|Ahead of print publication
Raised carbohydrate antigen 19-9 levels detect recurrences and impacts overall and disease-free survival in radically resected gallbladder cancer: A simple surveillance marker?
Sushma Agrawal1, Rajan Saxena2
1 Department of Radiotherapy, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India
2 Department of Surgical Gastroenterology, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India
|Date of Submission||27-May-2020|
|Date of Decision||29-Aug-2020|
|Date of Acceptance||15-Sep-2020|
|Date of Web Publication||23-Jul-2021|
Department of Radiotherapy, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow - 226 014, Uttar Pradesh
Source of Support: None, Conflict of Interest: None
Background: There are no established markers which can be used for surveillance after curative resection in gallbladder carcinoma (GBC). Though carbohydrate antigen 19-9 (CA 19-9) has low specificity as a diagnostic marker, its role as a surveillance marker has not been explored. The aim of this study is to evaluate the predictive ability of CA 19-9 as a surveillance marker to detect recurrences on follow-up.
Methods: A retrospective analysis of a prospectively maintained database of radically resected GBC who were either on observation or completed adjuvant therapy (chemotherapy or chemoradiation) were followed up 3 monthly with CA 19-9 and ultrasound (US) abdomen for the first 2 years and 6 monthly CA 19-9 and US for further 3 years. Patients with raised CA 19-9 and a recurrent lesion on US abdomen were confirmed with contrast-enhanced computed tomography (CECT) abdomen and fine-needle aspiration cytology (FNAC) of recurrent lesion to establish the diagnosis of recurrence. The performance of CA 19-9 levels (20 and more units/mL) for prediction of recurrence and its impact on survival was estimated.
Results: Out of sixty patients on follow-up, 40% recurred: loco-regional (16.7%) and distant metastases (23.4%). The sensitivity, specificity, positive predictive value, and negative predictive value of CA 19-9 in detecting recurrence were 79.1%, 97.2%, 95%, and 87.5%, respectively. The median disease-free survival was 56 months versus 15 months (P = 0.008, hazard ratio [HR]: 7.4 [1.3–40]) and the median overall survival was not reached versus 20 months (P = 0.000, HR: 10.7 [confidence interval 4.2–27.3]) for CA 19-9 levels less than and more than 20 ng/mL.
Conclusions: Based on the high positive and negative predictive value in our dataset, CA 19-9 can be used as a surveillance biomarker for follow-up of radically resected GBC. Raised levels of >20 ng/mL should be correlated with imaging findings and any suspicious lesion should be confirmed for recurrence by FNAC and CECT abdomen. Levels >20 ng/mL should be taken as a threshold for suspecting recurrence.
Keywords: Biomarker, gallbladder cancer, radical resection, surveillance, carbohydrate antigen 19-9
|How to cite this URL:|
Agrawal S, Saxena R. Raised carbohydrate antigen 19-9 levels detect recurrences and impacts overall and disease-free survival in radically resected gallbladder cancer: A simple surveillance marker?. J Can Res Ther [Epub ahead of print] [cited 2021 Dec 5]. Available from: https://www.cancerjournal.net/preprintarticle.asp?id=322162
| > Introduction|| |
Gallbladder carcinoma (GBC) is the most common cancer of the biliary tract with a high mortality rate. Radical resection is the only curative modality of treatment. Forty percent of patients recur after radical surgery. Adjuvant chemo-radiotherapy is advocated in patients with node-positive and positive resection margins. Adjuvant single-agent capecitabine has also recently shown to improve survival as compared to observation in the BILCAP study. The consensus statement on the management of GBC advocates 3–4 monthly computed tomography (CT) abdomen for surveillance in patients after radical surgery. Early detection and diagnosis of recurrent cancer is critical in initiating earlier salvage treatment and improvement of prognosis. Potential biomarkers for detecting recurrence after adjuvant therapy and impacting survival have not been mentioned in literature. The utility of carbohydrate antigen 19-9 (CA 19-9) and carcinoembryonic antigen has been explored only in preoperative setting. Another marker CA 242 has been found to have better sensitivity and specificity than CA 19-9 for diagnostic purposes. The combination of CA 19-9 with CA 242 yields a sensitivity and specificity of 73.7% and 84.6%, respectively. The drawback of CA 242 is that it is not widely available. A novel marker, serum CYFRA 21-1, has been found to have better sensitivity (93.7%) and specificity (96.2%) in GBC, but this study was a mixed bag of biliary tumors, and it did not impact survival. Preoperatively CA 19.9 levels are suggestive of prognostic subsets in secretors and has been used in a scoring system to decide resectability., However, how should we follow-up patients who have undergone radical resection? The available guidelines mention surveillance with radiology. None of the guidelines mention the availability of any proven marker suitable for surveillance. Availability of a biomarker would pick up early recurrences. In view of the high rate of recurrence, there is an urgent need to demonstrate the utility of a biomarker for early detection of recurrence.
Therefore, our aim is to evaluate the predictive ability of CA 19-9 as a surveillance biomarker, in detecting early recurrences in radically resected patients after completion of adjuvant therapy.
| > Methods|| |
This is a retrospective analysis of a prospectively maintained database. Case records of sixty consecutive patients with GBC who underwent extended cholecystectomy and registered between January 2011 and December 2016 under first author, were the participants of this analysis. All patients had normal CA 19.9 levels before surgery and before starting adjuvant therapy. After radical resection, Stage II patients (according to AJCC 7th Edition) were kept on observation. Stage III, node negative, and R0 resection (Radical resection with no microscopic residual disease) patients were offered four cycles of chemotherapy (doublet of cisplatin and gemcitabine) and R1 (radical resection with microscopic residual disease) and node-positive patients were offered concurrent chemo-radiation to a dose of 50.4 Gy/24 fractions in 5 weeks along with concurrent capecitabine (1250 mg/m2 daily with radiotherapy). After completion of adjuvant therapy, patients were kept on 3-monthly follow-up with serum CA 19-9 (ascertained by automated immunoassay [AIA Tosho]) and 6-monthly ultrasound (US) whole abdomen for the first 2 years and thereafter 6 monthly CA 19-9 and US abdomen for further 3 years. A threshold of >20 units was considered as the threshold for prediction of recurrence based on our previous publication. Patients with raised CA 19-9 (>20 units/mL) on follow-up, were advised US abdomen on that visit for any evidence of recurrence. Any evidence of recurrence was further confirmed with a CT scan of abdomen and fine-needle aspiration cytology (FNAC) of the recurrent lesion. Patients with no evidence of recurrence were kept on follow-up to evaluate the trend of rise in CA 19-9. For those who exhibited a rising trend in CA 19-9, a contrast-enhanced computed tomography (CECT) abdomen was advised. Recurrences were categorized as Loco-regional recurrence (LRR): in tumor bed and lymphatics) or distant (DM).
The sensitivity and specificity of CA 19-9 levels (<20 and 20 or more) for prediction of recurrence were computed. The association of prognostic data including gender, age, tumor, node, metastasis stage, histopathologic grade, adjuvant treatment received, patterns of recurrence, and survival with CA 19-9 levels was also evaluated for any evidence of significance.
The Kaplan–Meier analysis, with log-rank comparison when indicated, was used to calculate the correlation of Ca 19-9 levels with disease-free survival (DFS) and overall survival (OS). All survival outcome end points were calculated from the date of radical surgery. For analysis of DFS and OS, univariate and step-wise multivariate Cox proportional hazards regression models were used. In all analyses, factors whose P < 0.05 were considered statistically significant.
| > Results|| |
Patient and tumor characteristics
Sixty patients formed the study cohort, which comprised of 14 males and 46 females, with a mean age of 51 years (range 30–77 years). T status was as follows: T1 (8.3%), T2 (50%), T3 (33.3%), and T4 (8.3%); N status was as follows: N0 (31.7%), N1 (40%), N2 (16.7%), and unknown (1.7%). Resection status was as follows R0 (81.7%) and R1 (18.3%). Patient, tumor, and treatment characteristics are reported in [Table 1].
At a median follow-up of 46 months (interquartile range [IQR], 24–63 months), 24 patients (40%) recurred. Nearly 92% of the recurrences occurred within 36 months. Among these, twenty patients (83.3%) had CA 19.9 levels more than 20. Nineteen out of the 20 patients with raised CA 19-9 had evidence of recurrence on imaging (advised on the visit with raised CA 19-9), which was confirmed by FNAC. The median value of posttreatment CA 19-9 values at the time of recurrence was 25 (IQR, 9–75). The median number of CA 19-9 observations was 5 in those who recurred, and there was a rising trend of CA 19-9 levels in ten out of the twenty patients. The sensitivity of CA 19-9 in detecting recurrence was 79.1% and the specificity was 97.2%. The positive predictive value of CA 19-9 was 95% and the negative predictive value was 87.5%. The relapse rate was 27% in CTRT (chemoradiotherapy) group as compared to 40% in CT (chemotherapy) group and 50% in observation group. Out of 24 relapses, 46% were locoregional and 54% were distant. Almost 81% of the patients with LRR had Ca 19.9 levels >20 and 77% patients with distant metastases had CA 19-9 levels >20. There was no difference in the raised median CA 19-9 levels of patients with LRR or DM. There was no difference in the type of recurrence based on treatment modality. Prognostic factors such as age, sex, stage, grade, and treatment modality had no correlation with CA 19-9 levels.
The 5-year OS was 87% versus 12% in patients with CA 19-9 more than and less than 20, respectively (P = 0.000) [Figure 1]a and [Table 2]. The median OS was not reached versus 20 months in patients with Ca 19.9 more than and less than 20. No prognostic variable except CA 19-9 level was significant for OS on univariate analysis, and this was also significant on Cox-regression analysis (hazard ratio [HR]: 10.7; [confidence interval (CI): 4.2–27.3]).
|Figure 1: (a) Overall survival according to carbohydrate antigen 19-9 level less than and more than 20, (b) disease-free survival according to carbohydrate antigen 19-9 level less than and more than 20|
Click here to view
Overall, the median DFS was 39 months, 15 months for those who recurred, and 56 months who were disease free. The 5-year DFS was 83% versus 0% in patients with CA 19-9 more than and less than 20, respectively (P = 0.008) [Figure 1]b and [Table 3]. The median DFS was not reached versus 26 months in patients with CA 19-9 more than and less than 20, respectively. No prognostic variable except CA 19-9 level was significant for DFS on univariate analysis, and this was also significant in Cox-regression analysis (HR: 7.4 [1.3–40]).
| > Discussion|| |
There is no consensus on the biomarkers to be used for surveillance in resected GBC. Elevated preoperative CA 19-9 has been found to portend poor prognosis in patients undergoing resection in biliary cancers, but its utility as a surveillance marker after radical surgery in GBC has never been reported before., Our results reveal that raised CA 19-9 levels on surveillance has a sensitivity of 79.1%, a specificity of 97.2%, a positive predictive value of 95%, and a negative predictive value of 87.5% in detecting recurrence. The area under the receiver operating characteristic curve of this marker has been found to be 0.72 (95% CI: 0.49–0.91) in radically resected patients. Based on the high positive and negative predictive values of this test in a region with high prevalence of GBC, we propose that this test should be used as a surveillance marker to detect recurrences. It can be used as a diagnostic triage test marker, with a rise in CA 19-9, triggering further investigation (usually by CT imaging) and a confirmation by FNAC of the recurrent lesion, rather than initiation of therapy. We suggest a threshold of 20 units to detect 83% recurrences based on our earlier publication. In view of the majority of recurrences occurring within 3 years and a rising trend observed in 50% of patients, an increase in frequency of this test in the first 3 years, might detect recurrences earlier than manifest recurrences.
Rise in CA 19-9 (>20 units/mL) seems to be a relevant clinical endpoint for recurrence as it impacts survival. All patients on follow-up without any evidence of recurrence had normal CA 19-9 levels. Second, we also observed that rise in CA 19-9 was observed in recurrences after all the three modalities of treatment (observation, chemotherapy, and chemo-radiation) and hence it can be advised as a marker for follow-up in radically resected patients with or without adjuvant treatment. Based on our observations, raised CA 19.9 levels did not indicate the location, extent, or type of recurrence, nor did it have any association with the type of treatment.
Though serum CYFRA 21-1 has been reported to have better sensitivity and specificity in GBC, this analysis was based on 29 curatively resected and 3 advanced GBC. Additionally, this marker was not found to be prognostic. In comparison to this, our results are based on a homogenous group of sixty curatively resected GBC where CA 19-9 was found to be significantly affecting DFS and OS.
Reduction of CA 19-9 after initiation of salvage treatment was not evaluated because 50% of patients did not return for salvage therapy after detection of recurrence, the interval between diagnosis of recurrence and death was 3 to 9 months, repeated testing could not be afforded by all patients, and due to lack of health insurance. A prospective systematic study in a large sample can validate our findings. The relation of raised CA 19-9 in the presence of competing known risk factors such as stage, resection status, and node positivity might be more apparent in a large sample size. There is a need to arrive at a consensus regarding the required frequency of CA 19-9 testing to inform us regarding lead time before the detection of manifest recurrence. Demonstration of reduction of its raised level after initiation of salvage therapy is also needed.
To the best of our knowledge, there is no publication demonstrating the utility of CA 19.9 as a marker for surveillance. The strength of this manuscript is that the results are based on a homogenous population of radically resected GBC, where raised CA 19-9 levels have been demonstrated to detect recurrences and significantly impact clinically relevant end points such as DFS and OS.
| > Conclusions|| |
Based on the high positive and negative predictive value of CA 19.9 in detecting recurrences, we propose that it can be used as a surveillance biomarker for radically resected GBC. Raised levels more than 20 should be correlated with imaging findings and any suspicious lesion should be confirmed for recurrence by FNAC and CECT abdomen.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Randi G, Franceschi S, La Vecchia C. Gallbladder cancer worldwide: Geographical distribution and risk factors. Int J Cancer 2006;118:1591-602.
Agrawal S, Gupta PK, Rastogi N, Lawrence A, Kumari N, Maria Das KJ, et al
. Outcomes of adjuvant chemo-radiation and predictors of survival after extended cholecystectomy in gall bladder carcinoma: A single institution experience from north India. J Gastrointest Cancer 2015;46:48-53.
Primrose JN, Fox R, Palmer DH, Prasad R, Mirza D, Anthoney DA, et al
. Adjuvant capecitabine for biliary tract cancer: The BILCAP randomized study. The Lancet 2019;20:663-73.
Aloia TA, Járufe N, Javle M, Maithel SK, Roa JC, Adsay V, et al
. Gallbladder cancer: Expert consensus statement. HPB (Oxford) 2015;17:681-90.
Alvaro D. Serum and bile biomarkers for cholangiocarcinoma. Curr Opin Gastroenterol 2009;25:279-84.
Shukla VK, Gurubachan , Sharma D, Dixit VK, Usha . Diagnostic value of serum CA242, CA 19-9, CA 15-3 and CA 125 in patients with carcinoma of the gallbladder. Trop Gastroenterol 2006;27:160-5.
Rana S, Dutta U, Kochhar R, Rana SV, Gupta R, Pal R, et al
. Evaluation of CA 242 as a tumor marker in gallbladder cancer. J Gastrointest Cancer 2012;43:267-71.
Huang L, Chen W, Liang P, Hu W, Zhang K, Shen S, et al
. Serum CYFRA 21-1 in biliary tract cancers: A reliable biomarker for gallbladder carcinoma and intrahepatic cholangiocarcinoma. Dig Dis Sci 2015;60:1273-83.
Agrawal S, Lawrence A, Saxena R. Does CA 19-9 have prognostic relevance in gallbladder carcinoma (GBC)? J Gastrointest Cancer 2018;49:144-9.
Shukla PJ, Neve R, Barreto SG, Hawaldar R, Nadkarni MS, Mohandas KM, et al
. A new scoring system for gallbladder cancer (aiding treatment algorithm): An analysis of 335 patients. Ann Surg Oncol 2008;15:3132-7.
Hatzaras I, Schmidt C, Muscarella P, Melvin WS, Ellison EC, Bloomston M. Elevated CA 19-9 portends poor prognosis in patients undergoing resection of biliary malignancies. HPB (Oxford) 2010;12:134-8.
Wen Z, Si A, Yang J, Yang P, Yang X, Liu H, et al
. Elevation of CA19-9 and CEA is associated with a poor prognosis in patients with resectable gallbladder carcinoma. HPB (Oxford) 2017;19:951-6.
[Table 1], [Table 2], [Table 3]