(Gleason, 1977). The remainder of the tissue was snap frozen in liquid nitrogen and finally pulverised in a micro-dismembrator (Braun AG, Melsungen, Germany). The powdered tissue was then lyophilised to dryness and stored at -70°C until needed. For growth factor studies 70 mg of the lyophilised material was used whereas only 50 mg were required for steroid measurements.Extraction of growth factors Lyophilised samples were resuspended in 1 ml buffer (50 mmol 1 Tris, 5 mol 1' of phenyl-methylsulphonyl fluoride, 2 mmol I 1' of EDTA pH 7.4) containing 1,000 counts of either '251I-EGF or '25I-TGFx to monitor manipulative losses. The samples were homogenised with a glass/glass Dounce Homogeniser and left on ice for 1 h to equilibrate. The homogenates were then centrifuged, the supernatants saved and the resultant pellets were resuspended in 2 ml of enzyme mixture (230 U ml-' of collagenase, 125 U ml-' of hyaluronidase in 150 mmol 1' of NaCI) and incubated overnight at 37C; these conditions were found to be optimal and ensured maximum growth factor recovery. The digested pellets were subsequently sonicated (six cycles of 20s with 1 min cooling interval at an amplitude of 22 ,; MSE soniprep 150) the resultant mixture was centrifuged as before, the supernatant saved and the pellet was resuspended in 2 ml Tris buffer. Homogenisation, centrifugation and resuspension were then repeated twice with the supernatants always being saved. The supernatants were then pooled and the recoveries were assessed. The mean recovery of '251I-EGF added at the beginning of the extraction procedure was 72 ± 6% whilst that of 125I-TGFo was 69 ± 8%. The. supernatants were finally snap frozen, lyophilised to dryness and stored at -700 until required.