Skip to main content
Download PDF

Combination of tumor markers predicts progression and pathological response in patients with locally advanced gastric cancer after neoadjuvant chemotherapy treatment

BMC Gastroenterology volume 21, Article number: 283 (2021) Cite this article

Abstract

Background

The prognostic values of preoperative tumor markers (TMs) remain elusive in patients with locally advanced gastric cancer (LAGC) after neoadjuvant chemotherapy treatment (NACT). This study aimed to assess and establish a novel scoring system incorporating carcinoembryonic antigen (CEA), carbohydrate antigen 19-9 (CA19-9), carbohydrate antigen 72-4 (CA72-4) to enhance prognostic accuracy for progression-free survival (PFS) and pathological response (pCR).

Methods

Patients' data were retrospectively analyzed from December 2006 to December 2017 in our center. The cutoff value of TMs was determined using the time-dependent receiver operating test characteristics method. These three TMs were allocated 1 point each for the post neoadjuvant chemotherapy combination of tumor markers (post-NACT CTM) scores. The training group comprised 533 patients, responsible for full analysis, and the validation group comprised 137 patients based on the selection protocol.

Results

Of 533 enrolled patients, 138, 233, 117, and 45 patients scored 0, 1, 2, 3 respectively. The 3-year PFS rate Multivariate analysis revealed that post-NACT CTM score was an independent predictor of PFS (0 vs. 1, HR: 1.34, 95% CI: 0.92–1.96, P = 0.128; 0 vs. 2, HR: 2.03, 95% CI: 1.35–3.05, P = 0.001; 0 vs. 3, HR: 2.98, 95% CI: 1.83–4.86, P < 0.001). The time-dependent area under curve (AUC) revealed a consistent highest level for post-NACT CTM than other three single TMs. Lower post-NACT CTM score significantly correlated with higher pCR rate based on multivariate logistic regression (2/3 vs. 1, OR: 2.77, 95% CI: 0.90–8.53, P = 0.077; 2/3 vs. 0, OR: 4.33, 95% CI: 1.38–13.61, P = 0.012). A nomogram was formed with both internal and external validation.

Conclusions

The post-NACT CTM score system served as a strong independent predictor for PFS and pCR in LAGC patients who received NACT. Further population-based studies are required to confirm our results.

Peer Review reports

Introduction

Gastric cancer (GC) is the fifth most commonly diagnosed neoplasm, accounting for 5.7% of all cancers worldwide [1]. In China, GC is often diagnosed at advanced stage and has poor prognosis. Although surgery is the only curative approach for patients with locally advanced gastric cancer (LAGC), multimodal therapy has shown to be superior to surgery alone. Curative treatment LAGC usually entails neoadjuvant chemotherapy (NACT) followed by surgery and postoperative adjuvant chemotherapy [2].

Despite the introduced 8th AJCC post-neoadjuvant therapy stage (ypTNM) is effective in predicting long-term survival, the prognosis of GC may be affected by other individual factors like tumor differentiation, behavior and genetic abnormalities, etc. [3, 4].

Throughout these years, studies have revealed that levels of tumor markers (TMs) may reflect tumor burden in certain circumstances and can be used in staging, prognostication or prediction of response to therapy [5,6,7]. Based on the NACT modality, the benchmarked TMs levels were usually based on the pretreating settings, while most studies focused on the alteration of TMs level during preoperative treatment. The sad truth is, nearly half of patients are pathological poor-responders, changed values of TMs during NACT are often minimal [8, 9]. Moreover, the measurement of response for NACT is more likely to depend on pathological evaluations instead of changed TMs levels [10]. In this process, the clinical value of TMs in measuring residual tumor burden after NACT is neglected. As the prognostic values for preoperative TMs have been vastly investigated in previous studies, we have reasons to suppose that the post-NACT TMs values may have similar predictive strength in patients' survival as immediate feedback for residual tumor load [11].

However, a single TM value always has low rate of sensitivity and specificity and can be easily affected by noncancerous conditions [12]. To achieve a higher prognostic value and put TMs into applications, a combined diagnosis of TMs is to be at hand. The most commonly used tumor markers of gastric cancer were CEA, CA19-9, CA72.4, and CA125, sometimes plus AFP and CA242. Liu et al. once found the potential numeric association between the combination of tumor markers (CTM) and the intrinsic tumor parameters that could evaluate the probability of overall survival (OS) in patients with direct curative surgery [7]. The method was revealed to be effective with high feasibility, which can be potentially applied to patients undergoing NACT measuring the tumor load and predicting prognosis.

Given the considerations above, whether a benchmark value for TMs can classify the prognosis for LAGC patients with NACT is still in the mist. Hereby, we introduce a post-NACT combination of serum tumor markers (post-NACT CTM) involving CEA, CA 19-9, and CA 72-4. Other TMs were counted out following the result from a preliminary dataset (data not shown). We aim to see the clinical value in using absolute posttreatment levels of TMs to predict patients' risk of progression (primary outcome) and the rate of pathological response (secondary outcome). A nomogram was also plotted based on the independent predictors.

Methods

Patients

We obtained data from a prospective database of all patients who started NACT at the Peking University Cancer Hospital and Institute from December 1, 2006, to December 1, 2017. The determination of clinical stages, design for treatment route, preoperative assessment, and prompt intervention for adverse events were managed by the multidisciplinary team (MDT).

Study inclusion criteria included: (1) proven diagnosis of gastric adenocarcinoma by preoperative biopsy; (2) complete medical record and document with TMs record before NACT initiation; (3) no signs of distant metastasis at first visit; (4) curative gastrectomy was performed.

The exclusion criteria for the training dataset were as follows: (1) Incomplete post-NACT tumor marker record including CEA, CA19-9 and CA72-4; (2) patients who received preoperative radiotherapy, targeted therapy and interventional chemotherapy; (3) patients who received intraperitoneal chemotherapy or hyperthermia intraperitoneal chemotherapy before curative resection; (4) patients with R1/R2 resection; (5) patients with D0/D1/D1 + lymphadenectomy; (6) Prior history of gastrointestinal tumors; (7) Inconsistent of non-adenomas diagnosis confirmed by postoperative pathology. The availability of pre-NACT TMs data separated patients into a training group and validation group. Finally, 533 patients were eligible for main analysis, while 146 cases without pre-NACT TMs records served as the validation samples for the nomogram model (Fig. 1).

Fig. 1
figure 1

Selection of patients for inclusion

Full size image

Regimen and radical surgery

All patients received at least two cycles of chemotherapy, in the form of 5-Fu based combined regimens before surgery's radicalness. The majority of patients were treated with 5-Fu-based plus oxaliplatin doublets: SOX (oxaliplatin plus S-1) in 254 patients, CapeOX (oxaliplatin plus capecitabine) in 123 patients, and FOLFOX (oxaliplatin plus 5-Fu/4-Lv) in 105 patients. Some received 5-Fu-based plus paclitaxel doublets: PX (paclitaxel plus capecitabine) in 18 patients, PS (paclitaxel plus S-1) in 15 patients. The rest of the other regimens involving 18 patients were as follows: POS (paclitaxel, oxaliplatin, and S-1) in 10, CS (cisplatin plus S-1) in four, IRIS (irinotecan plus S-1) in two, EOX (epirubicin, oxaliplatin and capecitabine) in one, DCF (paclitaxel, cisplatin and 5-Fu) in one, respectively. Additional file 1: Table S1 described detailed dosing regimens. To assess the influence of the treatment duration, the three 14-day cycles of FOLFOX or POS were calculated as two 21-day cycles and were transformed based on the rounding strategies, consistent with the other 5-Fu-based regimens. Dosage reduction or withdrawal was applied in cases of severe adverse events during chemotherapy, as determined by clinicians. After two to three chemotherapy cycles, the antitumor effect was evaluated using abdominal computed tomography (CT). Basically, two or three alignment cycles were performed. The therapy was prematurely terminated in cases of disease progression. Otherwise, gastrectomy or continued NACT was considered after obtaining informed consent and approval from patients. Subtotal or total gastrectomy plus D2 lymphadenectomy was performed according to the Japanese Gastric Cancer Association (JGCA) guideline [13].

Data collection

The patient characteristics, including age, body mass index (BMI), gender, American Society of Anesthesiologists score (ASA), ECOG performance status, comorbidities, tumor location, tumor diameter (on short axis), differentiation grade, vascular involvement, posttherapy pathological (yp) TNM stage according to the 8th American Joint Committee on Cancer (AJCC) guideline, type of resection, complications graded by Clavien-Dindo classification system, total cycles of chemotherapy, date of NACT initiation, date of surgery, date of adjuvant chemotherapy initiation, date of progression [14, 15]. The value of three TMs in each patient was obtained at the time of the first visit and within 45 days before surgery.

Histopathology analysis

All pathological examinations were undertaken by two experienced gastrointestinal pathologists, who were blinded to the group assignment. We assessed efficacy by using the pathological complete response (pCR) rate according to National Comprehensive Cancer Network (NCCN) guidelines [16]. pCR was defined as the elimination of any viable residual tumor cell in the resected primary tumor and adjacent lymph nodes (ypT0N0).

Follow-up

Patients were followed up regularly via physical examination, radiological examination, endoscopic examination, and laboratory examination or telephone call when faced with the inconvenience. These examinations were performed quarterly during the first 2 years, then every 6 months until the fifth year, and then once a year.

Tumor markers and scoring methods

The level of CEA, CA19‐9, and CA72-4 was obtained via laboratory analysis of the patients' routine blood test at initial diagnosis with the upper normal values 5 ng/mL, 35 ng/mL and 6.9 ng/L, respectively. The optimal cutoff values for CEA, CA19-9 and CA72-4 were determined by the highest Youden index using time-dependent receiver operating curve (ROC) analysis with "survivalROC" package based on 3-year PFS with Kaplan–Meier method [17]. The output of 95% confidence interval and compared of AUC were based on the inverse probability of censoring weighting (IPCW) approach implemented in the package "timeROC" developed by Blanche et al. [18]. Because the analysis indicated that a TM level exceeding the cutoff value was associated with shorter PFS, each tumor marker was allocated 1 point of post-NACT CTM score.

Statistical analysis

Continuous variables were summarized as median (IQR) and were compared across groups using the Wilcoxon-rank-sum or Kruskal–Wallis test for two or more group comparisons for continuous variables. Categorical variables were analyzed using the Chi-squared test or Fisher's exact test. The areas under the ROC curve (AUCs) of the TMs for predicting 1-year, 3-year, and 5-year PFS were calculated and used for comparisons with other models, which has been described above. The relationships between clinical and pathological factors and long-term PFS were assessed using log-rank tests and Cox proportional hazard model. Tumor or treatment characteristics that achieved a P value < 0.10 in univariate analysis were included in the multivariate analysis. To examine parameters with high collinearity, we used pairwise relationship correlation coefficients (no coefficient greater than |0.4|) to assess collinearity among predictors. Based on the univariate and multivariate Cox regression analyses, we established nomograms predicting 1-, 3- and 5-year PFS. For the convenience of clinical application, the model was represented as a nomogram using R software's "rms" package [19]. The predictive accuracy of the model was internally validated with Harrell's C-statistic (C-index) using 1000 bootstrap repetitions ranging from which 0.5 (perfect discordance) to 1 (ideal concordance). A calibration plot was generated to examine the performance characteristics of nomograms. Logistic regression was used to evaluate time-independent outcomes. Testing for trends can be applied based on various statistical hypothesis when necessary. We conduct a Spearman correlation analysis to assess the relationship between factors identified in the risk assessment and post-NACT scores. Conventionally, we interpreted a correlation coefficient of < 0.3 as weak, 0.3–0.7 as moderate and > 0.7 as strong. For all analyses, P < 0.05 was considered statistically significant. Statistical analyses were performed using SE STATA (Stata Statistical Software, release 15.1; Stata Corp, College Station, TX, USA) or R (R version 3.6.2).

Results

Patients characteristics and establishment of the scoring system

The clinicopathological characteristics of the patients and their predicted 1-year and 3-year PFS KM rate with 95% CI based on each character are summarized in Table 1. The median follow-up time was 63 (IQR 38–87) months. Median post-NACT CEA, CA19-9 and CA72-4 were 2.42 (IQR: 1.63–4.21), 12.88 (IQR: 7.24–24.64) and 2.92 (IQR: 1.54–7.12), respectively.

Table 1 Demographic and clinicopathologic characteristics of the study population with 1-year and 3-year PFS
Full size table

The time-dependent ROC for the 3-year PFS was used to define the optimal cutoff of CEA, CA19-9, and CA72-4, with the AUC 0.592 (cutoff value: 5.72 ng/ml), 0. 620 (cutoff value: 15.00 ng/ml), and 0.597 (cutoff value: 2.60 ng/ml), respectively (Table 2). Thus, the combination of optimal cutoff values for the three tumor markers in identifying patients' prognosis was introduced as post-NACT CTM scores based on previous descriptions. Of the 533 patients, 140 (26.27%) patients scored 0, whilst 233 (43.71%), 116 (21.76%), and 44 (8.26%) patients scored 1, 2 and 3 respectively. Time-dependent ROC for 1-year, 3-year and 5-year PFS are presented in Additional file 2: Fig. S1.

Table 2 Performance of candidate TMs and post-NACT CTM for predicting risk of 3-year PFS
Full size table

We compare the AUC of the three tumor markers and the post-NACT CTM at 3-year PFS (Fig. 2a). The AUC of the combined diagnostic method was 0.677, significantly different from the AUC of the individual diagnostic method (P = 0.003, P = 0.011 and P < 0.001, respectively, Table 2). On the other hand, the AUC values were comparable for the addition of three tumor markers (CEA vs. CA19-9, P = 0.616; CA19-9 vs. CA72-4, P = 0.432; CEA vs. CA72-4, P = 0.740). Figure 2b plots the AUC for time-dependent ROC performance within 5 years at continuous-time points for the four measurements. It can be seen from the curve that the time-dependent AUC of post-NACT CTM keeps the highest for progression across all the time points while the AUC curve of CEA, CA19-9 and CA72-4 are in similar positions and get crossed over time, indicating that post-NACT CTM can better predict progression at a random exit time.

Fig. 2
figure 2

a Time-dependent ROC curves for the CEA, CA19-9, CA72-4 as well as the post-NACT CTM according to 3-year progression-free survival (PFS). The AUCs for the four indicators is presented in the plot. While all performed statistical significance compared to reference, Compared with the AUC in the CEA, CA19-9 and CA72-4 group, the AUC in combination group was significantly increased (vs. CEA, P = 0.003; vs. CA19-9, P = 0.011; vs. CA72-4 P < 0.001). b Accuracy of recurrence prediction as a function of follow-up months. The time-dependent AUC values for three TMs and post-NACT CTM were computed for each time point separated by month. The prediction accuracy was largely stable in the combination method except during the first three after surgery

Full size image

Performance of the post-NACT CTM for predicting the risk of progression

At the time of the analysis, 232 had experienced a recurrence. Kaplan–Meier curves for PFS and PFS stratified by post-NACT CTM are presented in Fig. 3. There is significant stratification between groups in PFS (log-rank P < 0.001, Ptrend < 0.001).

Fig. 3
figure 3

Kaplan–Meier estimates for Progression-free survival by post-NACT CTM group (log-rank test of equality, P < 0.001). The numbers below the graph indicate the number of subjects at each follow-up

Full size image

To further examine the prognostic values of post-NACT CTM with regard to the survival in all patients, the multivariate Cox proportional hazards model was formulated, adjusting for potential confounders based on P < 0.10 in BMI, ECOG, tumor location, diameter, differentiation, LVI, ypTNM stages, resection types, complications, and post-NACT CTM score in univariate analyses (Table 3). At multivariate analysis, LVI and type of resection were excluded from the variables because of their association with the ypN (r = 0.546, P < 0.001) and tumor location (r = − 0.402, P < 0.001), respectively. Of note, because of their association (r = 0.445, P < 0.001), ypT and ypN were integrated as ypTNM stage for further analysis. Results from the multivariate analysis indicated that post-NACT CTM score was independent predictor of PFS (0 vs. 1, HR: 1.37, 95% CI: 0.94–2.00, P = 0.099; 0 vs. 2, HR: 2.07, 95% CI: 1.37–3.10, P < 0.001; 0 vs. 3, HR: 3.11, 95% CI: 1.91–5.07, P < 0.001) along with ECOG, tumor location, diameter, ypTNM stage and adjuvant chemotherapy. The prognostic significance of post-NACT CTM was still performed well in patients with ypTNM stage either in 0/I (0 vs. 1, HR: 2.60, 95% CI: 0.69–9.73, P = 0.156; 0 vs. 2, HR: 4.84, 95% CI: 0.85–27.51, P = 0.075; 0 vs. 3, HR: 17.96, 95% CI: 2.06–156.58, P = 0.009) or in stage II/III (0 vs. 1, HR: 1.18, 95% CI: 0.80–1.75, P = 0.408; 0 vs. 2, HR: 2.07, 95% CI: 1.36–3.15, P = 0.001; 0 vs. 3, HR: 2.88, 95% CI: 1.74–4.76, P < 0.001).

Table 3 Univariate and multivariate analyses for progression-free survival using a Cox proportional hazards model
Full size table

Nomogram for prediction of PFS

To better predict the prognosis of clinical outcomes, a nomogram was established by involving all the independent prognostic factors above (Fig. 4). A higher total score reveals a higher probability of cancer progression. The nomogram showed that the ypTNM stage was the most significant predictor for PFS risk, followed by tumor diameter and post-NACT CTM. The C-index value of 0.762 (95% CI: 0.742–0.782) calculated for the nomogram model indicated a high prognostic prediction accuracy, outweighing the single value of ypTNM stage (C-index 0.706, 95% CI: 0.691–0.721) or post-NACT CTM (C-index 0.639, 95% CI: 0.620–0.658) based on proportional hazards model. The calibration plots for the probability of 1- and 3-year PFS also presented an optimal agreement between actual observation and the nomogram's prediction (Fig. 5a, b). The C-index in the validation group was 0.693 (95%CI, 0.630–0.756). The favorable calibration for 1-year and 3-year PFS were confirmed in the validation cohort (Fig. 5c, d).

Fig. 4
figure 4

Post-neoadjuvant chemotherapy treatment nomogram for predicting 1-, 3- and 5-year progression-free survival. ypTNM stage = 0 stands for complete pathological response; NACT, neoadjuvant chemotherapy; PFS, progression-free survival

Full size image
Fig. 5
figure 5

Calibration plots of the nomogram for 1-year in the training cohort (a), 3-year in the training cohort (b), 1-year in the validation cohort (a), 3-year in the validation cohort (d), progression-free survival. The x-axis represents nomogram-predicted survival, and the y-axis represents actual survival and 95% confidence intervals. Calibration curves analysis using 1000 bootstraps replicates predict survival in subcohorts of 1000 bootstrap samples. The dotted line represents the ideal correlation between predicted and actual PFS

Full size image

Performance of the post-NACT CTM for predicting pCR

pCR was confirmed in 34 cases (6.38%). The relationships between post-NACT CTM levels and clinicopathologic characteristics were summarized in Table 4. The higher level of post-NACT CTM scores was found significantly correlated with the positive lymph nodes number (P < 0.001), presence of ypN0 (P < 0.001) and pCR (P = 0.039). Univariate and multivariate analysis were performed by logistic regression to evaluate the predictive role of Post-NACT CTM, triplet therapy and NACT duration for pCR. The results showed that both duration of NACT (cycles ≥ 4, OR: 3.09, 95% CI: 1.30–7.33, P = 0.011) and lower post-NACT CTM score (2/3 vs. 1, OR: 2.75, 95% CI: 0.89–8.50, P = 0.078; 2/3 vs. 0, OR: 4.19, 95% CI: 1.33–13.14, P = 0.014) were strong predictors for pCR (Table 5). Simultaneously, we applied another multivariate model based on single tumor marker performance and found that CA72-4 revealed a significant association with pCR according to the cutoff value, while CEA and CA19-9 did not show any statistical difference in univariate regression (Table 5). The AUC values of post-NACT CTM score and CA72-4 were 0.628 (95% CI: 0.544–0.712, P = 0.003) and 0.601 (95% CI: 0.517–0.685, P = 0.019), respectively. The Spearman correlations were significant between post-NACT CTM and most predictor variables, but the correlation coefficients were low for all pairs (r < 0.25), which fortified the independence of prognostic ability of post-NACT CTM scores (Additional file 1: Table S2).

Table 4 Perioperative parameters and response information stratified by post-NACT CTM score after NACT in 533 patients
Full size table
Table 5 Univariate logistic regression analysis examining the influence of patient factors on pathologic complete response (pCR)
Full size table

Changes of tumor markers and their correlation with PFS and pCR

Median time from the initiation of NACT to the operation was 97 days (range: 78–119). However, no differences were found between the baseline and post-NACT concentrations of CEA, CA19-9 and CA72-4 either for the whole sample or stratified by ypTNM stage (Tables 6, 7 and 8). No efficient AUC of CEA, CA19-9 or CA72-4 was observed in pCR rate (Fig. 6a–c) or in the prediction of progression according to a ranking-based evaluation using ROC-AUC and time-dependent ROC-AUC (Fig. 6d).

Table 6 Treatment progress and changes in the serum carcinoembryonic antigen (CEA) level among different stages
Full size table
Table 7 Treatment progress and changes in the serum carbohydrate antigen 19-9 (CA19-9) level among different stages
Full size table
Table 8 Treatment progress and changes in the serum carbohydrate antigen 72-4 (CA72-4) level among different stages
Full size table
Fig. 6
figure 6

a–c ROC curves for the decrease of the three TMs to predict the pathological complete response (pCR) after neoadjuvant chemotherapy. The changes of TMs were compared with baseline level and are based on the following classification rules in CEA (a), CA19-9 (b), and CA72-4 (c) elevation > 50%, elevation > 20% but < 50%, elevation < 20%, decline < 20%, decline > 20% but < 50% and decline > 50% of the baseline levels. The figure shows the of these markers decrease of CEA (AUC: 0.544, 95% CI: 0.433–0.655), CA19-9 (AUC: 0.458, 95% CI: 0.371–0.545) and CA72-4 (AUC: 0.485, 95% CI: 0.384–0.586) was not statistically significant c. d The time-dependent AUC for prediction of risk of PFS for the level change in CEA, CA19-9, CA72-7 are also plotted. All three markers fluctuated around the reference line throughout the follow-up months, indicating there is no prognostic value for PFS among changes of CEA, CA19-9, and CA72-4 during NACT period

Full size image

Discussion

Unlike AFP for hepatic cell carcinoma or CA125 for ovarian tumors, the relationship between TMs and the prognosis of the GC was unclear. This is largely because there is no conclusive evidence that TMs expression levels can exactly reflect tumor burden. Of the three TMs in the present study, CEA is an oncofetal glycoprotein of the cell surface involved in intracellular adhesion during the fetal development of gastrointestinal tissue [20]. CEA's overexpression can be particularly observed in patients with various malignancies, including adenocarcinomas of the gastrointestinal tract (stomach, colon, rectum, and pancreas). It is commonly hypothesized that the overexpressed CEA protein occupies cell membranes' surface and prevents normal growth inhibition and cellular differentiation, which finally leads to tumor progression [21, 22]. CA19-9, on the other hand, is a sialylated Lewis blood group antigen but is widely expressed in gastrointestinal malignancies. The biological functions of CA19-9 are still poorly defined. As a mucin antigen with high molecular weight, CA19-9 is hypothesized to act as an anti-adhesive molecule contributing to tumor migrations and to locally inhibit T cell-mediated antitumor response [23, 24]. Moreover, the prominent predictive ability of CA19-9 for pancreatic cancer indicated the possibility to measure tumor biology from a quantitative view [11, 25]. CA72-4 was initially designated as a tumor-associated glycoprotein-72 which has been found in various epithelial malignancies tumors as well as benign gastritis [26, 27]. Although with obscure biological functions, CA72-4 appears to be more accurate than CEA and CA19-9 in detecting lymph node status and progression in GC [28]. Moreover, the serum level of CA72-4 correlating with the stage of tumor has been confirmed by previous studies [26, 29, 30]. Taken together, all these TMs are nonspecific for GC, but the elevation of each TM should mechanistically accompany the tumor growth, infiltration, and spreading into the blood circulation. With the cumulative clinical evidence, it is justifiable to suppose that CEA, CA19-9 and CA72-4 could be used conjunctively in reflecting tumor burden for LAGC and predicting prognosis [31, 32].

For TMs being as prognostic factors in LAGC patients undergoing NACT, there are two distinct questions to be addressed first. One problem is that treatment duration, dosage and chemosensitivity are not equalized in each patient. To eradicate confounders surrounding NACT settings, the post-NACT TMs values should be more logical with explicit clinical meaning used as benchmark value of TMs in predicting patients prognosis and tumor response rather than pretreatment values. The other question is the indication of changed TMs values during NACT. As current NACT protocols are far from satisfying in treating LAGC, the decreased values of TMs are not always magnificent and may not be easily interpreted [8, 9]. On the other hand, the alteration of TMs is considered to reflect the chemotherapy efficacy and chemosensitivity of the tumors, but not the residual tumor burden, although it can be employed to indicate prognosis to some degree [33, 34]. In the light of these, a post-NACT, comprehensive, cross-sectional benchmark value for TMs was introduced in the current study.

Because the low sensitivities and specificities, using "normal range" as cutoff values, restrict their clinical application, the predictions based on the serum TMs and combined method were investigated using time-dependent ROC curves for cumulative PFS. Relating to CA19-9 and CA72-4, the cutoff values were within the "normal range". This is reasonable as current ranges of TMs are mainly designed for screening the general population instead of predicting tumor progression in already diagnosed patients [35]. Fluctuations in TMs expression frequently occur in patients following NACT, which is also likely to require a reframing of TMs norms and standards. Ma et al. investigated a cohort of 154 LAGC patients with tumor regression grade 0–1 and found that CEA lower than 5.0 ng/ml after NACT greatly improved patients prognosis [36]. In other words, the criteria of each TM may be varied in terms of different usage.

We did not evaluate the efficacy of TMs on overall survival as some previous work did [32, 37]. Considering that patients' long-term survival relied on the treatment after recurrence, the results of post-NACT TMs on OS should be affected by miscellaneous factors and be over-interpreted. We also used dynamic AUC to illustrate the consistent advantages in post-NACT CTM for predicting the risk of progression. This avoids data-driven analysis and enhances transparency. Furthermore, the results of the log-rank test and Cox regression indicated that post-NACT CTM could classify patients into four independent groups with good discrimination of PFS.

Although nomographs were widely publicized, less work has been done on NACT patients following the 8th AJCC ypTNM system. The nomogram profiling for NACT population was firstly introduced in our previous study [38]. However, the information of TMs in our previous work was not available. Following that ypTNM stage, tumor location, and BMI were independent prognostic factors, the present work complemented our previous with a satisfactory C-index of 0.762 based on internal validation and good continuity in the validation cohort. Of note, the post-NACT CTM = 3 contributed the third-highest HR to the model. This indicates that the combined diagnosis of tumor marker serves as a potentially strong prognostic indicator for disease progression after NACT. Interestingly, as we conducted the correlation analysis, all the included covariates revealed either weak or non-significant associations with post-NACT CTM scores. The reason for its independent prognostic value might be credited to TMs' unique diagnostic advantages. In measuring residual tumor burden, not only TMs can reflect viable tumor mass, but it may indicate the amount of circulating tumor cells which can further forecast metastases [39, 40].

It has been confirmed that there is a strong association between pCR and long-term survival or recurrence [41, 42]. The pCR rate of gastric cancer patients after NACT is likely to be affected by the tumor location, differentiation, the Lauren classification and type of chemotherapy regimens and cycles [43]. In the realm of tumor markers, results are not conclusive. Either the baseline or the changes of TMs might be a potential indicator for complete pathological response. Sun et al. retrospectively reviewed 184 GC patients with NACT and found pretreatment CEA and CA72-4 change were associated with a higher response rate [33]. Chen et al. first proposed a nomogram method giving a systemic evaluation based on pretreatment parameters in patients receiving NACT and found only CEA had prognostic value on pCR [44]. Interestingly, as opposed to our studies, higher CEA (> 5.0 ng/ml) is associated with higher probability of pCR in their study. The authors hypothesized that elevated CEA is associated with heavier tumor load and faster tumor growth rate, and indicate that tumors are more susceptible to chemotherapy. We believe their hypothesis should be testified by the post-NACT TMs level under the premise that cross-sectional levels of TMs reflect residual tumor load. Contrary to Chen et al., the lower post-NACT level of CA72-4 and post-NACT CTM score significantly correlated with pCR rate, while CEA, CA19-9 and TMs’ change did not reveal statistical relevance. Our results advocate that post-NACT CTM reveals residual tumor burden which can furtherly predict tumor pathological response. However, there is another equally important possibility that should not be neglected. As NACT for gastric cancer is of limited benefit, usually with pCR rate less than 10%, studies on pCR prediction are based on fairly low rate of positive events and insufficient sample size [45,46,47]. We believe this could be the main reason that some of previous results are with low repeatability. To achieve a convincing result or model, either the sample size should multiply or a combined diagnosis method should be introduced, which could cut down the number of covariates and therefore stabilize variances. In the current study, on the premise of the capacity, we adopted the latter one as a solution.

As dynamic monitoring of TMs change is considered as a regular practice in screening progression or recurrence. We also conducted a comparison between the pretreatment TMs and post-NACT TMs, aiming to find clues on prognosis. Unfortunately, contrary to Sun et al., no statistical differences were found between the baseline and post-NACT levels of three TMs. Nor did we find altering of these TMs could serve as a prognostic indicator for pathological responses or disease progression. We suggest that the gastric adenocarcinoma is generally less responsive, resulting in over one-third of poor-responders after NACT, relatively short treatment duration may not meet the window that could reflect the treatment efficacy (around half of our patients received two cycles of NACT). It should be noted that surgical resection is the only curative approach for GC, and sometimes there might be tumor markers surge after initiation of chemotherapy [5, 48]. While predicting pathological response during NACT in TMs' change remained ambiguous, the post-NACT timepoint level is of clinical significance.

We acknowledge that there were some limitations to our study. First, the study is restricted by its single-center retrospective nature. Second, although the definition of cutoff values followed a systemic design, a population-based study is required to set our cutoff values more accurately. Third, although a validation set was formed, there lacked large prospective studies for validation, and some inflammatory and nutritional markers like neutrophil-to-lymphocyte ratio or prognostic nutrition index were not evaluated in our regression model [49]. Last but not least, the change of TMs may not be fully investigated in our study. We are preparing to conduct long-term surveillance with controlling of confounders more strictly to address this question further.

Conclusions

In conclusion, we have demonstrated that post-NACT CTM showed a favorable accuracy as an independent predictor of progression as well as pathological response in LAGC patients with neoadjuvant chemotherapy. As a simple but cost-effective method, we believe it points to a way of prognostic prediction and guiding treatment strategies in LAGC patients who received NACT. Limited by its retrospective nature and capacity, our issues serve as the triggers of yet prospective population-based studies to refine our criteria in the future.

Availability of data and materials

The datasets during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Abbreviations

AFP:

Alpha fetoprotein

AJCC:

American Joint Committee on Cancer

ASA:

American Society of Anesthesiologists

AUC:

Area under curve

BMI:

Body mass index

CA125:

Cancer antigen 125

CA19-9:

Cancer antigen 19-9

CA72-4:

Cancer antigen 72-4

CEA:

Carcinoembryonic antigen

CTM:

Combination of serum tumor markers

GC:

Gastric cancer

IPCW:

Inverse probability of censoring weighting

JGCA:

The Japanese Gastric Cancer Association

LAGC:

Locally advanced gastric cancer

LVI:

Lymph-vascular invasion

NACT:

Neoadjuvant chemotherapy

OS:

Overall survival

pCR:

Pathological complete response

PFS:

Progression free survival

ROC:

Receiver operating curve

TMs:

Tumor markers

References

  1. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424.

    Article  PubMed  Google Scholar 

  2. Petrillo A, Pompella L, Tirino G, Pappalardo A, Laterza MM, Caterino M, Orditura M, Ciardiello F, Lieto E, Galizia G, et al. Perioperative treatment in resectable gastric cancer: current perspectives and future directions. Cancers (Basel). 2019;11(3):399.

    Article  CAS  Google Scholar 

  3. Li Z, Wang Y, Shan F, Ying X, Wu Z, Xue K, Miao R, Zhang Y, Ji J. YpTNM staging after neoadjuvant chemotherapy in the Chinese gastric cancer population: an evaluation on the prognostic value of the AJCC eighth edition cancer staging system. Gastric Cancer. 2018;21(6):977–87.

    Article  PubMed  Google Scholar 

  4. Li Z, Xiao Q, Wang Y, Wang W, Li S, Shan F, Zhou Z, Ji J. A modified ypTNM staging system-development and external validation of a nomogram predicting the overall survival of gastric cancer patients received neoadjuvant chemotherapy. Cancer Manag Res. 2020;12:2047–55.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Jing R, Cui M, Ju S, Pan S: The Changes and Clinical Significance of Preoperative and Postoperative Serum CEA and CA19-9 in Gastric Cancer. Clin Lab 2020, 66(4).

  6. Huang ZB, Zhou X, Xu J, Du YP, Zhu W, Wang J, Shu YQ, Liu P. Prognostic value of preoperative serum tumor markers in gastric cancer. World J Clin Oncol. 2014;5(2):170–6.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Liu X, Qiu H, Liu J, Chen S, Xu D, Li W, Zhan Y, Li Y, Chen Y, Zhou Z, et al. Combined preoperative concentrations of CEA, CA 19–9, and 72–4 for predicting outcomes in patients with gastric cancer after curative resection. Oncotarget. 2016;7(23):35446–53.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Alderson D, Cunningham D, Nankivell M, Blazeby JM, Griffin SM, Crellin A, Grabsch HI, Langer R, Pritchard S, Okines A, et al. Neoadjuvant cisplatin and fluorouracil versus epirubicin, cisplatin, and capecitabine followed by resection in patients with oesophageal adenocarcinoma (UK MRC OE05): an open-label, randomised phase 3 trial. Lancet Oncol. 2017;18(9):1249–60.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  9. Smyth EC, Fassan M, Cunningham D, Allum WH, Okines AF, Lampis A, Hahne JC, Rugge M, Peckitt C, Nankivell M, et al. Effect of pathologic tumor response and nodal status on survival in the medical research council adjuvant gastric infusional chemotherapy trial. J Clin Oncol. 2016;34(23):2721–7.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  10. Westerhoff M, Osecky M, Langer R. Varying practices in tumor regression grading of gastrointestinal carcinomas after neoadjuvant therapy: results of an international survey. Mod Pathol. 2020;33(4):676–89.

    Article  PubMed  Google Scholar 

  11. Sakamoto Y, Miyamoto Y, Beppu T, Nitta H, Imai K, Hayashi H, Baba Y, Yoshida N, Chikamoto A, Baba H. Post-chemotherapeutic CEA and CA19-9 are prognostic factors in patients with colorectal liver metastases treated with hepatic resection after oxaliplatin-based chemotherapy. Anticancer Res. 2015;35(4):2359–68.

    PubMed  CAS  Google Scholar 

  12. Nagpal M, Singh S, Singh P, Chauhan P, Zaidi MA. Tumor markers: a diagnostic tool. Natl J Maxillofac Surg. 2016;7(1):17–20.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Japanese Gastric Cancer A: Japanese gastric cancer treatment guidelines 2014 (ver. 4). Gastric Cancer Off J Int Gastric Cancer Assoc Jpn Gastric Cancer Assoc 2017, 20(1):1–19.

  14. Amin MB, Greene FL, Edge SB, Compton CC, Gershenwald JE, Brookland RK, Meyer L, Gress DM, Byrd DR, Winchester DP. The eighth edition AJCC cancer staging manual: continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging. CA Cancer J Clin. 2017;67(2):93–9.

    Article  PubMed  Google Scholar 

  15. Clavien PA, Sanabria JR, Strasberg SM. Proposed classification of complications of surgery with examples of utility in cholecystectomy. Surgery. 1992;111(5):518–26.

    PubMed  CAS  Google Scholar 

  16. Ajani JA, D’Amico TA, Almhanna K, Bentrem DJ, Chao J, Das P, Denlinger CS, Fanta P, Farjah F, Fuchs CS. Gastric cancer, version 32016, NCCN clinical practice guidelines in oncology. J Natl Compr Cancer Netw. 2016;14(10):1286–312.

    Article  Google Scholar 

  17. Heagerty PJ, Zheng Y. Survival model predictive accuracy and ROC curves. Biometrics. 2005;61(1):92–105.

    Article  PubMed  Google Scholar 

  18. Blanche P, Dartigues JF, Jacqmin-Gadda H. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Stat Med. 2013;32(30):5381–97.

    Article  PubMed  Google Scholar 

  19. Harrell Jr FE: Regression modeling strategies: with applications to linear models, logistic and ordinal regression, and survival analysis: Springer; 2015.

  20. Nicholson BD, Shinkins B, Pathiraja I, Roberts NW, James TJ, Mallett S, Perera R, Primrose JN, Mant D. Blood CEA levels for detecting recurrent colorectal cancer. Cochrane Database Syst Rev. 2015;2015(12):CD011134.

    PubMed Central  Google Scholar 

  21. Ilantzis C, DeMarte L, Screaton RA, Stanners CP. Deregulated expression of the human tumor marker CEA and CEA family member CEACAM6 disrupts tissue architecture and blocks colonocyte differentiation. Neoplasia. 2002;4(2):151–63.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  22. Blumenthal RD, Leon E, Hansen HJ, Goldenberg DM. Expression patterns of CEACAM5 and CEACAM6 in primary and metastatic cancers. BMC Cancer. 2007;7(1):2.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  23. Gimmi CD, Morrison BW, Mainprice BA, Gribben JG, Boussiotis VA, Freeman GJ, Park SY, Watanabe M, Gong J, Hayes DF, et al. Breast cancer-associated antigen, DF3/MUC1, induces apoptosis of activated human T cells. Nat Med. 1996;2(12):1367–70.

    Article  PubMed  CAS  Google Scholar 

  24. Duffy MJ. CA 19-9 as a marker for gastrointestinal cancers: a review. Ann Clin Biochem. 1998;35(3):364–70.

    Article  PubMed  Google Scholar 

  25. Saad ED, Reis PT, Borghesi G, Machado MC, Simon SD, Tabacof J, Gansl RC. Further evidence of the prognostic role of pretreatment levels of CA 19-9 in advanced pancreatic cancer. Rev Assoc Med Bras (1992). 2010;56(1):22–6.

    Article  Google Scholar 

  26. Hu P-J, Chen M-Y, Wu M-S, Lin Y-C, Shih P-H, Lai C-H, Lin H-J. Clinical evaluation of CA72-4 for screening gastric cancer in a healthy population: a multicenter retrospective study. Cancers (Basel). 2019;11(5):733.

    Article  CAS  Google Scholar 

  27. Colcher D, Hand PH, Nuti M, Schlom J. A spectrum of monoclonal antibodies reactive with human mammary tumor cells. Proc Natl Acad Sci USA. 1981;78(5):3199–203.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  28. Chen XZ, Zhang WK, Yang K, Wang LL, Liu J, Wang L, Hu JK, Zhang B, Chen ZX, Chen JP, et al. Correlation between serum CA724 and gastric cancer: multiple analyses based on Chinese population. Mol Biol Rep. 2012;39(9):9031–9.

    Article  PubMed  CAS  Google Scholar 

  29. Spila A, Roselli M, Cosimelli M, Ferroni P, Cavaliere F, Arcuri R, Tedesco M, Carlini S, D’Alessandro R, Perri P, et al. Clinical utility of CA 72-4 serum marker in the staging and immediate post-surgical management of gastric cancer patients. Anticancer Res. 1996;16(4b):2241–7.

    PubMed  CAS  Google Scholar 

  30. Fernandes LL, Martins LC, Nagashima CA, Nagae AC, Waisberg DR, Waisberg J. CA72-4 antigen levels in serum and peritoneal washing in gastric cancer. Correlation with morphological aspects of neoplasia. Arq Gastroenterol. 2007;44(3):235–9.

    Article  PubMed  Google Scholar 

  31. Kotzev AI, Draganov PV. Carbohydrate antigen 19-9, Carcinoembryonic antigen, and carbohydrate antigen 72-4 in gastric cancer: is the old band still playing? Gastrointest Tumors. 2018;5(1–2):1–13.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  32. Guo J, Chen S, Li S, Sun X, Li W, Zhou Z, Chen Y, Xu D. A novel classifier based on three preoperative tumor markers predicting the cancer-specific survival of gastric cancer (CEA, CA19-9 and CA72-4). Oncotarget. 2018;9(4):4814–22.

    Article  PubMed  Google Scholar 

  33. Sun Z, Zhang N: Clinical evaluation of CEA, CA19-9, CA72-4 and CA125 in gastric cancer patients with neoadjuvant chemotherapy. World J Surg Oncol 2014, 12.

  34. Liu X, Meng X, Li Y, Chai W, Qian J, Tang H. Role of CA19-9 in the prognostic evaluation of SOX neoadjuvant chemotherapy for gastric cancer. Int J Clin Exp Pathol. 2018;11(11):5387–93.

    PubMed  PubMed Central  CAS  Google Scholar 

  35. Wang W, Chen XL, Zhao SY, Xu YH, Zhang WH, Liu K, Chen XZ, Yang K, Zhang B, Chen ZX, et al. Prognostic significance of preoperative serum CA125, CA19-9 and CEA in gastric carcinoma. Oncotarget. 2016;7(23):35423–36.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Ma F, Zhang Y, Peng L, Zhang Z, Yang W, Chai J, Zhang B, Ji S, Hua Y, Chen X et al: Which is the optimal management for locally advanced gastric cancer patients with TRG 0 and 1 after R0 resection? Annals of Translational Medicine 2020, 8(15).

  37. Lin JX, Wang W, Lin JP, Xie JW, Wang JB, Lu J, Chen QY, Cao LL, Lin M, Tu R, et al. Preoperative tumor markers independently predict survival in stage III gastric cancer patients: should we include tumor markers in AJCC staging? Ann Surg Oncol. 2018;25(9):2703–12.

    Article  PubMed  Google Scholar 

  38. Li Z, Xiao Q, Wang Y, Wang W, Li S, Shan F, Zhou Z, Ji J. A Modified ypTNM staging system-development and external validation of a nomogram predicting the overall survival of gastric cancer patients received neoadjuvant chemotherapy. Cancer Manag Res. 2020;12:2047–55.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Yu H, Ma L, Zhu Y, Li W, Ding L, Gao H. Significant diagnostic value of circulating tumour cells in colorectal cancer. Oncol Lett. 2020;20(1):317–25.

    PubMed  PubMed Central  CAS  Google Scholar 

  40. Zhao J-X, Liu L-R, Yang X-Y, Liu F, Zhang Z-G. Serum CA19-9 as a marker of circulating tumor cells in first reflux blood of colorectal cancer patients. Oncotarget. 2017;8(40):67918–32.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Li Z, Shan F, Wang Y, Zhang Y, Zhang L, Li S, Jia Y, Xue K, Miao R, Li Z, et al. Correlation of pathological complete response with survival after neoadjuvant chemotherapy in gastric or gastroesophageal junction cancer treated with radical surgery: a meta-analysis. PLoS ONE. 2018;13(1):e0189294–e0189294.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  42. Wan T, Zhang XF, Liang C, Liao CW, Li JY, Zhou YM. The prognostic value of a pathologic complete response after neoadjuvant therapy for digestive cancer: systematic review and meta-analysis of 21 studies. Ann Surg Oncol. 2019;26(5):1412–20.

    Article  PubMed  Google Scholar 

  43. Becker K, Langer R, Reim D, Novotny A, Meyer zum Buschenfelde C, Engel J, Friess H, Hofler H. Significance of histopathological tumor regression after neoadjuvant chemotherapy in gastric adenocarcinomas: a summary of 480 cases. Ann Surg. 2011;253(5):934–9.

    Article  PubMed  Google Scholar 

  44. Chen YH, Xiao J, Chen XJ, Wang HS, Liu D, Xiang J, Peng JS. Nomogram for predicting pathological complete response to neoadjuvant chemotherapy in patients with advanced gastric cancer. World J Gastroenterol. 2020;26(19):2427–39.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  45. Yoshikawa T, Tanabe K, Nishikawa K, Ito Y, Matsui T, Kimura Y, Hirabayashi N, Mikata S, Iwahashi M, Fukushima R, et al. Induction of a pathological complete response by four courses of neoadjuvant chemotherapy for gastric cancer: early results of the randomized phase II COMPASS trial. Ann Surg Oncol. 2014;21(1):213–9.

    Article  PubMed  Google Scholar 

  46. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49(12):1373–9.

    Article  CAS  PubMed  Google Scholar 

  47. Austin PC, Steyerberg EW. Events per variable (EPV) and the relative performance of different strategies for estimating the out-of-sample validity of logistic regression models. Stat Methods Med Res. 2017;26(2):796–808.

    Article  PubMed  Google Scholar 

  48. Kim HJ, Lee K-W, Kim YJ, Oh D-Y, Kim JH, Im S-A, Lee JS. Chemotherapy-induced transient CEA and CA19-9 surges in patients with metastatic or recurrent gastric cancer. Acta Oncologica. 2009;48(3):385–90.

    Article  PubMed  CAS  Google Scholar 

  49. Li Z, Li S, Ying X, Zhang L, Shan F, Jia Y, Ji J. The clinical value and usage of inflammatory and nutritional markers in survival prediction for gastric cancer patients with neoadjuvant chemotherapy and D2 lymphadenectomy. Gastric Cancer. 2020;23(3):540–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

Download references

Acknowledgements

None.

Funding

This work was funded by Beijing Municipal Health Commission (DFL20181103, ZYLX201701).

Author information

Author notes

  1. Zining Liu and Yinkui Wang have contributed equally to this work

Authors and Affiliations

  1. Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Gastrointestinal Cancer Center, Peking University Cancer Hospital and Institute, Beijing, 100142, China

    Zining Liu, Yinkui Wang, Fei Shan, Xiangji Ying, Yan Zhang, Shuangxi Li, Yongning Jia, Rulin Miao, Kan Xue, Zhemin Li, Ziyu Li & Jiafu Ji

Authors
  1. Zining Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Yinkui Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Fei Shan
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Xiangji Ying
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Yan Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Shuangxi Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Yongning Jia
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Rulin Miao
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Kan Xue
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Zhemin Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Ziyu Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Jiafu Ji
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Z.Y.L, Z.N.L and Y.K.W designed this study. Y.K.W, Z.N.L and Y.Z enrolled patients and collected clinical data. Z.N.L and X.J.Y conducted statistical analysis. Z.N.L is responsible for data visualization; The manuscript was drafted by Z.N.L and Y.K.W, and reviewed for content by Z.Y.L, J.F.J, F.S, Y.N.J, S.X.L, K.X, R.L.M, and Z.M.L. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Ziyu Li.

Ethics declarations

Ethics approval and consent to participate

The Ethics Committee of Peking University Cancer Hospital approved this study. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and later versions. Informed consent or substitute for it was obtained from all patients for being included in the study.

Consent for publication

Not applicable.

Competing interests

All authors declare that there are no conflicts of interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1

. Table S1. Dosage and schedule of the treatment regimen; Table S2. Spearman correlation analysis between the post-NACT CTM and potential prognostic factors

Additional file 2

. Fig. S1. Time-dependent ROC curves for progression-free survival (PFS) at 1st, 3rd, and 5th year time points are plotted for (A) CEA, (B) CA19-9, (C) CA72-4, (D) post-NACT CTM. The cutoff values for CEA, CA19-9, CA72-4 are 5.72 (sensitivity 27.49%, specificity 90.33%), 15.00 (sensitivity 54.9%, specificity 66.91%) and 2.60 (sensitivity 67.11%, specificity 53.90%) which are based on 3-year PFS using Kaplan–Meier method.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Liu, Z., Wang, Y., Shan, F. et al. Combination of tumor markers predicts progression and pathological response in patients with locally advanced gastric cancer after neoadjuvant chemotherapy treatment. BMC Gastroenterol 21, 283 (2021). https://doi.org/10.1186/s12876-021-01785-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12876-021-01785-7

Keywords

BMC Gastroenterology

ISSN: 1471-230X

Contact us