1. Introduction
Pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of cancer-related deaths, while PDAC is the 14th most common type of newly diagnosed cancer worldwide and the third most common cause of cancer-related deaths.[1] Although it is seen less often than other cancer types, its prognosis is extremely poor with a 5-year survival rate of < 5%. The vast majority of patients die within 4 to 6 months after diagnosis.[2] Of the patients who undergo surgical resection for curative purposes, recurrence develops in 76.6%. Disease-free survival has been reported to be 11.7 months.[3] Despite developments in testing and imaging methods, only 20% of pancreatic cancer patients have a chance to undergo surgery at the time of diagnosis because of the anatomic localization of the pancreas gland.[4] The pancreatic gland is known to be localized in close proximity to the main vascular structures in the retroperitoneal region. Similar to all diseases of the retroperitoneal region, there is a high probability that when it is symptomatic, the size has increased and/or there has been local invasion of the vascular structures. There are no symptoms specific to PDAC. Initially, nonspecific symptoms can be observed such as abdominal pain, nausea and vomiting, and at advanced stages, symptoms such as obstructive jaundice, weight loss, back pain, and cachexia.
As in other types of cancer, tumor markers are used in PDAC for the follow up of the disease, especially after surgery. With the exception of a few tumor markers (prostate specific antigen, alpha-fetoprotein, calcitonin), the remaining markers are not used for screening or early diagnosis purposes as they are not organ-specific. Moreover, for a marker to be able to be used as a screening test, it must have high sensitivity and specificity rates. The tumor markers most often used in PDAC are carbohydrate antigen 19-9 (Ca19-9) and carcinoembryogenic antigen (CEA).[5] The sensitivity and specificity of these 2 markers for pancreatic cancer have been examined in several studies. In particular, these studies have examined the intensity of Ca19-9 in the surroundings and whether or not Ca19-9 could be used as a diagnostic marker for pancreatic cancer. However, in these extensive studies, Ca19-9 was found to have a sensitivity of 79% and specificity of 82% for PDAC.[6] There is also an increased level of Ca19-9 in nonmalignant hepatobiliary and pancreatic diseases other than PDAC, as well as in other malignancies, including colorectal cancer. Additionally, 5% to 10% of the population has Lewis antigen negativity, and Ca19-9 production is extremely low in these patients.[7] The serum CEA level is elevated in 40% to 60% of patients with pancreatic cancer, and CEA is used more in the evaluation of response to treatment after surgical resection of PDAC. As studies have shown that CEA alone does not have sufficient sensitivity and specificity for PDAC, more accurate results are obtained when it is used together with Ca19-9.[8] For these reasons, these 2 markers are used more frequently in the follow up of the disease course than for screening and early diagnosis of PDAC. Studies which have examined the effect of tumor markers on PDAC prognosis have focused more on patients with locally advanced stage or metastatic disease. It has been shown there is a significant correlation between the tumor marker levels and poor prognosis in locally or systemic advanced stage patients. However, there is no significant correlation between prognosis and marker levels in early-stage patients.
The aim of this study was to examine the effect of the Ca19-9/CEA ratio in ductal adenocarcinoma of the pancreatic head on disease prognosis and mean survival. Despite the low sensitivity and specificity rates, this study aimed to increase the efficacy of 2 frequently used markers in combination. We examined whether an acceptable threshold value of this ratio would be a more useful marker than standard tumor markers in the estimation of prognosis and mean survival, especially in patients with early stage PDAC.
2. Materials and Methods
2.1. Study population
Patients who underwent pancreatectomy for PDAC diagnosis in the Surgical Oncology Department of Ankara University Medical Faculty Hospital between 2010 and 2020 were retrospectively screened. A total of 186 patients were identified. Twenty-seven patients who underwent distal pancreatectomy, 12 who underwent subtotal pancreatectomy, 8 who underwent total pancreatectomy, and 10 who had no pathological diagnosis of adenocarcinoma were excluded from the study. Finally, a total of 129 patients who underwent a Whipple procedure for the diagnosis of pancreatic head adenocarcinoma were included in the study.
2.2. Establishment of tumor marker groups and tumor marker ratio group
Blood samples were collected from the patients during the preoperative anesthesia preparation, and tumor markers were examined (CEA ng/mL, Ca19-9 U/mL). The Ca19-9/CEA ratio was calculated by dividing the Ca19-9 value by the CEA value.
2.3. Study design
Approval from the Ethics Committee was not required, as this was a retrospective study that did not use personal information (name, marital status, etc) and did not require any extra costs. The patient files were screened retrospectively, and the clinicopathological parameters and demographic data were recorded. First, the relationship between the CA19-9 and CEA levels and the Ca19-9/CEA ratio was examined. Second, the relationships between the tumor markers and the CA19-9/CEA ratio with lymph node metastasis and the number of metastatic lymph nodes were examined. ROC analysis was performed for lymph node metastasis and tumor size. In the grouping of tumor size, T1 and T2 were classified as early stage, and T3 and T4 as locally advanced stage. The CA19-9 and CEA levels were determined using the Cox regression analysis, and the Ca19-9/CEA ratio was determined using univariate and multivariate hazard ratios.
Finally, survival analysis was performed using the log-rank mean for CA19-9, CEA, and the Ca19-9/CEA ratio in early-stage patients. In this calculation, T1, T2, and N0, N1 group patients were accepted as early stage. Other patients were excluded from this analysis. The patients were first separated into 2 groups: those above and below the CEA and Ca19-9 reference values. Both markers were removed separately from the log-rank analysis. A reference value was determined according to the Ca19-9/CEA ratio for the differentiation of patients into 2 groups. The formula developed in this study was used to calculate this value. Taking the mean cutoff values determined for lymph node metastasis (lymph node metastasis present or absent) and tumor size (T1, T2 vs T3, T4), a new cutoff value was calculated for the Ca19-9/CEA ratio (cutoff value for T stage/2 + cutoff value for N stage/2). Thus, the effects on survival were compared between these 2 basal tumor markers and the ratio of the 2 markers.
2.4. Statistical analysis
IBM SPSS Statistics for Windows version 22.0 (IBM Corporation, Armonk, NY) was used for statistical analysis and calculations. Conformity of the measured data to normal distribution was assessed using the Kolmogorov-Smirnov test, and homogeneity was assessed using the Levene test. One way-ANOVA, Kruskal-Wallis test, and post hoc multiple comparison (Bonferroni) tests were used in the analyses of multiple groups. The Mann–Whitney U test was used for the analysis of 2 groups of measurements. Univariate and multivariate death hazard ratios (HRs) were calculated using the Cox regression hazard model. Mean survival analysis was performed using the Kaplan–Meier (log-rank) test. Spearman correlation test was used to analyze the dependence of the measured data. Cutoff sensitivity, and specificity values were determined using ROC curve analysis. Differences were considered statistically significant at P < .05.
3. Results
3.1. Patient characteristics
An evaluation was made of 129 patients who underwent a Whipple procedure for the diagnosis of PDAC, comprising 72 males and 57 females with a mean age of 57.78 years (range, 44–88 years). The Whipple procedure was performed with conventional surgery in 120 patients and laparoscopically in 9. Early postoperative surgical complications were present in 35.7% of patients. Complications were classified as Clavian Dindo type 1 in 83 patients (64.3%) and type 5 in 13 patients. Lymph node metastasis was present in 63 patients and the mean number of metastatic lymph nodes was 1.46. Early mortality (within the first 30 days) was observed in 14 patients. Among the 115 surviving patients, the mean follow-up period was 33.39 months (range, 0.1–129.3 months). For all the patients included in the study, the mean CEA value was determined as 4.64 ng/mL (range 0.38–133.5), the mean Ca19-9 value was 150.44 U/mL (range, 0.80–1902) and the mean CA19-9/CEA ratio was 44.77 (range, 0.61–356.37). The distribution of the patients demographic and clinicopathological data is shown in detail in Table 1.
Table 1 -
Demographic and clinical data of all the patients.
| Age, yr, mean ± SD, range |
57.78 ± 13.15 (44–88) |
| Gender: n (%) |
|
| Male |
72 (55.8%) |
| Female |
57 (44.2%) |
| Operation type: n (%) |
|
| Laparosopic |
9 (7%) |
| Open |
120 (93%) |
| Complications: n (%) |
|
| Absent |
83 (64.3%) |
| Present |
46 (35.7%) |
| Clavian Dindo: n (%) |
|
| Cdc 1 |
83 (64.3%) |
| Cdc 2 |
27 (20.9%) |
| Cdc 3 |
6 (4.7%) |
| Cdc 4 |
0 (0%) |
| Cdc 5 |
13 (10.1%) |
| Metastatic LN |
|
| Absent |
66 (51.2%) |
| Present |
63 (48.8%) |
| Metastatic node, number, mean ± SD, range |
1.46 ± 1.82 (0–7) |
| N Stage |
|
| N0 |
66 (51.2%) |
| N1 |
46 (35.7%) |
| N2 |
17 (13.1%) |
|
T Stage |
|
|
T1 |
24 (18.6%) |
|
T2 |
45 (34.9%) |
|
T3 |
51 (39.5%) |
|
T4 |
9 (7%) |
| Early mortality |
|
| Absent |
115 (89.1%) |
| Present |
14 (10.9%) |
| Survival |
|
| Exitus |
84 (65.1%) |
| Survived |
45 (34.9%) |
| Follow up time, mo, mean ± SD, range |
33.39 ± 31.93 (0.1–129.3) |
| Carcinoembryonic antigen (CEA), ng/mL, mean ± SD, range |
4.64 ± 12.63 (0.38–133.50) |
| Carbohydrate antigen 19–9, U/mL, mean ± SD, range |
150.44 ± 261.46 (0.80–1902.0) |
| Ca19-9/CEA ratio, mean ± SD, range |
44.77 ± 59.87 (0.61–356.37) |
Ca19-9 = carbohydrate antigen 19-9, CEA = carcinoembryogenic antigen.
3.2. Comparisons of the Ca19-9, CEA values and the Ca19-9/CEA ratio with clinicopathological factors
First, normality of distribution and homogeneity tests were applied to the CA19-9 and CEA values, and CA19-9/CEA ratio. The distribution was not equal in both the analyses. No significant difference was observed in the distribution of the CEA and Ca19-9/CEA ratio median values according to sex, and a homogenous distribution was observed. The median Ca19-9 value was found to be higher in males than females (93.75 U/mL vs 47.4 U/mL, P = .006). No significant difference was found in the distribution of the median values of Ca19-9, CEA, and Ca19-9/CEA ratio median values according to the operation type, development of complications, and the Clavian Dindo surgical complication classification. A statistically significant difference was found in the distribution of the Ca19-9 and the Ca19-9/CEA ratio median values according to lymph node metastasis, N grade, and T stage. The Ca19-9 levels and Ca19-9/CEA ratios were determined to be significantly higher in patients with lymph node metastasis than in those without (P < .001 for Ca19-9 and P < .001 for Ca19-9/CEA). A statistically significant relationship was determined between N grades and the distributions of the Ca19-9 level and Ca19-9/CEA ratio (P < .001 for Ca19-9 and P < .001 for Ca19-9/CEA). This relationship was due to the difference between the Ca19-9 value and the Ca19-9/CEA ratio of N2 patients and the Ca19-9 level and Ca19-9/CEA ratio of N0 and N1 patients. A statistically significant relationship was determined between T stages and the distributions of the Ca19-9 level and Ca19-9/CEA ratio (P < .001 for Ca19-9 and P = .001 for Ca19-9/CEA). This relationship was due to the difference between the Ca19-9 median value of T3 patients and the Ca19-9 median value of T1 and T2 patients. The Ca19-9/CEA ratio median value was different between T4 patients and T1, T2, and T3 patients, and the median value of T3 patients differed from that of T1 and T2 patients (Table 2).
Table 2 -
Analysis of tumor markers and
tumor markers ratio according to clinicopathological factor.
| Variables |
Groups |
| Ca19-9 U/mL |
CEA ng/mL |
Ca19-9/CEA |
| Value (median, range) |
P value |
Value (median, range) |
P value |
Value (median, range) |
P value |
| Gender |
|
|
|
|
|
|
| Male |
93.75 (3.9–1902) |
.006*
|
2.66 (0.5–43.07) |
.067*
|
27.56 (1.77–356.3) |
.107*
|
| Female |
47.4 (0.8–1429) |
|
1.76 (0.38–133.5) |
|
23.47 (0.61–172.5) |
|
| Operasyon type |
|
|
|
|
|
|
| Laparoscopic |
59.6 (6.6–185.3) |
.292*
|
1.65 (0.87–3.81) |
.210*
|
23.05 (3.46–52.49) |
.712*
|
| Open |
66.23 (0.8–1902) |
|
2.33 (0.38–133.50) |
|
26.9 (0.61–356.37) |
|
| Complications |
|
|
|
|
|
|
| Absent |
57.4 (0.8–1902) |
.473*
|
2.51 (0.5–43.07) |
.902*
|
23.09 (3.31–356.37) |
.762*
|
| Present |
76.33 (3.6–1429) |
|
2.25 (0.38–133.5) |
|
28.06 (0.61–259.34) |
|
| Clavian Dindo |
|
|
|
|
|
|
| Cdc 1 |
57.4 (0.8–1902) |
.737†
|
2.25 (0.38–133.5) |
.565†
|
28.06 (0.61–259.34) |
.740†
|
| Cdc 2 |
76.07 (3.6–812.53) |
|
1.83 (0.5–43.07) |
|
23.05 (3.31–356.37) |
|
| Cdc 3 |
77.75 (7.3–232) |
|
3.3 (1.11–5.07) |
|
18.46 (4.01–71.38) |
|
| Cdc 5 |
97.4 3.9–1429) |
|
3.53 (0.5–8.28) |
|
26.15 (4.88–304.54) |
|
| Metastatic LN |
|
|
|
|
|
|
| Absent |
41.63 (0.8–1429) |
<.001*
|
2.18 (0.39–43.07) |
.710*
|
16.18 (0.61–211.15) |
<.001*
|
| Present |
105.7 (7.4–1902) |
|
2.37 (0.38–133.50) |
|
37.46 (4.01–356.37) |
|
| N Stage |
|
|
|
|
|
|
| N0 |
41.63 (0.8–1429) |
<.001†
|
2.18 (0.39–43.07) |
.519†
|
16.18 (0.61–211.15) |
<.001†
|
| N1 |
79.91 (8.7–1208.6) |
|
2.2 (0.38–133.5) |
|
32.64 (4.01–304.54) |
|
| N2 |
249.7 (7.4–1902) |
|
2.67 (1.07–24.46) |
|
67.15 (6.92–356.37) |
|
|
T Stage |
|
|
|
|
|
|
|
T1 |
26.95 (0.8–148–5) |
<.001†
|
1.58 (0.61–8.34) |
.092†
|
10.38 (0.61–38.27) |
.001†
|
|
T2 |
62.2 (2.61–1208.6) |
|
2.65 (0.38–36.36) |
|
22.38 (1.77–71.20) |
|
|
T3 |
204 (2.9–1902) |
|
2.67 (0.39–133.5) |
|
50.2 (4.09–211.15) |
|
|
T4 |
152.27 (34.9–812.53) |
|
1.7 (0.5–4.61) |
|
137.11 (28.89–356.3) |
|
| Early mortality |
|
|
|
|
|
|
| Absent |
59.6 (0.8–1902) |
.180*
|
2.25 (0.38–133.5) |
.661*
|
26.68 (0.61–356.37) |
.276*
|
| Present |
113.55 (3.9–1429) |
|
3.31 (0.5–8.28) |
|
26.7 (4.88–304.54) |
|
| Survival |
|
|
|
|
|
|
| Exitus |
102.21 (3.9–1902) |
<.001*
|
2.7 (0.38–133.5) |
.012*
|
30.11 (1.79–356.37) |
<.001*
|
| Survived |
31.1 (0.8–251.5) |
|
1.71 (0.39–8.18) |
|
15.96 (0.61–154.76) |
|
Ca19-9 = carbohydrate antigen 19-9, CEA = carcinoembryogenic antigen.
* Mann–Whitney U test
† Kruskal-Wallis Test.
The Ca19-9, CEA values and Ca19-9/CEA ratios for lymph node metastasis and T stage were examined using ROC curve analysis. The CEA level was not determined to have sufficient statistical value in the differentiation of early or late T stage or lymph node metastasis. A Ca19-9 cutoff value of 64.45 was determined for lymph node metastasis (P < .001). In patients with a value above this threshold, there was an increased probability of lymph node metastasis with 63.5% sensitivity and 63.6% specificity. The cutoff value for the Ca19-9/CEA ratio was 27.18. This cutoff value had a sensitivity of 79.4% and a specificity of 80.3% for lymph node metastasis (Fig. 1). For T stage, the Ca19-9 cutoff value was found to be 72.76 with 62.8% sensitivity and 62.8% specificity (P < .001). The sensitivity and specificity values of the Ca19-9/CEA ratio were higher. A Ca19-9/CEA ratio > 29.77 showed 69.9% sensitivity and 70.9% specificity for the probability of the T3 and T4 stages (Fig. 2).
;)
Figure 1.: ROC curve analysis of CEA, Ca19-9, and Ca19-9/CEA ratio for lymph node metastasis. Ca19-9 = carbohydrate antigen 19-9, CEA = carcinoembryogenic antigen.
Figure 2.: ROC curve analysis of CEA, Ca19-9, and Ca19-9/CEA ratio for T1–T2 and T3–T4 stage status. Ca19-9 = carbohydrate antigen 19-9, CEA = carcinoembryogenic antigen.
When the correlation between the CA19-9 level, the Ca19-9/CEA ratio, and the number of metastatic lymph nodes was examined, a significant positive correlation was found between the number of metastatic lymph nodes and both parameters (Table 3). Each increase of 1 unit in the Ca19-9 level caused an increase of 0.002 in the number of metastatic lymph nodes (P < .001). Each increase of 1 unit in the Ca19-9/CEA ratio caused an increase of 0.013 in the number of metastatic lymph nodes (P < .001).
Table 3 -
Correlation & regression analysis between Ca19-9, CEA, Ca19-9/CEA ratio and metastatic sentinel lymph node number.
| Clinicopathological factors |
N |
rho |
P value |
| Ca19-9 |
129 |
0.421 |
<.001 |
| CEA |
129 |
0.029 |
.742 |
| Ca19-9/CEA |
129 |
0.528 |
<.001 |
| Regression |
| Dependent Variable |
Independent Variable |
B |
95% Cl for B |
t |
R |
P value |
| Metastatic lymph |
Ca19-9 |
0.002 |
0.001–0.003 |
3.786 |
0.318 |
<.001 |
| Node number |
Ca19-9/CEA |
0.013 |
0.008–0.018 |
5.404 |
0.432 |
<.001 |
Ca19-9 = carbohydrate antigen 19-9, CEA = carcinoembryogenic antigen.
3.3. Cox regression analysis according to the Ca19-9 and CEA values and the Ca19-9/CEA ratio
Univariate and multivariate Cox regression analyses were performed to investigate the effects of the Ca19-9 and CEA levels and the Ca19-9/CEA ratio on mean survival. The Ca19-9 level and Ca19-9/CEA ratio were found to be independent risk factors for mean survival [HR = 1.001, 95% confidence interval (CI): 1.00–1.001, P = .005 for Ca19-9 and HR = 1.004, 95% CI: 1.002–1.007, P = .001 for Ca19-9/CEA ratio]. The CEA level was not observed to have any significant positive or negative effects on mean survival (P = .066). In the multivariate analysis, with the exception of the Ca19-9/CEA ratio, no variable was found to make a significant contribution to the model. There was no statistically significant change in the HR (HR = 1.004, 95% CI: 1.001–1.007, P = .019). The Ca19-9/CEA ratio was a dependent and independent risk factor for mean survival compared with the Ca19-9 and CEA levels alone (Table 4).
Table 4 -
Univariate and Multivariate analysis of Ca19-9, CEA, and Ca19-9/CEA ratio for overall survival.
| Characteristics |
Univariate analysis |
Multivariate analysis |
| HR (95% Cl) |
P value |
HR (95% Cl) |
P value |
| CEA |
1.013 (0.99–1.027) |
.066 |
|
|
| Ca 19-9 |
1.001 (1.00–1.001) |
.005 |
|
|
| Ca19-9/CEA ratio |
1.004 (1.002–1.007) |
.001 |
1.004 (1.001–1.007) |
.019 |
Ca19-9 = carbohydrate antigen 19-9, CEA = carcinoembryogenic antigen, HR = hazard ratio.
3.4. The prognostic impact of Ca19-9, CEA and Ca19-9/CEA ratio for overall survival in early stage (T1 and T2) PDAC: Kaplan–Meier plots
Kaplan–Meier survival analysis was performed for early stage patients (Fig. 3). First, the patients were separated into 2 groups: those with CEA levels above and below 3ng/mL. The CEA reference value groups were examined using the log-rank analysis. No significant difference was found between the mean survival curves and CEA reference values in early stage PDAC (P = .780). The patients were then separated into 2 groups according to the Ca19-9 value < 37 U/mL or > 37 U/mL, and the groups were examined using the log-rank analysis. No significant differences were observed between the mean survival curves of the Ca19-9 reference value (P = .102). Finally, log-rank analysis was applied to the 2 groups with Ca19-9/CEA ratio above and below 28.475, and a statistically significant difference was observed between the groups. Patients with a Ca19-9/CEA ratio below the cutoff value of 28.475 had a mean survival of 93.161 months, and those with a value higher than the cutoff value had a mean survival of 28.541 months (P < .001).
Figure 3.: Log-rank analysis of Ca19-9, CEA, and Ca19-9/CEA ratio cutoff values in early-stage pancreas head ductal adenocarcinoma using the Kaplan–Meier method. Ca19-9 = carbohydrate antigen 19-9, CEA = carcinoembryogenic antigen.
4. Discussion
Although PDAC is seen at a lower frequency than other cancers of the gastrointestinal system, it has high rates of morbidity and mortality. The anatomical localization of the pancreatic gland places it in close proximity to the main vital vascular structures, such as the aorta, superior mesenteric artery and portal vein. The tumor naturally invades surrounding tissues, and because of its anatomic localization, patients may be inoperable or unresectable as a result of local invasion. In previous studies, 80% to 90% of pancreatic head cancers were observed to be inoperable at the time of diagnosis. Despite developments in imaging methods, 30% to 40% of patients with local invasion not determined preoperatively have been reported to show partial or full layer invasion during the operation.[9] Even in high-volume centers, R0 resection can be be performed in only 30% to 40% of patients because of these characteristics, and the disease recurs in 70% of these patients.[10] In addition, there is a risk of 0.7% to 3% mortality and morbidity at the non-insignificant rates of 36 % to 41%, and these rates are higher for locally advanced stage tumors.[11]
Owing to their poor prognostic characteristics, researchers have studied diagnostic methods and biomarkers that could be helpful in determining PDAC at an early stage. Smilar to other cancer types, various tumor markers are used in pancreatic cancer, of which especially Ca19-9, CEA, and Ca125 are at the forefront. The NIH Biomarker Definitions Working Group defined a biomarker as a characteristic marker (anatomic, serological, or physiological) that can be measured with high sensitivity and specificity for the evaluation of the normal biological processes, the pathological processes, and the biological responses to therapeutic interventions.[12] The heading emphasized with importance in this definition is its high sensitivity and specificity. Unfortunately, even CA19-9, which is the most frequently used biomarker in PDAC, has a sensitivity and specificity of 80%, and the sensitivity and specificity of CEA are even lower. Therefore, they are used as markers of treatment response rather than for screening and diagnosis.
In the current study, CEA was found to be high in 40% and Ca19-9 in 67% of patients. These values were similar to the current literature, and thus, the result obtained was one that is already well known. The majority of the high rates of false negativity of these tumor markers used for PDAC is the inadequate for early-stage cancer. From this starting point, the 2 tumor markers were combined for use in this study, with the idea of turning this disadvantage into an advantage. The inspiration for this combined use was the application of sentinel lymph node biopsy for breast cancer. When performing sentinel lymph node biopsy, the use of blue stain or lymphoscintigraphy-gamma probe is known, and studies have shown that blue stain used alone has 40% to 98% sensitivity and lymphoscintigraphy-gamma probe has 80% to 98%, when these 2 methods are used in combination, sensitivity has been shown to be as high as 96% to 98%.[13,14]
There are studies in the literature that have examined the effects of tumor markers on disease prognosis in PDAC. Some studies showed that in the early diagnosis of PDAC, the CEA rise has lower sensitivity and specificity than CA19-9.[15,16] Van Manen et al[17] stated that when the CEA value is combined with Ca19-9 it has a greater value as a prognostic marker in advanced PDAC. In several studies from China, combinations of markers have been used.[18] However, the markers used in these combinations are not those used in routine clinical practice. Therefore, in the current study, the Ca19-9 and CEA values were preferred for use, as they are studied from peripheral blood samples in almost all clinics and have higher sensitivity and specificity rates than other markers. Thus, an extremely practical method was used with the application of a simple mathematical formula, without incurring any extra cost.
Lymph node metastasis and tumor stage are the primary factors affecting the prognosis and mean survival of cancer patients. In this study, the relationships between the Ca19-9 and CEA levels and the Ca19-9/CEA ratio and the N grades and T stages were examined first. There was a significant relationship between lymph node metastasis and Ca19-9 but not with CEA. These results were consistent with those of Ermiah et al[19] However, the difference in the current study was that the relationship between the Ca19-9/CEA ratio and the N grade was also examined. An increase in the Ca19-9 level increased the probability of the patient being N2 compared to the probability of N0. Thus, an increase in the Ca19-9 level is not a prognostic factor for the differentiation of N0 and N1. However, the Ca19-9/CEA ratio shows this relationship in greater detail. An increase in the Ca19-9/CEA ratio increased the probability of the patient being N2 compared with the probability of being N1 and N0. In other words, with the Ca19-9 level, it can only be understood whether the patient is N2 or N0, but using the Ca19-9/CEA ratio, it can be understood whether the patient is N0, N1, or N2. When the correlations of the tumor markers and the Ca19-9/CEA ratio with the number of metastatic lymph nodes were examined with regression analysis, the Ca19-9/CEA ratio was determined to have a stronger correlation with the number of metastatic lymph nodes and regression results than the Ca19-9 and CEA levels alone. The success of the Ca19-9/CEA ratio in showing lymph node metastasis and number was greater than that of Ca19-9 and CEA alone. Similar to Ermiah et al, Ferrone et al[20] also examined the relationship between the Ca19-9 level and N and T stages in PDAC. In both studies, a positive relationship was found between an increase in the Ca19-9 level and T stage. In the current study, a significant relationship was determined between T stage and Ca19-9 level but not between T stage and CEA level. The relationship between Ca19-9 and T stage was due to the difference between the T3 group and the T1 and T2 groups. Thus Ca19-9 was seen to be a successful marker for the differentiation between T3, T1, and T2 patients. When the relationship between the Ca19-9/CEA ratio and T stage was examined, this relationship was found to be significantly expanded, as for N grades. The CA19-9/CEA ratio of T4 patients was significantly different from that of the other T groups. Thus, it was determined that the Ca19-9 ratio was a more accurate marker for the differentiation of T4 patients from all the other T stages than the Ca19-9 and CEA markers alone.
Finally, survival analysis was applied to the Ca19-9, CEA values and the CA19-9/CEA ratio in early stage PDAC. The reason for separating the patients as early stage was that tumor markers are known to be elevated in advanced stage cancers. The main handicap of the markers is low sensitivity and specificity at the early stage. The basis of this study was to use the 2 markers together to increase the low rates of early-stage cancer. Patients with T3 and T4 stages and N2 grades were excluded from the log-rank analysis applied to the tumor markers. No significant correlation was found between Ca19-9 and CEA levels and survival in early-stage cancer. However, a significant result was found in the mean survival analysis of the Ca19-9/CEA ratio. Patients with a value below the Ca19-9/CEA ratio cutoff value of 28.475 had a mean survival of 93.16 months and those with a value higher than the cutoff value had a mean survival of 28.54 months.
From these results, we confirmed our study hypothesis. The Ca19-9 and CEA levels alone do not have sufficient sensitivity and specificity, especially in early-stage pancreatic cancer. However, the Ca19-9/CEA ratio completely changed these results and required high accuracy. Although there are studies in the literature that have examined the relationship between tumor markers and survival, none of these have examined the relationship in the early-stage subgroup and have only confirmed a known fact.[21,22] Tumor markers will be high in advanced stage tumors, and the mean survival will be low. The results obtained in this study increased the value of this study.
This study had no bias, but did have the limitations inherent to all retrospective studies. Patients without the necessary data in the files for the parameters examined were excluded from the study, resulting in a relatively low number of patients.
5. Conclusion
In conclusion, the results of this study showed that by using a simple mathematical calculation to combine the 2 markers most often used in PDAC, but which have insufficient sensitivity and specificity, successful results can be obtained. This combination was determined to have higher accuracy rates in predicting tumor prognosis and in determining mean survival, which are particularly needed in early-stage cancers. Nevertheless, further studies with larger patient populations are required to support these results.
Author contributions
Conceptualization: Ebru Esen, Sevket Baris Morkavuk, Ali Ekrem Unal.
Data curation: Ebru Esen, Mehmet Aslan, Sevket Baris Morkavuk, Siyar Ersoz, Ibrahim Burak Bahcecioglu.
Formal analysis: Ebru Esen, Mehmet Aslan, Sevket Baris Morkavuk, Ibrahim Burak Bahcecioglu, Ali Ekrem Unal.
Investigation: Ebru Esen, Cem Azili, Siyar Ersoz, Ibrahim Burak Bahcecioglu.
Methodology: Ebru Esen, Sevket Baris Morkavuk, Ali Ekrem Unal.
Software: Ebru Esen, Sevket Baris Morkavuk, Ibrahim Burak Bahcecioglu.
Supervision: Ali Ekrem Unal.
Writing – original draft: Ebru Esen, Mehmet Aslan, Sevket Baris Morkavuk, Cem Azili, Siyar Ersoz, Ibrahim Burak Bahcecioglu, Ali Ekrem Unal.
Writing – review & editing: Ebru Esen, Mehmet Aslan, Sevket Baris Morkavuk, Cem Azili, Siyar Ersoz, Ibrahim Burak Bahcecioglu, Ali Ekrem Unal.
References
[1]. Hepatology, T. L. G. Pancreatic cancer: a state of emergency? Lancet Gastroenterol Hepatol. 2021;6:81.
[2]. Michaud DS. Epidemiology of pancreatic cancer. Minerva Chir. 2004;59:99–111.
[3]. Nimura Y, Nagino M, Takao S, et al. Standard versus extended lymphadenectomy in radical pancreatoduodenectomy for ductal adenocarcinoma of the head of the pancreas. J Hepato-Biliary-Pancreat Sci. 2012;19:230–41.
[4]. Gillen S, Schuster T, Meyer zum Büschenfelde C, et al. Preoperative/neoadjuvant therapy in pancreatic cancer: a systematic review and meta-analysis of response and resection percentages. PLoS Med. 2010;7:e1000267.
[5]. Takaori K, Bassi C, Biankin A, et al. International association of pancreatology (IAP)/European pancreatic club (EPC) consensus review of guidelines for the treatment of pancreatic cancer. Pancreatology. 2016;16:14–27.
[6]. Goonetilleke KS, Siriwardena AK. Systematic review of carbohydrate antigen (CA 19-9) as a biochemical marker in the diagnosis of pancreatic cancer. Eur J Surg Oncol. 2007;33:266–70.
[7]. Magnani JL, Steplewski Z, Koprowski H, et al. Identification of the gastrointestinal and pancreatic cancer-associated antigen detected by monoclonal antibody 19-9 in the sera of patients as a mucin. Cancer Res. 1983;43:5489–92.
[8]. Haas M, Heinemann V, Kullmann F, et al. Prognostic value of CA 19-9, CEA, CRP, LDH and bilirubin levels in locally advanced and metastatic pancreatic cancer: results from a multicenter, pooled analysis of patients receiving palliative chemotherapy. J Cancer Res Clin Oncol. 2013;139:681–9.
[9]. Costello E, Greenhalf W, Neoptolemos JP. New biomarkers and targets in pancreatic cancer and their application to treatment. Nat Rev Gastroenterol Hepatol. 2012;9:435–44.
[10]. Büchler MW, Kleeff J, Friess H. Surgical treatment of pancreatic cancer. J Am Coll Surg. 2007;205:S81–6.
[11]. Kostov D, Kobakov G, Yankov D. Involvement of regional lymph nodes in patients with pancreatic head adenocarcinoma. Surg Chron. 2015;20:265–9.
[12]. Atkinson AJ Jr, Colburn WA, DeGruttola VG, et al. Biomarkers and surrogate endpoints: preferred definitions and conceptual framework. Clin Pharmacol Ther. 2001;69:89–95.
[13]. Simmons R, Thevarajah S, Brennan MB, et al. Methylene blue dye as an alternative to isosulfan blue dye for sentinel lymph node localization. Ann Surg Oncol. 2003;10:242–7.
[14]. Gutman F, Sanson A, Piquenot JM, et al. Intra-individual comparison of sentinel lymph node scintigraphy on the day of injection and on the following day in breast cancer. Nucl Med Commun. 2006;27:5–9.
[15]. Xing H, Wang J, Wang Y, et al. Diagnostic value of CA 19-9 and carcinoembryonic antigen forpancreatic cancer: a meta-analysis. Gastroenterol Res Pract. 2018;2018:8704751.
[16]. Zhang Y, Yang J, Li H, et al. Tumor markers CA19-9, CA242 and CEA in the diagnosis of pancreaticcancer: a meta-analysis. Int J Clin Exp Med. 2015;8:11683–91.
[17]. Van Manen L, Groen JV, Putter H, et al. Elevated CEA and CA19-9serum levels independently predict advanced pancreatic cancer at diagnosis. Biomarkers. 2020;25:186–93.
[18]. Kim H, Kang KN, Shin YS, et al. Biomarker panel for the diagnosis of pancreatic ductal adenocarcinoma. Cancers (Basel). 2020;12:1443.
[19]. Ermiah E, Eddfair M, Abdulrahman O, et al. Prognostic value of serum CEA and CA19-9 levels in pancreatic ductal adenocarcinoma. Mol Clin Oncol. 2022;17:1–10.
[20]. Ferrone CR, Finkelstein DM, Thayer SP, et al. Perioperative CA19-9 levels can predict stage and survival in patients with resectable pancreatic adenocarcinoma. J Clin Oncol. 2006;24:2897–902.
[21]. Wentz SC, Zhao ZG, Shyr Y, et al. Lymph node ratio and preoperative CA 19-9 levels predict overall survival and recurrence-free survival in patients with resected pancreatic adenocarcinoma. World J Gastrointest Oncol. 2012;4:207–15.
[22]. Wu L, Huang P, Wang F, et al. Relationship between serum CA19-9 and CEA levels and prognosis of pancreatic cancer. Ann Transl Med. 2015;3:328.
