Summary A biochemical response index comprising ESR, CEA and CA 15.3 was evaluated in 67 patients with systemic breast cancer treated by chemotherapy; 55 were assessable by UICC criteria and the response index (96% of all UICC assessable patients). Marker changes at 2 and 4 months showed a highly significant correlation with the UICC assessed response at 3 and 6 months (P<0.001); sensitivity 100%, specificity 87%; positive predictive value 85%; negative predictive value 100%.This index was then used to select out truly responsive patients and to prospectively direct their chemotherapy. Twenty-six responding (biochemical/clinical) patients were randomised to discontinue cytotoxics after 6 months and move to maintenance hormones (n = 13) or continue chemotherapy whilst the biochemical markers kept falling or remained within the normal range. Biochemical progression prompted a change of chemotherapy. Continuous chemotherapy in biochemically defined responders was associated with a significant lengthening of remission duration and an improved quality of life and survival. We are now using the index to routinely direct chemotherapy and select out true responders for maintenance chemotherapy.As yet there is no single ideal tumour marker for breast cancer and no established role. Combinations of serum markers, including carcinoembryonic antigen (CEA), have been investigated in an attempt to increase the sensitivity of detecting metastases (Franchimont et al., 1976; Coombes et al., 1988;Cowen et al., 1978;Cove et al., 1979). Very few studies have looked at combinations in measuring response to therapy. On retrospective and prospective analyses we have shown that changes in CEA, ESR and CA 15.3 individually correlate with therapeutic response in patients with metastatic breast cancer treated by first-line hormones (Williams et al., 1990;Robertson et al., 1992). In a prospective study, 93% of patients were assessable with a sensitivity for response of 92% and specificity 82% (Robertson et al., 1992). This present study examines the index in patients receiving chemotherapy and describes our experience of using objective biochemical assessment to direct individual patient therapies. (British Breast Group, 1974). Comparisons were made between assessments by UICC criteria with a minimum duration of remission or statis disease of 6 months and the changes in the biochemical markers measured at 6-8 and 12-16 weeks. In analysing the correlation between biochemical marker movement and UICC assessed response, objective responders and static disease were combined into a non-progressive disease group and compared with those patients that showed disease progression.
Patients and methods
SixtyBiochemical assessment of response The biochemical score was calculated in the same manner as described for the first-line endocrine studies (Williams et al., 1990;Robertson et al., 1992). Namely, any change in marker whilst the patient is on therapy is related to the pretreatment value. A cut-off for each marker of the mean + 2 s.d. of the nor...