The Challenge of Pneumatosis Intestinalis: A Contemporary Systematic Review

Purpose: Pneumatosis intestinalis is a radiological finding with incompletely understood pathogenesis. To date, there are no protocols to guide surgical intervention. Methods: A systematic review of literature, according to PRISMA criteria, was performed. Medline and PubMed were consulted to identify articles reporting on the items “emergency surgery, pneumatosis coli, and pneumatosis intestinalis” from January 2010 up to March 2022. This study has not been registered in relevant databases. Results: A total of 1673 patients were included. The average age was 67.1 ± 17.6 years. The etiology was unknown in 802 (47.9%) patients. Hemodynamic instability (246/1673–14.7% of the patients) was associated with bowel ischemia, necrosis, or perforation (p = 0.019). Conservative management was performed in 824 (49.2%) patients. Surgery was performed 619 (36.9%) times, especially in unstable patients with bowel ischemia signs, lactate levels greater than 2 mmol/L, and PVG (p = 0.0026). In 155 cases, surgery was performed without pathological findings. Conclusions: Many variables should be considered in the approach to patients with pneumatosis intestinalis. The challenge facing the surgeons is in truly identifying those who really would benefit and need surgical intervention. The watch and wait policy as a first step seems reasonable, reserving surgery only for patients who are unstable or with high suspicion of bowel ischemia, necrosis, or perforation.


Introduction
In 1783, Johann Georg Du Vernoy described, for the first time, pneumatosis intestinalis (PI), subsequently named Pneumatosis cystoides intestinalis by Mayer in 1825 [1].PI is a physical or radiographic finding; it suggests the presence of gas in the bowel wall.PI is distributed throughout the digestive tract involving the subserous and/or submucosa of the small and large bowel.The typical location of pneumatosis intestinalis is the descending and sigmoid colon [2].
The true incidence of PI is unknown but the growing use of CT scans has contributed to the increased detection of this radiographic finding [3][4][5].
To date, there are no protocols to guide surgical intervention.Several factors are related to different management approaches to PI [6].
First, the clinical significance of PI can vary from benign findings to pathologic lifethreatening bowel ischemia and necrosis [7,8].
Second, the pathogenesis of PI is poorly understood.PI can be the clinical manifestation of several diseases (IBD, especially Crohn's disease, immune reactions or infections, bacterial abscesses, suppurative cholangitis, and other conditions that may require surgical treatment as bowel obstruction, pseudo-obstruction, malignancies, diverticulitis, and paralytic ileus).
PI has been also related to portal venous gas (PVG), especially in case of intestinal ischemia requiring emergency surgery.
PI and its wide range of clinical manifestations and etiologies represent a challenge for physicians, and especially for surgeons, the choice of the right treatment is not so easy.
The treatment depends on several factors.Suspected etiology and clinical and radiological presentation are the main factors behind the choice of treatment.Treatment can vary from simple drug discontinuation to open abdomen [2][3][4][5].
Literature reports about PI are typically case reports or small case series.Only a few cohort studies with a high number of included patients have been reported.
A systematic review of the literature of the last years was performed to evaluate the factors behind the choice of treatment and the real need for surgery in patients with pneumatosis intestinalis.

Methods Design
An extensive bibliographic search of the literature was performed according to modified PRISMA 2020 guidelines (Figure 1).The study was not registered.
Second, the pathogenesis of PI is poorly understood.PI can be the clinical manifestation of several diseases (IBD, especially Crohn's disease, immune reactions or infections, bacterial abscesses, suppurative cholangitis, and other conditions that may require surgical treatment as bowel obstruction, pseudo-obstruction, malignancies, diverticulitis, and paralytic ileus).
PI has been also related to portal venous gas (PVG), especially in case of intestinal ischemia requiring emergency surgery.
PI and its wide range of clinical manifestations and etiologies represent a challenge for physicians, and especially for surgeons, the choice of the right treatment is not so easy.
The treatment depends on several factors.Suspected etiology and clinical and radiological presentation are the main factors behind the choice of treatment.Treatment can vary from simple drug discontinuation to open abdomen [2][3][4][5].
Literature reports about PI are typically case reports or small case series.Only a few cohort studies with a high number of included patients have been reported.
A systematic review of the literature of the last years was performed to evaluate the factors behind the choice of treatment and the real need for surgery in patients with pneumatosis intestinalis.

Design
An extensive bibliographic search of the literature was performed according to modified PRISMA 2020 guidelines (Figure 1).The study was not registered.All stages of study selection, data abstraction, and quality assessment were carried out independently by three reviewers (M.G. and A.A.).Any disagreements were resolved by consulting two other reviewers (F.C., G.P.).
Medline and PubMed were consulted in order to identify articles reporting the item "emergency surgery" from January 2010 up to March 2022 and then the Boolean operators "AND" and "OR" were used to mesh it with the following mesh terms: "pneumatosis coli", "pneumatosis intestinalis", "acute mesenteric ischemia".Additional articles were searched by manual identification from the key articles.
We decided to include only papers from 2010 analyzing only a limited period of time.This choice was made to reduce the diagnostic and treatment biases of the past decades related to medical breakthroughs.We aim to take a picture of the etiology, diagnosis, and treatment of PI to understand why and when surgery must be performed or avoided.
Inclusion criteria: pneumatosis intestinalis of the small bowel and large bowel, articles in the English language.In the case of multiple papers from the same group of authors, an effort was made to identify duplicate papers.In the final dataset, every paper on pneumatosis intestinalis (cohort studies, retrospective and prospective studies) is included, also case reports and case series with complete data were included in the paper.
Exclusion criteria: Cases were excluded if the studies reported incomplete data or if the studies were not available in the English language or performed not in humans.Reviews were excluded.
Data relevant to the items of interest were abstracted.Several parameters were recorded and analyzed: gender, mean age, etiology, laboratory tests including cultural exams, symptoms, assessment of hemodynamic status (stable or unstable patients) diagnostic tests (colonoscopy, CT-Scan), location of PI, presence of pneumoperitoneum or portal vein gas (PVG) at diagnosis or delayed, treatment (conservative, surgical) and follow-up.Primary or secondary outcomes were analyzed.
Data analysis was performed using IBM SPSS Statistics 26.0.Univariate and multivariate analyses were performed.
Statistical analysis was obtained for the main descriptive indexes.Quantitative data are expressed as mean or median ± standard deviation (SD).The qualitative data were elaborated as absolute frequencies, relative frequencies, cumulated frequencies, and percentages.
All factors were deemed to be statically significant at a p-value of less than 5% (p < 0.05).

General Characteristics
After the assessment of abstracts and papers according to the inclusion criteria, 188 articles were included (Figure 1, Table 1).A total of 1673 patients with pneumatosis intestinalis were included in the study, 773 (46.2%) were males and 581 (34.7%) were females.Gender was not reported in the remaining 319 (19.0%).
Bacterial etiology was reported only in a few cases.Strongyloides stercoralis and Clostridium difficile were identified in 7 and 3 cases, respectively.Demographic, pathological features, and etiology are detailed in Tables 2 and 3.The most common symptom was abdominal pain with distension in 396 patients (23.6%).Hemodynamic instability was found in 246 (14.7%) patients.
CT-scan of the abdomen was the most common diagnostic test in 1673 patients (100.0%).Plain X-ray was performed in 459 (27.4%) cases, colonoscopy was performed in 49 cases (2.3%).
PI of the small bowel was the most common site in 610 (36.4%) cases followed by colon and rectum in 497 (29.7%) cases.
The whole colon was involved in 107 (6.3%) patients.In 13 papers (566 patients, 33.8%), the exact location of PI was not reported.

Therapy
Conservative management was the most common treatment in 824 (49.2%) cases.Surgery was performed in 619 (36.9%) patients.Treatment was not reported in 230 (13.7%) cases.
Bowel rest, fluid administration, and antibiotics were the most common conservative treatments in 266 (15.8%) patients.
Every patient with PI related to IBD flare was successfully treated with a high dose of mesalamine and prednisone.PI caused by chemotherapeutic agents, monoclonal antibody drugs, and alpha-glucosidase inhibitors for diabetes were treated successfully with therapy discontinuation in 96 cases (11.7%).
Surgery was performed 619 (36.9%) times.Data about the surgical treatment of 227 (36.6%) patients were not reported.
Among the 619 who underwent surgery, bowel resection was the most common treatment in 237/619 (38.2%) cases.
Death was reported in 390 (23.3%) cases, and 293 (75.1%) occurred in patients with critical conditions at hospital admission or during the first day after admission, where only supportive therapy was given.A total of 41 (10.5%) deaths were related to other causes.During the follow-up of the 155 cases treated with laparoscopic/laparotomy exploration alone without bowel resection were not reported as deaths.
Surgical management was significantly higher in unstable patients, with bowel ischemia signs, lactate levels greater than 2 mmol/L, and PVG (p = 0.0026).
Hemodynamic instability was reported in 246 patients (14.7%).Data about the clinical status of patients have not been reported in 309 (18.4%) patients.
Higher mortality was significantly related to unstable patients, lactate levels greater than 2 mmol/L, and bowel ischemia signs (p = 0.031) but not with PVG (p > 0.05).

Discussion
Pneumatosis intestinalis is a radiological sign that shows several diagnostic and treatment issues.
Treatment can be a lifesaving decision and often the timing for surgical intervention is wrong.Clinical evolution of PI can often be unpredictable, it is responsible for a difficult treatment decision-making process that requires careful evaluation of every variable.
PI can be divided into primary PI (15% of all PI cases) and secondary PI representing 85% of cases.PI can be also divided into pathologic and asymptomatic PI [197].
Secondary PI has been attributed to endoscopic procedures, immunological disturbances, bowel mucosal disruptions, and intra-abdominal pathologies.
Pneumatosis intestinalis is a radiographic phenomenon produced by underlying diseases, which can vary widely.The pattern or extent of PI does not necessarily correlate with the severity of the symptoms or of the underlying disease.The same etiology can lead to both asymptomatic or pathologic PI, the PI severity depends on several factors, but there are no specific findings for pathological and asymptomatic PI [193,195,198,199].
The etiology, both for primary and secondary PI, remains unclear.More than 60 causative diseases and conditions have been identified, but the specific pathophysiology remains unknown [11,16,20,26,127,138,173].
Two pathogenetic hypotheses have been proposed, the mechanical and bacterial theories.
The mechanical theory hypothesizes that gas dissects into the bowel wall from the bowel lumen to some mechanism, causing increased overpressure, such as a bowel obstruction.
The bacterial theory proposes that gas-forming organisms produce gas within the bowel wall, entering the submucosa through mucosal rents or increasing mucosal permeability.
Different laboratory tests (CRP, LDH, and CPK) were reported to be elevated in the case of PI, especially in bowel ischemia, but their role in the diagnosis of pathologic PI is limited because they can be also elevated in systemic inflammatory reactions [47,152,162,192,200].
The patient's personal history is mandatory in order to discover an underlying cause of PI, as suggested by our results where PI etiology was identified in 52% of the patients (recent endoscopy, diabetes therapy, steroid therapy, IBD, etc.).
Many studies have attempted to create algorithms for PI management.These algorithms may be difficult to apply clinically, especially when the patient requires immediate evaluation.Several studies have investigated the role of risk factors (hypotension, peritonitis, renal failure, serum lactate levels, older age) as predictors of a compromised bowel and the probable need for surgery [5,153,154,193,195,201].
The benign causes of PI usually result in mild or even no abdominal symptoms.In these patients, there are often no CT abnormalities other than the diagnosis of pneumatosis intestinalis.
CT findings can lead to an overtreatment of patients with PI.Portal venous gas has been traditionally associated with bowel necrosis, but our results do not suggest that PVG is always related to bowel ischemia.Among the 556 patients with PVG, 33.6% underwent surgery for bowel ischemia, necrosis, or perforation .
Peritoneal symptoms are usually reported in patients with life-threatening causes of PI.
The treatment of pneumatosis intestinalis must focus on the underlying disease rather than on the radiographic sign itself.Surgery could be avoided when a non-organic etiology has been discovered.In this study, bowel resection was performed in 149 patients (48.3%) due to organic disease (volvulus, intussusception, Ogilvie's syndrome, bowel obstruction, etc.).In 155/392 (39.5%) cases, surgery was performed without the identification of intraoperative pathological findings.
The treatment decision-making should be based on different points of view: the clinical status of patients, the presence of an underlying condition, the need for emergency surgery, and the possibility of simple observation and re-evaluation .
The timing and the decision process are crucial for the patient's outcome.The first step remains patient physical examination.Unstable patients with signs of sepsis and symptoms of shock are most often associated with mesenteric ischemia, bowel necrosis, or bowel obstruction, as suggested by our findings.The outcome for these patients is most unfavorable among patients with PI.Surgical exploration has been performed in almost all cases of instability.
The second step is the identification of an underlying disease that may guide the treatment choice.An accurate anamnesis is fundamental to identifying and treating several diseases or conditions related to PI.
The third step includes the need for surgery.This is the sore point of PI treatment.
When an organic disease has been identified (bowel obstructions, intussusception, or volvulus) surgery remains the main treatment option, and also unstable patients could benefit from surgical exploration.
For stable patients without organic disease, a watchful waiting approach may be more indicated.The possibility of simple observation and re-evaluation should be considered, especially in stable patients with unknown etiology.
Instrumental findings of PVG and massive PI alone in stable patients are not mandatory for surgery.Another approach to stable patients could consist of initial laparoscopic exploration in patients with one or more signs of bowel ischemia or necrosis.Laparoscopy as the first step could avoid unnecessary laparotomy.

Conclusions
Our findings suggest and confirm the challenges associated with the appropriate treatment of patients with pneumatosis intestinalis.Many variables should be considered in the approach to patients with pneumatosis intestinalis.The treatment of patients with pneumatosis intestinalis is a lifesaving decision and the timing for surgical intervention is crucial.Accurate personal history of patients is fundamental for the management.Considering the wide range of causes and outcomes of pneumatosis intestinalis, the watch and wait policy as a first step could be reasonable in selected cases.
Surgery remains mandatory in unstable patients and when an organic disease has been identified.Surgical options should be explored, especially laparoscopic exploration in non-responders to conservative management with high suspicion of bowel ischemia and necrosis.It is important to recognize pneumatosis intestinalis as a clinical sign and not as a diagnosis.

Table 1 .
Papers included in the literature systematic review.

Table 2 .
Demographic and pathological features of the studied population.

Table 3 .
Etiology of Pneumatosis Intestinalis.: Pneumatosis Intestinalis.* Percentage refers to the total of patients for respective etiology. PI