www.impactjournals.com/oncotarget/
Oncotarget, 2017, Vol. 8, (No. 8), pp: 14058-14067 Clinical Research Paper
A novel grade-lymph node ratio model predicts the prognosis of the advanced gastric cancer patients after neoadjuvant radiotherapy Jianjun Liu1,*, Mingxue Su2,*, Jing Wang3,*, Gan Zhang1, Jing Zhou1, Anqing Zhang1, Zixue Ren1, Xucai Zheng1, Shikai Hong1, Shengying Wang1 and Rongxin Zhang1 1
Department of Head - Neck and Thoracic Surgery, Anhui Provincial Cancer Hospital, West branch of Anhui Province Hospital, Hefei, China 2
Department of Infectious Disease Epidemiology, Lu’an People’s Hospital, Lu’an, China
3
Department of Urologic Surgery, Anhui Provincial Cancer Hospital, West branch of Anhui Province Hospital, Hefei, China
*
These authors have contributed equally to this work
Correspondence to: Rongxin Zhang, email: [email protected] Correspondence to: Shengying Wang, email: [email protected] Keywords: neoadjuvant radiotherapy; gastric cancer; survival analysis; SEER Received: August 23, 2016
Accepted: October 05, 2016
Published: October 11, 2016
ABSTRACT Although local advanced gastric cancer (AGC) could benefit from neoadjuvant radiotherapy (NRT), there are few studies evaluating patients’ survival after NRT. In current study, we aimed to investigate the value of prognostic factors in AGC patients after NRT and to evaluate whether post-therapy pathological characteristics were predictive factors in these patients. We retrospectively analyzed AGC patients who underwent NRT from Surveillance, Epidemiology, and End Results (SEER) Database. The patients’ clinical and post-therapy pathological characteristics were analyzed. The best cutoff points for continuous variables were identified by X-tile. The discrimination of risk factors were compared by receiver operating characteristic (ROC) curve. As a result, 1,429 AGC patients were included into this study. In the multivariate analysis, the lymph nodes status and histology grade were significant risk factors for DSS (disease special survival). Then, we propose a novel Grade-lymph node Ratio (GR) staging system for the AGC patients’ survival prognosis. Clearly, the new G-R staging system has a more-accurate 3-year and 5-year DSS prediction than the AJCC staging system (p = 0.001, 0.007, respectively). In conclusions, the current large, general population-based study demonstrated that the G-R staging system resulting in more-accurate DSS prediction. It could be regarded as a reliable classification for AGC patients after NRT in future.
INTRODUCTION
not obvious in early stage, most patients were diagnosed with advanced stage in most countries. Recently, the MAGIC trials demonstrated that the neoadjuvant chemotherapy improved survival for AGC patients. [5] Meanwhile, several phase I/II trails also demonstrated a survival benefit for AGC patients after NRT. [6-8] Indeed, the neoadjuvant therapy can achieve a clinical downstaging before resection and increase the possibility of R0 resection for AGC patients. [9-11] As a result, the neoadjuvant therapy were widely used for potentially resectable AGC before surgery. Based on those promising results, the neoadjuvant
Gastric cancer is the fourth most common malignancy and second leading cause of cancer-related death worldwide. [1] In 2014, approximately 22,220 new cases were diagnosed and 10,990 deaths attributed to gastric cancer in United States. Although the incidence has declined recently, the 5-year survival was less than 30%. [2, 3] Completed resection with lymph node dissection is the only potential curative treatment for resectable cancer. [4] However, the cancer symptoms usually were www.impactjournals.com/oncotarget
14058
Oncotarget
Table 1: Characteristic of patients from SEER database Characteristic
Patients(n = 1,429) NO.
Age (years) Median Range Gender Male Female Race White Black AI API Unknown Tumor size (cm) Median(n = 1027) Range Tumor location Cardia Fundus Body Antrum Pylorus Lesser curvature Greater curvature Overlapping Unknown Grade Well differentiated Moderately differentiated Poorly differentiated Undifferentiated Depth of invasion Mucosa or submucosa Proper muscle Subserosa Serosa Adjacent invasion Unknown Number of positive LN. 0 1 to 2 3 to 6 7 to 15 16 or more Positive LN (Mean±SD) Total LN (Mean±SD)(n=1411) AJCC Stage IA www.impactjournals.com/oncotarget
% 60.8±10.4 14 to 88
1206 223
84.4 15.6
1278 61 18 70 2
89.4 4.3 1.3 4.9 0.1 4.6±3.4 0.1 to 50
1299 9 17 32 4 29 9 18 12
90.9 0.6 1.2 2.2 0.3 2.0 0.6 1.3 0.8
62 510 827 30
4.3 35.7 57.9 2.1
134 215
9.4 15.0
667 260 125 28
46.7 18.2 8.7 2.0
757 342 231 81 18
53.0 23.9 16.2 5.7 1.3 1.8±3.3 14.3±9.9
100
7.0 14059
Oncotarget
IB IIA IIB IIIA IIIB IIIC Unknown Adjuvant radiotherapy Yes No
146 410 307 196 173 69 28
10.2 28.7 21.5 13.7 12.1 4.8 2.0
60 1369
4.2 95.8
Abbreviations: AI, American Indian or Alaska Native; API, Asian or Pacific Islander; LN, lymph node; AJCC, American Joint Committee on Cancer. therapy may be a potential standard treatments for AGC patients. However, the prognostic value of post-therapy pathologic characteristic were still unclear, especially in patients underwent NRT. [10, 12-15] In current study, we aimed to investigate the value of prognostic factors in AGC patients and to evaluate whether post-therapy pathological characteristics were predictive factors of survival in these patients.
the inclusion criteria were analyzed in current study. The patients’ characteristics were listed in Table 1. Overall, there were 1,206 male patients and 223 female patients, and the median age was 60.8 years old. There were also 60 (4.2%) patients who received adjuvant radiotherapy after curative resection. The mean number of positive lymph node was 1.8±3.3, and the mean total examined number of lymph node was 14.3±9.9. With a median follow-up of 31.9 months, there were 760 (53.2%) patients died before the analysis of the present study, and 102 (13.4%) patients of them were died because of other causes.
RESULTS Patients and demographics
Survival and lymph node ratio categories
A total of 1,429 patients between January 1998 and November 2013 in the SEER database who met all
The median DSS for all patients was 35 months, and the 1-year DSS, 3-year DSS, 5-year DSS were 83.0%, 48.9% and 39.0%, respectively. Lymph node ratio was
Figure 1: X-tile analysis identified the best cutoff points for mLNR A., and validated by Kaplan-Meier Curve B. Abbreviation: mLNR, metastatic lymph node ratio.
www.impactjournals.com/oncotarget
14060
Oncotarget
Table 2: Prognostic factors for DSS of all the patients Univariate analysis
Multivariate Analysis
HR
95% CI
p
Age
1.005
0.998 to 1.013
0.167
Gender (Female / Male)
0.879
0.708 to 1.091
0.243
Race API/ nonAPI
95% CI
p
0.062 0.762
0.573 to 1.014
Location
0.939
Antrum
ref
Fundus
0.771
0.220 to 2.707
Body
1.255
0.520 to 3.029
Cardia
1.336
0.771 to 2.315
Pylorus
1.464
0.417 to 5.139
Lesser curvature
1.462
0.678 to 3.155
Greater curvature
1.371
0.488 to 3.847
Overlapping
1.547
0.661 to 3.620
Grade Well /Moderately differentiated Poorly differentiated/ Undifferentiated Total LN
0.714
0.608 to 0.838
1.002
0.994 to 1.010
0.584
Tumor Size (n = 1027)
1.001
0.999 to 1.004
0.414
Oncotarget, 2017, Vol. 8, (No. 8), pp: 14058-14067 Clinical Research Paper
A novel grade-lymph node ratio model predicts the prognosis of the advanced gastric cancer patients after neoadjuvant radiotherapy Jianjun Liu1,*, Mingxue Su2,*, Jing Wang3,*, Gan Zhang1, Jing Zhou1, Anqing Zhang1, Zixue Ren1, Xucai Zheng1, Shikai Hong1, Shengying Wang1 and Rongxin Zhang1 1
Department of Head - Neck and Thoracic Surgery, Anhui Provincial Cancer Hospital, West branch of Anhui Province Hospital, Hefei, China 2
Department of Infectious Disease Epidemiology, Lu’an People’s Hospital, Lu’an, China
3
Department of Urologic Surgery, Anhui Provincial Cancer Hospital, West branch of Anhui Province Hospital, Hefei, China
*
These authors have contributed equally to this work
Correspondence to: Rongxin Zhang, email: [email protected] Correspondence to: Shengying Wang, email: [email protected] Keywords: neoadjuvant radiotherapy; gastric cancer; survival analysis; SEER Received: August 23, 2016
Accepted: October 05, 2016
Published: October 11, 2016
ABSTRACT Although local advanced gastric cancer (AGC) could benefit from neoadjuvant radiotherapy (NRT), there are few studies evaluating patients’ survival after NRT. In current study, we aimed to investigate the value of prognostic factors in AGC patients after NRT and to evaluate whether post-therapy pathological characteristics were predictive factors in these patients. We retrospectively analyzed AGC patients who underwent NRT from Surveillance, Epidemiology, and End Results (SEER) Database. The patients’ clinical and post-therapy pathological characteristics were analyzed. The best cutoff points for continuous variables were identified by X-tile. The discrimination of risk factors were compared by receiver operating characteristic (ROC) curve. As a result, 1,429 AGC patients were included into this study. In the multivariate analysis, the lymph nodes status and histology grade were significant risk factors for DSS (disease special survival). Then, we propose a novel Grade-lymph node Ratio (GR) staging system for the AGC patients’ survival prognosis. Clearly, the new G-R staging system has a more-accurate 3-year and 5-year DSS prediction than the AJCC staging system (p = 0.001, 0.007, respectively). In conclusions, the current large, general population-based study demonstrated that the G-R staging system resulting in more-accurate DSS prediction. It could be regarded as a reliable classification for AGC patients after NRT in future.
INTRODUCTION
not obvious in early stage, most patients were diagnosed with advanced stage in most countries. Recently, the MAGIC trials demonstrated that the neoadjuvant chemotherapy improved survival for AGC patients. [5] Meanwhile, several phase I/II trails also demonstrated a survival benefit for AGC patients after NRT. [6-8] Indeed, the neoadjuvant therapy can achieve a clinical downstaging before resection and increase the possibility of R0 resection for AGC patients. [9-11] As a result, the neoadjuvant therapy were widely used for potentially resectable AGC before surgery. Based on those promising results, the neoadjuvant
Gastric cancer is the fourth most common malignancy and second leading cause of cancer-related death worldwide. [1] In 2014, approximately 22,220 new cases were diagnosed and 10,990 deaths attributed to gastric cancer in United States. Although the incidence has declined recently, the 5-year survival was less than 30%. [2, 3] Completed resection with lymph node dissection is the only potential curative treatment for resectable cancer. [4] However, the cancer symptoms usually were www.impactjournals.com/oncotarget
14058
Oncotarget
Table 1: Characteristic of patients from SEER database Characteristic
Patients(n = 1,429) NO.
Age (years) Median Range Gender Male Female Race White Black AI API Unknown Tumor size (cm) Median(n = 1027) Range Tumor location Cardia Fundus Body Antrum Pylorus Lesser curvature Greater curvature Overlapping Unknown Grade Well differentiated Moderately differentiated Poorly differentiated Undifferentiated Depth of invasion Mucosa or submucosa Proper muscle Subserosa Serosa Adjacent invasion Unknown Number of positive LN. 0 1 to 2 3 to 6 7 to 15 16 or more Positive LN (Mean±SD) Total LN (Mean±SD)(n=1411) AJCC Stage IA www.impactjournals.com/oncotarget
% 60.8±10.4 14 to 88
1206 223
84.4 15.6
1278 61 18 70 2
89.4 4.3 1.3 4.9 0.1 4.6±3.4 0.1 to 50
1299 9 17 32 4 29 9 18 12
90.9 0.6 1.2 2.2 0.3 2.0 0.6 1.3 0.8
62 510 827 30
4.3 35.7 57.9 2.1
134 215
9.4 15.0
667 260 125 28
46.7 18.2 8.7 2.0
757 342 231 81 18
53.0 23.9 16.2 5.7 1.3 1.8±3.3 14.3±9.9
100
7.0 14059
Oncotarget
IB IIA IIB IIIA IIIB IIIC Unknown Adjuvant radiotherapy Yes No
146 410 307 196 173 69 28
10.2 28.7 21.5 13.7 12.1 4.8 2.0
60 1369
4.2 95.8
Abbreviations: AI, American Indian or Alaska Native; API, Asian or Pacific Islander; LN, lymph node; AJCC, American Joint Committee on Cancer. therapy may be a potential standard treatments for AGC patients. However, the prognostic value of post-therapy pathologic characteristic were still unclear, especially in patients underwent NRT. [10, 12-15] In current study, we aimed to investigate the value of prognostic factors in AGC patients and to evaluate whether post-therapy pathological characteristics were predictive factors of survival in these patients.
the inclusion criteria were analyzed in current study. The patients’ characteristics were listed in Table 1. Overall, there were 1,206 male patients and 223 female patients, and the median age was 60.8 years old. There were also 60 (4.2%) patients who received adjuvant radiotherapy after curative resection. The mean number of positive lymph node was 1.8±3.3, and the mean total examined number of lymph node was 14.3±9.9. With a median follow-up of 31.9 months, there were 760 (53.2%) patients died before the analysis of the present study, and 102 (13.4%) patients of them were died because of other causes.
RESULTS Patients and demographics
Survival and lymph node ratio categories
A total of 1,429 patients between January 1998 and November 2013 in the SEER database who met all
The median DSS for all patients was 35 months, and the 1-year DSS, 3-year DSS, 5-year DSS were 83.0%, 48.9% and 39.0%, respectively. Lymph node ratio was
Figure 1: X-tile analysis identified the best cutoff points for mLNR A., and validated by Kaplan-Meier Curve B. Abbreviation: mLNR, metastatic lymph node ratio.
www.impactjournals.com/oncotarget
14060
Oncotarget
Table 2: Prognostic factors for DSS of all the patients Univariate analysis
Multivariate Analysis
HR
95% CI
p
Age
1.005
0.998 to 1.013
0.167
Gender (Female / Male)
0.879
0.708 to 1.091
0.243
Race API/ nonAPI
95% CI
p
0.062 0.762
0.573 to 1.014
Location
0.939
Antrum
ref
Fundus
0.771
0.220 to 2.707
Body
1.255
0.520 to 3.029
Cardia
1.336
0.771 to 2.315
Pylorus
1.464
0.417 to 5.139
Lesser curvature
1.462
0.678 to 3.155
Greater curvature
1.371
0.488 to 3.847
Overlapping
1.547
0.661 to 3.620
Grade Well /Moderately differentiated Poorly differentiated/ Undifferentiated Total LN
0.714
0.608 to 0.838
1.002
0.994 to 1.010
0.584
Tumor Size (n = 1027)
1.001
0.999 to 1.004
0.414
