In 430 stage I-II breast cancer patients the cost-benefit of investiga
tions during follow-up have been studied. Median follow-up time was 8
years and 128 patients had relapsed, 91 with metastatic disease. High
costs of routine chest X-ray, limited skeletal X-ray and bone scan exa
minations were associated with low incidence of diagnosed relapses not
suspected otherwise. In the eight blood analyses examined, increases
of more than 10 mm/h in erythrocyte sedimentation rate (ESR), 20 U/l i
n gamma-glutamyltransferase (GT) or 60 U/l in alkaline phosphatase (AL
P) resulted in a combined sensitivity of 55% and specificity of 91% fo
r relapses with distant metastases. Elevation of at least two blood te
sts gave a combined sensitivity of 31% and a specificity of 98%. The i
mportance of using individual reference values in screening for recurr
ences is emphasised. Symptomatic relapse or relapse detected at interv
al visits were not independent prognostic factors. The blood tests ALP
, ESR and GT were strong predictors of survival measured from relapse
which increase their legitimacy in follow-up. A more frequent follow-u
p for patients with 4+ involved nodes is proposed: three visits annual
ly the first 5 years vs. two visits annually for the others. We conclu
de that history, clinical examination, ALP, ESR and GT are sufficient
as a baseline screening for relapse in breast cancer patients.