Influence of implantation site and track geometry on the extrusion rate and pharmacology of testosterone implants

BACKGROUND Implantation of testosterone pellets under the lateral abdominal wall skin is an old but popular and effective form of androgen replacement therapy. Extrusion of one or more pellets remains the most frequent adverse event. OBJECTIVE To determine whether an alternative implantation site (hip) and/or track geometry (two vs. four tracks) would reduce extrusion rates compared with the standard of a four‐track abdominal site. Additionally, the study aimed to evaluate the effects of site and track geometry on other adverse effects (bruising, infection) and the pharmacology of testosterone release from the implants. DESIGN A prospective, parallel‐group unmasked study design in a single centre. The primary objective was to evaluate sites, with evaluation of track geometry a subordinate objective made necessary by anatomical differences. Accordingly, androgen deficient men requiring testosterone implantation with the standard dose (four 200 mg pellets) were randomized into one of three groups (ratio 1 : 1 : 2): a four‐track abdomen site (standard), a two‐track abdomen site or a two‐track hip site. The pharmacological substudy was to evaluate the impact of site and track geometry on testosterone implant pharmacology by monthly hormone assays following implantation. PATIENTS Two hunded and forty‐six implantation procedures involving 96 androgen deficient men. MEASUREMENTS The primary end‐point, extrusion rate per procedure, and secondary end‐points (bruising or infection post procedure) were evaluated prospectively by self‐report from the participants, and verified when they returned next for implantation. The pharmacology substudy involved monthly blood sampling for hormone (total and free testosterone, LH, FSH) measurements. RESULTS The extrusion rate was significantly higher [odds ratio (OR) = 2·6, 95% confidence interval (CI) 1·1–7·1] for the hip site (15/125, 12%) compared with the abdominal site (6/121, 5%). Track geometry made no significant difference (OR = 1·05, 95% CI 0·2–5·4) to the extrusion rate. No secondary end‐points (bruising, infection) were significantly different according to either site or track geometry. One operator who performed the implant procedures had significantly less primary and secondary adverse events than the other operators (P = 0·006). Neither implantation site, nor track geometry influenced pharmacokinetics [peak plasma total and free testosterone concentrations and net hormone release (area‐under‐curve, AUC)] or pharma