November 2004

This is a summary of a presentation made at the 6th International NLN Conference: New Frontiers in Lymphedema Research and Therapy, Sparks/Reno, NV October 2004.

Credible estimates of the incidence of lymphedema are difficult to find. Few consistent surveys appear in the literature as the occurrence of lymphedema is so dependent on patient genetic predisposition, the patient's general health and lymphatic system health, the nature and extent of lymphatic trauma and lack of a standard clinical definition of the condition.

An extensive literature search was performed on over 1900 references to primary and secondary lymphedema. Over 200 references which cited the incidence of lymphedema from any cause were abstracted and the relevant statistics collected into a matrix citing the reference, cause of the lymphedema, appropriate statistic, number of cases, year of treatment, length of study and lymphedema measure used.

The majority of the estimates relate to breast cancer treatment protocols, but the survey includes pelvic and inguinal treatment protocols as well as estimates of primary lymphedema incidence. Differences in lymphedema estimates between references is demonstrated to be caused by changing methods of diagnosis and treatment during studies, lack of standard measurement and grading criteria, delays in development of swelling, therapeutic interventions during the study, physicians' viewpoints and knowledge, inadequate contemporary documentation, selection criteria of patients for study and non-use of actuarial estimates.

Onset of lymphedema is shown to vary as a function of the method of measurement and the causative therapeutic procedure. Toxic effects of radiotherapy do not become fully evident until many years after treatment. Using sensitive lymphoscintigraphic measures of lymphedema, Campisi 2003 shows early effects of breast cancer treatment at 3-6 months (range <1 to 24 months). The delayed effects of radiotherapy are demonstrated [Pierquin 1986] with median onset at 7 (range 2-37) months with surgery alone, 12 (1-52) months with surgery and radiation and 25 (6-156) months with radiation alone. Other researchers demonstrate medians between 1 and 2 years, with maximum times of onset of 3 to 10 years for cohorts with a mix of treatments.

Swelling after breast cancer treatment can occur at a number of sites, and measurement of swelling at one site such as the forearm, upper arm or entire arm and hand results in an underestimation of the incidence of lymphedema. Arm swelling may account for only about half of the patient-reported swelling [Bosompra et al 2002]. Other reported sites include the breast, chest, underarm and back. But measurement of these sites is very difficult, and so remain largely unreported. Breast lymphedema incidences of 70% using measurement of dermal swelling have been demonstrated [Rönkä 2004] while clinical examination detects only 35% in the same cohort.

Changes in the mix of breast cancer surgery and radiotherapy over the last 50 years have resulted in a change in the incidence of lymphedema, since each therapy has a different associated morbidity. Halsted Radical Mastectomies with and without radiotherapy, the standard until the 1970's, resulted in upper limb lymphedema rates of 22-44% without and with radiotherapy. With the ascendancy of the less radical Modified Radical Mastectomy in the 1970's and 1980's lymphedema rates fell to 19-29% without and with radiotherapy [Schünemann & Willich 1997]. The 1990's brought Breast Conserving Surgery from a small percentage to approximately half of the surgeries performed [Yoshimoto et al 2004] with a further drop in upper limb lymphedema rates to 7-10% without and with radiotherapy [Schünemann & Willich 1997].

Breast lymphedema started to receive attention in 1982 with Kissin reporting clinical rates of 8% and Clarke reporting rates of 41% using skin measurements. Recent reports estimate the rates at 1-9% based on subjective reporting [Fehlauer 2003][Højris 2000], 10-19% based on clinical examination [Fehlauer 2003][Goffman 2004] 20-48% [Rönkä 2004][Senofsky 1991] and 30-70% based on skin thickness measurement [Rönkä 2004].

Lower limb lymphedema rates are likewise a strong function of the extent of the surgery and radiation used for treatment of reproductive and pelvic cancers, as well as lower limb melanomas. Whereas there are many different methods commonly used to evaluate upper limb swelling, there are very few methods reported to measure lower limb swelling. Lower limb lymphedema is reported in medical records only when it is severe enough that compression is not adequate, or it causes disablement. Reported lower limb lymphedema ranges from zero [Coblenz 2002] to 60-80% [Balzer 1993][James 1982][Papachristou 1977] with many reports between these extremes.

Lymphedema of the genitals has been reported as 2-5% [Gaarenstroom 2003][Nelson 2004] and 18% (combined with lower limb) [Lieskovsky 1980]. Genital lymphedema among users of pneumatic pumps on the lower limb has been reported at 43% [Boris 1998].

Prevalence of primary lymphedema has been estimated as 1.15/100,000 persons under 20 years [Smeltzer 1985]

This systematic review of lymphedema references results in an estimate of lymphedema incidence overall and by causative factor.