This study aimed to verify whether it is reasonable to distinguish AEML and sRE-D by the characteristics of the proximal mucosal damage and to clarify the characteristics of AEML in comparison with sRE-D. Compared to the sRE-D group, the AEML group had more comorbidities, most of which were infections. The fact that many patients in the AEML group had been treated for malignancies was also considered to be a reason for their susceptibility to comorbid conditions, particularly infections. The significantly higher WBC and CRP levels in the AEML group than in the sRE-D group may have been due to the higher prevalence of coexisting diseases, such as infections. The possibility of AEML should be considered in patients with worsening systemic conditions, including infections.
The AEML group showed significantly higher lactate levels and acidemia than the sRE-D group. Lactate elevation and acidemia are caused by anaerobic metabolism and lactate production due to tissue hypoxia. The AEML group had more gastric and duodenal ulcers than the sRE-D group, suggesting that mucosal injury may have occurred because of impaired circulation, which may have been due to worsening of the patients’ general condition because of coexisting disease. Additionally, the fact that the use of NSAIDs was significantly higher in the AEML group supports this hypothesis, as the use of NSAIDs may contribute to impaired circulation [11].
The BUN and creatinine levels were also significantly higher in the AEML group. Compared with those in the sRE-D group, patients in the AEML group were more likely to exhibit renal dysfunction and volume loss due to deterioration of their general condition. In particular, volume loss resulted in a tendency for the development of microvascular insufficiency.
The blood glucose levels were significantly higher in the AEML group. This is because patients with AEML are under physical and psychological stress, as well as because some patients had coexisting diabetic ketoacidosis (DKA) and hyperosmolar hyperglycemic syndrome (HHS). In particular, DKA and HHS are diseases that cause severe circulatory insufficiency, which may lead to esophageal mucosal damage and the development of AEML.
The possibility of AEML been caused by gastric acid reflux has been reported in previous studies [1–4]. However, in the present study, the use of PPI was higher in the AEML group (<45%), and esophageal hiatal hernia was less common in the sRE-D group. Since atrophic gastritis was also comparable, it can be considered that there was no significant difference in gastric acid secretion capacity between the two groups. Compared to the sRE-D group, AEML development seemed to be more related to inadequate peripheral circulation than to gastric acid secretion.
The frequency of endoscopic hemostasis in the AEML group was 7.8%, which was significantly lower than that in the sRE-D group. Bleeding was less frequent in the AEML group, but this did not mean that endoscopic procedures were unnecessary. Patients with AEML improve quickly with conservative treatment [1, 2, 4, 7]. If there is no bleeding on upper endoscopy, it is important to improve the general condition of the patient, including the treatment of comorbidities.
In the AEML group, a case of stenosis was observed after healing. Thus, the incidence of stenosis was only 1.6% in our study, whereas previous reports have shown an incidence of 4% [4]. The most common timing for the stenosis is at 7‒10 days after onset [5], but it can also occur within 1 month, as in this case. Therefore, it is necessary to perform follow-up at an appropriate time, considering the risk of stenosis.
In this study, we differentiated sRE-D from AEML by the presence of spiculation on the proximal edge of the esophageal mucosa, as reported by Sakata et al. [7]. The characteristics of proximal mucosal damage in AEML were classified into two types as shown in Figure 6, but many of them changed gradually. The reason for this is that the proximal mucosal damage in sRE is clearly defined, and the involvement of gastric acid reflux may be strong in the clear boundary. However, this has not been proved yet. The squamocolumnar junction has a clear border, but the proximal mucosal damage is gradual, which is typical of AEML.
The reason for the significantly higher number of older women in the sRE-D group was considered to be the effect of hunchback. Older women are more prone to osteoporosis and hunchback than men. Chronic acid reflux and decreased clearance due to hunchback may predispose older women to sRE-D.
The novel findings of this study are as follows: 1) the distinction between sRE-D and AEML, which is more relevant in clinical practice; 2) suggesting that microcirculatory disturbances may be more relevant to the development of AEML than acid reflux, compared with sRE-D; 3) two types of erosions can be distinguished in AEML; and 4) the possibility of stenosis in AEML.
This was a retrospective, single-center study, which may limit the generalizability of the results. It should also be noted that the statistically significant difference in AEML in this study is only a result of comparison with sRE-D and may not be a significant feature of AEML itself.