Multiple strictures of the small bowel induced by nonsteroidal anti-inflammatory drugs (NSAIDs), also known as diaphragm disease, were first described by Lang et al. in 1988 [1]. Diaphragm disease is characterized by multiple, thin, concentric septa-like mucosal projections or broad circumferential ridges that narrow the small bowel lumen to a ‘pinhole.’ This can result in a subacute intestinal obstruction that sometimes requires laparotomy. Long-term use of NSAIDs is considered to be the most important causative factor of this disease entity, but several studies have reported cases with multiple diaphragm-like strictures in the small bowel that are not associated with the chronic use of NSAIDs [2]. With the widespread use of NSAIDs, an elderly society, and the advent of recent modalities such as capsule endoscopy (CE) and doubleballoon endoscopy (DBE), the incidence of diaphragm disease is on the rise and seems likely to increase in the future [34].

We report 3 cases of diaphragm disease requiring surgery from May 2012 to July 2016. We reviewed the clinical features of the disease by radiologic, endoscopic, operative, and pathologic findings.

The Institutional Review Board of Daegu Catholic University Medical Center is waived for case report.

Diaphragm disease is now considered to be an enteropathy secondary to NSAID-induced injury. NSAIDs are the most commonly prescribed drugs in the world, and while their adverse effects on the stomach and the duodenum are widely recognized, their effects on the small bowel are underreported. Although 50%–70% of patients on long-term NSAIDs develop certain forms of small bowel enteropathy, the incidence of NSAID-induced diaphragm disease is only around 2% [5]. This underreporting is probably due to their subclinical courses. Furthermore, most clinicians are unaware of the fact that patients taking NSAIDs can develop these strictures.

The exact pathogenesis of diaphragm disease remains poorly understood. However, damage induced by NSAIDs has been discussed extensively [14678]. Therefore, it is thought that a reduction in the mucosal synthesis of prostaglandins induced by NSAIDs results in damage to enterocytes and villous microcirculation, and disrupts mucosal integrity. The subsequent inflammatory reaction leads to circumferential ulceration that may result in mucosa vulnerable to bacteria or toxin. In the healing phase, submucosal granulation tissue matures into collagenous scar tissue, and these rings of scar tissue contract, like drawstrings across the bowel lumen, eventually forming diaphragm-like strictures. The microscopic feature of diaphragm disease is not pathognomonic, but usually shows multiple shallow ulcers, submucosal fibrosis and/or a variable degree of accompanying stromal inflammation, and less evident for the chronicity of crypt inflammation related to the inflammatory bowel disease or connective tissue disorder. Of note, it appears that NSAIDs produce these diaphragm-like strictures only in a few susceptible patients. Recently, it has been found that a specific cytochrome P450 2C9 polymorphism, which causes a decreased breakdown of NSAIDs, was significantly associated with an increased risk of diaphragm disease [9].

The relative risks of the different NSAIDs are not very clear, and many studies have shown that selective COX-2 inhibitors may be significantly less injurious to gastrointestinal (GI) tract than traditional NSAIDs [9]. Dosage is one factor affecting the plasma concentration of NSAIDs. In general, diaphragm disease has been associated with high doses taken daily [1]. Diaphragm disease can occur anywhere along the whole GI tract. However, the majority of instances are located in the small intestine. Many reports state that small bowel diaphragm disease is located predominantly in the ileum [28]. This may be because of differences in the bacterial flora and immune system between the jejunum and ileum. Colonic diaphragm disease has more recently been reported and most often has been related to slow-releasing or enteric-coated NSAID forms [78]. Because these NSAIDs have a longer half-life, they are more likely to reach the colon before they are entirely digested.

Definite diagnosis is challenging because symptoms are often nonspecific, and conventional radiological techniques are inaccurate. Therefore, it is important for the physician who is aware of the disease to consider the possibility of diaphragm disease. Laboratory tests may reveal microcytic anemia and hypoalbuminemia [4]. Plain abdominal radiographs are usually unhelpful. Barium studies may show the diaphragms, but they are as easily overlooked as thin-walled diaphragms resembling exaggerated plicae circulares [24]. A CT scan may show a degree of obstruction but is unable to identify the thin diaphragms. The advent of CE and DBE may facilitate evaluation of the small bowel. Although the diagnosis of small bowel diaphragm disease has improved with the introduction of CE, there is a risk of capsule retention if strictures are present [238]. In all our cases, CE was performed, and all the capsules were retained by circumferential strictures. Under these circumstances, it usually requires an exploratory laparotomy or DBE to remove the retained capsule, so it should be used cautiously. If an obstruction is suspected, DBE might be a better option; this also has the advantage of allowing biopsies for histology. DBE is a valuable and minimally invasive technique for the detection of diaphragm disease, and endoscopic treatment is possible [10]. Despite these advances, a correct preoperative diagnosis is made at laparotomy in most cases.

At laparotomy, the meticulous palpation of the suspicious lesions is essential for making the diagnosis. The lesions may be missed unless they are specifically sought because they affect only the mucosa and submucosal layers leaving an intact serosa, and the diaphragm may feel slightly thickened. Therefore, the role of laparoscopy appears to be limited because the bowel usually has a normal appearance [38]. In our case, we failed to identify diaphragm disease by laparoscopy. Through mini-laparotomy incision, we can decide on the extent of the diseased segment with palpation. Additionally, it is important to examine the whole intestine carefully because there is potential for the existence of concomitant strictures in other regions. If the diagnosis is uncertain, intraoperative enteroscopy may be used to explore and assess the extent of the lesion. The management of diaphragm disease in most reported cases entails segmental resection of the involved intestine because multiple lesions are located close together along the intestine [8]. Although not common, stricturoplasty may be an alternative choice for scattered lesions that are impossible to resect. Recently, with the development of DBE, endoscopic balloon dilation could be considered as an alternative option for diaphragm disease [10]. The risk of intestinal perforation with endoscopic balloon dilation would be low because of the fibrotic features of the diaphragm disease.

The prognosis would be good if the offending NSAIDs could safely be withdrawn. The benefit of continuing with NSAIDs may outweigh the risk of GI injury for some patients; therefore, this needs to be considered on a case-by-case basis [3]. The use of prostaglandin derivatives (such as misoprostol) may protect against NSAID-induced GI damage; therefore, their concomitant use should be considered in patients who are particularly at risk of NSAID-associated GI complications where NSAIDs cannot be withdrawn [4]. A previous study has reported that a recurrence rate as high as 50% after surgery may be caused by the surgeon's failure to appreciate the extent of the lesions at the initial operation or a true recurrence due to continued use of NSAIDs [1].

In conclusion, the diagnosis of diaphragm disease is possible with a thorough understanding and careful interpretation of the disease. Especially, clinicians should be aware that diaphragm disease might be a cause of small bowel obstruction in patients receiving NSAIDs. It is important to confirm the diagnosis by meticulous palpation because the outer appearance shows only subtle changes. Further studies are needed to evaluate the pathogenesis, the risk factors, newer diagnostic methods, and treatment including medication.