Comparison of the prognostic difference between ypTNM and equivalent pTNM stages in esophageal squamous cell carcinoma based on the 8th edition of AJCC classification

Objective: With the separate ypTNM stage groupings established in the 8th edition of AJCC staging system for esophageal squamous cell cancer (ESCC), we aimed to evaluate the prognostic difference between ypTNM stage and equivalent pTNM stage. Methods: ESCC patients with surgery alone (cohort 1) and patients with neoadjuvant therapy plus surgery (cohort 2) were enrolled in the study. Results: In p0, pIb, pIIa, pIIb, pIIIa, pIIIb and pIVa stages of cohort 1, the 5-year DFS and OS rates were 100/100%, 80.5/86.2%, 58.9/57.8%, 51.1/52.7%, 36.3/35.8%, 21.5/22.6% and 11.9/18.0%. In ypI, ypII, ypIII and ypIVa stages of cohort 2, the 5-year DFS and OS rates were 60.9/67.0%, 44.3/52.1%, 48.4/43.2% and 0. Patients in ypI stage had a tendency of poorer survival compared with those in pI stage (P=0.024 for DFS, P=0.067 for OS). There was no significant difference in terms of DFS (P=0.335) or OS (P=0.903) between ypII and pII. Patients in ypIII stage had a tendency of better survival compared with those in pIII stage (P=0.015 for DFS, P=0.059 for OS). Patients in ypIVa stage exhibited a significantly poorer OS compared with those in pIVa stage (P=0.038). Conclusions: With down-staged tumor after neoadjuvant therapy, survival of ypI was closed but not reached to the prognosis of equivalent pI, prognosis of ypII was similar to equivalent pII, and survival of ypIII tended to be better compared with equivalent pIII. However, without down-staged ypIVa tumor, the prognosis was worse compared with equivalent pIVa, indicating those patients were primary resistant to prescribed neoadjuvant therapy.


Introduction
Esophageal cancer (EC) is the ninth most commonly diagnosed cancer and the sixth most common cause of cancer death globally [1]. Within the so-called EC belt, China alone contributes more than Ivyspring International Publisher half of the global EC cases [2,3]. The Chinese National Central Cancer Registry (NCCR) called for data in 2015 from all population-based cancer registries and showed that EC, with estimated new cases of 477,900 and death of 375,000, ranks as the third most common cancer and the fourth leading cause of cancer death in China [4]. The overall 5-year survival rate of EC ranges from 15% to 25% [5]. The better outcomes are associated with early diagnosis. However, many Chinese patients are diagnosed in locally advanced stages with poorer outcomes [6]. Therefore, it remains a challenge to improve the survival of patients with locally advanced tumor in China.
In recent years, neoadjuvant therapy, including chemotherapy and/or radiotherapy, is commonly used as an adjunct to surgical resection in patients with locally advanced EC, leading to down-staged tumor, increased resection rate and improved survival [1,7,8]. With the widespread use of neoadjuvant therapy, it clearly creates an extra demand on prognostic evaluation. The 7th edition of American Joint Committee on Cancer (AJCC) staging system is widely used for prognosis stratification throughout the world in EC patients with surgery alone [9][10][11]. Some studies believe that the ypTNM classification describing the presence and anatomical extent of vital tumor cells is also an essential prognostic factor in EC patients with neoadjuvant therapy [12,13]. However, many investigators have argued that this ypTNM stage largely loses its prognostic strength in EC patients with neoadjuvant therapy [14,15], partly because that the database which is used to establish the 7th edition of AJCC ypTNM staging system includes patients who undergo surgical resection alone and excludes patients who receive neoadjuvant therapy followed by surgery.
An 8th edition primer of the AJCC staging system for EC was published in January 2017 [16]. New to the 8th edition is stage grouping of 7,773 patients treated with neoadjuvant therapy from 33 Worldwide Esophageal Cancer Collaboration (WECC) institutions, and the separate stage groupings are established for patients with neoadjuvant therapy [17,18]. With the new TNM staging system, the intriguing question is whether there is prognostic difference between neoadjuvant categories (ypTNM) and equivalent pathologic categories (pTNM). At present, no related data comparisons have been reported using the new staging system.
In the present study, we presented data from 779 esophageal squamous cell carcinoma (ESCC) patients who underwent surgery (cohort 1) or received neoadjuvant therapy plus surgery (cohort 2) from a single cancer center. We aimed to evaluate the prognostic difference between ypTNM stage in cohort 2 and equivalent pTNM stage in cohort 1 based on the 8th edition of AJCC staging system.

Patient selection
A total of 779 patients with biopsy-proven primary ESCC who were treated with curative intent at Zhongshan Hospital, Fudan University were enrolled in this retrospective analysis. According to the different treatment modalities, these patients were divided into two groups (cohort 1 and 2). All patients included for analysis fit the following inclusion criteria: 1) pathologically confirmed ESCC; 2) received R0 resection; 3) without neoadjuvant therapy for cohort 1; with neoadjuvant therapy for cohort 2. The main exclusion criteria were past or current history of malignancy other than the ESCC, and lost in follow-up. In cohort 1, a total of 604 patients only underwent surgical resection without neoadjuvant therapy from January 2007 to November 2010. In cohort 2, a total of 175 patients with locally advanced EC received neoadjuvant therapy (neoadjuvant chemotherapy or radiochemotherapy) in combination with surgical resection from November 2009 to December 2015.
Patients were initially diagnosed through endoscopy, and tumor staging was confirmed with computed tomography (CT) scan of the thorax and abdomen. Endoscopic ultrasound (EUS) and positron emission tomography (PET) were not routinely used. Patients were followed up with a clinical examination every 3 months for the first year, every 6 months for the second year, and every 6-12 months thereafter. Patients who did not go to our hospital were contacted by telephone to obtain follow-up data.
This study was approved by the Institutional Review Board of Zhongshan Hospital, Fudan University in accordance with the Declaration of Helsinki. Informed consent was obtained from all patients before treatment.

Clinicopathologic evaluation
All hematoxylin and eosin-stained glass slides were reviewed by two experienced pathologists. Following clinicopathologic features were collected, including tumor grade (G category), invasive depth (pT category or ypT category), the number of positive lymph nodes (pN category or ypN category), and the presence of vessel (lymphatic and venous) invasion and nerve invasion. Other demographic characteristics, including age, gender and tumor location, were also recorded.

Statistical analyses
Disease-specific overall survival (OS) was calculated from the date of operation to the date of death or most recent follow-up. For tumor specific survival, an event was defined by death of ESCC. For disease free survival (DFS), the first occurrence of locoregional recurrence, distant metastasis or death by ESCC was defined as an event. OS and DFS were estimated using the Kaplan-Meier method. The log rank test was used to compare survival curves.
Patients in ypI stage (cohort 2) had a significantly poorer DFS and a potential poorer OS compared with those in pI stage (cohort 1) (P=0.024 for DFS, P=0.067 for OS). However, no difference in survival was observed between patients in ypI (cohort 2) and pII (cohort 1) stages (P=0.228 for DFS, P=0.057 for OS). Moreover, there was no significant difference in DFS (P=0.335) or OS (P=0.903) between ypII (cohort 2) and pII (cohort 1). Patients in ypIII stage (cohort 2) had a significantly better DFS and a potential better OS compared with those in pIII stage (cohort 1) (P=0.015 for DFS, P=0.059 for OS) but a poorer survival compared with those in pII stage (cohort 1) (P=0.010 for DFS, P=0.001 for OS). Patients in ypIVa stage (cohort 2) had a significantly poorer OS and a potential poorer DFS compared with those in pIVa stage (P=0.038 for OS, P=0.133 for DFS) (Figure 2).

Discussions
In recent years, neoadjuvant therapy has come to be included in potentially curative treatment of EC prior to surgical resection [7,8,20,21]. The prognostic strength of new ypTNM stage based on the 8th Figure 1. There was no significant difference in terms of survival between cohort 1 and cohort 2 (a, DFS; b, OS).

Figure 2.
Survival comparisons within "the same stage". Patients in ypI stage (cohort 2) had a significantly poorer DFS and a potential poorer OS compared with those in pI stage (cohort 1). There was no significant difference in DFS or OS between ypII (cohort 2) and pII (cohort 1). Patients in ypIII stage (cohort 2) had a significantly better DFS and a potential better OS compared with those in pIII stage (cohort 1). Patients in ypIVa stage (cohort 2) had a significantly poorer OS and a potential poorer DFS compared with those in pIVa stage. edition of AJCC ESCC staging system needs to be evaluated [9,22]. Because squamous cell carcinoma is the most common histological type of EC in China [3,23], we believed that more data from Chinese patients should be assembled to analyze the prognostic significance. However, at present, few studies have reported in China [17]. In this retrospective study, we aimed to determine the prognosis of two cohorts of patients undergoing surgery alone or receiving neoadjuvant therapy followed by surgery, and these patients staged with the recently issued 8th edition of the AJCC TNM staging system. Moreover, we elucidated the prognostic difference between ypTNM and equivalent pTNM.

The prognosis of two cohorts
In our cohort 1 (surgery alone), the 1-, 3-and 5-year DFS rates were 79.5%, 51.1% and 46.4%, respectively, with a median survival time of 39.0 months. The 1-, 3-and 5-year OS rates were 88.9%, 57.3% and 47.6%, respectively, with a median survival time of 48.0 months. The survival rates obtained in our study were very similar with previous reports. Chen et al. have collected data from 2,011 ESCC patients who underwent surgical resection alone, and shown that the 1-, 3-and 5-year OS rates are 83.5%, 57.4% and 47.4%, respectively, with a median survival time of 51.0 months [24].
ypTNM staging is new to the 8th edition, and survival for ypTNM stage groups differs from that for comparable pTNM stage groups [16,18]. In ypI, ypII, ypIII and ypIVa stages of our cohort 2, the 5-year DFS and OS were 60.9/67.0%, 44.3/52.1%, 48.4/43.2% and 0, respectively. Previous studies have indicated that ypTNM stage grouping is a superior predictor of outcome in stage III ESCC patients undergoing neoadjuvant therapy followed by radical esophagectomy, and patients with down-staged tumors after neoadjuvant therapy experience improved survival compared with patients without response [25,26]. In our cohort 2, a total of 175 patients were evaluated with cIII-IVa stage disease before the treatment, neoadjuvant therapy was associated with significant tumor down-staging, and 92 cases (52.5%) were down-staged to ypI (62) or ypII (30).
To understand the real-world survival status of patients with different ypTNM stages, we conducted the equivalent stage comparisons between cohort 2 and cohort 1. The DFS of ypI was poorer than that of pI. There was no significant difference in survival between ypI/II and pII. The DFS of ypIII was better than that of pIII. The OS of ypIVa was poorer than that of pIVa. In WECC's data, the survival of ypTNM stage group I is equivalent to pTNM stage group IIB, but the survival for ypTNM stage group II is intermediate between pTNM stage groups IIB and IIIA, and survival for ypTNM stage groups IIIA through IVB is equivalent to the same pTNM stage groups. There were some differences between our study and WECC's conclusion partly due to the changes of the new AJCC ypTNM staging system [17]. Therefore, it is necessary to further investigate the survival difference between ypTNM and equivalent pTNM in ESCC using the 8th edition of AJCC classification in order to improve the management and counseling of EC patients receiving neoadjuvant therapy.
In the present study, we compared the survival of patients receiving neoadjuvant therapy with that of patients with equivalent pathologic categories receiving surgery alone based on separate ypTNM stage in the 8th edition of AJCC staging system. Taken together, survival of ypI was closed but not reached to equivalent pI, prognosis of ypII was reached to equivalent pII, and survival of ypIII tended to be better compared with equivalent pIII. The prognosis of ypIVa was worse compared with equivalent pIVa, indicating those patients were primary resistant to prescribed neoadjuvant therapy.