Retrospective analysis of predictive factors for pathological complete response after neoadjuvant chemotherapy in gastric cancer

Abstract

BACKGROUND

Gastric cancer is a malignant tumor with high morbidity and mortality worldwide. Neoadjuvant chemotherapy (NAC), defined as chemotherapy administered before the primary treatment (usually surgery) to reduce tumor size and control micrometastases, has emerged as a crucial therapeutic strategy for locally advanced gastric cancer. Pathological complete response (pCR), characterized by the absence of viable tumor cells in the resected specimen after neoadjuvant treatment, is recognized as a strong predictor of favorable prognosis. However, the factors influencing the achievement of pCR remain incompletely understood.

AIM

To identify and analyze the predictive factors associated with achieving pCR after NAC in gastric cancer patients, thereby providing evidence-based guidance for clinical decision-making.

METHODS

A retrospective analysis was performed on 215 patients from Shandong Cancer Hospital and Tai’an Central Hospital with locally advanced gastric cancer who underwent NAC followed by radical surgery at our hospital between January 2015 and December 2023. Comprehensive clinical and pathological data were collected, including age, gender, tumor location, Lauren classification, clinical staging, chemotherapy regimens, number of chemotherapy cycles, and baseline hematological indicators. The baseline hematological indicators included neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio, albumin level, carcinoembryonic antigen (CEA), and carbohydrate antigen 19-9. Univariate and multivariate logistic regression analyses were employed to determine the independent predictive factors for pCR.

RESULTS

Among 215 gastric cancer patients, 41 (19.1%) achieved pCR after NAC. Multivariate analysis identified five independent predictive factors for pCR: Lauren intestinal type [odds ratio (OR) = 3.28], lower clinical T stage (OR = 2.75), CEA decrease ≥ 70% after NAC (OR = 3.42), pre-treatment NLR < 2.5 (OR = 2.13), and ≥ 4 chemotherapy cycles (OR = 2.87). The fluorouracil, leucovorin, oxaliplatin, docetaxel regimen achieved the highest pCR rate (27.5%), and oxaliplatin-containing regimens were superior to cisplatin-containing regimens (22.3% vs 12.7%, P = 0.034). Patients with both low NLR and platelet-to-lymphocyte ratio had the highest pCR rate (33.8%), while those with both high inflammatory markers had the lowest rate (10.7%). Earlier clinical stage disease (cT3N+ vs cT4N+: 28.6% vs 13.0%) and lower lymph node burden were associated with higher pCR rates.

CONCLUSION

The achievement of pCR after NAC in gastric cancer patients is closely associated with Lauren intestinal type, lower clinical T stage, a significant decrease in CEA after chemotherapy, low pre-treatment NLR, and an adequate number of chemotherapy cycles.

Key Words: Gastric cancer; Neoadjuvant chemotherapy; Pathological complete response; Predictive factor

Core Tip: This study included patients with pathologically confirmed gastric adenocarcinoma who received neoadjuvant chemotherapy and completed radical gastrectomy + D2 dissection. Unified tumor regression grade assessment defines pathological complete response as the absence of residual primary tumor and lymph node metastasis. The system collected demographic data, comorbidities, tumor location and clinical stage, Lauren classification, human epidermal growth factor receptor 2/microsatellite instability/programmed death-ligand 1, baseline and dynamic laboratory indicators imaging and endoscopic responses, chemotherapy regimens/cycles/dose intensities and adverse reactions. With pathological complete responses the primary outcome, univariate and multivariate logistic regression were used to screen independent predictors.

INTRODUCTION

Gastric cancer is one of the malignant tumors with high incidence and mortality rates worldwide, particularly with a high disease burden in East Asia. Despite continuous advances in diagnostic and treatment technologies, the overall prognosis of gastric cancer remains unfavorable, especially for patients with locally advanced disease. Surgery remains the cornerstone of gastric cancer treatment; however, surgery alone has limited efficacy for locally advanced patients, with high risks of recurrence and metastasis[1-3].

Over the past two decades, multimodal treatment strategies have gradually become the mainstream approach for locally advanced gastric cancer. Among these, neoadjuvant chemotherapy (NAC) has demonstrated significant advantages in multiple clinical studies. NAC can not only downstage tumors and improve radical resection rates but also potentially eliminate micrometastases and improve overall patient survival. Key clinical trials such as MAGIC, FNCLCC/FFCD, and FLOT4-AIO have confirmed the survival benefits of NAC in patients with locally advanced gastric cancer[4-6].

Pathological complete response (pCR) refers to the absence of viable tumor cells in the surgical specimen after neoadjuvant treatment. Multiple studies have shown that patients achieving pCR typically have longer disease-free survival and overall survival, making it an important indicator of treatment success and favorable prognosis. However, the incidence of pCR after NAC in gastric cancer patients is relatively low, usually between 10%-30%, and the factors influencing pCR achievement are complex and diverse, not yet fully clarified[7-9].

Previous studies have indicated that tumor biological characteristics (such as histological type, differentiation degree, Lauren classification), clinical stage, pre-treatment inflammatory indicators, tumor marker levels, as well as chemotherapy regimens and cycle numbers may all influence pCR achievement. However, these research results show certain heterogeneity, and most studies have limited sample sizes, making it difficult to form consistent conclusions. Therefore, identifying clinical factors and biomarkers that can accurately predict pCR after NAC has important clinical significance for optimizing treatment strategies and achieving individualized treatment[10-13].

Furthermore, with the advent of precision medicine, searching for biomarkers that can predict treatment response has become a research hotspot. NAC has significant toxic side effects, and some patients may respond poorly to specific regimens, which not only delays the optimal treatment timing but may also increase patient suffering and economic burden. Therefore, accurately identifying patient populations likely to benefit from NAC before treatment is important for clinical decision-making and optimizing resource allocation. Based on this, domestic and international scholars are actively exploring various predictive tools including imaging features, hematological indicators, and molecular biological markers, hoping to establish more precise prediction models to guide clinical practice[14-16].

In recent years, potential value has been shown in emerging biomarkers such as changes in tumor markers after NAC, dynamic changes in inflammatory indicators before and after treatment, and immune microenvironment status in predicting treatment response. In particular, the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR), as indicators reflecting systemic inflammation and immune status, have shown the ability to predict treatment response and prognosis in various tumors. However, the predictive value of these indicators in gastric cancer NAC still needs further validation through more large-sample, prospective studies. This study aims to establish a prediction model integrating multi-dimensional information by analyzing the relationship between various clinicopathological factors and pCR, providing more reliable decision-making tools for clinical practice, and ultimately achieving individualized and precise treatment for gastric cancer patients.

MATERIALS AND METHODS

Study design and patient selection

This study employed a retrospective cohort study design, collecting clinical data from locally advanced gastric cancer patients who received NAC followed by radical surgery at Shandong Cancer Hospital and Tai’an Central Hospital from January 2015 to December 2023. The study subjects were patients aged 18-75 years, with gastric adenocarcinoma confirmed by gastroscopic biopsy pathology, with preoperative clinical staging of cT3-4N + M0 or cTanyN + M0 (according to the American Joint Committee on Cancer 8^th edition tumor-node-metastasis staging system), who received at least 2 cycles of NAC followed by radical gastrectomy (with D2 Lymph node dissection), and had complete clinical data including pre- and post-chemotherapy imaging evaluations, hematological examinations, and pathology reports. Patients were excluded if they had previously received gastric surgery or radiotherapy, had a history of other primary malignant tumors, had severe cardiac, hepatic, or renal dysfunction or other severe systemic diseases, experienced severe adverse reactions during NAC leading to treatment interruption, had incomplete surgical specimen pathology reports, died within 30 days after surgery, or had missing follow-up data. Based on postoperative pathological results, patients were divided into a pCR group and a non-pCR group. pCR was defined as the absence of viable tumor cells in both the primary tumor site and all resected lymph nodes (ypT0N0M0), assessed based on complete pathological examination following standardized D2 Lymph node dissection. All patients underwent D2 Lymph node dissection as per standard surgical protocol, and pathologists performed comprehensive histological examination of all resected lymph nodes to determine the presence of viable tumor cells. The non-pCR group included patients with viable tumor cells found in the primary site and/or any of the resected lymph nodes after surgery. This standardized approach to pCR assessment, involving systematic examination of all harvested lymph nodes rather than selective sampling, ensures the reliability and consistency of pCR classification across all study participants.

Treatment regimen

All patients received platinum-based (oxaliplatin or cisplatin) NAC regimens, mainly including fluorouracil, leucovorin, oxaliplatin, docetaxel (FLOT), S-1, oxaliplatin (SOX), capecitabine, oxaliplatin (XELOX), or docetaxel, cisplatin, fluorouracil (DCF). The number of chemotherapy cycles was determined based on the patient's clinical response and tolerance, typically ranging from 2 cycles to 6 cycles. After completing NAC, patients underwent reassessment and received radical gastrectomy within 2-4 weeks after chemotherapy completion.

Observation indicators and study endpoints

This study collected patients’ demographic data [age, gender, body mass index (BMI)], tumor characteristics [primary site location, Lauren classification, histological differentiation degree, clinical staging (clinical T stage, N stage), pre-treatment hematological indicators NLR, PLR, albumin level, carcinoembryonic antigen (CEA), carbohydrate antigen 19-9 (CA19-9)], and other baseline characteristics; it also recorded treatment-related indicators such as NAC regimen, number of chemotherapy cycles, post-chemotherapy imaging response evaluation (RECIST 1.1 criteria), and changes in tumor marker levels before and after chemotherapy; pathological assessment included yp tumor-node-metastasis staging (American Joint Committee on Cancer 8^th edition), pCR status (defined as no viable tumor cells found in the primary site and all resected lymph nodes), lymph node metastasis, lymphovascular invasion, and perineural invasion.

Statistical analysis

SPSS 25.0 and R software version 4.0.5 were used for statistical analysis. Continuous variables were expressed as mean ± SD or median (interquartile range) and compared between groups using t-test or Mann-Whitney U test; categorical variables were expressed as frequency (percentage) and compared using χ² test or Fisher’s exact test. Univariate and multivariate logistic regression analyses were used to determine independent predictive factors for pCR, with results expressed as odds ratio (OR) and 95% confidence interval (CI). For continuous variables, optimal cutoff values were determined through receiver operating characteristic curve analysis. Based on the independent predictive factors identified by multivariate analysis. All statistical tests were two-sided, with P < 0.05 considered statistically significant.

RESULTS

Patient baseline characteristics

From January 2015 to December 2023, a total of 215 Locally advanced gastric cancer patients who received NAC followed by radical surgery were included in the study. Among them, 41 patients (19.1%) achieved pCR, while 174 patients (80.9%) did not achieve pCR. There were no significant differences between the two groups in demographic characteristics such as age (median age: 61.3 years in the pCR group vs 60.8 years in the non-pCR group), gender ratio (male: 63.4% in the pCR group vs 65.5% in the non-pCR group), and BMI (mean BMI: 22.4 kg/m² in the pCR group vs 22.1 kg/m² in the non-pCR group) (P > 0.05). Regarding tumor characteristics, the proportion of Lauren intestinal type was significantly higher in the pCR group than in the non-pCR group (63.4% vs 31.6%, P < 0.001), while diffuse type (24.4% in the pCR group vs 46.6% in the non-pCR group) and mixed type (12.2% in the pCR group vs 21.8% in the non-pCR group) were more prevalent in the non-pCR group. Tumor location distribution showed that antral tumors were slightly more common in the pCR group (43.9% vs 38.5%), but the difference was not statistically significant (P = 0.218). Patients in the pCR group had significantly lower clinical T and N stages than those in the non-pCR group, with cT3 stage patients accounting for 53.7% in the pCR group compared to 34.5% in the non-pCR group (P = 0.022); cN1 stage patients accounted for 48.8% in the pCR group compared to 31.0% in the non-pCR group (P = 0.038, Table 1).

Table 1 Comparison of baseline characteristics between pathological complete response and non-pathological complete response groups, n (%).

Characteristics	pCR group (n = 41)	Non-pCR group (n = 174)	P value
Demographic characteristics
Age (years), median	61.3	60.8	0.742
Gender			0.796
Male	26 (63.4)	114 (65.5)
Female	15 (36.6)	60 (34.5)
BMI (kg/m2), mean	22.4	22.1	0.631
Tumor characteristics
Lauren classification			< 0.001
Intestinal	26 (63.4)	55 (31.6)
Diffuse	10 (24.4)	81 (46.6)
Mixed	5 (12.2)	38 (21.8)
Tumor location			0.218
Cardia	10 (24.4)	46 (26.4)
Body	12 (29.3)	57 (32.8)
Antrum	18 (43.9)	67 (38.5)
Diffuse	1 (2.4)	4 (2.3)
Clinical staging
Clinical T stage			0.022
cT3	22 (53.7)	60 (34.5)
cT4	19 (46.3)	114 (65.5)
Clinical N stage			0.038
cN1	20 (48.8)	54 (31.0)
cN2	14 (34.1)	70 (40.2)
cN3	7 (17.1)	50 (28.8)

BMI: Body mass index; pCR: Pathological complete response.

Comparison of baseline hematological indicators

Patients in the pCR group had significantly lower pre-treatment NLR and PLR compared to those in the non-pCR group, with median values of 2.1 vs 3.4 (P = 0.002) and 117.6 vs 156.3 (P = 0.008), respectively. Receiver operating characteristic curve analysis determined the optimal cutoff values for NLR and PLR to be 2.5 and 135, with areas under the curve of 0.74 (95%CI: 0.65-0.83) and 0.68 (95%CI: 0.59-0.77), respectively. Multivariate analysis revealed that NLR < 2.5 (OR = 2.86, 95%CI: 1.54-5.32, P = 0.001) and PLR < 135 (OR = 2.41, 95%CI: 1.28-4.53, P = 0.006) were independent factors predicting pCR. Pre-treatment albumin levels did not differ significantly between the two groups (P = 0.137), but patients with albumin < 35 g/L had significantly lower pCR rates (15.2% vs 32.7%, P = 0.024). Pre-treatment CEA and CA19-9 Levels also showed no significant differences between the two groups, but patients in the pCR group had markedly greater decreases in tumor markers after chemotherapy compared to the non-pCR group (CEA decrease rate: 78.4% vs 48.2%, P = 0.003; CA19-9 decrease rate: 82.1% vs 53.6%, P = 0.001), and tumor marker decrease ≥ 50% was significantly associated with higher pCR rates (OR = 3.14, 95%CI: 1.72-5.76, P < 0.001, Figure 1).

Figure 1 Receiver operating characteristic curve analysis of neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio for predicting pathological complete response. Figure shows the receiver operating characteristic curves of neutrophil-to-lymphocyte ratio and platelet-to-lymphocyte ratio for predicting pathological complete response, with neutrophil-to-lymphocyte ratio (areas under the curve = 0.74) demonstrating better predictive value than platelet-to-lymphocyte ratio (areas under the curve = 0.68), and optimal cutoff values of 2.5 and 135, respectively. NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; ROC: Receiver operating characteristic; AUC: Areas under the curve; CI: Confidence interval.

Relationship between NAC regimen and pCR

Analysis showed significant differences in pCR rates among different chemotherapy regimens (P = 0.023), with the FLOT regimen showing the best efficacy (27.5%), followed by SOX (18.2%) and XELOX (15.4%) regimens, while the DCF regimen had the lowest pCR rate (11.8%). Comparison of drug combinations revealed that oxaliplatin-containing regimens had significantly higher pCR rates than cisplatin-containing regimens (22.3% vs 12.7%, P = 0.034). The number of chemotherapy cycles was also an important factor affecting pCR, with patients receiving ≥ 4 cycles of chemotherapy showing markedly better pCR rates than those receiving < 4 cycles (24.8% vs 11.2%, P = 0.006). These results suggest that FLOT regimen, oxaliplatin-containing drug combinations, and chemotherapy cycles (≥ 4 cycles) may be key elements for optimizing NAC regimens for gastric cancer (Figure 2).

Figure 2 Comparison of pathological complete response rates among different neoadjuvant chemotherapy approaches in gastric cancer. This figure compares pathological complete response rates in gastric cancer patients receiving neoadjuvant chemotherapy, demonstrating that fluorouracil, leucovorin, oxaliplatin, docetaxel regimen, oxaliplatin-containing compounds, and ≥ 4 chemotherapy cycles achieved significantly higher pathological complete response rates (P < 0.05) than their counterparts. The results highlight key factors that may optimize pathological complete response in gastric cancer treatment, with fluorouracil, leucovorin, oxaliplatin, docetaxel showing the highest response rate (27.5%) among all evaluated regimens. P value was shown as a P < 0.05. FLOT: Fluorouracil, leucovorin, oxaliplatin, docetaxel; SOX: S-1, oxaliplatin; XELOX: Capecitabine, oxaliplatin; DCF: Docetaxel, cisplatin, fluorouracil; CI: Confidence interval; pCR: Pathological complete response.

Univariate logistic regression analysis

Univariate analysis showed multiple factors significantly associated with pCR. Regarding tumor characteristics, Lauren intestinal type showed the strongest correlation (OR = 3.75, 95%CI: 1.82-7.73, P < 0.001); in terms of clinical staging, low clinical T stage (OR = 2.93, 95%CI: 1.37-6.28, P = 0.005) and low clinical N stage (OR = 2.41, 95%CI: 1.15-5.06, P = 0.020) were both favorable factors for achieving pCR; tumor marker changes showed that CEA decrease ≥ 70% after NAC (OR = 3.86, 95%CI: 1.84-8.12, P < 0.001) and CA19-9 decrease ≥ 60% (OR = 2.57, 95%CI: 1.23-5.35, P = 0.012) both suggested higher pCR probability; among baseline inflammatory indicators, pre-treatment NLR < 2.5 (OR = 2.68, 95%CI: 1.41-5.09, P = 0.003) and PLR < 135 (OR = 2.24, 95%CI: 1.18-4.25, P = 0.014) also showed significant positive correlations with pCR; additionally, treatment regimen characteristics including oxaliplatin-containing chemotherapy regimens (OR = 1.97, 95%CI: 1.04-3.74, P = 0.037) and chemotherapy cycles ≥ 4 (OR = 2.61, 95%CI: 1.28-5.32, P = 0.008) similarly contributed to increased pCR rates (Figure 3A).

Figure 3 Forest plots. A: Univariate analysis of factors associated with pathological complete response in gastric cancer. This forest plot illustrates nine significant predictive factors for achieving pathological complete response after neoadjuvant chemotherapy in gastric cancer patients, with Lauren intestinal type and carcinoembryonic antigen decrease ≥ 70% showing the strongest associations (odds ratios of 3.75 and 3.86, respectively). All identified factors demonstrated statistically significant positive correlations with pathological complete response achievement (P < 0.05), providing potential clinical indicators to predict treatment response; B: Independent predictors of pathological complete response following neoadjuvant chemotherapy in gastric cancer. This forest plot illustrates five independent predictors of pathological complete response after neoadjuvant chemotherapy in gastric cancer patients, with carcinoembryonic antigen decrease ≥ 70% (odds ratio = 3.42) and Lauren intestinal type (odds ratio = 3.28) showing the strongest associations. All identified factors demonstrated statistically significant positive correlations with pathological complete response achievement (P < 0.05), providing potential clinical indicators to predict treatment response and optimize patient selection for neoadjuvant therapy. CI: Confidence interval; OR: Odds ratio; CEA: Carcinoembryonic antigen; CA19-9: Carbohydrate antigen 19-9; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; pCR: Pathological complete response.

Multivariate logistic regression analysis

Multivariate logistic regression analysis included all variables with P < 0.1 in the univariate analysis, resulting in five factors being identified as independent predictors of achieving pCR. Lauren intestinal type (OR = 3.28, 95%CI: 1.47-7.32, P = 0.004) and CEA decrease ≥ 70% after NAC (OR = 3.42, 95%CI: 1.53-7.64, P = 0.003) demonstrated the strongest predictive power, followed by chemotherapy cycles ≥ 4 (OR = 2.87, 95%CI: 1.29-6.38, P = 0.010) and low clinical T stage (OR = 2.75, 95%CI: 1.21-6.24, P = 0.016), while pre-treatment NLR < 2.5 (OR = 2.13, 95%CI: 1.06-4.29, P = 0.034) was statistically significant but had relatively weaker predictive strength. These findings provide reliable predictive tools for clinical assessment of gastric cancer patients' response to NAC (Figure 3B).

Relationship between inflammatory indicators and pCR

Pre-treatment inflammatory indicator levels showed significant correlation with pCR achievement after NAC in gastric cancer. Research analysis demonstrated that NLR and PLR were both important predictive markers. Specifically, patients with NLR < 2.5 had a pCR rate of 29.6%, significantly higher than patients with NLR ≥ 2.5 (13.4%) (P = 0.003); patients with PLR < 135 had a pCR rate of 26.1%, higher than patients with PLR ≥ 135 (13.9%) (P = 0.017). The interaction between the two inflammatory indicators was even more pronounced, with patients having both low NLR (< 2.5) and low PLR (< 135) showing the highest pCR rate (33.8%), while those with both high NLR (≥ 2.5) and high PLR (≥ 135) showing the lowest pCR rate (10.7%) (P < 0.001), presenting a clear gradient trend; groups with abnormal single indicators showed intermediate pCR rates (low NLR/high PLR group: 22.4%, high NLR/Low PLR group: 16.9%, Table 2).

Table 2 Relationship between inflammatory indicators and pathological complete response in gastric cancer.

Group	NLR	PLR	pCR rate (%)	P value	Clinical significance
Individual NLR analysis
Low NLR group	< 2.5	-	29.6	0.003	Significantly higher pCR
High NLR group	≥ 2.5	-	13.4	-	Lower pCR rate
Individual PLR analysis
Low PLR group	-	< 135	26.1	0.017	Significantly higher pCR
High PLR group	-	≥ 135	13.9	-	Lower pCR rate
Combined analysis
Optimal group	< 2.5	< 135	33.8	< 0.001	Highest pCR rate
Single low NLR	< 2.5	≥ 135	22.4	-	Intermediate pCR rate
Single low PLR	≥ 2.5	< 135	16.9	-	Intermediate pCR rate
Poor prognosis group	≥ 2.5	≥ 135	10.7	-	Lowest pCR rate

NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; pCR: Pathological complete response.

Impact of chemotherapy regimen and cycle number

The choice of NAC regimen and the number of chemotherapy cycles significantly influenced pCR rates in gastric cancer patients. Analysis indicated that oxaliplatin-containing chemotherapy regimens (FLOT, SOX, XELOX) generally performed better than cisplatin-containing regimens (DCF), with pCR rates of 22.3% and 12.7%, respectively (P = 0.034). Among these, the FLOT regimen demonstrated the best efficacy across various cycle numbers, with a pCR rate of 27.5%, while SOX (18.2%) and XELOX (15.4%) regimens were next, and the DCF regimen showed the lowest efficacy (11.8%). The number of chemotherapy cycles was also a key factor affecting pCR achievement, with patients receiving ≥ 4 cycles of chemotherapy having significantly higher pCR rates (24.8%) than those receiving < 4 cycles (11.2%) (P = 0.006), and showing a clear dose-response relationship: The pCR rate was 11.2% in the 2-3 cycle group, rose to 21.6% in the 4-5 cycle group, reached 28.3% in the 6-cycle group, and was 27.9% in the > 6 cycle group, indicating that pCR rates increased with the number of chemotherapy cycles but plateaued after exceeding 6 cycles, suggesting that a reasonable number of NAC cycles should be 4-6 cycles, considering the balance between efficacy and toxicity (Table 3).

Table 3 Influence of chemotherapy regimens and cycles on pathological complete response rates.

Factor	Category	pCR rate (%)	P value
Chemotherapy regimen	Platinum compound type
	Oxaliplatin-containing (FLOT, SOX, XELOX)	22.3%	0.034
	Cisplatin-containing (DCF)	12.7%	-
	Specific regimen		0.023
	FLOT	27.5%
	SOX	18.2%
	XELOX	15.4%
	DCF	11.8%
Chemotherapy cycles	General comparison
	≥ 4 cycles	24.8%	0.006
	< 4 cycles	11.2%	-
	Detailed cycle ranges	-	N/A
	2-3 cycles	11.2%
	4-5 cycles	21.6%
	6 cycles	28.3%
	> 6 cycles	27.9%

FLOT: Fluorouracil, leucovorin, oxaliplatin, docetaxel; SOX: S-1, oxaliplatin; XELOX: Capecitabine, oxaliplatin; DCF: Docetaxel, cisplatin, fluorouracil; pCR: Pathological complete response; N/A: Not applicable.

DISCUSSION

Gastric cancer, as one of the most common malignant tumors globally, has a particularly high disease burden in East Asia. Despite advances in diagnostic and treatment technologies, the prognosis for patients with locally advanced gastric cancer remains unfavorable. NAC, as an important component of multimodal treatment strategies, has become a key measure for improving the prognosis of patients with locally advanced gastric cancer by downstaging tumors, increasing radical resection rates, and eliminating micrometastases. pCR is considered the “gold standard” for evaluating the effectiveness of neoadjuvant treatment, and multiple studies have confirmed that patients achieving pCR have significantly longer disease-free survival and overall survival. However, the pCR rate after NAC for gastric cancer is relatively low and influenced by complex factors, posing challenges for clinical decision-making[17-19].

Tumor biological characteristics have important effects on chemosensitivity. Lauren classification, as an important basis for gastric cancer pathological classification, reflects different growth patterns and biological behaviors of tumors. Intestinal-type gastric cancer typically has more regular glandular structures with tighter intercellular adhesion, while diffuse-type presents as infiltration of poorly differentiated individual tumor cells with reduced expression of cell adhesion molecules. These differences may lead to variations in the penetration and sensitivity of chemotherapy drugs between the two types of tumors. Intestinal-type gastric cancer cells usually proliferate more rapidly, while most chemotherapy drugs primarily target rapidly dividing cells, which may explain the greater sensitivity of intestinal-type gastric cancer to chemotherapy[20-22].

Inflammatory responses play complex roles in tumor occurrence, development, and treatment response. NLR and PLR, as markers of systemic inflammation, can reflect the status of the tumor microenvironment. High NLR and PLR may indicate the predominance of pro-tumor inflammatory environments, including increased tumor-associated macrophages, myeloid-derived suppressor cells, and other immunosuppressive cells, which can produce various cytokines and growth factors, promote tumor angiogenesis, suppress anti-tumor immune responses, and potentially reduce the effectiveness of chemotherapy drugs. Conversely, low NLR and PLR may reflect a more favorable immune microenvironment, facilitating the anti-tumor effects of chemotherapy drugs[23-25].

Tumor burden is a key factor influencing treatment response. Patients with low clinical T and N stages have a lower total number of tumor cells, presenting less resistance to overcome with treatment; additionally, early-stage tumors typically have better blood supply, allowing better penetration of chemotherapy drugs. As tumors progress, hypoxic regions and heterogeneous clones may develop within the tumor, with cells in these regions often being insensitive to chemotherapy or even developing resistance. Furthermore, advanced tumors often have more cancer stem cells, which have intrinsic resistance to conventional chemotherapy[26-28].

Tumor markers have unique value in monitoring treatment response. CEA, as a commonly used serum marker for gastrointestinal tumors, can partially reflect changes in tumor burden through its level changes. Significant decreases in CEA after chemotherapy suggest effective tumor cell killing, which is closely related to pathological response. Dynamic changes in markers have more clinical significance than single measurements, providing early information about tumor response to treatment and helping to adjust treatment strategies in a timely manner[29-31].

The choice of chemotherapy regimen and dosing intensity are equally crucial. This study found that oxaliplatin-containing regimens (especially FLOT) were superior to cisplatin-containing regimens, which may be related to the different mechanisms of action and pharmacokinetic properties of oxaliplatin and cisplatin, with the former having stronger cytotoxic effects and relatively more manageable adverse reactions. The number of chemotherapy cycles represents the time and cumulative dose of drug exposure, and a sufficient number of cycles (≥ 4 cycles) ensures sustained drug action, preventing tumor cells from recovering and ultimately leading to cell death. However, the study showed that pCR rates did not increase significantly after exceeding 6 chemotherapy cycles, suggesting the existence of a therapeutic benefit plateau, where excessive prolongation of chemotherapy may increase toxicity with limited benefit[32,33]. This study confirmed that Lauren intestinal type, low clinical T stage, CEA decrease ≥ 70% after treatment, low baseline NLR, and ≥ 4 chemotherapy cycles are independent predictive factors for achieving pCR after NAC in gastric cancer. For patients with multiple favorable factors (such as Lauren intestinal type, low NLR < 2.5, low PLR < 135), we recommend actively advocating for the FLOT regimen and ensuring ≥ 4 chemotherapy cycles to maximize pCR achievement rates. For patients with poor predictive factors, clinicians should consider direct surgery or adjusted treatment strategies. Regarding integration with emerging biomarkers, we suggest future studies combine our predictive model with molecular markers such as microsatellite instability status and programmed death-ligand 1 expression to construct more comprehensive individualized treatment decision tools. We will also discuss how to incorporate these predictive factors into existing clinical pathways and treatment guidelines.

Limitations

This retrospective study has several limitations including inherent selection bias, relatively small sample size (n = 215), and potential confounding factors that limit causal inference. The analysis lacks modern molecular biomarkers (microsatellite instability, programmed death-ligand 1, human epidermal growth factor receptor 2) and inadequately controls for important clinical variables such as performance status and comorbidities. The single-country cohort and extended study period (2015-2023) may limit generalizability due to population differences and evolving treatment protocols. Future prospective, multi-center studies with larger cohorts, comprehensive biomarker analysis, and long-term survival data are needed to validate these predictive factors.

CONCLUSION

These findings help identify patient populations most likely to benefit from NAC, providing a basis for clinical decision-making. For example, NAC can be actively recommended for patients with multiple favorable factors; while for patients expected to benefit little, direct surgery or adjusted treatment strategies can be considered.