Abstract
Background: CA19-9 was evaluated as a prognostic marker for colorectal cancer and whether it could be helpful in addition to surveillance using CEA was also tested. Materials and Methods: Serum CA19-9 levels were measured preoperatively in 1109 patients and monitored at 3-month intervals for the first 2 years postoperatively, and at 6-month intervals thereafter in 700 patients. Results: Preoperative high CA19-9 was independent prognostic factor for recurrence. Among patients with recurrence, 21.4% had a high postoperative CA19-9. High postoperative CA 19-9 levels were more likely in patients with high preoperative levels. Postoperative CA19-9 increased more in patients with a peritoneal recurrence than in those with liver metastasis (p=0.002). Among patients with recurrence, CA19-9 increased in 7.8% of the patients with a normal follow-up CEA. Conclusion: Postoperative CA19-9 was more frequently elevated when peritoneal recurrence occurred. Data on CA19-9 levels provided 7.8% of additional information in predicting recurrence in this study.
Tumor markers have been used as prognostic factors in the management of cancer patients. Generally, they are not useful for diagnosis because of their low specificity and sensitivity. However, such markers are more useful than screening in the follow-up of patients with malignant diseases. Carcinoembryonic antigen (CEA) is the most commonly used tumor marker in patients with colorectal cancer, and elevated concentrations of CEA have been reported to be associated with poor postoperative prognosis in such patients (1-5). The tumor marker carbohydrate antigen (CA) 19-9 is often used in combination with CEA to manage patients with colorectal cancer, even though CA19-9 has been reported to be a less sensitive marker (6). Although the American Society of Clinical Oncology guidelines suggest that there is insufficient evidence for using CA19-9 in the management of patients with colorectal cancer (7), several authors have described the prognostic significance of CA19-9 (3, 8-11). There are also reports indicating the possible usefulness of CA19-9 in monitoring recurrence rates (12-14), whereas others have shown contradictory results (15-17). In addition, the utility of CA19-9 in predicting prognosis is controversial, as is the clinical usefulness of simultaneous evaluations of CEA and CA19-9. In the present study, whether serum CA19-9 could be a prognostic factor in predicting the prognosis and tumor recurrence for patients with colorectal cancer was evaluated, and whether it could offer any useful information in addition to serum CEA levels during postoperative surveillance was also adressed.
Patients and Methods
The study population consisted of 1,109 patients who received regular postoperative surveillance among 1,452 who underwent curative resection for colorectal cancers. Preoperative serum CEA and CA19-9 levels were measured in the Department of Surgery of Kyungpook National University Hospital, Deagu, Korea between June 1996 and June 2006. The median follow-up period was 48 months (range 1-156). Serum CEA and CA19-9 concentrations were measured, with CEA concentrations >7.0 ng/mL and CA19-9 concentrations >37 U/mL being regarded as elevated. Patients with a synchronous metastatic disease or undergoing palliative resection and those with carcinoma in situ, familial adenomatous polyposis, or pathology other than an adenocarcinoma were excluded, as were patients with T1 stage cancer treated by local excision.
Patients were followed-up routinely at 3-month intervals for the first 2 years, and at 6-month intervals thereafter. At each visit, CEA levels were assayed, a full history was taken and a physical examination was performed. Colonoscopy was performed within 6 months to 1 year following surgery and then 3 years later. Chest radiographs and abdominopelvic computer tomography (CT) scans were performed at 6 months and then at yearly intervals. For 700 patients, serum CA19-9 was regularly followed up postoperatively with the same schedule as for the surveillance of serum CEA. CT or positron emission tomography (PET) scans were performed on patients with increased serum CEA concentrations, or when they were symptomatic. Diagnosis of recurrence was confirmed by biopsies or from histopathology of resected specimens. Otherwise, recurrence was documented from the first clinical or radiological signs of a disease that showed an unrelenting course leading to tumor progression and/or death. Criteria for the establishment of recurrent disease included histological confirmation, palpable disease, or disease evident radiologically with subsequent clinical progression and supportive biochemical data, particularly increased CEA levels. The time to recurrence was determined as the date of the follow-up visit at which recurrence was discovered, or by reviewing hospital records for patients who were admitted.
Clinicopathological characteristics of patients with normal and high preoperative serum carbohydrate antigen (CA) 19-9 concentrations.
The clinical and pathological characteristics of patients with and without elevated preoperative serum CA19-9 levels were compared, as well as those with and without recurrence. Clinicopathological variables between the groups were compared using unpaired Student's t-tests or one-way analysis of variance (ANOVA) using least significant difference (LSD) multiple comparisons. The two groups were compared by cross-table analysis using Fisher's exact test. The Cox proportional hazards model was applied for multivariate analysis to determine any important predictors of recurrence. Variables were included in the multivariate analysis only if p was less than 0.05 in the univariate analysis. Recurrence-free survival rates were compared by the Kaplan—Meier method using a log-rank test. The confidence intervals (CIs) were set at 95% and the significance level at p<0.05.
Results
Characteristics of patients relative to preoperative serum CA19-9 levels and prognostic significance of preoperative serum CA19-9 levels. Patients with high preoperative serum CA19-9 levels (>37.0 U/mL) were significantly more likely to show lymphatic invasion, perineural invasion, more advanced tumor stage, colon cancer and recurrence than patients with normal preoperative serum CA19-9 levels. Sex, histological differentiation and vascular invasion did not differ significantly between patients with normal and high preoperative serum CA19-9 (Table I). The 3-year disease-free survival rate was significantly lower in patients with high preoperative serum CA19-9 (p<0.0001) and was significantly lower in patients with American Joint Committee on Cancer (AJCC) stage II and III disease who also showed high preoperative serum CA19-9 levels, when considering each TNM stage (Figure 1).
Risk factors for recurrence based on multivariate analysis.
Multivariate analysis was performed to estimate the independent predictive value of CA19-9 elevation against recurrence (Table II). Among 1,109 patients who had undergone regular follow-up, recurrences occurred in 250 (22.6%). Cox proportional hazards analysis with the disease-free period was used as a dependent variable to treat recurrence as a time-to-event outcome. The significant predictive value was confirmed with CA19-9 (odds ratio, OR=1.75, p<0.0001), with pT and pN categories, and with a preoperative high serum CEA.
The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of preoperative serum CA19-9 levels were calculated as markers for recurrence. When preoperative serum CA19-9 level was above 37 U/mL, it was considered a positive test. Sensitivity and specificity were 23.6% and 88.6%, respectively. The PPV and NPV of preoperative serum CEA level for recurrence were 37.6% and 79.9%, respectively. For preoperative CEA, the sensitivity and specificity were 38.8% and 83.5%, respectively. The PPV and NPV of preoperative serum CEA level for recurrence were 40.6% and 82.4%, respectively. When considered as positive if either CEA or CA19-9 was higher than the accepted normal value, the sensitivity, specificity, PPV and NPV of the combined tumor markers were 50%, 77.4%, 39.6% and 84%, respectively.
Disease-free survival curves stratified by preoperative serum CA19-9 concentration. Patients with American Joint Committee on Cancer (AJCC) stage I tumors showed no significant differences in preoperative serum CA19-9 levels (p=0.255). Patients with AJCC stage II (p=0.0028) and III (p=0.0015 tumors and a preoperative serum CA19-9 concentration >37 U/mL showed a poorer prognosis than did those with normal preoperative CA19-9 concentrations.
Correlation of post-operative serum CA19-9 levels with recurrence. There were 154 conformed recurrences among 700 patients (22%) with regular follow-up for CA19-9 after surgery. Among these 154 recurrences, a postoperative elevation of CA19-9 measured before the diagnosis of recurrence was seen in 33 (21.4%). The proportion of patients with a high postoperative serum CA19-9 was significantly higher among those who also had a high preoperative serum CA19-9 level.
Regarding recurrence site, patients with liver metastases and peritoneal recurrence were significantly more likely to have an elevated serum CA19-9. When patients had a peritoneal recurrence, the proportion of patients with elevated postoperative CA19-9 was much higher than among those with liver metastases. In contrast, postoperative serum CEA was more frequently elevated in patients with liver metastases (Table III).
The sensitivity, specificity, PPV and NPV were 76.3%, 84.7%, 76.3% and 84.7%, respectively, for postoperative CEA level and 21.4%, 95.8%, 58.9% and 81.2%, respectively, for postoperative serum CA19-9 level. When classified according to the preoperative serum CA19-9 level, the follow-up serum CA19-9 level showed higher PPV, NPV and specificity in patients with high preoperative serum CA19-9 (44.7%, 91.2%, 70.8% and 74.7%, respectively) than in patients with normal preoperative serum CA19-9 (13.8%, 97.3%, 50% and 82.2%, respectively). When postoperative serum CEA and CA19-9 monitoring were considered together, CA19-9 was elevated in 12.9% of patients with recurrence and normal postoperative CEA (Table IV).
Relationship between postoperative CA19-9 and CEA concentration, and recurrence site, and preoperative tumor marker levels in patients with recurrence.
Discussion
In the present study of 1,109 patients with colorectal cancer, high preoperative serum CA19-9 concentration was a marker of poor prognosis, together with CEA levels. Both serum CEA and CA19-9 concentrations have been used as factors indicative of malignant potential and/or poorer prognosis in such patients (18, 19). One study found that that only preoperatively elevated CEA was significantly correlated with poor prognosis, whereas the CA19-9 level was not a predictor of prognosis (20). In contrast, another study reported that preoperatively elevated CA19-9 was an independent predictor of subsequent tumor recurrence in patients with node-negative colorectal carcinomas who had been treated with curative resection (6). Several factors might account for such confusing and even contradictory results. Some reports had an insufficient sample size, and different methods for detecting CA19-9 were used. In addition, differences in tumor stage distribution across study populations may have influenced the outcomes. In one report on patients with stage II colorectal cancer, those with elevated levels of both CEA and CA19-9 were reported to have significantly poorer prognoses than those with normal CEA and CA19-9 concentrations (21). Others have demonstrated that both CEA and CA19-9 levels have prognostic impact in all patient groups but without significant between-stage differences (3, 10, 22, 23). It was found that preoperative CA19-9 levels were not prognostic in patients with stage I disease, which may have been because of the small number of patients in this study with both stage I colorectal cancer and high preoperative serum CA19-9 concentrations. CA19-9 levels may also be related to tumor stage, and may therefore be relatively less effective in predicting outcomes in patients with early-stage disease.
Relationship between postoperative serum CEA and CA19-9 in patients with recurrent disease.
CA19-9 is a widely used marker for colorectal cancer in clinical practice. More than 80% of surgeons also use other tumor markers for postoperative surveillance of colorectal cancer in Japan. Postoperative serum CA19-9 was regularly monitored in 63.1% of the patients. However, available evidence about the utility of CA19-9 for postoperative surveillance is inconsistent as there are contradictory reports about the usefulness of CA19-9 in monitoring recurrence rates (12-17, 24-28). The presented results showed that only 21.4% of patients diagnosed with recurrence by an imaging method, such as CT or PET, had high serum CA19-9 concentrations before the diagnosis of a recurrence. Only 18.8% of patients with a normal preoperative serum CA19-9 level had a high concentration before recurrence, compared with 58.9% of patients with high preoperative serum CA19-9. The sensitivity, PPV and NPV for postoperative serum CA19-9 screening were much lower than for CEA.
Some authors have suggested that CA19-9 should be used in combination with CEA, as it could increase sensitivity in detecting tumor recurrence. Contrary to this suggestion, in the present study, a positive CA19-9 but negative CEA elevation was seen in 7.8% of patients with recurrence and in only 1.7% of all patients. Others have also reported a low frequency (1-11%) of CA19-9 elevation without increase in CEA in patients with recurrence (18, 19, 24, 28), suggesting CA19-9 has limited value as a complementary marker. These results suggest that surveillance with CA19-9 in combination with CEA may be of little practical use in the routine follow-up of patients with colorectal cancer.
In the present study, postoperative serum CA19-9 elevation was better related to peritoneal recurrence than to liver metastasis. The correlation between postoperative CEA elevation and recurrence has been reported to be greater for liver metastases than for recurrent disease at other locations (29, 30). However, a correlation between CA19-9 level and the peritoneal recurrence of colorectal cancer has been reported only rarely (31). The presence of CA19-9 in tumors was reported to be related to tumor cell adhesion (32) and to tumor cell-induced platelet aggregation (33). Moreover, one study reported that peritoneal tumor cells of both colorectal and gastric cancer bind strongly to the anti-CA19-9 monoclonal antibody (34). Therefore, it was thought that colorectal cancer expressing CA19-9 might have a higher possibility of peritoneal recurrence based on the results of this study and others in the literature.
This study had several limitations. It was retrospective in design, although the data were collected prospectively. Thus, the analysis of postoperative serum CA19-9 levels, especially in the interval between its postoperative elevation and any diagnosis of recurrence, may have had an inadvertent bias. In addition, only 63.1% of patients were simultaneously assayed for serum CEA and CA19-9 concentrations postoperatively. Therefore, any prognostic effect of the combination of postoperative serum CA19-9 and CEA measurements might have been underestimated. It was found that patients with high preoperative serum CA19-9 levels were more likely to show higher postoperative serum CA19-9 concentrations and that patients with peritoneal recurrence had higher CA19-9 concentrations than did those with liver metastases. However, the reason for this could not determined nor could any beneficial effect on the treatment of recurrent disease arising from these correlations be found.
In conclusion, it was found that the preoperative serum CA19-9 level was an important prognostic marker for the recurrence of colorectal cancer. Postoperative serum CA19-9 concentration was more elevated when the preoperative serum CA19-9 concentration was high. The diagnostic accuracy of postoperative serum CA19-9 level in predicting recurrence was much lower than that of CEA, and only 7.8% of patients with recurrence showed a high postoperative CA19-9 level together with a normal postoperative CEA. However, postoperative CA19-9 elevation was correlated with peritoneal recurrence, and hence warrants further study to determine its utility as a tumor marker.
- Received December 3, 2008.
- Revision received February 25, 2009.
- Accepted April 9, 2009.
- Copyright© 2009 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- Treatment Outcome of Resection of Disseminated Peritoneal Metastases from Colorectal Cancer
- CA19-9 Concentration After First-line Chemotherapy Is Prognostic Predictor of Metastatic Colon Cancer
- Carbohydrate Antigen 19-9 Predicts Synchronous Peritoneal Carcinomatosis in Patients with Colorectal Cancer
- Post-chemotherapeutic CEA and CA19-9 Are Prognostic Factors in Patients with Colorectal Liver Metastases Treated with Hepatic Resection After Oxaliplatin-based Chemotherapy