Tumor size classification of the 8th edition of TNM staging system is superior to that of the 7th edition in predicting the survival outcome of pancreatic cancer patients after radical resection and adjuvant chemotherapy

The 8th edition of TNM staging system has been released and it incorporates many changes to the T and N classifications for pancreatic cancer. Comparative study between the 7th and 8th edition of TNM staging system from Asian population has not been reported yet. This study aimed to compare the 7th and 8th edition of staging system for pancreatic cancer by using a cohort of pancreatic cancer patients from China after R0 pancreaticoduodenectomy and adjuvant chemotherapy. The results showed according to the pT classification of 7th edition, pT3 was predominant (87.25%), however, the new edition led to a more equal distribution of pT classification. pT1, pT2 and pT3 was 27.45%, 56.86% and 15.69%, respectively. According to the new pN classification, 18.63% of the patients were pN2. The pT classification in the 8th edition was significantly superior to that in the 7th edition at stratifying patients by overall survival. The pN classification in the 8th edition failed to show an advantage over the 7th edition in stratifying patients by overall survival. Therefore, the new pT classification, but not the new pN classification, showed a significant advantage over the previous edition at predicting the overall survival of pancreatic cancer patients.

Comparison of the 7 th and 8 th editions of the TNM staging system for patients. Stages pT1 and pT2 in the 7 th edition were well matched with those in the 8 th edition, but only 18/89 (20.2%) stage pT3 cases in the 7 th edition were matched with those in the 8 th edition (Table 3) 60.78% of patients had lymph node metastasis, and according to the new pN classification, 18.63% of these patients had metastasis in more than 3 lymph nodes (pN2) ( Table 4). Stages IA and IB in these two editions were well matched. According to the 7 th edition, 33.3% and 59.8% of patients were stage IIA (T3N0M0) and IIB (T1-3N1M0), respectively. In the new edition, only 5.9% of the patients were stage IIA (T3N0M0); 22.5%, 43.1%, and 16.7% of the patients were stage IB (T2N0M0), IIB (T1-3N1M0) and III (T1-3N2M0), respectively. Moreover, 6/34 (17.7%) stage IIA cases in the 7 th edition were matched with those in the 8 th edition, and the others were characterized as stage IA and IB by the 8 th edition. A total of 17/61 (27.9%) stage IIB cases in the 7 th edition were considered stage III by the 8 th edition ( Table 5).
The ability of the 7 th and 8 th editions of the TNM staging system to stratify patients by overall survival. After multivariate analysis, CA19-9, tumor size larger than 4 cm (pT3 of the 8 th edition), poor differentiation and positive lymph node metastasis were independent risk factors for poor survival, but pT classification in the 7 th edition and the number of positive lymph nodes (pN1 and pN2 classification in the 8 th edition) failed to stratify patients by survival; this finding indicated that pT classification in the 8 th edition was superior to that in the 7 th edition. However, pN classification in the 8 th edition did not show significant superiority to that in the 7 th edition at stratifying patients by overall survival (Table 6, Fig. 1). In addition, the 8 th edition of the TNM staging system successfully stratified stage I patients from those at other stages based on overall survival, which the 7 th edition could not do (Fig. 2).

Discussion
Since the 1 st edition of the TNM staging system was published in 1977, the AJCC/UICC has released a total of eight editions. During the last 40 years, considerable improvements have been achieved in many malignancies, such as melanoma, gastric cancer, colorectal cancer, breast cancer and lung cancer. However, despite tremendous efforts, a diagnosis of PDAC remains devastating [15][16][17][18] . There has not been any change in the TNM staging system for PDAC in the last three editions. Finally, in the new 8 th edition, many changes in both T and N classification  for PDAC have been incorporated, but whether these changes make the 8 th edition superior to the 7 th edition at stratifying patients by prognosis remains unclear.
The role of the previous T classification in predicting the prognosis of PDAC patients is controversial 19 . It is difficult to accurately ascertain the T classification according to the previous editions; in the 8 th edition, only tumor size was considered, significantly increased the accuracy of the pathological assessment 7,10 . The number of positive lymph nodes was adopted to define the N classification in the new edition, but the role of this factor remained unclear [20][21][22] . Allen et al. 7 analyzed 2318 cases of PDAC after resection from 3 large volume centers in America and found that the 8 th edition increased the reproducibility of the T3 classification among different centers and that the N classification in the 8 th edition was able to discriminate the prognosis of patient subgroups. Kamarajah et al. 10 collected data from the Surveillance, Epidemiology and End Results (SEER) database from 2004 to 2013 and analyzed 8960 pancreatic cancer patients without metastasis who underwent surgical resection. The results showed that the 8 th edition allowed for finer stratification of patients according to the extent of nodal involvement. Although these two studies had a large number of patients, there were some limitations: (1) all the data were from America; (2) the time interval during which patients were enrolled was more than 10 years; and (3) information on adjuvant treatment was missing. More recently, the results of two validation studies from Germany were inconsistent with the significance of the N classification in the 8 th edition. Welsch et al. 9 reported a cohort of 256 PDAC patients who underwent curative resection from 2005 to 2015, and the results showed that the new N and T classifications both better discriminated PDAC patients by survival. Schlitter et al. 8 reported two cohorts of 523 PDAC patients from Germany who underwent surgery in two hospitals over two decades (1991-2006; 2007-2014). They found that the T classification in the 7 th edition could not discriminate patient prognosis, whereas the 8 th edition showed substantial success in stratifying patients by prognosis. The new N classification failed to show high clinical relevance in either cohort.
Herein, we report the first validation of the 8 th edition in an Asia population. To test the reliability of stratification based on the new TNM staging system, we enrolled patients according to rigorous criteria to avoid confounding factors to the maximal extent possible: (1) only patients treated after 2010 were enrolled; (2) there was a long follow-up period; (3) all cases achieved R0 resection; (4) only patients with tumors located in the head of the pancreas were enrolled; and (5) all the enrolled patient underwent at least 3 cycles of adjuvant chemotherapy based on gemcitabine. The median age of the patients was 60 years, which was a little younger than that of the above reports from Germany and America. There were slightly more male patients than female patients (1.8:1.0), which was in accordance with the ratio reported in the above studies from Germany. CA19-9, CA242 and CEA were elevated in 72.5%, 40.5% and 20.6% of the patients, respectively. Some previous studies have reported that elevated CA19-9 and CA242 are risk factors for poor prognosis 23,24 ; in this study, we also found that elevated CA19-9 and CA242 were associated with poor prognosis.    Overall, 60.7% of the patients had lymph node metastasis, and 42.2% and 18.5% of the patients were pN1 and pN2, respectively. The pN classification in the 7 th edition successfully stratified patient based on survival, but the new pN classification did not show an advantage over the 7 th edition in discriminating patients by prognosis. In the 7 th edition, 87.3% of patients were pT3, whereas 54.9% of patients were pT2 in the 8 th edition, which was similar to the percentage reported in the above studies from America and Germany. pT3 classification in the 7 th edition was redefined as pT1, pT2 and pT3 classification in the 8 th edition, with a predominance of pT2 (51.2%); only 20.2% of pT3 classifications in the 8 th edition corresponded with those in the 7 th edition. pT classification in the 7 th edition failed to stratify patients by survival. However, the prognosis of patients classified by the 8 th edition as pT1-pT2 or pT3 was significantly different, suggesting that the new T classification was superior to the previous one at stratifying patients by survival. Although the 8 th edition of the TNM staging system did not stratify patients by survival for all stages, it could discriminate stage I patients from those at other stages by survival, which the 7 th edition of the staging system failed to do, further showing the superiority of the new edition.
In conclusion, this study is the first to validate the 8 th edition of the TNM staging system for PDAC in an Asian population after its release. Compared to the validation studies from America and Germany, this study was performed at only a single center with a relatively small number of patients, which may weaken the  Table 6. Univariate analysis and multivariate analysis of the overall survival of the patients. *P < 0.05; **P < 0.01. reliability of the results. As previously stated, to validate the value of the TNM staging system, we enrolled patients according to rigorous criteria to minimize possible confounding factors that may affect the efficacy of the TNM staging system. The results supported that the new pT classification had a substantial advantage over the previous edition in predicting the overall survival of PDAC patients undergoing R0 pancreaticoduodenectomy and gemcitabine-based adjuvant chemotherapy. In this study, the new pN classification failed to show superiority over the previous edition in stratifying patients by overall survival. To further validate the superiority of the new edition of the TNM staging system at stratifying PDAC patients in Asia (China) by survival, a multicenter study with a large number of patients will be needed.

Methods
Patients and follow-up. In total, 102 consecutive cases of PDAC were enrolled from January 2010 to October 2014 according to the following inclusion criteria: (1) the final pathological examination confirmed PDAC; (2) none of the patients underwent neoadjuvant treatment; (3) radical R0 pancreaticoduodenectomy was achieved (microscopic margin > 1 mm); (4) all the patients underwent at least three cycles of gemcitabine-based adjuvant chemotherapy; (5) information on postoperative survival time was available, and patients who died within 3 months after surgery were excluded; and (6) all the patients signed the informed consent form. The clinicopathological information, including age, gender, tumor size, differentiation, perineural invasion (PNI), micro-cancerous embolus, CA19-9, CA242, CEA, pT classification, pN classification and TNM stage of the 7 th and 8 th editions, was extracted from the PACS system. After operation, the patients were followed up every 3∼6 months by outpatient clinic visits or telephone calls until patient death. Overall survival (OS) was defined as the survival time after surgery. All the patients agreed to donate bio-specimens for scientific research and publication and signed the informed consent form before surgery. The study was approved by the ethics committee of Peking Union Medical College Hospital.
Statistics. IBM SPSS Statistics software version 22.0 was applied for statistical analysis. Overall survival was analyzed using the Kaplan-Meier method, and the values were compared using the log-rank test. Multivariate analysis was performed using the Cox proportional hazard model. A P value less than 0.05 indicated statistical significance.
Data Availability. The primary data is available if reasonable requested.
Ethical approval. All procedures involving human participants were in accordance with the ethical standards of the ethical committee of Peking Union Medical College Hospital and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. All of the patients signed the informed consent for the scientific use of their information or bio-specimen before surgery.