the nonrandomized References design of our work is a major limitation. Is it possible that the lead time created bias when the frequencies of incident- 1. Cheville AL, McGarvey CL, Petrek JA, Russo SA, Thiadens SR, based LE were calculated? We counted LE events devel- Taylor ME (2003) The grading of lymphedema in oncology clinical trials. Semin Radiat Oncol 13(3):214–225. doi:10.1016/ oping after breast cancer surgery only if the LE was S1053-4296(03)00038-9 diagnosed within a group. For example, in the surveillance 2. Hayes SC, Johansson K, Stout NL, Prosnitz R, Armer JM, group, advanced LE was attributed to breast cancer only if Gabram S, Schmitz KH (2012) Upper-body morbidity after breast the disease was diagnosed early by means of surveillance cancer: incidence and evidence for evaluation, prevention, and management within a prospective surveillance model of care. or was clinically diagnosed while the woman in question Cancer 118(8 Suppl):2237–2249. doi:10.1002/cncr.27467 was in fact in the group. However, for women diagnosed 3. Casley-Smith JR (1995) Alterations of untreated lymphedema early during surveillance, the initial clinical diagnosis and it’s grades over time. Lymphology 28(4):174–185 would have been made at an unknown time within the 4. Fu MR, Chen CM, Haber J, Guth AA, Axelrod D (2010) The effect of providing information about lymphedema on the cog- study period. Thus, the lead time played no role when we nitive and symptom outcomes of breast cancer survivors. Ann calculated LE rates. We believe, therefore, that the calcu- Surg Oncol 17(7):1847–1853. doi:10.1245/s10434-010-0941-3 lations for the two groups are free of such bias. 5. Cornish BH, Thomas BJ, Ward LC, Hirst C, Bunce IH (2002) A new technique for the quantification of peripheral edema with We evaluated the effects of a surveillance program application in both unilateral and bilateral cases. Angiology compared to self-reporting/clinical evaluation. Our results 53(1):41–47 provide a rationale for the future randomized clinical trials 6. Stout Gergich NL, Pfalzer LA, McGarvey C, Springer B, Gerber required to validate our program in patients at high risk of LH, Soballe P (2008) Preoperative assessment enables the early diagnosis and successful treatment of lymphedema. Cancer LE. We conclude that surveillance reduces the rate of LE 112(12):2809–2819. doi:10.1002/cncr.23494 developing after surgery. The magnitude of the benefit was 7. Fu MR, Axelrod D, Cleland CM, Qiu Z, Guth AA, Kleinman R, modest. Most importantly, the apparent benefit afforded Scagliola J, Haber J (2015) Symptom report in detecting breast by optimized patient care may be lost unless patients cancer-related lymphedema. Breast cancer 7:345–352. doi:10. are empowered in terms of self-assessment and 2147/BCTT.S87854 123 276 Breast Cancer Res Treat (2016) 160:269–276 8. Shi S, Lu Q, Fu MR, Ouyang Q, Liu C, Lv J, Wang Y (2016) 19. Springer BA, Levy E, McGarvey C, Pfalzer LA, Stout NL, Psychometric properties of the Breast Cancer and Lymphedema Gerber LH, Soballe PW, Danoff J (2010) Pre-operative assess- Symptom Experience Index: the Chinese version. Eur J oncol ment enables early diagnosis and recovery of shoulder function in Nurs 20:10–16. doi:10.1016/j.ejon.2015.05.002 patients with breast cancer. Breast Cancer Res Treat 9. York SL, Ward LC, Czerniec S, Lee MJ, Refshauge KM, Kil- 120(1):135–147. doi:10.1007/s10549-009-0710-9 breath SL (2009) Single frequency versus bioimpedance spec- 20. Chan DN, Lui LY, So WK (2010) Effectiveness of exercise troscopy for the assessment of lymphedema. Breast Cancer Res programmes on shoulder mobility and lymphoedema after axil- Treat 117(1):177–182. doi:10.1007/s10549-008-0090-6 lary lymph node dissection for breast cancer: systematic review. 10. Lasinski BB, McKillip Thrift K, Squire D, Austin MK, Smith J Adv Nurs 66(9):1902–1914. doi:10.1111/j.1365-2648.2010. KM, Wanchai A, Green JM, Stewart BR, Cormier JN, Armer JM 05374.x (2012) A systematic review of the evidence for complete 21. Norman SA, Localio AR, Kallan MJ, Weber AL, Torpey HA, decongestive therapy in the treatment of lymphedema from 2004 Potashnik SL, Miller LT, Fox KR, DeMichele A, Solin LJ (2010) to 2011. PMR 4(8):580–601. doi:10.1016/j.pmrj.2012.05.003 Risk factors for lymphedema after breast cancer treatment. 11. Ridner SH, Dietrich MS, Kidd N (2011) Breast cancer treatment- Cancer Epidemiol Biomark Prev 19(11):2734–2746. doi:10.1158/ related lymphedema self-care: education, practices, symptoms, 1055-9965.EPI-09-1245 and quality of life. Support care cancer 19(5):631–637. doi:10. 22. Kim M, Kim SW, Lee SU, Lee NK, Jung SY, Kim TH, Lee ES, 1007/s00520-010-0870-5 Kang HS, Shin KH (2013) A model to estimate the risk of breast 12. International Society of L (2013) The diagnosis and treatment of cancer-related lymphedema: combinations of treatment-related peripheral lymphedema: 2013 consensus document of the Inter- factors of the number of dissected axillary nodes, adjuvant national Society of Lymphology. Lymphology 46(1):1–11 chemotherapy, and radiation therapy. Int J Radiat Oncol Biol 13. Schmitz KH, Ahmed RL, Troxel A, Cheville A, Smith R, Lewis- Phys 86(3):498–503. doi:10.1016/j.ijrobp.2013.02.018 Grant L, Bryan CJ, Williams-Smith CT, Greene QP (2009) 23. Sakorafas GH, Peros G, Cataliotti L, Vlastos G (2006) Lym- Weight lifting in women with breast-cancer-related lymphedema. phedema following axillary lymph node dissection for breast N Engl J of Med 361(7):664–673. doi:10.1056/NEJMoa0810118 cancer. Surg Oncol 15(3):153–165. doi:10.1016/j.suronc.2006.11. 14. Osborne RH, Elsworth GR, Whitfield K (2007) The Health 003 Education Impact Questionnaire (heiQ): an outcomes and eval- 24. Coen JJ, Taghian AG, Kachnic LA, Assaad SI, Powell SN (2003) uation measure for patient education and self-management Risk of lymphedema after regional nodal irradiation with breast interventions for people with chronic conditions. Patient Educ conservation therapy. Int J Radiat Oncol Biol Phys Couns 66(2):192–201. doi:10.1016/j.pec.2006.12.002 55(5):1209–1215 15. Ostby PL, Armer JM, Dale PS, Van Loo MJ, Wilbanks CL, 25. Kaur JS (2014) How should we ‘‘empower’’ cancer patients? Stewart BR (2014) Surveillance recommendations in reducing Cancer 120(20):3108–3110. doi:10.1002/cncr.28852 risk of and optimally managing breast cancer-related lym- 26. McCorkle R, Ercolano E, Lazenby M, Schulman-Green D, phedema. J Pers Med 4(3):424–447. doi:10.3390/jpm4030424 Schilling LS, Lorig K, Wagner EH (2011) Self-management: 16. Cornish BH, Chapman M, Thomas BJ, Ward LC, Bunce IH, Hirst enabling and empowering patients living with cancer as a chronic C (2000) Early diagnosis of lymphedema in postsurgery breast illness. CA Cancer J Clin 61(1):50–62. doi:10.3322/caac.20093 cancer patients. Ann N Y Acad Sci 904:571–575 27. Aujoulat I, Marcolongo R, Bonadiman L, Deccache A (2008) 17. Soran A, Ozmen T, McGuire KP, Diego EJ, McAuliffe PF, Reconsidering patient empowerment in chronic illness: a critique Bonaventura M, Ahrendt GM, DeGore L, Johnson R (2014) The of models of self-efficacy and bodily control. Soc Sci Med importance of detection of subclinical lymphedema for the pre- 66(5):1228–1239. doi:10.1016/j.socscimed.2007.11.034 vention of breast cancer-related clinical lymphedema after axil- 28. Maunsell E, Lauzier S, Brunet J, Pelletier S, Osborne RH, lary lymph node dissection; a prospective observational study. Campbell HS (2014) Health-related empowerment in cancer: Lymphat Res Biol 12(4):289–294. doi:10.1089/lrb.2014.0035 validity of scales from the Health Education Impact Question- 18. Cheville AL, Nyman JA, Pruthi S, Basford JR (2012) Cost con- naire. Cancer 120(20):3228–3236. doi:10.1002/cncr.28847 siderations regarding the prospective surveillance model for breast cancer survivors. Cancer 118(8 Suppl):2325–2330. doi:10. 1002/cncr.27473 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Breast Cancer Research and Treatment Pubmed Central

Use of a prospective surveillance model to prevent breast cancer treatment-related lymphedema: a single-center experience

Breast Cancer Research and Treatment , Volume 160 (2) – Sep 24, 2016

Loading next page...
 
/lp/pubmed-central/use-of-a-prospective-surveillance-model-to-prevent-breast-cancer-GfuYtnP9wU

References (30)

Publisher
Pubmed Central
Copyright
© The Author(s) 2016
ISSN
0167-6806
eISSN
1573-7217
DOI
10.1007/s10549-016-3993-7
Publisher site
See Article on Publisher Site

Abstract

Breast Cancer Res Treat (2016) 160:269–276 DOI 10.1007/s10549-016-3993-7 CLINICAL TRIAL Use of a prospective surveillance model to prevent breast cancer treatment-related lymphedema: a single-center experience 1 2 3 3 • • • • Eun Joo Yang Soyeon Ahn Eun-Kyu Kim Eunyoung Kang 2 1 4 • • Youngmi Park Jae-Young Lim Sung-Won Kim Received: 9 September 2016 / Accepted: 19 September 2016 / Published online: 24 September 2016 The Author(s) 2016. This article is published with open access at Springerlink.com Abstract LE during follow-up. Of these, 126 (19.7 %) were in the Purpose Breast cancer patients undergoing axillary lymph surveillance group and 77 (24.3 %) in the HC group. The node dissection (ALND) are at risk of lymphedema (LE). overall 5-year cumulative incidence of LE (greater than Successful management of LE relies on early diagnosis stage 3) was 25 (95 % CI 15.4–34.6) (6.4 %) in the SLYM using sensitive modalities. In the current study, we group and 48 (95 % CI, 15.4–34.6) (15.1 %) in the HC explored the effectiveness of a surveillance program for group. In the SLYM group, poor compliance had a sig- lymphedema management (SLYM) compared to standard nificant impact on LE incidence (OR = 2.98, P = 0.002). care. Low level of self-monitoring and insight scores were sig- Methods Breast cancer patients who underwent ALND in nificantly related to LE incidence (OR = 1.31, P = 0.025) Seoul National University Bundang Hospital from January after adjusting for age, body mass index, the type of sur- 2008 to December 2015 were included in this prospective gery chosen, radiation therapy, and chemotherapy. With a study. The SLYM commenced in May 2011. The LE cut-off of 29.5 days from operation to the first visit to the outcomes of patients treated prior to initiation of the SLYM LE clinic, the sensitivity was 60 % and the specificity 61 % were compared with those of patients after SLYM in terms of predicting a LE event. implementation. Conclusions Surveillance improves LE prevention com- Results A total of 707 patients were included, 390 in the pared to clinical evaluation. The first visit to the LE clinic SLYM group and 317 in the historical control (HC) group. should be made within 1 month after surgery. In the first A total of 203 patients (28.7 %) had episodes of all-stage year, visits should be made at intervals of less than 3 months. Keywords Lymphedema  Surveillance  Breast cancer Electronic supplementary material The online version of this article (doi:10.1007/s10549-016-3993-7) contains supplementary Compliance material, which is available to authorized users. & Eun Joo Yang graceloves@gmail.com Introduction Department of Rehabilitation Medicine, Seoul National Breast cancer survivors are at increased risk for the University College of Medicine, Seoul National University development of breast cancer-related lymphedema Bundang Hospital, 300 Gumi-dong Bundang-gu, Seongnam-si, Gyeonggi-do 463-707, Korea (BCRL), a chronic, debilitating, and disfiguring condition that is progressive and requires lifelong self-management Medical Research Collaborating Center, Seoul National University Bundang Hospital, Seongnam-si, Korea of symptoms. Lymphedema (LE) is caused by a disruption of the lymphatic system that, in the initial stages, leads to Department of Surgery, Seoul National University Bundang Hospital, Seoul National University College of Medicine, fluid accumulation in the interstitial tissue spaces, and Seongnam-si, Korea eventually manifests clinically as swelling of the arm, breast, shoulder, neck, or torso [1]. Early assessment and Daerim St. Mary’s Hospital, Seoul, Korea 123 270 Breast Cancer Res Treat (2016) 160:269–276 intervention may be important to correct subtle subclinical undergone breast cancer surgery with axillary lymph node LE that, if left untreated, may progress to chronic and dissection (ALND). This was thus a single-center obser- severe LE. Previous studies suggested that regular vational study. Clinical characteristics, demographic surveillance of upper-body morbidities such as LE should information, treatment details, and LE status upon follow- be integrated into the routine postoperative care of women up were all recorded in our clinical database warehouse with breast cancer, as early diagnosis potentially con- system. A surveillance program for LE management tributes to more effective management, and prevention of (SLYM) commenced in May 2011. We sought to identify progression of troublesome conditions [2]. patients with early-stage LE and institute management at Detection and management of early-stage LE may pre- that time. vent progression to chronic disabling disease [3] and may The 767 patients were divided into three groups: one enable cost-effective conservative intervention. Fu et al. [4] historical comparison (HC) group treated prior to imple- found that patient education on the early signs and symp- mentation of the surveillance program (thus, from 2008 toms of upper-body morbidity, in particular disease pro- through 2010; HC group, n = 317); those who were gression, was important. Bioimpedance spectroscopy (BIS) screened from 2010 to 2011 (group B, n = 60); and a assesses changes in extracellular fluid levels and can current group (treated from 2011 through 2015) who were identify such changes in limbs prior to clinical presentation participating in the surveillance program (SLYM group, (thus before the condition becomes non-pitting [fibrotic] n = 390). [5]). A short trial showed that compression garments The primary aim of our study was to compare the event- effectively treated subclinical LE [6]. A systematic review free survival rate between the HC and SLYM groups. This revealed that compression garments and bandages reduced was defined as survival without advanced LE. LE was the volume of cancer-related LE. defined by reference to the guidelines of the International A prospective surveillance model may be useful to Society of Lymphology (ISL) consensus document. These detect BCRL at an early stage, when the opportunities to guidelines feature a staging system based on the amount of reduce risk or slow progression are optimal. A surveillance swelling and the condition of the skin and tissues, and can program would allow healthcare providers to detect BCRL be used to identify disease progression and severity and the symptoms early, affording better opportunities to prevent potential for successful treatment. We defined the event of progression to the subclinical stage and to institute con- interest as advanced LE (greater than stage 3). The sec- tinuous care plans from the inpatient to the outpatient ondary aim was to measure the incidence of LE of any settings [6]. Few rigorous comparative research studies stage during follow-up. have been performed on patients with BCRL, compro- Three physicians defined LE stages by reference to the mising the development of evidence-based assessment of, ISL criteria. Inter-rater agreement between physicians A (a and treatments for, hundreds of thousands of women who psychiatrist, EJY), B (a surgeon, SWK), and C (a surgeon, have, or are at risk for the development of, BCRL. EYK) was analyzed by calculation of kappa coefficients. Therefore, we hypothesized that a surveillance program We used simple randomization to select representative featuring the use of extracellular water (ECW) ratio to samples from the pre-surveillance (May 2010 to April detect subclinical LE might be effective to prevent the 2011) (n = 45) and post-surveillance (May 2012 to April development of LE of stages 2 and 3 after surgery. 2013) (n = 50) periods. Kappa coefficients were 0.87 and The purpose of our study was thus to evaluate the effi- 0.93 in the pre- and post-surveillance period, respectively. cacy of a surveillance program including ECW ratio ICCs were 0.84 and 0.92 in the pre- and post-surveillance measurement in terms of detection of subclinical LE; we period, respectively. Overall agreements (%) were 81.2 and explored whether such detection might prevent the devel- 83.7 % in the pre- and post-surveillance period, opment of advanced LE (stages 2 and 3) after surgery. respectively. Surveillance protocol Patients and methods The SLYM program was implemented in May 2011 to Study design identify high-risk LE patients who would benefit from comprehensive surveillance by a transdisciplinary team, We accessed the database of the Seoul National University Bundang Hospital and extracted records made from Jan- with an emphasis on early detection and prevention of LE. A care plan was initiated immediately after surgery for all uary 1, 2008, through to December 31, 2015. We collected patients who underwent ALND to identify patients at high data on women aged 19–99 years who were newly diag- nosed with stage 1–3 unilateral breast cancer and who had risk of LE. 123 Breast Cancer Res Treat (2016) 160:269–276 271 To detect subclinical LE, we used a reliable and valid up days, were calculated. The cut-off intervals were instrument, the breast cancer and lymphedema symptom 3 months (good compliance B 3 months; poor compliance experience index [7, 8] and a multi-frequency bioelectrical [3 months). impedance analyzer (BIA) (Inbody S10 Biospace, Bio- To assess health-related empowerment, patients were space Co. Ltd., Korea; Model JMW140), according to the asked to complete the Health Education Impact Question- manufacturer’s instructions. The BIA estimates bodily naire (HeiQ) at their first visit. The HeiQ is a well-vali- composition by comparing conductivity differences dated, widely used instrument developed in Australia, between various tissues; these reflect the biological char- containing 40 questions exploring eight different domains acteristics of the tissues. Electrodes are placed at eight (health-directed behavior, positive and active life engage- precisely defined tactile points prior to multi-segmental ment, emotional well-being, self-monitoring and insight, frequency analysis. A total of 30 impedance measurements constructive attitudes and approaches, skill and technique are obtained at six different frequencies (1, 5, 50, 250, 500, acquisition, social integration and support, and health ser- and 1000 kHz) from the following five locations: the right vice navigation) [14]. We analyzed the self-monitoring and and left arms, the trunk, and the right and left legs. Each insight dimension scores (ranges 1.0–4.0) and categorized single-frequency bioimpedance ratio is expressed as a them into low (1.0–2.0) and high level (3.0–4.0). Z /Z ratio, yielding LE index values unaffected limb affected limb [1[9]. Historical control group (HC group) The inclusion criterion for compressive intervention was a diagnosis of subclinical LE. The diagnostic criteria In the HC group, LE was diagnosed when patients com- included subjective LE symptoms and an ECF ratio[1.066 plained of subjective symptoms, or by clinicians of our at the dominant arm, or a ratio[1.106 at the non-dominant multidisciplinary team. Patients who complained and those arm. The garment provided was a 20–30 mm-Hg com- with swelling on the operative site (noted by clinicians) pression sleeve (with a gauntlet) fitted by a trained nurse. If were referred to the LE clinic for further evaluation and the volume increased by [3 %, both a compression gar- management. If LE was confirmed, CDT was applied. ment and education on manual lymphatic massage were prescribed for 4 weeks. Statistical considerations If the clinical LE stage is greater than two, complete decongestive therapy (CDT) performed by specially The actuarial rates of irreversible LE were calculated using trained LE therapists, is considered to be the international the Kaplan–Meier method. All statistics were calculated standard of care [10, 11]. Such therapy seeks to move from the date of surgery. To isolate the effect of the SLYM program, our calculation of the high risk-free survival rate lymphatic fluid to an area from which it may drain, thus reducing swelling. The therapy features manual lymph in the SLYM group included only survival without drainage, progressive active and action-assisted exercises, advanced (over stage 3) LE. Cox’s proportional hazard and compression therapy [12]. model was used to compare the high risk-free survival rates If LE recurred, we checked the status thereof using the between groups after adjusting for age, body mass index, same tools. Patients with stage 0 or 1 subclinical LE the number of cycles of chemotherapy given, and the commenced progressive strengthening exercises; they were radiotherapy field. Only variables that were significant instructed in modified progressive weight-lifting exercises upon univariate analysis were included in multivariate [13]. Patients participated in a supervised program run in analysis. Comparisons between groups were performed the hospital and continued the program at home. with the aid of the log-rank test. P values \0.05 were Strengthening exercises commenced with the aid of a deemed to be statistically significant. All statistical tests thera-band. After checking shoulder strength using a were two-sided and were performed with the aid of SPSS manual muscle strength test, an appropriate thera-band was (version 17.0) software (SPSS Inc., Chicago, IL). prescribed, as was the number of exercise repetitions. This number was progressively increased if no symptoms were evident after three sessions. If the fluid volume decreased Results after intensive treatment, the patients returned to 3-month surveillance until 1 year after surgery, and were then Patient characteristics screened at 6-month intervals. Compliance with the surveillance program was catego- Patient and treatment characteristics are summarized in rized as good or poor by reference to the intervals between Table 1. The median age at the time of surgery was visits to the LE clinic. The times from the day of operation 47.9 years (range, 23–89 years). Breast-conserving surgery to the first visit to the LE clinic, and those between follow- (BCS) was performed on 138 (35.4 %) and mastectomy on 123 272 Breast Cancer Res Treat (2016) 160:269–276 Table 1 Patient and treatment HC group SLYM group, characteristics (N = 317), n (%) (N = 390), n (%) Age at diagnosis (years) 48.6 ± 11.7 (25–82) 47.6 ± 10.7 (23–89) BMI 23.3 ± 9.8 24.3 ± 9.5 Dominant side 132 (41.6 %) 207 (53.0 %) Histopathologic stage I(%) 92 (29.0 %) 109 (28.0 %) II (%) 168 (53.0 %) 215 (55.0 %) III (%) 57 (18.0 %) 66 (17.0 %) Breast surgery BCS 28 (8.8 %) 138 (35.4 %) Mastectomy 289 (91.1 %) 252 (64.6 %) Radiotherapy Not done 41 (12.9 %) 47 (12.0 %) Breast only 174 (54.9 %) 211 (54.0 %) Breast and SCRT (%) 102 (32.2 %) 140 (36.0 %) Chemotherapy Doxetaxel 223 (70.3 %) 280 (71.9 %) Cycle of doxetaxel 5.1 ± 0.3 6.1 ± 0.2 HC historical control, SLYM surveillance program for lymphedema management, BCS breast-conserving surgery, SCRT supraclavicular radiation therapy 252 (64.6 %) patients in the SLYM group, and 28 (8.8 %) (15.1 %) in the HC group (Fig. 1). The median interval and 289 (91.9 %) patients in the HC group. Chemotherapy from surgery to initial swelling in patients with advanced (doxetaxel) was administered to 280 patients (71.9 %) in LE was 25.4 months (95 % CI 18.6–32.1) in the SLYM the SLYM group and 223 (70.3 %) in the HC group. group and 20.7 months (17.0–24.5 months) in the HC Radiotherapy was prescribed for 351 patients (90.0 %) in group. the SLYM group, of whom 211 (54.0 %) received breast or The overall 5-year cumulative incidence of LE (any chest-wall irradiation alone, and for 276 in the HC group stages) was 126 (95 % CI 106.8–145.2 (32.3 %) in the (87.1 %), of whom 174 (54.9 %) received breast or chest- SLYM group and 145 (95 % CI 127.6–162.4) (45.7 %) in the HC group. The median interval from surgery to initial wall irradiation alone. swelling in patients with any stage of LE was 16.9 months Incidence and time course of LE in the SLYM group and 18.4 months in the HC group. Of the 77 affected patients, LE first occurred within 1 year of Initially, 203 patients (28.7 %) were found to have devel- diagnosis in 97 and 80 % of patients, and within 3 years in oped LE of any stage during follow-up. Of these, 126 89 and 62 %, in the SLYM and HC group, respectively. (19.7 %) were in the SLYM group and 77 (24.3 %) in the HC group. Of these patients, 131 (18.5 %) in whom LE had Risk factors for LE resolved or improved at the next follow-up (thus after 6 months) were defined as having reversible LE. Of these Upon multivariate analysis, the following treatment-related patients, 101 (33.6 %) were in the SLYM group and 30 factors were significantly correlated with an increased risk (9.5 %) in the HC group. The initial stage of the first of LE: chemotherapy with docetaxel (hazard ratio (HR) swelling episode was two in 179 patients (25.3 % of all 4.98; P = 0.001) and radiotherapy on breast with SCRT patients; 116 in the SLYM group and 63 in the HC group) (HR 1.20; P = 0.045) (Table 2). The HR of the patho- and three in 24 patients (3.4 %) (Ten in the SLYM group logical stage was 1.52, and lacked significance. and 14 in the HC group). A total of 48 (6.8 %) patients who were initially of stage 2 progressed to stage 3 (15 in the Risk factors for LE in the SLYM group SLYM group and 33 in the HC group). The overall 5-year cumulative incidence of advanced Table 3 shows the results of multivariate logistic regres- LE (greater than stage 3) was 25 (95 % CI 15.4–34.6 sion modeling. Model 1 computed coefficients for a LE (6.4 %) in the SLYM group and 48 (95 % CI 15.4–34.6) event using demographic characteristics such as age, body 123 Breast Cancer Res Treat (2016) 160:269–276 273 Fig. 1 Kaplan–Meier with log- Kaplan-Meier survival estimates rank analysis for the irreversible lymphedema in the surveillance group compared the historical control group 0 20 40 60 Months Historical Surveillance group N at risk 12 mo 24 mo 36 mo 48 mo 60 mo Surveillance 144 94 57 49 1 Probability 97 93 89 80 80 (%) Historical 119 85 47 18 2 Probability 80 66 62 58 58 (%) Table 2 Multivariate analysis of risk factors associated with lym- Model 3 incorporated compliance with the surveillance phedema (N = 707) program in addition to demographic and treatment char- Clinical characteristics HR 95 % CI P value acteristics. Chemotherapy and radiation therapy remained significantly associated with LE events (OR = 5.55, Age (C60 years) 0.03 0.01–0.05 0.014 P = 0.041; OR = 2.01, P = 0.042, respectively). Poor BMI (C25 kg/m ) 1.60 0.69–2.75 0.255 compliance had a significant impact on LE events Dominant side 1.75 0.45–8.63 0.432 (OR = 2.98, P = 0.002). The low level of self-monitoring Histopathologic stage (CII) 1.52 0.73–3.11 0.321 and insight were significantly related to LE events Type of surgery (mastectomy) 1.17 0.42–3.29 0.766 (OR = 1.31, P = 0.025) (Fig. 2). Radiation therapy (breast with SCRT) 2.01 1.05–3.03 0.045 We explored the sensitivity and specificity of a cut-off Chemotherapy (taxel) 4.98 1.93–12.87 0.001 value (days) from operation to the first visit to the LE Surveillance protocol 0.31 0.17–0.56 \0.001 clinic in terms of predicting a LE event. With a cut-off of HR hazard ratio, CI confidence interval, SCRT supraclavicular radi- 29.5 days, the sensitivity was 60 % and the specificity ation therapy 61 %. However, the area under the ROC curve was Cox proportional hazards model 0.6417, thus affording poor diagnostic utility. However, 29.5 days can be used as a reference value (Supplemen- tary Table). mass index (BMI), and the dominant side. None of age, BMI, or the operative site was significantly associated with a LE event. Model 2 incorporated cancer and treatment character- Discussion istics, such as histopathological stage, type of breast sur- gery, radiotherapy, and chemotherapy (docetaxel), in The frequency of advanced LE after breast cancer surgery addition to demographic characteristics. Chemotherapy and was reduced on the introduction of a LE surveillance radiotherapy on breast with SCRT were significantly rela- program. The program afforded ten percentage points of ted to the development of LE events (OR = 5.56, the total reduction in advanced LE among women invited P = 0.032; OR = 2.02, P = 0.243, respectively). to participate. The first visit to the LE clinic should be 0.00 0.25 0.50 0.75 1.00 274 Breast Cancer Res Treat (2016) 160:269–276 Table 3 Multiple logistic regression analysis of patients with breast cancer in surveillance group (N = 390) Variables Model 1 Model 2 Model 3 Age 1.01 (0.98–1.05) 1.01 (0.98–1.05) 1.02 (0.98–1.05) BMI 1.02 (0.97–1.06) 1.02 (0.97–1.05) 1.03 (0.98–1.05) Dominant side 1.05 (0.96–1.16) 1.06 (0.96–1.16) 1.06 (0.97–1.15) Type of surgery (mastectomy) 1.70 (0.86–6.55) 1.67 (0.88–6.11) * * Radiation therapy (breast with SCRT) 2.02 (1.06–3.11) 2.01 (1.05–3.10) * * Chemotherapy (taxel) 5.56 (2.00–9.11) 5.55 (1.99–9.01) Poor compliance (interval of follow-up [3 months) 3.16 (1.36–6.89) Low grade of self-monitoring and insight (score B2) 1.31 (1.03–3.24) Values are odds ratio (95 % CI) Model 1: Age, BMI, and dominant side Model 2: Model 1 ? histopathologic stage, type of surgery, radiation therapy, chemotherapy Model 3: Model 2 ? compliance, self-monitoring, and insight SCRT supraclavicular radiation therapy * P \ 0.05 and is more sensitive and specific than conventional methods [16]. BIS allows earlier recognition of subclinical LE, before the disease is clinically evident upon tape measurement of water displacement. Soran et al. [17] reported that periodic monitoring of women at high risk of LE, using BIS, facilitated early detection and timely intervention. Currently, the lack of data on appropriate assessment frequency and the assessor skills required compro- mise analyses of the cost-effectiveness of prospective 0.00 0.25 0.50 0.75 1.00 surveillance models for breast cancer survivors [18]. One 1-Specificity previous study [17] recorded measurements preoperatively, Model 1 ROC area=0.5538 Model 2 ROC area=0.5906 at 3–6 months after surgery, and then annually for 5 years. Model 3 ROC area=0.6807 Reference Other studies assessed patients five times: preoperatively Fig. 2 ROC curve of lymphedema prediction in surveillance group and every 3 months for the following year [6, 19]. Clearly, such variations affect costs. The incremental benefits made within 1 month after surgery. In the first year, visits afforded by surveillance at 3- versus 4- or 6-month inter- should be made at intervals of less than 3 months. vals warrants study [18]. We explored the sensitivity and Early LE detection is important, and can prevent specificity of time from surgery to the onset of surveillance development of irreversible disease. Surveillance can pre- in terms of predicting a LE event. Surveillance should vent progression of subclinical LE [6]. No evidence-based commence within 1 month of surgery; this should be method for early detection of subclinical LE is yet avail- considered when planning clinical interventions and the able. Relevant clinical trials, using various surveillance follow-up schedule. A previous study also suggested that protocols, are in progress [15]. However, few rigorous the first postoperative assessment visit should take place comparative studies on patients with BCRL have appeared, within 1 month after surgery [20]. compromising the development of evidence-based assess- Of the patient-, treatment-, and disease-related factors, ment and treatment for women who have LE or are at risk chemotherapy with doxetaxel was associated with an of LE development [15]. We found that the hazard ratio for increased risk of LE. Neither age nor BMI correlated with LE development was three times higher in the control than development of a LE event. An earlier study found that the surveillance group. these factors were, in fact, important [21], but another We used BIS to measure extracellular fluid volumes; study found no such correlations [22]. Radiation therapy of BIS records the responses to an applied electrical current, the breast, and SCRT, were significantly associated with Sensitivity 0.00 0.25 0.50 0.75 1.00 Breast Cancer Res Treat (2016) 160:269–276 275 LE development. Regional irradiation is considered to be a self-monitoring. The take-home message is that the LE significant risk factor for LE [23]. Irradiation of the breast surveillance program achieved an absolute reduction of ten and SCRT increased the risk of LE compared to that percentage points in the frequency of advanced LE devel- associated with breast irradiation alone [24]. The question oping after breast cancer surgery. of whether chemotherapy is a risk factor for the develop- Acknowledgments This research was supported by a grant of the ment of lymphedema remains controversial. Patients Korea Health Technology R&D Project through the Korea Health receiving docetaxel-based chemotherapy were at an Industry Development Institute (KHIDI), funded by the Ministry of increased risk of developing LE [21]. Health & Welfare, Republic of Korea (Grant No. L-2010-52-1). Adherence to self-management regimens must be Compliance with ethical standards improved. Adherence to the surveillance program was checked by recording the intervals between visits to the in- Conflict of interest All the other authors declare that they have no hospital program. The self-monitoring and insight scales conflict of interest. capture the ability of an individual to monitor a medical Ethical approval All procedures performed in studies involving condition, triggering physical responses that create insight human participants were in accordance with the ethical standards of and appropriate self-management [25]. We found that the institutional and/or national research committee and with the 1964 adherence and self-monitoring were significantly associ- Helsinki declaration and its later amendments or comparable ethical ated with LE development. Visits should be at less than standards. 5-month intervals. After adjusting for cancer and treatment Informed consent Informed consent was not obtained from all factors, compliance with the surveillance program individual participants included in the study. remained significant. Cancer patients require health-related empowerment to manage the challenges of cancer and to Open Access This article is distributed under the terms of the control their lives [26, 27]. Especially in patients at high Creative Commons Attribution 4.0 International License (http://crea tivecommons.org/licenses/by/4.0/), which permits unrestricted use, risk of LE, fear or denial operates against disease preven- distribution, and reproduction in any medium, provided you give tion and management. All surveillance programs must seek appropriate credit to the original author(s) and the source, provide a to empower patients; this is becoming increasingly link to the Creative Commons license, and indicate if changes were important [28]. made. Caution must be exercised before seeking to generalize our results. Although our choice of a matched cohort may minimize the risk of systematically overestimating the effects of the surveillance program, the nonrandomized References design of our work is a major limitation. Is it possible that the lead time created bias when the frequencies of incident- 1. Cheville AL, McGarvey CL, Petrek JA, Russo SA, Thiadens SR, based LE were calculated? We counted LE events devel- Taylor ME (2003) The grading of lymphedema in oncology clinical trials. Semin Radiat Oncol 13(3):214–225. doi:10.1016/ oping after breast cancer surgery only if the LE was S1053-4296(03)00038-9 diagnosed within a group. For example, in the surveillance 2. Hayes SC, Johansson K, Stout NL, Prosnitz R, Armer JM, group, advanced LE was attributed to breast cancer only if Gabram S, Schmitz KH (2012) Upper-body morbidity after breast the disease was diagnosed early by means of surveillance cancer: incidence and evidence for evaluation, prevention, and management within a prospective surveillance model of care. or was clinically diagnosed while the woman in question Cancer 118(8 Suppl):2237–2249. doi:10.1002/cncr.27467 was in fact in the group. However, for women diagnosed 3. Casley-Smith JR (1995) Alterations of untreated lymphedema early during surveillance, the initial clinical diagnosis and it’s grades over time. Lymphology 28(4):174–185 would have been made at an unknown time within the 4. Fu MR, Chen CM, Haber J, Guth AA, Axelrod D (2010) The effect of providing information about lymphedema on the cog- study period. Thus, the lead time played no role when we nitive and symptom outcomes of breast cancer survivors. Ann calculated LE rates. We believe, therefore, that the calcu- Surg Oncol 17(7):1847–1853. doi:10.1245/s10434-010-0941-3 lations for the two groups are free of such bias. 5. Cornish BH, Thomas BJ, Ward LC, Hirst C, Bunce IH (2002) A new technique for the quantification of peripheral edema with We evaluated the effects of a surveillance program application in both unilateral and bilateral cases. Angiology compared to self-reporting/clinical evaluation. Our results 53(1):41–47 provide a rationale for the future randomized clinical trials 6. Stout Gergich NL, Pfalzer LA, McGarvey C, Springer B, Gerber required to validate our program in patients at high risk of LH, Soballe P (2008) Preoperative assessment enables the early diagnosis and successful treatment of lymphedema. Cancer LE. We conclude that surveillance reduces the rate of LE 112(12):2809–2819. doi:10.1002/cncr.23494 developing after surgery. The magnitude of the benefit was 7. Fu MR, Axelrod D, Cleland CM, Qiu Z, Guth AA, Kleinman R, modest. Most importantly, the apparent benefit afforded Scagliola J, Haber J (2015) Symptom report in detecting breast by optimized patient care may be lost unless patients cancer-related lymphedema. Breast cancer 7:345–352. doi:10. are empowered in terms of self-assessment and 2147/BCTT.S87854 123 276 Breast Cancer Res Treat (2016) 160:269–276 8. Shi S, Lu Q, Fu MR, Ouyang Q, Liu C, Lv J, Wang Y (2016) 19. Springer BA, Levy E, McGarvey C, Pfalzer LA, Stout NL, Psychometric properties of the Breast Cancer and Lymphedema Gerber LH, Soballe PW, Danoff J (2010) Pre-operative assess- Symptom Experience Index: the Chinese version. Eur J oncol ment enables early diagnosis and recovery of shoulder function in Nurs 20:10–16. doi:10.1016/j.ejon.2015.05.002 patients with breast cancer. Breast Cancer Res Treat 9. York SL, Ward LC, Czerniec S, Lee MJ, Refshauge KM, Kil- 120(1):135–147. doi:10.1007/s10549-009-0710-9 breath SL (2009) Single frequency versus bioimpedance spec- 20. Chan DN, Lui LY, So WK (2010) Effectiveness of exercise troscopy for the assessment of lymphedema. Breast Cancer Res programmes on shoulder mobility and lymphoedema after axil- Treat 117(1):177–182. doi:10.1007/s10549-008-0090-6 lary lymph node dissection for breast cancer: systematic review. 10. Lasinski BB, McKillip Thrift K, Squire D, Austin MK, Smith J Adv Nurs 66(9):1902–1914. doi:10.1111/j.1365-2648.2010. KM, Wanchai A, Green JM, Stewart BR, Cormier JN, Armer JM 05374.x (2012) A systematic review of the evidence for complete 21. Norman SA, Localio AR, Kallan MJ, Weber AL, Torpey HA, decongestive therapy in the treatment of lymphedema from 2004 Potashnik SL, Miller LT, Fox KR, DeMichele A, Solin LJ (2010) to 2011. PMR 4(8):580–601. doi:10.1016/j.pmrj.2012.05.003 Risk factors for lymphedema after breast cancer treatment. 11. Ridner SH, Dietrich MS, Kidd N (2011) Breast cancer treatment- Cancer Epidemiol Biomark Prev 19(11):2734–2746. doi:10.1158/ related lymphedema self-care: education, practices, symptoms, 1055-9965.EPI-09-1245 and quality of life. Support care cancer 19(5):631–637. doi:10. 22. Kim M, Kim SW, Lee SU, Lee NK, Jung SY, Kim TH, Lee ES, 1007/s00520-010-0870-5 Kang HS, Shin KH (2013) A model to estimate the risk of breast 12. International Society of L (2013) The diagnosis and treatment of cancer-related lymphedema: combinations of treatment-related peripheral lymphedema: 2013 consensus document of the Inter- factors of the number of dissected axillary nodes, adjuvant national Society of Lymphology. Lymphology 46(1):1–11 chemotherapy, and radiation therapy. Int J Radiat Oncol Biol 13. Schmitz KH, Ahmed RL, Troxel A, Cheville A, Smith R, Lewis- Phys 86(3):498–503. doi:10.1016/j.ijrobp.2013.02.018 Grant L, Bryan CJ, Williams-Smith CT, Greene QP (2009) 23. Sakorafas GH, Peros G, Cataliotti L, Vlastos G (2006) Lym- Weight lifting in women with breast-cancer-related lymphedema. phedema following axillary lymph node dissection for breast N Engl J of Med 361(7):664–673. doi:10.1056/NEJMoa0810118 cancer. Surg Oncol 15(3):153–165. doi:10.1016/j.suronc.2006.11. 14. Osborne RH, Elsworth GR, Whitfield K (2007) The Health 003 Education Impact Questionnaire (heiQ): an outcomes and eval- 24. Coen JJ, Taghian AG, Kachnic LA, Assaad SI, Powell SN (2003) uation measure for patient education and self-management Risk of lymphedema after regional nodal irradiation with breast interventions for people with chronic conditions. Patient Educ conservation therapy. Int J Radiat Oncol Biol Phys Couns 66(2):192–201. doi:10.1016/j.pec.2006.12.002 55(5):1209–1215 15. Ostby PL, Armer JM, Dale PS, Van Loo MJ, Wilbanks CL, 25. Kaur JS (2014) How should we ‘‘empower’’ cancer patients? Stewart BR (2014) Surveillance recommendations in reducing Cancer 120(20):3108–3110. doi:10.1002/cncr.28852 risk of and optimally managing breast cancer-related lym- 26. McCorkle R, Ercolano E, Lazenby M, Schulman-Green D, phedema. J Pers Med 4(3):424–447. doi:10.3390/jpm4030424 Schilling LS, Lorig K, Wagner EH (2011) Self-management: 16. Cornish BH, Chapman M, Thomas BJ, Ward LC, Bunce IH, Hirst enabling and empowering patients living with cancer as a chronic C (2000) Early diagnosis of lymphedema in postsurgery breast illness. CA Cancer J Clin 61(1):50–62. doi:10.3322/caac.20093 cancer patients. Ann N Y Acad Sci 904:571–575 27. Aujoulat I, Marcolongo R, Bonadiman L, Deccache A (2008) 17. Soran A, Ozmen T, McGuire KP, Diego EJ, McAuliffe PF, Reconsidering patient empowerment in chronic illness: a critique Bonaventura M, Ahrendt GM, DeGore L, Johnson R (2014) The of models of self-efficacy and bodily control. Soc Sci Med importance of detection of subclinical lymphedema for the pre- 66(5):1228–1239. doi:10.1016/j.socscimed.2007.11.034 vention of breast cancer-related clinical lymphedema after axil- 28. Maunsell E, Lauzier S, Brunet J, Pelletier S, Osborne RH, lary lymph node dissection; a prospective observational study. Campbell HS (2014) Health-related empowerment in cancer: Lymphat Res Biol 12(4):289–294. doi:10.1089/lrb.2014.0035 validity of scales from the Health Education Impact Question- 18. Cheville AL, Nyman JA, Pruthi S, Basford JR (2012) Cost con- naire. Cancer 120(20):3228–3236. doi:10.1002/cncr.28847 siderations regarding the prospective surveillance model for breast cancer survivors. Cancer 118(8 Suppl):2325–2330. doi:10. 1002/cncr.27473

Journal

Breast Cancer Research and TreatmentPubmed Central

Published: Sep 24, 2016

There are no references for this article.