The trusted source for
healthcare information and
ABSTRACT & COMMENTARY
Synopsis: Recurrent gastric cancer after primary resection is both common and difficult to treat. One reason is that routine follow-up strategies often miss early recurrence, and, by the time of diagnosis, effective treatment is not available. In the current analysis, 3 tumor markers (CEA, CA 19-9, and CA 72-4) were shown, in aggregate, to be highly sensitive indicators of disease recurrence (87%). Furthermore, CA 72-4 proved also to be highly specific (97%), in that elevations were highly predictive of recurrent disease. Thus, routine use of these tumor markers would likely result in earlier recognition of recurrent gastric cancer, and this may have clinical value if effective treatments can be developed for this earlier stage of recurrent disease.
Source: Marrelli D, et al. Am J Surg. 2001;181:16-19.
Despite curative intent surgery, gastric cancer commonly recurs. The diagnosis of recurrent disease is often delayed and subsequent treatment is frequently ineffective. Earlier recognition of recurrence may allow more effective therapy. The purpose of the current study was to evaluate the effectiveness of the serum tumor markers CEA, CA 19-9, and CA 72-4 in the early diagnosis of recurrence of gastric cancer. One hundred thirty-three patients who had potentially curative surgery at the University of Sienna (Italy) between the years 1988-1995 were included in this analysis. Serum tumor markers were obtained preoperatively, 1 week after surgery, and at regular intervals thereafter.
Preoperatively, CEA was elevated in 16% of patients, CA 19-9 in 35%, and CA 72-4 in 20%. The overall combined sensitivity was 51%. Of the 133 patients included in the study, 75 (56%) had tumor recurrence with a mean time interval of 18 ± 15 months. Of the 75 patients, CEA was elevated at the time of recurrence in 33 (sensitivity 44%), CA 19-9 was elevated in 42 (sensitivity 56%), and CA 72-4 was elevated in 38 (sensitivity 51%). An increase in at least 1 of the markers was observed in 65 patients (overall sensitivity, 87%). The increase in tumor markers preceded (by an average of 5 months) clinical diagnosis of recurrence in 46 cases. In 19, tumor markers became elevated at a time when recurrence was evident by clinical parameters.
In the 58 patients who did not develop recurrent disease during this analysis, CA 72-4 remained low in all but 2 (specificity, 97%). On the other hand, both CEA and CA 19-9 yielded a higher number of false-positives (CEA in 12 cases and CA 19-9 in 15 cases), and, therefore a lower specificity (79% and 74%, respectively). The positive predictive value of CEA, CA 19-9, and CA 72-4 for tumor recurrence was 73%, 74%, and 95% respectively.
Marrelli and colleagues concluded that the combined assay of CEA, CA 19-9, and CA 72-4 may be useful for diagnosis of recurrence of gastric cancer, but they warned that only CA 72-4 positivity should be considered a specific predictor of tumor recurrence.
COMMENT by William B. Ershler, MD
The striking findings in this report are 2-fold. First, all patients with preoperative elevations who relapsed did so with an associated reappearance of the tumor markers. Of those who recurred with normal markers (n = 10), all had normal markers prior to tumor resection. Accordingly, one potential application of this finding is to develop a follow-up strategy in which patients with positive preoperative markers simply have periodic reassessment of the panel of markers and proceed with more thorough clinical radioimaging studies only at the time of reappearance of an elevated level. On the other hand, the follow-up protocol for patients showing negative preoperative levels would include periodic scanning or other instrumental examinations in addition to the tumor markers.
The second important finding is the relatively high specificity of the CA 72-4 assay, as has been reported by other groups.1,2 Whereas both CEA and CA 19-9 frequently yielded false-positive results, CA 72-4 was shown to be highly specific (97%), and a pathological increase in serum levels was highly predictive of tumor recurrence.
Although inherently appealing, the value of early recognition of recurrent gastric cancer remains to be established. That is because treatment programs for recurrent disease have not typically been successful. Possibly, early recognition will identify some patients with resectable disease, particularly those with local recurrence at the gastric stump. Furthermore, although currently available chemotherapy regimens have met with little success, it is possible that an earlier diagnosis, with presumably a lower tumor burden, might offer more favorable treatment outcomes. One might envision, for example, initiation of an aggressive treatment regimen at the first sign of elevated CA 72-4. Certainly, this is a question that may effectively be answered by careful clinical investigation.
1. Guadagni F, et al. Cancer Res. 1992;52:1222-1227.
2. Heptner G, et al. Scand J Gastroenterol. 1989;24: 745-750.