Ms Brunelle is a certified cardiovascular and pulmonary specialist (CCS) and a certified lymphedema therapist (CLT). Dr Barrio is a Fellow of the American College of Surgeons. We reviewed with interest the clinical practice guideline (CPG) developed by Levenhagen et al, on behalf of the Oncology Section of the American Physical Therapy Association (APTA). We applaud the effort to provide a framework for diagnosis of upper quadrant lymphedema secondary to cancer (SUQL). As stated in the article, the incidence of SUQL varies widely but is high. SUQL has a significant, negative impact on quality of life, and the burden of lifelong care is substantial. There is a distinct need for research agreement on diagnostic criteria and definitions, as the Lymphedema Research Program at Massachusetts General Hospital (MGH) has championed in the past.1 This lack of consensus remains an ongoing challenge for patients, clinicians, and researchers, as highlighted by the guideline authors. This CPG attempts to clarify the evidence base for diagnostic methods for SUQL. We thank the authors for inviting feedback from a wider audience, and we acknowledge that this CPG was not piloted in clinical practice. As oncology clinicians and researchers, we have several comments, and we hope that our input will be helpful to shape a timely revised edition of this CPG. We recommend, based on current research, a CPG recommending prospective surveillance for SUQL, including subjective and objective measures, at baseline and regular intervals. The CPG authors state that “diagnostic definitions of lymphedema that require a consistent volume increase, and thus being at ISL [International Society of Lymphology] stage II or above, have the potential to hamper efforts to intervene at the early stages.” Throughout the history of the Lymphedema Research Program at MGH, clinically apparent lymphedema has been defined as a 10% relative volume change (RVC) or weight-adjusted change (WAC) (as compared to baseline) in research publications.2 The MGH team has developed and validated the RVC and WAC formulas for patients undergoing unilateral and bilateral breast cancer surgery, respectively. These formulas consider baseline measures, measurements of contralateral arm volume, and, for the WAC, patient weight when the contralateral arm volume cannot be used for patients with bilateral surgery.2,3 Clinically, MGH has always included subjective measures as well as clinical examination as part of their screening program, as successful early diagnosis incorporates all 3 elements. We do not agree that this definition of lymphedema hampers efforts to intervene at the early stages, as this does not indicate that RVCs < 10% do not warrant examination and potentially intervention. Researchers and clinicians have discussed and incorporated edema of 3% to 10% into their research,4,5 defining this amount of edema as “subclinical edema” or “mild edema” or by other terms. There is good movement toward emphasizing screening and early intervention in the literature. MGH has gone to great effort to develop and sustain a successful prospective surveillance program, which has been highlighted in detail recently.6 This program screens patients longitudinally, starting with a preoperative baseline, throughout and beyond their treatment for breast cancer. Incorporating knowledge of well-established risk factors to define a patient's risk helps guide decision making for patients who are at risk of edema progression in a subclinical stage, where symptoms may be apparent without volume change, or in early cases of edema (RVC or WAC of 3% or more).4 We do not believe that a definition of clinically apparent breast cancer–related lymphedema (BCRL) at 10% hampers early intervention efforts, unless providers are not mindfully screening for early signs and/or symptoms of lymphedema or considering individual risk. There is work to be done to better define how we as clinicians best define and intervene for the population with edema of 3% to < 10%. A comprehensive prospective surveillance program is a current recommendation by ISL,7 the National Lymphedema Network (NLN),8 and the American Society of Breast Surgeons (ASBrS),9 and we strongly recommend a prospective surveillance program be incorporated into any updated CPG document. We recommend, based on current evidence, a CPG highlighting the need for baseline measures to allow for accurate diagnosis of SUQL. The guideline authors discuss baseline limb volume measurements or bio-impedance spectroscopy in several areas of the paper; however, they make no clear recommendation regarding the importance of baseline measurements in the diagnosis of SUQL. The authors mention a paper by Stout et al,5 which rightly noted that preoperative assessment enables the early diagnosis and successful treatment of lymphedema. The MGH Lymphedema Research Program has developed formulas for limb volume relative to baseline volume, which are utilized in research and clinically.2,3 These formulas may be used with limb volume calculated from circumferential measures or from water displacement or perometry. RVC and WAC compare volume changes relative to baseline, regardless of body weight changes. The MGH team recently analyzed the effect on diagnosis of BCRL by using or not using a baseline limb volume measurement. Without a true preoperative baseline measurement, lymphedema was mis-diagnosed up to 50% of the time.10 ISL, the ASBrS Panel, and the NLN recommend preoperative baseline measurements.7-9 We recommend that a baseline measure be incorporated into any updated CPG document. We recommend that absolute measures or ratios of limb volume be used only when there is no recommended baseline measurement—and then only in conjunction with subjective data and clinical examination. It should be clarified that any absolute volume ratios, differences, or RVC or WAC formulas may be used with any measurement of limb volume (volume calculated from tape measurement circumferences, perometry, or water volume displacement), and such differences/ratios are not isolated to circumferential measurements taken by tape measure. We were encouraged to read the recommendation, “Circumferential measurement taken at any single site along the upper extremity, and specifically a ≥ 2 cm difference, should not be utilized as a diagnostic criterion for upper extremity lymphedema due to poor accuracy.” Despite its lack of validity and reliability, this method continues to be utilized in research and practice, and we applaud the authors for clarifying this point. The authors make the following recommendations: “A volume ratio of 1.04 may be indicative of upper extremity lymphedema,” with a Grade B recommendation strength, and “Calculated volume differential between sides (≥200 mL) will help rule in lymphedema,” with a Grade B recommendation strength. Although we note the authors’ careful wording, these methods are inherently erroneous ways to diagnose lymphedema, given known arm asymmetry, differences in body habitus, and lack of incorporation of a baseline measurement. The MGH team examined baseline asymmetry in arm volume in a cohort of 1028 patients undergoing treatment for breast cancer. Of these patients, 2.9% had arm volume asymmetry greater than 10%, and 28.3% had arm asymmetry greater than 5%.10 Using a volume ratio of 1.04, a volume differential of ≥ 200 mL, or even an absolute arm volume difference of 5% or 10% to diagnose lymphedema—in absence of a baseline measure, or without clinical and subjective data—is erroneous. This might need to be done, in conjunction with subjective measures and clinical examination when a baseline does not exist; however, we believe that this CPG recommendation should be revisited and clarified. We disagree with the following CPG recommendation: “Perometry may be used for assessment of volume but not as a diagnostic tool for upper extremity lymphedema. This technique is limited in clinical utility.” We strongly recommend that this statement be clarified. According to the authors, this statement is based on “Expert Opinion,” not on the evidence base. We feel compelled to clarify the evidence on perometry. As previously discussed, the MGH Lymphedema Research Team has established feasibility and sustainability of a prospective surveillance program that utilizes perometry.6 This program has been in place since 2006, and a perometer was chosen in this setting given its efficiency, high interrater reliability, and ability to facilitate the maintenance of a database for clinical and research purposes. Perometry fits extremely well into clinic workflow, and, although the perometer does have a space need, a suitable, small space on a treatment table was secured. To our knowledge, perometry has gained popularity for these reasons, and prospective screening programs have been developed at Memorial Sloan Kettering Cancer Center, The University of Texas MD Anderson Cancer Center, and Beth Israel Deaconess Medical Center, to name a few, obtaining and relying on perometry as their objective measure. We acknowledge that the equipment is expensive, and each facility should make clinical utility decisions independently given individual needs and resources. One should not conclude that clinical utility is limited, however, when many such prospective surveillance programs utilizing perometry, with established feasibility and sustainability, are in practice. Perometry is a reliable, valid, and accurate method of diagnosing lymphedema. The original perometer validation study was done in 1997 by Stanton et al,11 and therefore was not included in the CPG’s summary of research, as it was outside the date range of 2000 to 2015. The perometer also was validated as a measurement of hand volume by Lee in 2011,12 which was not included in the CPG’s review. To be considered as having high diagnostic accuracy, perometry should identify those with lymphedema as having lymphedema (sensitivity) and identify those without lymphedema as not having lymphedema (specificity). In a well-cited 2009 study, Stout et al5 prospectively measured limb volume using perometry in 196 women, preoperatively and at 3-month intervals after surgery. If an increase > 3% in upper limb volume developed compared with the preoperative volume, a diagnosis of lymphedema was made, and a compression garment intervention was prescribed for 4 weeks. In commenting on this study, Stout13 stated, “The Perometer device is sensitive enough to capture this slight change in limb volume, although it occurs before clinical signs of swelling are apparent. Based on our study, the Perometer was effectively able to detect a minimum 3% limb volume change when a preoperative measurement was taken and patients were monitored at regular intervals over time.” In a more recent study, again not included in the CPG given that it was outside of the 2000 to 2015 range, Dylke et al14 assessed the arms of women treated for breast cancer with and without a previous diagnosis of lymphedema, as well as healthy controls, using lymphoscintigraphy, bioimpedance spectro-scopy, and perometry. The reference standard was dermal backflow determined from lymphoscintigraphy imaging, and, for those with established dermal backflow, perometry had a sensitivity of 96% and a specificity of 78%. As previously discussed, the screening program at MGH utilizes perometry, and patients are measured at preoperative baseline, as well as every 3 to 8 months throughout follow-up. MGH has screened more than 4500 women at this point, incorporating subjective measures and clinical examination alongside perometry screening. Knowledge of known risk factors is incorporated into referrals for lymphedema treatment versus further surveillance when volume exceeds that of known baseline. Patients with RVC or WAC of > 3% to ≤ 10% within 3 months of surgery and patients with RVC or WAC of 5% to ≤ 10% are at higher risk of progression of edema than those without edema.4 We know from our screening programs, as well as the research base, that perometry is sensitive in picking up small volume changes, and it is a large reason why we screen, diagnose, and treat edema early. Finally, the authors state that there are discrepancies in the literature when attempting to use perometry as a diagnostic method. The discrepancies they discuss are not specific to perometry, but rather to diagnostic criteria for lymphedema diagnosis, including utilizing a 200 mL difference or a 10% difference. They cite 3 MGH studies2,15,16 to describe these perometry discrepancies; however, this indicates a lack of understanding of the literature. The MGH team found that a relative volume ratio of volumes of the affected and unaffected arm at time of measurement and baseline were the most appropriate to quantify asymmetry and temporal changes. They developed the RVC formula, which is not significantly associated with hand dominance, age, or time since diagnosis.2 The authors insinuate that there was a disadvantage of perometry when the team later reported an absolute change in arm volume (200 mL) correlated to a relative arm volume change between 2.9% and 15.7% based on body size. In fact, after validating the RVC formula, MGH analysis did demonstrate that an absolute change in arm size of 200 mL or 2 cm corresponds to a considerable range of relative arm volume changes depending on initial body size. This was an issue of determining that using absolute values to diagnose lymphedema is inaccurate; it has nothing to do with perometry. We do not recommend utilizing an absolute increase of 200 mL or 2 cm as diagnostic criteria for lymphedema diagnosis. Temporal change in the absolute volume of the unaffected arm is correlated with body size, but relative arm volume change is not. Therefore, relative arm volume change can be used to accurately quantify lymphedema in a range of body sizes and with changes in body size over time. Again, it should be noted that any discussion of the use of a 200 mL or 3%, 5%, or 10% absolute difference to diagnose lymphedema applies to all measures, not just to perometry. This is a discussion about diagnostic criteria, not measurement methodology; therefore, these discrepancies should not be attributed to perometry, as was the case in the CPG document. The authors correctly stated that “this highlights the need to consider underlying dominant versus nondominant arm asymmetry and reinforces the importance of baseline measurements,” as we previously discussed. We would not wish to dissuade practitioners from using perometry as a diagnostic method, and we would like the authors to revisit this in any updated CPG document. We would encourage clinicians and researchers to consider all measurement methods for screening and diagnosis, so that each facility can make an informed decision to incorporate a method that is reliable, valid, feasible, and sustainable for that facility. No one mode of measurement will suit every facility. We recommend clarification of the evidence base for bioimpedance spectroscopy (BIS). Prior to longstanding measurement methods, such as water displacement and circumferential measures, BIS is listed first, and it is the most comprehensively reviewed measurement method in this CPG document. There have been several studies over the past decade examining the use of BIS for screening and diagnosis in breast cancer-related lymphedema. However, there is limited evidence regarding its diagnostic accuracy for edema < 10%. The studies quoted by the CPG authors carry significant limitations. The MGH Lymphedema Research Team recently published a review paper on BIS,17 concluding that bioimpedance is an accurate diagnostic tool for preexistent lymphedema; however, there is not enough evidence to conclude that BIS detects BCRL earlier than other measurement techniques. Studies do not provide long-term follow-up or adequate clinical diagnosis to prove that this was in fact early detection and not a false positive result. A number of studies report high rates of false negative for BIS.18-20 The MGH group stated that the concept of bio-impedance detecting swelling before it is visible by volume change is innovative, and, if proven accurate, has the potential to enhance early diagnosis of BCRL. They recommended longitudinal studies to allow the validation of BIS as an early detector for BCRL. They concluded that it is possible bioimpedance can be used in conjunction with volumetric techniques to yield maximum benefit for all patients, and that this idea should also be explored in further research. A study by Fu et al,21 which formed the basis for the authors’ main recommendation in the CPG for BIS use for lymphedema diagnosis, has significant limitations. Fu concluded that: “if no preoperative assessment is available, an L-Dex score of > 7.1 is diagnostic of lymphedema.” The company that developed the BIS states that L-Dex scores < 10 without a baseline measurement are within normal range, as the authors also state within the CPG. Fu et al did not perform baseline measurements of limb volume on their patients and used a lymphedema diagnostic criteria of 200 mL limb volume difference between the affected and unaffected arm. This is a diagnostic technique fraught with error and a major limitation of the study.16 The document did not discuss a further prospective validation study of BIS by Barrio et al in 2015.19 This study used volume displacement to attempt to validate BIS in patients with early-stage breast cancer who are risk for BCRL. This study showed poor correlation between bioimpedance and volume displacement. L-Dex performed poorly as a predictor of lymphedema, with only 4 out of 25 patients with an abnormal L-Dex progressing to lymphedema (sensitivity 31%, positive predictive value 18%). This study used baseline measures of volume displacement and BIS, and relative arm volume difference from baseline, as diagnostic criteria for lymphedema diagnosis. Finally, BIS cannot be used to diagnose late-stage lymphedema. Bioimpedance analysis is a measure of extracellular fluid, and, as such, it does not therefore capture tissue changes, such as fi-brosis or adipose infiltration, seen in late-stage lymphedema.21 Regardless, the authors state that BIS “may be utilized as a diagnostic tool” in later stage lymphedema. This information is not included in the summary chart of recommendations for later-stage lymphedema. We recommend revisiting the evidence base for BIS and updating it for a timely revision of the CPG document. Although BIS may be a good tool for the screening and diagnosis of BCRL, more research is needed, especially in the populations with early- and late-stage BCRL. The concept of BIS detecting swelling before it is visible by volume change is innovative, and, if proven accurate, has the potential to enhance early diagnosis of BCRL. It is the belief of all of the authors of this letter that lymphedema diagnosis, management, and treatment will ultimately require a multimodality approach and the development of multidisciplinary teams and centers. It is important that as clinicians and researchers, we are moving in this direction, and this is worthy of mention in such a document. No one health care professional can develop a prospective surveillance model in a vacuum. For example, just as Physical and Occupational Therapy and Radiation Oncology have teamed up at Massachusetts General Hospital, so have Physical and Occupational Therapy, Surgical Oncology, Plastic Surgery, Body Imaging, and Nuclear Medicine at Beth Israel Deaconess Medical Center. These relationships can be and are synergistic. The CPG document does not encourage such an approach. The end of the document discusses some of the most objective diagnostic criteria available for lymphedema including lymphography, magnetic resonance imaging, and lymphoscintigraphy. The document does not encourage these modalities, and, in fact, cites limitations, including “need to refer to another for and of these and therefore methods are not recommended to be incorporated into clinical practice for diagnosing SUQL at this time.” for example, is most performed by and can with diagnosis and guide management, and also may help areas that might benefit most from A of health care many are an interest in this and the development of relationships these will allow to the diagnosis and of lymphedema. As such, a approach should be encouraged in such a CPG document. Although are included throughout the for each recommendation, 2 does not cite specific for and It would be helpful for the to be able to refer to the as a guide for this table to how studies were and attributed to develop the CPG. In we applaud the attempt to a CPG document for the diagnosis of upper quadrant lymphedema secondary to cancer. There is a body of research in this and this CPG should be as there are more recent available than those in the and several of the do not accurately the current evidence base. should be based on clinical practice. We strongly recommend the need for prospective including baseline measures, and of measurement method based on and of in to updated of We would the to be in a timely of this CPG document.
No takes yet. Share an insight, caveat, or question.
Brunelle et al. (2018) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: