Session Item

To assess the rate and predictors of lower extremity lymphedema (LEL) after radiotherapy (RT) following radical prostatectomy (RP) + pelvic lymph node dissection (PLND) for prostate cancer.  

This is a cross sectional study on patients (pts) treated with adjuvant or salvage RT after RP+PLND and with a minimum 2-year follow-up. LEL was defined as a volume difference of ≥10% between limbs determined using circumferential measurements with a flexible non-stretch tape at study follow up examination. The onset of LEL was then retrospectively assessed. The following potential predictors of the endpoint were investigated at logistic regression: age (continuous); body mass index (BMI) (continuous); exercise level according to the International Physical Activity Questionnaire (low vs medium/high); smoking history (yes vs no); cigarette pack/year (continuous); hypertension (yes ns no); vascular comorbidity (yes vs no); diabetes (yes vs no); PLND (yes vs no); number of examined nodes (continuous); whole pelvis radiotherapy (WPRT) (yes vs no); time between RP and RT (continuous); planning target volume (PTV) volume (continuous); PTV/BMI (continuous). Statistical significance was claimed for p values <0.05.

101 pts accepted study enrollment and were examined/interviewed. Median time from surgery to RT was 36.1 months (mths) (IQR: 15.0-68.3 mths) and median time from RT to the date of study examination was 51.1 months (IQR: 36.8-65.3 mths). All pts underwent RP & prostatic fossa RT, 70 pts (69.3%) underwent PLND with the removal of a median number of 12.5 nodes (IQR: 8-17.2) and 69 pts underwent WPRT (68.3%). 14 pts developed LEL (13.9%, 95%CI: 8.4-21.9%). Most of the pts (92.8%) developed unilateral LEL. Three pts dated the onset of LEL before RT while in the remaining pts LEL occurred after RT. The median time from RT to LEL was 4 mths (IQR: -0.5/17.3). The latest event was recorded 25.4 months after RT completion. At multivariable analysis (MVA) diabetes mellitus (OR=32.8, p=0.02), the time between surgery and RT (OR=0.966, p=0.039) and exercise (OR=0.03, p=0.002) were independently correlated to the risk of LEL. Smoking had a borderline effect (OR=4.8, p=0.052). The number of examined nodes was highly correlated to LEL at univariate analysis (OR=1.066, p=0.025) but disappeared at MVA (p=0.719). Interestingly, the distribution of examined nodes was statistically different between pts with low (median N=12) vs medium/high (N=5) exercise (p=0.034) suggesting exercise level being more a consequence of the extent of pelvic surgery rather than a cause of LEL.

Clinically detectable LEL involves a minority of pts after RT; diabetes is a predisposing factor while awaiting RT delivery has a protective effect favoring salvage over adjuvant RT. The role of physical exercise along with the extent of pelvic surgery needs to be investigated prospectively.