Case Report: Delayed complete transection of the sigmoid colon after severe pelvic fracture leading to septic shock

1 Introduction

High-energy pelvic ring injuries place a significant burden on trauma centers and often occur in conjunction with severe concomitant injuries (1). Colorectal injury is uncommon in this setting but may have serious clinical consequences (2). If bowel injury is missed or diagnosed late, fecal contamination may transform an initially closed pelvic fracture into a functionally open injury. Mortality after open pelvic ring injury has been reported to approach 50%, with late deaths commonly attributed to wound sepsis and multiple organ dysfunction (3, 4). Here, we describe an unusual case of delayed complete transection of the sigmoid colon following high-energy pelvic trauma. Complete transection is not a distinct clinical entity merely because it is anatomically more extensive than limited perforation, and both require timely source control once fecal contamination occurs. The value of the present case lies instead in its unusually severe and concealed diagnostic course. In a patient who could not be examined reliably, CT on post-injury days 1–3, FAST, and repeated ultrasonography remained non-diagnostic, while some laboratory indices temporarily improved. The bowel injury became apparent only on day 6 after abrupt clinical deterioration. This case illustrates the discordance that may arise between a high-risk injury mechanism, serial negative examinations, transient biochemical improvement, and continued progression of an uncontrolled structural source.

2 Case presentation

2.1 Clinical presentation

In June 12, 2025, a 57-year-old woman was brought to the emergency department of our hospital two hours after sustaining multiple pelvic fractures in a traffic accident. Upon presentation, the patient was in poor condition, suffering from hemorrhagic shock. At emergency admission, her temperature was 38.5 °C, pulse rate was 115 beats/min, respiratory rate was 28–32 breaths/min, and blood pressure was 97/68 mmHg. During the first 15 h, urine output was 700 mL; total infused volume was 3,539 mL, and gastric decompression output was 800 mL. On abdominal palpation, the abdomen was soft, with no tenderness or rebound tenderness; the abdominal contour and bowel sounds were normal. Additionally, a preliminary abdominal CT scan showed no evidence of significant bowel injury (Figure 1A). The patient had previously been in good health and had no relevant medical history. Emergency FAST demonstrated only a small amount of scattered abdominopelvic fluid. Due to severe pelvic ring injuries accompanied by hemorrhagic shock, the patient gradually slipped into a coma and was immediately transferred to the ICU for further treatment. The patient's detailed clinical course is shown in Table 1.

Table 1. Clinical timeline.

Timeline of events, interventions, and findings.

2.2 Examinations and diagnosis

Emergency CT demonstrated a Young–Burgess lateral compression type III pelvic ring injury, characterized by fractures of the left sacrum, left ischium, and bilateral pubic rami, together with right sacroiliac joint disruption. A concomitant fracture of the left acetabulum was also identified (Figures 1F,G). The patient also had traumatic intracranial hemorrhage, acute respiratory failure, contusions of the liver, right kidney, and spleen, multiple rib and sternal fractures, a severe left-elbow soft-tissue contusion with an open wound, and a small hemoperitoneum. Laboratory tests showed a serum lactate concentration of 5.6 mmol/L, a procalcitonin (PCT) level of 4.29 ng/mL, and an interleukin-6 (IL-6) concentration of 2,348 pg/mL; meanwhile, the white blood cell count was 8.0 × 10⁹/L, and the high-sensitivity C-reactive protein (hsCRP) was within the normal range at 6.38 mg/L. Cefuroxime sodium was initiated empirically for pulmonary contusion with suspected concomitant pulmonary infection in the setting of elevated inflammatory markers. Comprehensive bedside ultrasonography after ICU admission showed no specific sign of bowel injury and no increase in the small-volume abdominopelvic fluid. However, on the second day after the injury, the patient's condition deteriorated significantly; a neurological examination revealed a Glasgow Coma Scale (GCS) score of 3. Within 24 h of admission to the ICU, the patient's APACHE II score was 41, with a predicted mortality rate of 92%. Bowel sounds became reduced (1/min). Repeat chest and abdominal CT scans of the patient showed no significant changes compared to the scans taken at admission (Figure 1B); there was no increase in intra-abdominal hemorrhage, and no signs of bowel injury were found. Laboratory testing showed a white blood cell count exceeding 20 × 10⁹/L, with neutrophils accounting for 89.5%. High-sensitivity C-reactive protein had risen to 186.3 mg/L, procalcitonin to 39.0 ng/mL, and interleukin-6 to more than 5,000 pg/mL. Arterial blood gas analysis was consistent with metabolic acidosis, showing a pH of 7.32, base excess of −7 mmol/L, and lactate elevation to 4.2 mmol/L. These abnormalities reflected worsening systemic inflammation and progressive organ dysfunction. Therefore, empirical treatment with meropenem was initiated. On June 13, tachypnea progressed to 40 breaths/min, and endotracheal intubation with mechanical ventilation was required. By post-injury day 3, bowel sounds were no longer audible. PCT, IL-6, and some hepatic and renal indices decreased from their early peaks. Although delayed intestinal injury was considered possible, serial abdominal CT scans over 3 consecutive days showed no evidence of bowel wall perforation or extraluminal gas (Figures 1A–C). The hemoperitoneum and retroperitoneal hematoma were considered likely related to the concomitant pelvic and acetabular fractures and liver, splenic, and renal injuries, with no significant increase in bleeding. Diagnostic abdominal paracentesis was performed within the first 3 days after admission, but only a minimal amount of intraperitoneal fluid was obtained, which was insufficient for diagnostic evaluation. Following aggressive resuscitation, although severe organ dysfunction persisted, including renal failure, the patient's hemodynamic status improved and inflammatory markers decreased. Therefore, exploratory laparotomy was not considered indicated at that time. The enteral nutrition was withheld from the day of injury (June 12) because of hemorrhagic shock and impaired gastrointestinal function, and continuous gastric decompression was instituted. On post-injury day 3, after the NRS-2002 score exceeded 3, parenteral nutrition with intravenous lipid emulsion, glucose, and amino acids was initiated; enteral nutrition remained withheld. On post-injury day 6, these markers rose abruptly again (Figures 2A,B), accompanied by high fever, progressive abdominal distension, and worsening hemodynamic instability. During the preceding 18 h, fluid intake totaled 5,753 mL whereas urine output was 31 mL. Urgent CT then demonstrated extraluminal gas near the sigmoid colon (Figure 1D). Emergency laparotomy confirmed complete sigmoid transection (Figure 1E).

2.3 Therapeutic intervention

Upon ICU admission, mechanical ventilation, restrictive fluid resuscitation, and blood transfusion were started to stabilize respiratory and hemodynamic function. In the initial stages of the damage control strategy, we performed closed reduction of the left hip dislocation, applied tibial skeletal traction to the left lower extremity, and used a pelvic fixation belt for temporary pelvic stabilization. Supportive care also included stress-ulcer prophylaxis and nebulized airway treatment. On post-injury day 3, the patient developed renal failure with anuria and was treated with continuous renal replacement therapy (CRRT); hepatic failure also developed, and hepatoprotective therapy was administered. After a repeat CT demonstrated free intraperitoneal gas on post-injury day 6 (Figure 1D), emergency exploratory laparotomy was performed. Intraoperatively, approximately 1,500 mL of mixed hemorrhagic and feculent intraperitoneal fluid was found, and complete transection of the sigmoid colon was confirmed, with active bleeding from the transected ends, partial stump necrosis, and ongoing fecal spillage (Figure 1E). Partial sigmoid colectomy, sigmoid colostomy, abdominal and pelvic debridement, drainage, and sigmoid suspension were performed to achieve source control and prevent postoperative volvulus or torsion. Given septic shock, extensive contamination, and nonviable bowel ends, primary anastomosis was not selected. Postoperatively, the patient was transferred to the ICU. The patient had severe multiple organ damage, and the estimated Injury Severity Score (ISS) was 66, indicating extremely severe trauma with a very low survival rate. On postoperative day 1, weak gastrointestinal motility permitted only small-volume glucose water through a nasogastric tube as a trophic tolerance assessment; amino-acid-based parenteral nutrition continued. On postoperative day 2, after no obvious gastric retention following water administration, peptide-based enteral nutrition was initiated through the nasogastric tube and combined with parenteral nutrition. Marked gastric retention on postoperative day 13 prompted temporary cessation of the enteral formula. Feeding was gradually advanced as gastrointestinal function recovered, and fruit juice and rice porridge were introduced approximately one month after colectomy. Vancomycin was added on June 18 to broaden Gram-positive coverage during septic deterioration. As the patient was in a state of septic shock and had risk factors for invasive fungal infection, the antimicrobial regimen was further broadened and adjusted to include vancomycin, imipenem-cilastatin, and caspofungin. Pelvic purulent-fluid and ascitic-fluid specimens collected on June 18 grew Enterococcus faecium. Caspofungin was changed to micafungin on June 21 because hepatic dysfunction worsened; vancomycin was dose-adjusted by therapeutic drug monitoring and antibacterial agents by estimated glomerular filtration rate. Later, microbiology included Candida glabrata in urine, and Stenotrophomonas maltophilia plus Enterococcus faecium in blood cultures. Respiratory cultures grew Stenotrophomonas maltophilia and Enterobacter cloacae. Despite repeated dressing changes, the severe left-elbow contusion wound continued to produce copious purulent drainage. Culture of the wound exudate grew Mucor circinelloides. Invasive mucormycosis was therefore suspected, and liposomal amphotericin B was initiated. Antimicrobial dosing was guided by serum drug concentrations, organ function, CRRT, and multidisciplinary consultation (Table 2). The patient's signs of sepsis gradually resolved, and hepatic and renal function improved (Figures 2C,D). CRRT was discontinued on July 5 after urine output improved, and furosemide was used to facilitate diuresis. By post-injury day 34, systemic sepsis and organ dysfunction had improved sufficiently for definitive pelvic fixation; this did not indicate eradication of all local infection. Infection had not been eradicated: gray purulent pelvic drainage persisted, and microbiological examination identified Proteus penneri and Escherichia coli. Culture-directed treatment was continued. Because local pelvic infection remained a concern, two cannulated screws and four Steinmann pins connected to external rods were used to form a hybrid internal–external construct, providing stability while limiting the amount of internal hardware (Figure 1H). Following definitive fixation of the pelvic ring and acetabular fractures, markers of systemic infection continued to decline. The antimicrobial regimen was subsequently adjusted over time to include levofloxacin, cefoperazone-sulbactam, and piperacillin-tazobactam. About 3 weeks later, this was changed to cefuroxime for maintenance therapy. The timeline of antimicrobial therapy during hospitalization is presented in Table 2.

Table 2. Antimicrobial therapy and documented microbiological timeline during hospitalization.

IV, intravenous. Antimicrobial dosing was guided by serum drug concentrations, organ function, CRRT, and multidisciplinary consultation.

2.4 Follow-up and outcomes

Multidisciplinary management contributed to the patient's survival despite a prolonged and complicated clinical course. Following a 97-day ICU stay, the patient's clinical condition had improved sufficiently for transfer to the orthopedic ward. During the subsequent 20-day stay on the orthopedic ward, the anterior pelvic ring–acetabular external fixation construct was removed, after which she was discharged. At discharge, significant infection-related sequelae remained, including severe bilateral lower-extremity weakness (left lower limb, MRC grade 1; right lower limb: proximal muscles, MRC grade 1, and distal muscles, MRC grade 0), consistent with critical illness polyneuropathy/myopathy secondary to sepsis and prolonged critical illness. Additional complications included localized infection-related bone resorption at the pelvic fracture site (Figure 1E), a persistent sigmoid colostomy, and a retained tracheostomy tube. At the 7-month follow-up after pelvic ring and acetabular fracture fixation, muscle strength proximal to the knees had improved to MRC grade 3 in both lower limbs, whereas distal strength had improved to MRC grade 1. A localized pelvic infectious focus persisted (Figure 1I).

3 Discussion

Several diagnostic challenges were evident in retrospect. Early abdominal assessment may have been confounded by several factors, including the presence of a retroperitoneal hematoma, and impaired responsiveness related to the concomitant head injury, all of which could have reduced the sensitivity of abdominal examination for detecting peritonitis (2). Initial CT may also be falsely reassuring; once free gas becomes visible, full-thickness bowel necrosis may already have occurred (5, 6). The early elevation of inflammatory markers was initially considered in the differential diagnosis to possibly reflect pulmonary or urinary tract infection. However, early sputum, blood, and urine cultures showed no bacterial growth after 5 days. All positive microbiological findings were obtained after sigmoid colon transection had been diagnosed at laparotomy on June 18. In retrospect, the available evidence did not support pulmonary or urinary tract infection as the explanation for the early inflammatory abnormalities. Bowel sounds were normal on admission, became reduced on post-injury day 2, and were absent from post-injury day 3 onward. Following repeat surgical assessment on post-injury day 3, diagnostic abdominal paracentesis was performed, but no diagnostically useful intraperitoneal fluid was obtained. The general surgery team considered that the dependent pelvic location and partial containment of the hemoperitoneum may have limited the diagnostic yield. However, in patients with high-energy pelvic trauma, particularly when clinical assessment is limited, persistent or increasing suspicion should prompt close observation, serial clinical and laboratory reassessment, and consideration of repeat contrast-enhanced CT. Surgical exploration may be warranted if the patient deteriorates or significant bowel injury is suspected despite negative or equivocal imaging findings.

To our knowledge, very few case reports have described sigmoid colon injury associated with pelvic fracture. Santana-Cabrera et al. reported a delayed sigmoid colon perforation 5 days after a major pelvic fracture caused by a motor vehicle collision, with no bowel injury detected on initial CT (7). Shapiro and Wolverson described retroperitoneal sigmoid colon perforation secondary to fracture-dislocation of the sacroiliac joint (8). More recently, delayed bowel injury associated with pelvic fracture has also been reported in other unusual clinical contexts (9). In the present case, the delayed bowel injury was confirmed intraoperatively as complete transection of the sigmoid colon, rather than a limited perforation or bowel-wall laceration. We agree that complete transection is not a separate clinical entity simply because it is more extensive than limited perforation. Both can cause fecal contamination and sepsis and require prompt source control. The present case is clinically distinctive because a severe injury evolved under unusually concealed conditions despite contemporary serial CT, FAST, bedside ultrasonography, and intensive monitoring. It became apparent only after progression to complete loss of bowel continuity, stump necrosis, extensive fecal contamination, septic shock, and multiorgan dysfunction. The clinical contribution of this case therefore lies not in the anatomical term “complete transection” alone, but in the mismatch between repeated negative CT and ultrasonographic findings, temporary improvement in inflammatory markers, and the continued progression of the bowel injury. Following high-energy pelvic trauma, delayed ischemic bowel injury should be considered even when the initial abdominal examination reveals no abnormalities.

Following high-energy pelvic trauma, even if an initial abdominal examination reveals no abnormalities, delayed ischemic bowel injury should be considered. In this case, the fracture site may have exerted traction or shearing forces on the mesenteric vessels, leading to bowel-wall ischemia, which in turn caused delayed necrosis and perforation (5). In addition, we considered that, following high-energy pelvic trauma, mesenteric compression by hematoma together with severe hemorrhagic shock may have aggravated intestinal hypoperfusion and microcirculatory dysfunction, leading to ischemic injury and ultimately contributing to delayed bowel perforation. No evidence of direct penetration or impingement was identified during the surgical procedure. Partial necrosis of the transected ends was compatible with progressive ischemia after mesenteric traction or shearing. Because no angiographic or targeted histopathological evidence demonstrated mesenteric vascular disruption, an initial direct mechanical component cannot be excluded.

Notably, the laboratory and imaging course also requires cautious interpretation. PCT, IL-6, and some organ-function indices partially declined after their early peaks; these changes were supportive rather than decisive and did not demonstrate source control. The June 18 CT was prompted by high fever, progressive abdominal distension, hemodynamic deterioration, and renewed laboratory abnormality, rather than by a single biomarker value. Similarly, FAST and bedside ultrasonography were useful for monitoring fluid and associated injuries but could not exclude bowel or mesenteric injury. The World Society of Emergency Surgery guidelines emphasize that a negative or equivocal initial CT does not exclude bowel or mesenteric injury in a high-risk patient. Close observation, serial clinical and laboratory assessment, and repeat CT are appropriate when suspicion persists; surgical exploration should be considered when the patient deteriorates or clinical findings indicate significant bowel injury (5). The Bowel Injury Prediction Score (BIPS) assigns one point each for abdominal tenderness, a white blood cell count of at least 17,000/mm³, and CT evidence of significant mesenteric injury. A score of 2 or more is associated with an increased risk of blunt bowel injury requiring operative management (10). In this case, the utility of BIPS was limited because abdominal tenderness could not be reliably assessed and early serial CT examinations showed no significant mesenteric injury. Although the patient's overall condition improved following resuscitation, with decreasing inflammatory markers and stabilization of hemodynamics, marked leukocytosis and the progressive loss of bowel sounds remained concerning, highlighting the need for continued observation and serial clinical reassessment.

This case supports a practical escalation principle: In a clinically non-evaluable patient, serial negative imaging indicates only that specific signs are absent at the time of examination. When a high-risk mechanism, evolving gastrointestinal findings, and the systemic trajectory remain inconsistent with those results, clinicians should reconsider the working diagnosis, obtain renewed surgical review, and change the diagnostic approach. Depending on physiological stability, options include diagnostic peritoneal sampling, diagnostic laparoscopy, or exploratory laparotomy rather than repeated observation or antimicrobial escalation alone.

Several limitations remain: this single retrospective case cannot establish an LC-III-specific mechanism; the timing and vascular basis of the transection were not directly demonstrated.

4 Conclusion

The clinical value of this case does not lie in complete sigmoid transection as an anatomical distinction alone. It shows that severe structural bowel injury may progress despite serial negative CT, FAST, bedside ultrasonography, and transient biochemical improvement in a clinically non-evaluable patient. Persistent discordance between a high-risk injury mechanism and the clinical course should prompt renewed surgical assessment and, when noninvasive evaluation remains inconclusive, consideration of invasive diagnosis or operative exploration. Repeated clinical reassessment and timely control of the source of infection can reduce the risk of further deterioration.

5 Patient perspective

The patient and her family expressed their gratitude for the care they received during the hospital stay. They particularly acknowledged the contribution of the multidisciplinary team, including the surgical, intensive care, and nursing teams, in managing her complex condition.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Dimitrios Kehagias, Leeds Teaching Hospitals, United Kingdom

Reviewed by: Fuminori Yamaji, Gifu University, Japan

Irfan Amir Kazi, University of Missouri, United States