Gastric dysplasia, also known as gastric adenoma, is a precancerous stomach lesion. Low-grade gastric dysplasia (LGD) is a relatively common diagnosis in South Korea due to the widespread implementation of a national cancer screening program. Although not all cases of LGD progress to cancer, a cohort study of >400000 patients reported progression to cancer in 1 of 19 patients within a 20-year period [1]. Interestingly, a Korean cohort study reported a higher rate of progression to cancer than has been reported in Western studies. Among 26 patients who were followed-up for an average of 58 months, four (15.4%) developed gastric cancer, and another four showed progression from LGD to high-grade gastric dysplasia (HGD) [2]. In South Korea, the high incidence of gastric cancer and the wide availability of physicians with extensive experience in performing endoscopic procedures has resulted in endoscopic resections being proactively performed in patients with LGD.

The three main types of endoscopic treatment for LGD are endoscopic submucosal dissection (ESD), endoscopic mucosal resection (EMR), and argon plasma coagulation (APC). ESD and EMR directly remove the lesion, whereas APC ablates the affected tissue region. ESD or EMR should be considered first-line treatment options for patients with the possibility of receiving a final pathological diagnosis of early gastric cancer (EGC), whereas APC may be a better alternative for patients with LGD who are less likely to be diagnosed with EGC. APC ablation has many advantages over other treatment options, including fewer technical issues, shorter procedure duration, and fewer complications [3]. These advantages make APC a valuable treatment option for older patients who have difficulty undergoing lengthy endoscopic procedures and for those with an increased risk of bleeding (e.g., due to the use of anticoagulants). However, because APC does not involve a pathological evaluation, it does not allow immediate confirmation of the complete eradication of dysplastic cells. In a previous study, 21.2% of patients who underwent APC had local recurrence, suggesting that dysplastic cells may remain after the procedure [4]. Although several studies have focused on local recurrence after APC, none have pathologically evaluated the complete eradication of dysplasia after APC. Therefore, this study used pathological examinations to investigate the incidence and associated causative factors of residual dysplasia after APC.

From March 2020 to February 2021, a total of 82 patients were enrolled. Of these, 28 were assigned to the 45 W group, and the remaining 54 were equally distributed between the 60 W and 80 W groups, with 27 patients in each group. However, 14 patients, seven from each of the 60 W and 80 W groups, were excluded from the study (11 patients declined to undergo follow-up endoscopies, two were found to have cancer during endoscopy, and one withdrew from the study for personal reasons). The two patients who were excluded from the study because of cancer detection were diagnosed with gastric cancer during APC. These cancers were synchronous lesions; one patient underwent surgical resection, and the other underwent endoscopic resection. Therefore, data were collected from 68 patients. Due to the exclusion of some patients, the APC power groupings were assigned differently: 28 patients were assigned to the 45 W group and 20 were assigned to each of the 60 W and 80 W groups. The mean age of the study participants was 66.31 years (range, 45–84 years), and 47 (69.1%) were men. H. pylori infection was confirmed in 29 (42.6%) patients (Table 1). At the 24-h follow-up endoscopy, ulceration at the procedure site was observed in all patients (Fig. 1); biopsy results revealed residual lesions in 13 patients (19.1%). Each residual lesion was associated with a positive Ki-67 index (Table 2). The other clinical factors investigated were not statistically significant.

At the three-month follow-up endoscopy, residual lesions were found in three of the 55 patients (5.5%; 13 patients were lost to follow-up after three months owing to non-visits). The presence of residual dysplasia at three months was associated with both the Ki-67 index and the presence of residual lesions at 24 h. When the risk of residual lesions was analyzed according to APC output power (using 60 W as the reference point), the occurrence of residual lesions at the three-month follow-up was higher when APC was performed at an output power of <60 W (p=0.043). Patients without residual lesions at the 24-h follow-up did not have any residual lesions at the three-month follow-up. Of the 13 patients with residual lesions at the 24-h follow-up, two were lost to follow-up, eight had no residual lesions, and three had residual lesions at three months (Figs. 2 and 3). The three patients with residual lesions at the three-month follow-up endoscopy underwent additional APC ablation at an output of 80 W. Follow-up endoscopy was conducted 15 months later in 44 (44/68, 64.7%) patients, including three who received additional APC ablation; no local recurrence was observed in any of them. Among the study participants, none reported procedure-related complications, including bleeding or perforation.

APC is technically simpler and more cost-effective than ESD or EMR; it is also associated with shorter hospital stays. However, it has a higher local recurrence rate. Only one study has directly compared the local recurrence rate after APC with that after ESD and EMR, whereas other studies have reported relatively high local recurrence rates after APC (6.3% and 21.1%) [3-6]. At our hospital, the local recurrence rates from 2011 to 2015 were 1.3% after ESD and 4.2% after EMR, which were both lower than the post-APC recurrence rate (5.5%) in the current study [7]. We performed a follow-up endoscopy 24 h after APC to visually and pathologically examine the presence of residual lesions. In all patients, ulcers were observed during the endoscopy, but no residual lesions were visually detected at 24 h. However, biopsies performed at the ulcer base revealed remnant dysplasia in 19.1% of the patients, confirming our hypothesis (Fig. 1 and Table 2). Ki-67 staining was performed to confirm the viability of the remaining dysplastic cells after APC. The Ki-67 index is a cell proliferation marker, and Ki-67 expression is detected in cells during the active phases of the cell cycle (G1, S, G2, and mitosis) but not in resting cells [8]. We predicted that in the presence of residual and viable dysplastic cells after APC, active cell division would occur and that Ki-67 expression would be detected. As we expected, the results showed that Ki-67 positivity in the 24-h follow-up biopsy was statistically significant and correlated with the presence of dysplasia at both 24 h and three months after APC.

Ahn et al. [4] investigated the efficacy of APC for the treatment of gastric dysplasia and identified the non-lifting sign and low APC output power as risk factors for local recurrence, with a relatively high recurrence rate of 21.2% (15 of 71 patients). This high recurrence rate may have been due to the inclusion of patients with HGD and severe fibrosis in their study. The present study also identified low APC output power as a risk factor for remnant lesion development after three months, consistent with the study of Ahn et al. [4] Given that an output power of 40 W was identified as a risk factor for local recurrence in previous studies, the present study involve APC performed at output power settings of 45, 60, and 80 W. Ex vivo studies using human gastric tissues and porcine colon have shown that longer duration APC procedures and higher output powers can increase the extent of mucosal injury [9-11]. However, in this study, a correlation between APC procedure duration and the presence of residual lesions was not evident. The primary reason for this discrepancy may be the difference in the study design (ex vivo studies vs. patient studies). Ex vivo studies are conducted using non-moving tissues, whereas the tissues of patients are constantly moving (because of breathing, slight movements, and other factors), resulting in discontinuous APC application. In this study, the APC procedure duration was defined as the time that elapsed between the start and end of APC. The present results were expected to differ from those of previous studies because APC was not continuously applied in this study. In addition, a saline injection was administered before APC to avoid complications (e.g., perforations). Previous studies using porcine models showed that saline injections are effective for minimizing muscle damage [12-14]. Considering that Ahn et al.’s [4] study, in which the non-lifting sign was identified as a risk factor for local recurrence, a possibility exists that endoscopists may apply insufficient energy through the tissue owing to concerns about complications. In a previous study, the complication rate of APCs performed after saline injection was 1.4%–4.6% [3-6]. No postprocedural complications occurred in the present study. One study that reported a high complication rate did not involve hospitalization of the patients [6]. In the present study, the patients underwent hospitalization, fasting, and intravenous infusion of proton pump inhibitors. These differences may account for the discrepancy in the observed complication rates.

An important and novel finding of the present study is that the coagulation effects of APC persisted for >24 h. Of the 13 patients who had dysplastic cells 24 h after the procedure, 11 were available for follow-up endoscopy three months later. Among these, only three patients continued to exhibit dysplasia. Although previous studies have not evaluated the duration of tissue damage caused by APC, the present results indicate that the effects of APC persist for at least 24 h. Of the 13 patients who exhibited dysplastic glands in tissue samples taken 24 h post-procedure, eight showed no signs of dysplasia in tissue samples collected three months later. Examination of the 24-hour post-procedure tissue slides for these eight patients, as depicted in Fig. 3, revealed that the glandular morphology was altered compared with the initial biopsy, suggesting that although some dysplastic glands may persist after APC, the coagulation effect generally continues, ultimately leading to the destruction of most glands. Considering that the three patients with residual dysplasia at three months were all treated at an output power of 45 W and that residual lesions were not observed in patients treated at an output power of ≥60 W, our results imply that the higher APC output power results in a prolonged duration of tissue damage.

Gastric dysplasia is a precancerous condition. A recent meta-analysis reported that in 25% of LGD cases, the diagnosis was upgraded to advanced dysplasia or EGC after endoscopic resection [15]. From this perspective, APC is a demanding procedure because the pathological results cannot be verified. However, dysplasia, especially LGD, is pathologically limited to the epithelium and, unlike cancer, does not invade downward [16,17]. As demonstrated in this and previous studies, APC at an appropriate output power can effectively ablate the mucosal layer. Previous studies have shown that HGD, surface redness, depressions, and lesions >2 cm are more likely to be diagnosed as cancer after endoscopic resection [18-20]. Except for cases involving these high-risk lesions, APC can be safely performed for other (low-risk) lesions.

This study had some limitations. First, the study had a small sample size and was conducted at a single institution. Second, a significant number of patients withdrew from the study, largely because of the discomfort associated with the second-look endoscopy, leading to a disproportionate distribution of patients across the groups. The primary reason for this appears to be the lack of financial resources to compensate participants for their involvement in the study. Finally, logistic regression analysis was unsuccessful in determining the odds ratios for the risk factors of residual dysplasia. Although the analysis was attempted using SPSS, only error values were obtained, and the statistician concluded that the small sample size precluded proper analysis.

In conclusion, APC is a safe and effective treatment option for LGD. Although some lesions may persist immediately after the procedure, most disappear due to the prolonged effects of the high output power used during the procedure. Due to its low risk of complications and short procedure time, APC is a viable treatment option for high-risk patients with multiple comorbidities and LGD, serving as an alternative to EMR and ESD. If Ki-67-positive cells are detected in the remnant tissue following APC, additional interventions should be considered.