In China, over 80% of gastric cancer cases are at clinical stage III or beyond, making them not a local problem but a regional problem. Although local and regional disease control are frequently, but not always, feasible via R0 resection and extended lymphadenectomy, surgical treatment cannot reduce the extraregional recurrence risks that specifically result from hematogenous (visceral) and trans-serosal (peritoneal) progression.
Several approaches to perioperative adjuvant therapy of gastric cancer have shown survival benefits, including postoperative chemoradiation, postoperative systemic chemotherapy, postoperative intraperitoneal chemotherapy, and perioperative systemic chemotherapy [
While preliminary studies have shown the clinical use of neoadjuvant chemotherapy to improve prognoses in advanced gastric cancer (AGC), the optimal regime remains to be determined. Most neoadjuvant chemotherapy regimens were based on fluorouracil (5-FU) and cisplatin. In recent years, the FOLFOX chemotherapy regimen (oxaliplatin, leucovorin, and 5-FU) has been established as a staple of colorectal cancer chemotherapy [
This retrospective analysis included consecutive patients with AGC admitted to the Department of Surgical Oncology, Zhongnan Hospital of Wuhan University from January 1, 2006 to January 30, 2013. Major inclusion criteria were age from 20 to 80 years, WHO performance status 0 to 2, histologically proven gastric carcinoma, no absolution contraindication to surgery and no evidence of distant metastases, as evaluated by CT, chest radiography, and ultrasonography, no history of prior gastric surgery; no previous chemotherapy or radiotherapy, no uncontrolled infectious or cardiac disease; adequate hepatic and renal function, and no previous or other concurrent malignant tumors.
A proportion of these patients, who were diagnosed as resectable T3 or T4 cancers with or without nodal involvement (as determined by CT), underwent preoperative chemotherapy with a combination regimen (FOLFOX4) consisting of oxaliplatin, leucovorin, and 5-fluorouracil (5-FU) followed by curative surgical resection and postoperative chemotherapy mainly FOLFOX (neoadjuvant group). The adjuvant group selection was divided into two steps. During the first step, patients with T3/T4 gastric cancer treated with surgery first and followed by adjuvant FOLFOX chemotherapy were selected by surgeon investigators from their records. During the second step, the principal investigator double-checked the medical records of the potentially eligible patients by recontacting the investigators to ensure that the eligibility criteria had been applied homogeneously. During double-checking, the investigator was blinded to the detailed characteristics of the patients in the neoadjuvant group. We adopted a 1 : 1.5 ratio of exposed: unexposed proportion. This study was approved by the institutional review board of the hospital and informed consent was obtained from each patient.
In the neoadjuvant group, patients received 2 to 6 cycles of preoperative FOLFOX4 regiments consisting of oxaliplatin at 85 mg/m2 on day 1 and leucovorin at 200 mg/m2 administered intravenously for 2 hours followed by 5-FU at 400 mg/m2 as a bolus followed infusion 5-FU at 600 mg/m2 by continuous infusion for 22 hours on days 1 and 2. Antiemetics and granulocyte-colony stimulating factor (G-CSF) were prescribed when required. Before each cycle of chemotherapy, the tumor markers including CEA, CA125, and CA199 and a complete blood count were obtained and blood urea nitrogen, electrolyte, serum creatinine level, and liver function were determined. Adverse events graded according to the National Cancer Institute Common Toxicity Criteria version 4.0 [
Assessment of response to neoadjuvant therapy was based on reduction of primary tumor size measured by CT scan (RECIST evaluation criteria) [
Resection of the gastric tumor was performed within 1 week after hospital admission for patients in the adjuvant group and 3 to 4 weeks after completion of chemotherapy in the neoadjuvant chemotherapy group. After laparotomy, the extent of dissection and whether the surgical procedure was likely to be curative (R0) were decided; R1 indicated microscopic evidence of tumor cells at the margin of the resection, whereas R2 indicated macroscopic evidence of tumors beyond the margin of the resection. For patients with R0 resections, a D2 lymphadenectomy was performed according to the Japanese gastric cancer treatment guidelines 2010 (ver.3) [
All patients were followed until death or until the date of last followup as of 30 January, 2013. The presence of a relapse was determined by appropriate imaging studies. Overall survival (OS) was defined as the interval between commencement of treatment and death. Progression-free survival (PFS) was measured from commencement of treatment to occurrence of an event—relapse or death—whichever came first. Data on patients who were event-free were censored on the date of last followup.
Immunolocalization of Ki-67, c-erbB-2, CD34, and MMP9 (matrix metalloproteinase-9) was performed using avidin-biotin-peroxidase complex (ABC) method. Briefly, tissue slides were first deparaffinized in xylene, ethanol, and water; then, the slides were pretreated in 0.01 M citrate buffer (pH 6.0) and heated in a microwave oven (98°C) for 15 min. For staining, endogenous peroxidase activity was blocked by immersing in 3% hydrogen peroxide/methanol buffer for 20 min at room temperature. After washing in PBS, the slides were incubated with the primary antibodies for Ki-67 (MAB-0129, Maixin Biotechnology, China, working solution), c-erbB-2 (RMA-0555, Maixin Biotechnology, China, working solution), CD34 (MAB-0034, Maixin Biotechnology, China, working solution), and MMP9 (sc13595, Santa Cruz, USA, dilution 1/300) overnight at 4°C. Then the sections were washed with PBS and incubated with polymerase auxiliaries for 20 min. After washing in PBS, the sections were incubated with the biotinylated secondary antibody for 60 min at room temperature, and finally DAB (diaminobenzidine) was visualized. As a negative control, primary antibody was replaced with Tris-buffered saline on sections that were proven to be positive for Ki-67, c-erbB-2, CD34, and MMP9 in preliminary experiments.
All slides were independently observed by two investigators who were blinded to the clinicopathological characteristics of patients. A consensus score was agreed for each slide by the investigators.
The percentage of positively stained cells was calculated after 100 cells were counted in more than 5 high-power (×400) fields. The following definitions were made: Ki-67: more than 10% positive staining in nuclei was defined as positive staining; c-erbB-2 and MMP9: more than 10% positive staining in cytoplasm was defined as positive. For MVD, assessment involved the initial identification of highly vascular areas by scanning the entire section at low magnification (×100), defined as areas having the highest density of CD34-positive cells; microvessels were counted (×400 field) by light microscopy in each of the five most vascularized areas, and necrotic and ulcerated areas were avoided [
All data analyses were performed using the SPSS statistical software program, version 19.0 (SPSS Inc., Chicago, IL, USA) for Windows. Categorical data were compared using chi-square tests. Tumor markers, number of nodes harvested, and metastatic lymph nodes were compared using nonparametric Wilcoxon test. The Kaplan-Meier analysis was used to estimate survival rates and analyzed by Log rank test. The Cox proportional-hazards model was used to calculate the hazard ratios. Univariate analysis of associations was determined using Spearman’s correlation analysis.
Between January, 2006, and January, 2012, 58 patients were recruited, 23 in the neoadjuvant group and 35 in adjuvant group. As shown in Table
Patient demographics and tumor characteristics.
Items | Neoadjuvant group |
Adjuvant group |
|
---|---|---|---|
Median age (yr) (range) | 58 (34–79) | 57 (31–80) | 0.760 |
Gender: |
|||
Male | 15 (65.2) | 22 (62.9) | 0.855 |
Female | 8 (34.8) | 13 (37.1) | |
Anatomic location: |
|||
Proximal | 4 (17.4) | 3 (8.6) | 0.490 |
Body | 7 (30.4) | 9 (25.7) | |
Distal | 12 (52.2) | 23 (65.7) | |
Tumor grade: |
|||
Well/moderately differentiated | 6 (26.1) | 8 (22.8) | 0.701 |
Poorly differentiated | 11 (47.8) | 20 (57.1) | |
Mucinous/signet ring cell cancer | 4 (17.4) | 6 (17.2) | |
Others | 2 (8.7) | 1 (2.9) | |
Pretreatment clinical T-stage (CT): |
|||
Stage 3 | 17 (73.9) | 28 (80.0) | 0.587 |
Stage 4 | 6 (26.1) | 7 (20.0) |
Serum levels of CEA, CA125, and CA199 were significantly decreased after neoadjuvant chemotherapy (
Tumor marker responses after the neoadjuvant chemotherapy.
Items | Preneoadjuvant chemotherapy |
Postneoadjuvant chemotherapy |
|
---|---|---|---|
CEA median (range) | 0.84 (0.05–14.40) | 0.70 (0.17–13.90) | 0.030 |
CA125 median (range) | 0.77 (0.25–14.26) | 0.45 (0.19–6.28) | 0.003 |
CA199 median (range) | 0.24 (0.05–59.46) | 0.20 (0.00–25.31) | 0.005 |
CEA: carcinoembryonic antigen; CA125: cancer antigen 125; CA199: carbohydrate antigen 199.
The R0 resections rate was 82.6% (19/23) in the neoadjuvant group and 80.0% (28/35) in the adjuvant group (
Surgical and pathological results.
Items | Neoadjuvant group |
Adjuvant group |
|
---|---|---|---|
Margins: |
|||
R0 | 19 (82.6) | 28 (80.0) | 0.807 |
R1/2 | 4 (17.4) | 7 (20.0) | |
Postoperative complication: |
|||
Anastomotic leak | 1 (4.3) | 0 (0) | 0.507 |
Respiratory infection | 2 (8.7) | 3 (8.6) | |
Postoperative hemorrhage | 1 (4.3) | 1 (2.9) | |
Postoperative bowel obstruction | 0 (0) | 2 (5.7) | |
Pathological staging | |||
T-stage: |
|||
T1 | 1 (4.3) | 2 (5.7) | 0.335 |
T2 | 5 (21.7) | 2 (5.7) | |
T3 | 4 (21.7) | 8 (22.9) | |
T4 | 13 (52.1) | 23 (65.7) | |
N-stage: |
|||
N0 | 8 (34.8) | 2 (5.7) | 0.034 |
N1 | 3 (13.0) | 9 (25.7) | |
N2 | 4 (17.4) | 10 (28.6) | |
N3 | 8 (34.8) | 14 (40.0) | |
Number of nodes harvested: |
16 (0–49) | 13 (3–40) | 0.886 |
Number of metastatic lymph nodes: |
3 (0–14) | 6 (0–27) | 0.040 |
According to the Japanese Gastric Cancer Research pathological evaluation standard of chemotherapy effects, no patient had grade 3 pathological response, 8 (34.8%) had grade 2 response, 10 (43.5%) had grade 1b response, 3 (13.0%) had grade 1a response, and 2 (8.7%) had grade 0 response. The median number of dissected lymph nodes was similar in both arms, (16 (0–49) in neoadjuvant group versus 13 (3–40) in adjuvant group). The median number of positive lymph nodes was 3 (0–14) in the neoadjuvant group versus 6 (0–27) in the adjuvant group (
The most common toxicities were hematological and gastrointestinal toxicity. In all cases, the toxic effects resolved after treatment completed and no treatment was terminated because of toxicity. There was no clinically significant difference in the incidence of the toxic effects associated with the chemotherapy between two groups (Table
Adverse effects associated with preoperative and postoperative chemotherapy.
Adverse events | Neoadjuvant group |
Adjuvant group |
|
---|---|---|---|
Neutropenia: |
|||
Grades 1-2 | 6 (26.1) | 7 (20.0) | NS |
Grades 3-4 | 3 (13.0) | 3 (8.8) | |
Nausea and vomiting: |
|||
Grades 1-2 | 4 (17.4) | 8 (22.9) | NS |
Grades 3-4 | 1 (4.3) | 2 (5.7) | |
Liver toxicity: |
|||
Grades 1-2 | 1 (4.3) | 2 (5.7) | NS |
Grades 3-4 | 0 (0) | 0 (0) | |
Stomatitis: |
|||
Grades 1-2 | 2 (8.7) | 1 (2.9) | NS |
Grades 3-4 | 0 (0) | 0 (0) | |
Diarrhea: |
|||
Grades 1-2 | 2 (8.7) | 2 (5.7) | NS |
Grades 3-4 | 0 (0) | 2 (5.7) | |
Neurologic effects: |
|||
Grades 1-2 | 4 (17.4) | 2 (5.7) | NS |
Grades 3-4 | 0 (0) | 0 (0) | |
Skin effects: |
|||
Grades 1-2 | 1 (4.3) | 1 (2.9) | NS |
Grades 3-4 | 0 (0) | 0 (0) |
The median followup was 26.0 months (10.0–61.0 months) in the neoadjuvant group and 31.0 months (15.0–72.0 months) in the adjuvant group. The median OS for patients in the neoadjuvant and adjuvant groups were 29.0 months (95% CI, 25.3–32.7 months) and 22.0 months (95% CI, 18.2–25.8 months), respectively. The median PFS were 26.0 months (95% CI, not reached) and 18.0 months (95% CI, 14.4–21.6) respectively. The 3-year OS was 73.9% (95% CI, 54.6%–93.2%) in the neoadjuvant group and 40% (95% CI, 30.1%–49.9%) in the adjuvant group (
Kaplan-Meier analysis of the neoadjuvant group and adjuvant group. OS (overall survival) and PFS (progression-free survival) between the two groups ((a), (b)); OS between the two groups without metastasis (c); OS between the two groups with metastasis (d); ((e), (f)) Kaplan-Meier analysis of the Ki-67 positive and negative patients in the neoadjuvant group. Ki-67 negative patients had greater OS benefits than Ki-67 positive patients in the neoadjuvant group (e). There was also a trend towards better PFS benefits in Ki-67 negative patients than positive patients, although the difference did not reach statistical significance (f).
Multivariate Cox proportional hazards analysis indicated that neoadjuvant chemotherapy (hazards ratio [HR] = 0.202 (95% CI, 0.072–0.570),
A total of 20 specimens were suitable for immunohistochemical analyses in the neoadjuvant group and 31 in the adjuvant group.
Ki-67 positive rate in the neoadjuvant group (60.0%, 12/20) was slightly lower than that in the adjuvant group (74.2%, 23/31;
Immunohistochemical staining of Ki-67 ((a) brown stain in nuclei), c-erbB-2 ((b) brown stain in membrane and cytoplasm), MMP-9 ((c) brown stain in membrane and cytoplasm) and microvessels. (d) Original magnification 100x; insets 400x. All tissues were adenocarcinoma of GC.
The positive expression rate of c-erbB-2 in the neoadjuvant group (10.0%, 2/20) was similar to that of the adjuvant group (16.1%, 5/31;
The positive expression rate of MMP-9 in the neoadjuvant group (55.0%, 11/20) was similar to that in the adjuvant group (67.7%, 21/31;
The median MVD value was 19. When this median value was determined as the cut-off point; the high MVD rate in the neoadjuvant group (45.0%, 9/20) was similar to that in the adjuvant group (51.6%, 16/31;
The prognosis of AGC is poor despite seemingly curative resection [
The use of oxaliplatin, 5-FU, and leucovorin (FOLFOX) in this setting has been under investigation, which prompted this study to investigate the effectiveness of neoadjuvant FOLFOX in the treatment of gastric cancer. Since 2001, the FOLFOX regimen had become one of the most common treatments for AGC. Al-Batran et al. [
Clinical response rate is an important indicator of judging the efficacy of chemotherapy; however, a variety of adverse events caused by chemotherapy should also be observed. Pre-operative ECF has been previously reported to achieve response rates of 49% to 56% [
The survival benefits of neoadjuvant chemotherapy for AGC warrant further discussion. In this study, the three-year OS in the neoadjuvant group was 73.9% (95% CI, 54.6%–93.2%) as opposed to 40.0% (95% CI, 30.1%–49.9%) in the adjuvant group. This finding is consistent with two recent large scale randomized trials. MAGIC and FNLCC trials [
We also investigated the potential biomarkers that may further evaluate the efficacy of the neoadjuvant chemotherapy. We focused on the expression of Ki-67, c-erbB-2, MMP-9, and CD34 in postoperation samples in order to investigate their reactions towards neoadjuvant chemotherapy and analyze their prognostic and predictive potential in the patients treated with FOLFOX4 neoadjuvant regimen. A lower Ki-67 positive expression rate was observed in the neoadjuvant group implying that FOLFOX4 chemotherapy may have influenced the tumor cell proliferation in gastric cancer. We also show that nuclear expression of Ki-67 is negatively associated with histopathological response to neoadjuvant chemotherapy implying that Ki-67 could be a promising predictive marker. However, the pretreatment Ki-67 level still needs to be investigated in order to further assess the role of Ki-67 as a predictive factor of responsiveness to chemotherapy or other biological therapies. Moreover, we have also shown that nuclear Ki-67 expression correlates with poor OS (
Our study is limited by the retrospective design and small sample size. However, we have provided evidence that the FOLFOX4 regimen as neoadjuvant chemotherapy could reduce lymph node metastasis and improve survival without increasing adverse events in patients with AGC. We also have provided evidence that postneoadjuvant chemotherapy Ki-67 level could be a promising predictive biomarker in patients with advanced gastric cancer who receive FOLFOX4 regimen. To further validate FOLFOX4 regimen that is likely to have important clinical implications for patients receiving neoadjuvant chemotherapy for AGC, randomized controlled trial with larger sample size will be done in the near future in our department based on the preliminary but promising results in this study.
The authors declare that there is no conflict of interests regarding the publication of this paper.
This study was supported by the National Natural Science Foundation of China (no. 81230031/H18) and the National Science and Technology Major Project (no. 2012ZX10002012-12).