The effect of microacinar growth patterns on survival after radical su
rgery for rectal cancer was assessed in 138 consecutive patients. All
had Dukes and Jass staging and a median follow-up of 95 (range 11-180)
months. Tumour acini were classified according to size (microacinar,
28; macroacinar, 110). Patients with microacinar tumours had a signifi
cantly reduced 5-year survival rate compared with those with macroacin
ar lesions (43 and 68 per cent respectively, P = 0.004). When examined
in the presence of other histological factors, acinar size had indepe
ndent prognostic value (relative risk 2.37, P = 0.006). This was true
even when the Dukes and Jass classifications were included in the mode
l (relative risk 2.08, P = 0.02 and 1.95, P = 0.03 respectively). Hist
ological classification of rectal tumours into microacinar and macroac
inar types adds prognostic value to the Dukes and Jass classifications
and may be a more objective criterion than conventional histological
grading.