ENDOSCOPIC CHARACTERISTICS OF PATIENTS WITH COMPLETE PATHOLOGICAL RESPONSE AFTER NEOADJUVANT CHEMOTHERAPY FOR GASTRIC AND ESOPHAGOGASTRIC JUNCTION ADENOCARCINOMAS

ABSTRACT 
Background:
 Gastric and esophagogastric junction adenocarcinoma are responsible for approximately 13.5% of cancer-related deaths. Given the fact that these tumors are not typically detected until they are already in the advanced stages, neoadjuvancy plays a fundamental role in improving long-term survival. Identification of those with complete pathological response (pCR) after neoadjuvant chemotherapy (NAC) is a major challenge, with effects on organ preservation, extent of resection, and additional surgery. There is little or no information in the literature about which endoscopic signs should be evaluated after NAC, or even when such re-evaluation should occur. 
Aim: To describe the endoscopic aspects of patients with gastric and esophagogastric junction adenocarcinomas who underwent NAC and achieved pCR, and to determine the accuracy of esophagogastroduodenoscopy (EGD) in predicting the pCR. 
Methods: A survey was conducted of the medical records of patients with these tumors who were submitted to gastrectomy after NAC, with anatomopathological result of pCR. 
Results: Twenty-nine patients were identified who achieved pCR after NAC within the study period. Endoscopic responses were used to classify patients into two groups: G1-endoscopic findings consistent with pCR and G2-endoscopic findings not consistent with pCR. Endoscopic evaluation in G1 was present in an equal percentage (47.4%; p=0.28) in Borrmann classification II and III. In this group, the predominance was in the gastric body (57.9%; p=0.14), intestinal subtype with 42.1% (p=0.75), undifferentiated degree, 62.5% (p=0.78), Herb+ in 73.3% (p=0.68). The most significant finding, however, was that the time interval between NAC and EGD was longer for G1 than G2 (24.4 vs. 10.2 days, p=0.008). 
Conclusion: EGD after NAC seems to be a useful tool for predicting pCR, and it may be possible to use it to create a reliable response classification. In addition, the time interval between NAC and EGD appears to significantly influence the predictive power of endoscopy for pCR.


INTRODUCTION
Despite a reduction in the incidence of gastric adenocarcinoma (GC) in the last decade, it remains the third most common cause of cancer-related death in the world, with an estimated 783,000 deaths per year 5 .The incidence of adenocarcinoma of the esophagogastric junction (GEJ) has increased markedly in Western countries in recent years 5,4 . Population analyses in the United States have reported an almost 2.5-fold increase in incidence since the 1970s 14 . In Brazil, GC is the 5 th most common cancer overall (3 th among men and 5 th among women) 15 . For each year from 2020 to 2022 it has been projected that there will be 13,360 new cases in men and 7,870 new cases in women per hundred thousand people 15  Currently, the recommended therapeutic approach for locally advanced tumors of the stomach and GEJ is perioperative chemotherapy 3,7 . Several studies have demonstrated this strategy to yield increased rates of complete resection, downstaging, overall survival, and progression-free survival 3,8 . It has also been noted that some tumors exhibit better responsiveness than others 7,8 . Therefore, diagnostic methods that can predict a complete pathological response (pCR) have important clinical implications 7,22 . Several combined chemotherapy regimens have shown good efficacy for GC and non-resectable GEJ allowing a potential curative gastrectomy 22 . Among these regimens, the most used today is the FLOT scheme, composed of fluorouracil, leucovorin, oxaliplatin, and docetaxel 2 . In patients with locally advanced lesion above clinical stage tumor (cT) 2 or compromised lymph nodes, neoadjuvant chemotherapy (NAC) can increase the likelihood of curative surgery, with complete tumor response rates around 10% and increased rates of both progression-free and overall survival 9,21 .
In addition, NAC can offer treatment options for patients for whom surgery is risky, such as those with more advanced disease progression 4 . The identification of patients with pCR after NAC could, in the future, become a tool to select those who really benefit from adjuvant chemotherapy and perhaps even in the suppression of surgery to patients at high risk for the procedure, becoming a useful tool in the decision multidisciplinary therapy 4 . In general, the morbidity rate of radical gastrectomies is around 33.5% and mortality between 0.6% to 4.7% 11,16 . According to a 2015 study the overall survival of stage III/IV patients who underwent NAC and who obtained pCR was similar to those with stage I/II who did not receive NAC 6 . Recently published data demonstrated that pathological staging was better than conventional staging at predicting responsiveness to and survival after neoadjuvancy 7 . Other studies have also indicated that location in GEJ and TNM are associated with a worse prognosis 8 .
Preoperative endoscopic evaluation of patients undergoing NAC is recommended in many services, but as of the writing of this paper there has been no published description of endoscopic findings in these patients and the ideal time interval between NAC and surgery remains unclear.
Therefore, in the present study we investigated the following questions: 1) What are the endoscopic features that support detection of pCR following NAC?

2)
What is the diagnostic accuracy and sensitivity of esophagogastroduodenoscopy (EGD) in the assessment of pCR following NAC?
and 3) What time interval between NAC and EGD supports optimal response prediction?

METHODS
This is a retrospective study in a single center specializing in cancer treatment. BC-IV, diffusely infiltrative lesion, with no limit between the tumor and the normal mucosa.
After the histopathological diagnosis patients were staged using a All surgical specimens were analyzed by two independent pathologists using the tumor regression score as recommended by the National Comprehensive Cancer Network 1 . According to this scale, a score of zero indicates complete response and comprises no viable cancer cells, including in lymph nodes. All of the 29 patients included in this study showed scores of zero and complete anatomopathological responses.

Objectives and definitions
The primary objectives of this study were to describe the endoscopic aspects of patients with gastric or GEJ adenocarcinomas submitted to NAC who exhibited pCR, and to determine the accuracy of EGD in predicting this response in these patients.     Table 1.

Characteristics of patients and tumors are summarized in
Although there were more males enrolled than females, genders were distributed similarly across groups (p=0.22).
The G1 group compared to G2 group, more patients presented with GC (73.7%, p=0.67) cT 3 (68.4%, p=0.72), and absent lymph node status (68.4%, p=0.59). After EGD, the lesions were classified in equal proportions as BC-II and BC-III (47.4%, p=0.28), more in the gastric body (57.9%, p=0.14) and the histopathological study showed that, the intestinal subtype was present in 42.1% The ideal time calculated for performing EGD after NAC was 8.5 days. At this time, a sensitivity of 93% and specificity was reached: 66% (CI: 0.6-1.0). EGD after NAC showed accuracy in predicting endoscopic complete response (eCR) and sensitivity, G1, in 65.5% of the analyzed cases. All cases underwent surgical treatment with partial gastrectomy, total gastrectomy with or without distal esophagectomy associated with D2 lymphadenectomy, with a finding of pCR in the surgical specimen in an average of 43.8 days in both groups, with the average in G1 being 52, and in G2 39 days after endoscopic control.

DISCUSSION
GC is one of the most common neoplasms in the world, with high rates of incidence and mortality 20 . The most common location is the gastric antrum although the incidence of GEJ tumors is gradually increasing 13 . GEJ tumors are very prevalent worldwide and are among the most aggressive tumors of the digestive tract 5 . Furthermore, in most western countries they are not diagnosed until the more advanced stages, when isolated surgical treatment is less effective 6 . The low rate of early gastric cancer is related to the lack of specific symptoms 19 . Advanced tumors exhibit considerable metastatic potential and generally worse prognosis, indicating a need for combined systemic and local treatments to reduce the risk of tumor recurrence 12 .
Complementary treatments associated to radical surgery are being more frequently indicated and have demonstrated significant efficacy 2,7,19 .
For advanced cancers, the most successful treatments are combined chemotherapy with surgery 2,6,20 . However, there are few tools to restaging patients before the surgical procedure 10 .
This study describe the pre-operative endoscopic findings of 29 patients with GC and GEJ cancer who were submitted to NAC and who achieved pCR after surgery. Methods for assessing tumor response and metastases after chemotherapy include endoscopic ultrasound, CT, and PET-CT, but these techniques show low accuracy and the possibility of over or under-staging 17,18 .
The study included EGD performed by two independent senior endoscopists, after the neoadjuvant treatment. The exams were performed within 30 days of the end of NAC, and surgery was performed 4-8 weeks after the end of NAC.
The EGD was able to predict pCR in 65.5% of cases (G1). In addition, the time interval between the end of NAC and the performance of EGD2 was significantly shorter for the group in which EGD was unable to predict pCR (G2) and for each additional day there was a 12% increase in the probability of predicting pCR.
The ideal time calculated to perform EGD was 8.5 days after ending NAC, at which point sensitivity reached 93%. The presence of active lesions in G2 may have been due to inflammatory responses that occur during normal healing of the mucosa. In these cases, EGD performed later may have revealed scarring compatible with that observed in G1.

CONCLUSION
The eCR, determined by the presence of endoscopic scar, reddish or whitish, without active lesions after NAC, was consistent with the pCR. EGD after NAC showed accuracy in predicting eCR and sensitivity in 65.5% of the cases analyzed. The minimum time interval for performing EGD after the end of NAC was 8.5 days. Respecting this interval may increase the possibility of predicting pCR with endoscopic evaluation and supports optimal response prediction