The diagnostic unreliability of classic physical signs of lymphedema

Arjun Jayaraj, MD, Seshadri Raju, MD, Corbin May, MS, and Nicholas Pace, MS, Jackson,Miss. Journal of Vascular Surgery: Venous and Lymphatic Disorders. 2019.

Objective: In most communities, the diagnosis of lymphedema in the lower extremity currently rests on clinical signs. Lymphoscintigraphy, which is objective, is performed infrequently to confirm the clinical suspicion. Given absence of a curative option for lymphedema, it is essential to obtain an accurate diagnosis before committing the patient to lifelong conservative therapy. The aim of this study was to evaluate the diagnostic accuracy of clinical signs in comparison to lymphoscintigraphy, the current objective standard.

Methods: Retrospective review of contemporaneously collected data of 636 consecutive limbs with swelling (318 left, 318 right) that underwent initial evaluation during a 12-month period between 2016 and 2017 was performed. All limbs were assessed for classic clinical signs of lymphedema including dorsal hump of the foot, square toes, Kaposi-Stemmer sign, and nonpitting edema. Lymphoscintigraphy was routinely performed for objective evaluation. The 436 patients who underwent the study were scored positive for lymphedema on the basis of transit time delay for the radioisotope in minutes, presence of dermal backflow, presence of collateral channels, intensity of uptake in the main channel and lymph nodes, number of nodes in the groin, and presence of popliteal nodes. Analysis was carried out to determine sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of the clinical signs in determining whether a patient had lymphedema. In addition, regression analysis was carried out to evaluate the predictive value of different clinical signs in determining lymphedema. Patients with positive clinical signs but with normal findings on lymphoscintigraphy who did not have a medical cause for swelling underwent workup to determine a possible venous cause.

Results: Of 636 limbs with swelling, 436 (69%) underwent lymphoscintigraphy, the findings of which were normal in 178 (41%) and abnormal in 258 (59%). Of the 636 swollen limbs, 96 (15%) had clinical signs of lymphedema; 95% had dorsal hump, 37% had square toes, 32% had presence of Kaposi-Stemmer sign, and 12% had nonpitting edema. Of these 96, lymphoscintigraphy was performed on 66 (69%); 45 of 66 (68%) patients with clinical signs were positive for lymphedema; the remaining 32% were normal. Conversely, among 258 swollen limbs with abnormal findings on lymphoscintigraphy, only 45 (17%) had one or more of the clinical signs. Sensitivity and specificity of clinical signs in predicting lymphoscintigraphy-confirmed lymphedema were 17% and 88%, respectively. Overall accuracy was 47%. Of the clinical signs, only the Kaposi-Stemmer sign was a significant predictor of lymphedema (odds ratio, 7.9; P¼ .02). In patients with positive clinical signs but normal findings on lymphoscintigraphy, venous obstruction was the most common cause of swelling.

Conclusions: Clinical signs of lymphedema appear to be unreliable in making a correct diagnosis of lymphedema in one-third of patients. Conversely, in lymphoscintigraphy-confirmed lymphedema, only 17% had positive clinical signs. Of the clinical signs, only Kaposi-Stemmer sign has some predictability in determining lymphoscintigraphy-confirmed lymphedema. Venous obstruction is the most common cause of clinical signs in patients without lymphedema. Routine use of lymphoscintigraphy is recommended in patients to make an objective diagnosis of lymphedema.