Clinical parameters and outcomes of necrotizing soft tissue infections secondary to gastrointestinal fistulas

Gastrointestinal (GI) fistula represents a relatively rare yet serious condition in clinical scene [1‐3]. Despite recent advances in the field of surgical infections, the incidence and mortality of intestinal fistulas is particularly high [4‐6]. Infectious complications secondary to GI fistula can range from localized infection to lethal sepsis. Even more serious is that once the uncontrolled spillage of intestinal content into the abdominal wound, with ongoing sepsis and usually accompanies extensive necrosis of the abdominal wall fascia [6]. There is no doubt that necrotizing soft-tissue infections (NSTIs) of the abdominal origin are a rare but serious in-hospital complication [7].

NSTIs are a group of life-threatening skin infections associated with high rates of both morbidity and mortality [8‐10]. However, there are a few well-documented cases of NSTIs secondary to gastrointestinal fistula [11‐13]. GI fistula associated NSTIs has special clinical characters compared to other types of NSTIs [14]. These critically ill surgical people’s conditions change rapidly, and there are serious disorders in the host homeostasis. Despite the severity of this complication, there is no generally accepted therapeutic approach to GI fistulas associated NSTIs. Brafa et al [15] reported that used an abdominoplastic advancement technique for delayed primary closure of a cancer patient affected by necrotizing fasciitis after 2 months of serial debridement and negative pressure wound therapy (NPWT).

However, most studies on GI fistulas associated NSTIs are retrospective case series and focus on predictors of mortality and length of hospitalization [16‐18]. Very few studies have examined surgical procedures or clinical outcomes. We report a large series of NSTIs secondary to GI fistulas at a tertiary-referral center. This report analyzes the surgical treatment of GI fistula associated NSTIs, and also examines related factors that affect patient’s prognosis.

Patients with GI fistula associated NSTIs at the Research Institute of General Surgery in Jinling Hospital between 2014 and 2017 were identified from the prospectively maintained gastrointestinal fistula database. The study was conducted in accordance with the ethical principles of the Helsinki Declaration and was approved by the Jinling Hospital Ethics Committee. Written informed consent was obtained from all individual participants included in the study. The patients included in the database have been followed by reviewing their medical information and by directly contacting the patients.

All patients were included in this study according to the following inclusion criteria: (1) age between 18 and 65 years; (2) diagnosis of GI fistula associated NSTIs according to clinical, radiological and histological findings. Patients who died or discharged within 48 h of admission was excluded. Patients with incomplete or missing date were also excluded.

NSTIs are a group of aggressive soft tissue infections characterized by widespread necrosis of the fascia and subcutaneous tissue with initial sparing of the skin and muscle [24]. Options of the treatment of NSTIs depend on the etiology of the infection and anatomic location of the process. When NSTIs occur are in association with gastrointestinal fistulas, they are often associated with peritoneal contamination, intra-abdominal abscesses, fluid loss, and evisceration [11, 12, 25]. As previously noted, early identification of sepsis, source control, and prompt antibiotic administration remains the mainstay of treatment in abdominal wall necrotizing fasciitis [26].

As shown by Fig. 2, a 27-year-old male was taken up for traffic accident for polytrauma and the patient was successfully treated with fluid replacement, broad-spectrum antibiotics, and debridement of necrotic tissue, followed by reconstructive surgery. Computed tomography (CT) scans showed necrotizing soft tissue infection of the abdomen and perineum (Fig. 2a). Necrotizing had also affected the soft tissues below the skin, causing stretching of abdominal wall down to the right hip area (Fig. 2b). On local examination, the skin around the abdominal incision site was edematous, indurate and large area of skin defect (Fig. 2c and d). Under local anesthesia, serial surgical debridement and change of regular dressing were performed at last 3 times a week. When the patient was taken for surgical debridement (Fig. 2e), cover the wound defect with chitosan sponge dressing to remove exudates and to promote wound healing (Fig. 2f), the severe infection subsided with daily wound irrigation and fresh granulation tissue gradually formed (Fig. 2g). At 32th day, her condition was significantly improved (Fig. 2h) and scar tissue eventually formed after discharge (Fig. 2i).

Different types of microorganisms can cause NSTIs. In our study, the majority of cases begin with an existing infection, most frequently in the abdomen, perineum or skin. Previous studies show mortality in patients with NSTIs was significantly associated with the presence of Vibrio spp in wound cultures and Streptococcus group A in blood cultures [27]. NSTIs caused by gastrointestinal fistula have special clinical characters compared with other origins. NSTIs in GI fistula patients have its own microbial distribution that all infections originated from the patients’ endogenous microflora. In the present study, Klebsiella pneumoniae were the most common pathogens isolated from wound culture, which is consistent with the results of previous studies [28]. Therefore, such high-risk groups should be given timely identification and bacteriologic diagnosis. Such patients will benefit from appropriate antibiotic administration. Further study is needed to determine the best combination strategy and optimum duration of antimicrobial therapy in these patients.

The key tasks in the surgical management of patients with NSTIs secondary to gastrointestinal fistula are source control. It is apparent that removal of infected necrotic tissue by drainage and debridement plays an important role in eliminating the trigger of an ongoing inflammatory response. It is borne in mind that intra-abdominal and extra-abdominal infections should receive the same attention. If the intra-abdominal disorders can be appropriately managed, the soft tissue infection also can be treated effectively [13]. However, some patients often have life-threatening conditions, such as hemodynamic instability and shock. When the clinician may first focus only on resuscitation, may result in a significant delay in diagnosis or an increase in clinical infection. Currently, open abdomen management became an effective treatment option of abdominal disaster, including intra-abdominal infections, gastrointestinal fistula, abdominal compartment syndrome (ACS) and wound dehiscence [29, 30]. Seternes et al. has reported 7 patients with necrotizing fasciitis undergo open abdomen treatment [31]. In our study, eight patients have undergone open abdomen to treat abdominal wall necrotizing fasciitis. Therefore, necrotizing fasciitis is an indication for open abdomen in some cases, such as difficult wound close and extensive tissue defects.

The use of NPWT in the wound management of NSTIs has been well documented. This technology facilitates the inflammation elimination and wound healing process, which involved mechanisms included reducing edema, removing infectious materials and exudates, and increasing blood supply [29, 32]. Our results are consistent with the positive outcomes of the previous studies. However, this method has some disadvantages and limitations such as abdominal skin loss after drainage and irrigation [33]. Additionally, skin margin necrosis due to excess tension in suture traction might be a potential complication. In our study, 11 patients had local skin grafting procedure and 16 patients underwent delayed abdominal wall reconstruction closure, because the defect size was reduced to small and medium after NPWT.

The mortality rate of NSTI is still high, and the overall mortality rate is between 25 and 73%, despite the use of modern powerful antimicrobial drug and advances in the nursing care [32, 34, 35]. Arif and her colleagues’ results showed that NSTIs-related deaths are associated with more diabetes and obesity when compared to other fatal diseases [32]. The other main prognostic factors of these patients include advanced age, poor nutrition, concomitant diseases and immunosuppressed host and nosocomial infection. However, it is quite different from surgical patients, especially in patients with GI fistula. After the appearance of fistula, continuous chemical irritation of digestive effluent can severely compromise skin integrity and exacerbate the spread of infection. In-hospital mortality in our cohort of patients was associated with multiple organ dysfunction syndrome, shock, polymicrobial infection, abdominal hemorrhage, multidrug-resistant organisms and source control failure. The above reasons are similar to the causes of death in patients with intra-abdominal infections [36], which is determined by the characteristics of our patients. Thus, any measure to prevent NSTIs development or control its drainage should be attempted. In our practice we treat all patients with washing and drainage, multiple debridement in 17 patients, open abdomen in 8 patients, NPWT in 33 patients and hyperbaric oxygen in patients. Finally, early source control (< 48 h) was achieved in 22 (56.4%) patients. This comprehensive strategy is also necessary to treat NSTIs.

Our research also has many disadvantages. As a retrospective study in single center, our patients are relatively complex surgical patients who have undergone multiple operative procedures and hospitalizations before NSTIs occurs, so that some of our pre-hospital information is incomplete. Furthermore, there are multiple infection sites of fistula combined with NSTIs, and the therapeutic effect of the drug may be confused. For example, the time of use of antibiotics and the effect of dosage on the final outcome of patients. In addition, we only focus on patients with surgical infections, which may make it difficult for our research to be extended to all clinical scenarios. However, at present, there are limited researches on NSTIs secondary to GI fistulas, and our study might have some implications for this special populations.

The management of GI fistulas associated NSTIs challenge due to the presence of primary disease and the formation of fistulas, as well as variations in host system responses. Sepsis is the most important factor associated with the death of GI fistulas combined with NSTIs. Therefore, it is crucial to prevent the occurrence and deterioration of sepsis from the early source control. Our data may assist providing enlightenment for quality improvement in these special populations.