Delayed sacroiliac joint infection after robot-assisted percutaneous iliac screw fixation for tile type C2 pelvic fractures: A case report and literature review

Abstract

Background

Percutaneous iliosacral screw fixation has become the gold standard for surgical stabilization of acute pelvic ring disruptions with unstable posterior pelvic injuries. Wound infection is uncommon owing to shorter operative time, reduced soft-tissue damage, less blood loss, and lower infection rates. Delayed sacroiliac joint infection after robot-assisted percutaneous iliosacral screw fixation is even rarer.

Case presentation: We report a case of delayed sacroiliac joint infection after robot-assisted percutaneous iliosacral screw fixation for a tile type C2 pelvic fracture. A 43-years-old man developed recurrent low back pain and intermittent low-grade fever that occurred 10 months postoperatively and lasted 2 months. Symptoms temporarily improved with anti-inflammatory treatment but subsequently worsened, accompanied with chills and high fever (body temperature up to 39.0°C). Computed tomography revealed bone erosion and screw loosening, and inflammatory markers were elevated. The patient underwent open debridement and removal of the screws, followed by implantation of vancomycin- and gentamicin-loaded calcium sulfate beads. Cultures were negative; however, frozen section and pathology confirmed acute and chronic inflammatory infiltrations. The patient responded well to antibiotics and achieved full recovery.

Conclusions

Delayed sacroiliac joint infection after robot-assisted percutaneous iliosacral screw fixation is a rare complication. Diagnosis can be challenging owing to the presence of atypical symptoms and negative cultures. Open debridement with removal of fixation and targeted antibiotic therapy can result in successful outcomes.

Background

Percutaneous iliosacral screw fixation has become the gold standard for surgical stabilization of acute pelvic ring disruptions with unstable posterior pelvic injuries.^1–3 This technique offers several advantages, including a shorter operative time, reduced soft-tissue damage, less blood loss, and lower infection rates.^4–6 The combination of anterior subcutaneous internal fixation (INFIX) and percutaneous iliosacral screws is a minimally invasive, biomechanically stable, and effective method for managing vertically and rotationally unstable pelvic ring injuries. It achieves satisfactory fracture reduction with few complications and good functional and radiological outcomes. ⁷ Reported infection rates following percutaneous iliosacral screw fixation are low, ranging from 0% to 3.8%, and most cases involve only superficial infection.^8–10 Delayed deep sacroiliac joint infection after robot-assisted percutaneous iliosacral screw fixation is extremely rare, especially in tile C-type pelvic ring injuries. To the best of our knowledge, no such case has previously been reported in the English or Chinese literature. We describe an unusual case of delayed deep sacroiliac joint infection occurring more than 12 months postoperatively.

The reporting of this study conforms to the Case Report (CARE) guidelines. ¹¹

Case presentation

A man in his early 40s was admitted to Guangdong Provincial Hospital of Chinese Medicine, Zhuhai, China, in March 2022 with a closed pelvic fracture (Tile 61-C2.2) (Figure 1(a) to (c)), accompanied with a comminuted fracture of the left tibiofibular shaft, bilateral multiple rib fractures, a sternum fracture, multiple fractures of the right foot, and transverse fractures of the first to fourth lumbar vertebrae following a motor vehicle accident. There were no skin lesions or bleeding around the hips or back at admission. He denied any history of trauma or previous surgery.

Figure 1.

Preoperative axial pelvis CT scan reveals a right trans-sacral fracture-dislocation, combined with ipsilateral fractures of the superior and inferior pubic rami and a contralateral fracture of the superior pubic ramus. The injury is classified as Tile 61-C2.2 according to the Tile classification. (a) Axial pelvis CT shows the right trans-sacral fracture-dislocation (dark blue arrow). (b) Axial pelvis CT shows fractures of bilateral superior pubic rami, with displacement on the right side and no displacement on the left side. (c) Axial pelvis CT shows fracture of the right inferior pubic ramus. (d) Axial pelvis CT shows fusion of the left sacroiliac joint (orange arrow). CT: computed tomography.

Five days after admission, the patient underwent TiRobot-assisted percutaneous iliosacral screw fixation. Under general anesthesia with endotracheal intubation, he was placed in the supine position on a transparent operating table with a bolster under the pelvis. Closed reduction of the pelvic fracture was achieved under an external fixation frame (The Pelvic Minimally Invasive Reduction System, Shijiazhuang High-tech Zone Yicheng Technology Co., Ltd. Heibei China).

The procedure was truly percutaneous, with two separate stab incisions of approximately 1.5 cm each, made over the posterior superior iliac spine. Using TiRobot navigation (TINAVI Orthopedic Surgical Robot System, Beijing TINAVI Medical Technologies Co., Ltd), S1 and S2 screws were precisely inserted. The screws used were cannulated, partially threaded, titanium iliosacral screws (manufacturer: Double Medical, China): the S1 screw measured 7.3 × 90 mm and the S2 screw measured 7.3 × 135 mm. No bone grafting or cement augmentation was performed during the initial surgery. Accuracy was confirmed using three-dimensional (3D) C-arm fluoroscopy (CIOS Spin Arcadis Orbic 3 D, Siemens, Germany). Subsequently, INFIX was performed using an internal fixator.

Postoperative day 1 pelvis radiograph (Figure 2(a) to (c)) and computed tomography (CT) (Figure 3(a) to (d)) image indicated excellent reduction of pelvis fracture after fixation with S1/S2 sacroiliac screw in the posterior column and an internal fixator in the anterior column. The incisions healed well without infection. The reduction was excellent according to the Tornetta and Matta criteria.

Figure 2.

(a) AP, (b) outlet, (c) inlet radiography obtained 1 day postoperatively show excellent reduction of the Tile 61-C2.2 pelvis fracture after fixation with S1/S2 sacroiliac screws in the posterior column and anterior subcutaneous internal fixation (INFIX) in the anterior column. AP: anteroposterior.

Figure 3.

(a and b) Axial and (c and d) pelvic CT scans on postoperative day 1 show well-positioned S1 and S2 sacroiliac screws. CT: computed tomography.

One week later, the patient underwent minimally invasive plate osteosynthesis (MIPO) for the left tibia and fibulà fractures and open reduction and internal fixation (ORIF) for the bilateral foot fractures under general anesthesia. These surgical wounds also healed normally.

Three months later, the anterior subcutaneous internal fixator was removed, and the patient gradually resumed walking and weight-bearing. The patient resumed routine work and full weight-bearing without any discomfort at 10 months postoperatively. Pelvic CT performed 10 months postoperatively showed no evidence of infection (Figure 4(a) to (c)).

Figure 4.

Axial pelvic CT scan shows well-positioned (a) S1 sacroiliac screw and (b and c) S2 sacroiliac screw without any sign of erosion in and around the sacroiliac joint at 10 months postoperatively. CT: computed tomography.

Beginning at 10 months postoperatively, the patient presented with recurrent diffuse pain in the lower back and right gluteal region, accompanied with intermittent low-grade fever lasting 2 months. Symptoms were partially relieved with oral nonsteroidal anti-inflammatory drugs; however, the symptoms worsened subsequently, accompanied with severe pain and a high fever (body temperature up to 39.0°C). He was readmitted to our hospital for further evaluation.

Investigations

Pelvis CT imaging at 12 months postoperatively revealed partial bone resorption at the right sacroiliac joint and loosening of the S1 iliac screw, with pronounced peri-screw osteolysis (Figure 5(a) to (d)). Laboratory examinations revealed the following results: erythrocyte sedimentation rate (ESR), 64 mm/h; C-reactive protein (CRP) level, 101.46 mg/L; white blood cell count, 10.44 × 10⁹/L (76.7% neutrophils, 13.4% lymphocytes); hemoglobin level, 115 g/L; alanine aminotransferase level, 93 U/L; aspartate aminotransferase level, 71 U/L; alkaline phosphatase level, 253.2 U/L; gamma-glutamyl transferase level, 143.8 U/L; and uric acid level, 527.6 μmol/L. Magnetic resonance imaging (MRI) revealed peri-S1 iliac screw fluid collections, bone marrow edema extending beyond the screw tract, and surrounding soft-tissue edema and inflammation, visible as T2 hyperintensity in the iliopsoas muscle (Figure 6(a) to (d)). Urinalysis, chest radiography, electrocardiography (ECG), and echocardiography yielded normal results.

Figure 5.

Pelvic CT scan performed 12 months postoperatively shows evidence of S1 sacroiliac screw loosening and partial erosion in as well as around S1 sacroiliac screw. (a) Axial pelvic CT scan shows significant edema of the right iliacus muscle (yellow arrow) compared with the contralateral side. (b) Coronal pelvic CT scan shows significant bone erosion around the S1 sacroiliac screw (red arrow). (c) Axial pelvic CT scan shows significant ilium erosion (light green arrow) posterior to the sacroiliac joint. CT: computed tomography.

Figure 6.

(a) Axial and (b) coronal pelvic MRIs performed 12 months postoperatively show high signal intensity of the liopsoas soft-tissue (purple arrow) and the sacroiliac joint (purple arrow) on spin echo T2-weighted images at the level of the S1 screw. (c and d) Axial pelvic MRI performed 12 months postoperatively shows similar high signal intensity of the liopsoas soft-tissue (green arrow) and the sacroiliac joint (green arrow) on spin echo T2-weighted images at the level of the S2 screw. MRI: magnetic resonance imaging.

Treatment

Postoperative infection was diagnosed in the right sacroiliac joint. Five days after admission, open debridement was performed under general anesthesia. An 8-cm longitudinal incision was made 2 cm medial to the posterior superior iliac spine for access to the posterior sacroiliac joint, and a 5-cm incision was made at the original gluteal site for drainage and removal of the S1 and S2 screws. Necrotic and inflammatory tissues were debrided, and vancomycin- and gentamicin-impregnated calcium sulfate granules, along with autologous bone graft harvested from the right iliac crest, were implanted into the sacroiliac joint space.

Intraoperative inspection confirmed extensive necrotic and inflammatory tissues (Figure 7(a)). Cultures for bacteria, anaerobes, and fungi were negative. Frozen section analysis was performed, showing ≥5 neutrophils per high-power field in at least 5 separate microscopic fields, based on the criteria described by Mirra et al. ¹² and Athanasou et al. ¹³ This indicated the presence of an infection. Pathology confirmed acute and chronic inflammatory cell infiltrations (Figure 8(a)).

Figure 7.

A significant amount of inflammatory tissue located posterior to the right sacroiliac joint found during the operation.

Figure 8.

(Hematoxylin eosin staining, magnification: 200×) Microscopic examination of bone, bone marrow, and fibrous tissue from the right sacroiliac joint shows degeneration and necrosis with hyperplasia of fibrous granulation tissue and infiltration by abundant lymphocytes (▴), plasma cells (^), and neutrophils (↑).

Antibiotic therapy

Preoperatively, ceftriaxone (2 g daily) was initiated empirically for targeting common Gram-negative pathogens. After intraoperative findings and negative cultures were obtained, antibiotics were switched to vancomycin (1 g daily) for 10 days (for Methicillin-resistant Staphylococcus aureus (MRSA) coverage) and then oral levofloxacin (0.5 g daily) for 4 weeks (for atypical organisms and biofilm penetration). This regimen was based on local institutional protocol for suspected implant-related infection and is consistent with Infectious Diseases Society of America (IDSA) guidelines for culture-negative prosthetic joint infections.

Outcome and follow-up

Postoperatively, the patient demonstrated clinical improvement. He recovered without complications. The 2-years follow-up showed excellent outcomes (Figure 9(a) to (f)), with a Majeed pelvic fracture functional score of “excellent,” a visual analog scale (VAS) score of 1, and resumption of sports activities and routine work. ¹⁴ There were no signs of surgical incision infection at 2 years postoperatively (Figure 9(e) and (f)).

Figure 9.

(a to d) Excellent function of the right sacroiliac joint and hip joint was observed 2 years after postoperative of debridement and removal of S1/S2 screws. (e and f) No signs of surgical site infection at 2 years postoperatively following debridement and S1/S2 screws removal.

Discussion

Percutaneous iliosacral screw fixation is an effective treatment modality for tile C-type pelvic fractures, as supported by previous studies.^15–17 Smaller studies have also reported low infection rates. Cheng et al. have reported an incidence of 1.6% in the iliosacral screw group, ¹⁸ while Stover et al. observed surgical site infection in eight of 236 patients (3.4%) in a multicenter analysis. ¹⁹ Boudissa et al. have reported no wound infections (0%) among early complications. ²⁰ All of these studies involved only early superficial infection. With the use of navigation techniques for guidewire placement and intraoperative, 3D imaging, the rate of wound infection can be further reduced with a shorter operative times, less soft-tissue damage, and lower blood loss. Deep sacroiliac joint infections occurring >10 months postoperatively after robot-assisted percutaneous iliosacral screw fixation are therefore extremely rare. Unlike early infections that typically occur within 3 months, in this case, infection was observed after a prolonged period of 10 months and differed from common superficial infections, making the diagnosis challenging.

Robot-assisted navigation allows real-time, 3D visualization and precise guidewire placement, potentially reducing the number of guidewire passes, operative time, and soft-tissue disruption. These factors may theoretically lower the risk of surgical site infections. However, current literature does not demonstrate a statistically significant difference in the infection rates between robot-assisted and conventional fluoroscopy-guided techniques. Verbeek et al. ⁶ have reported similar infection rates between the two methods, suggesting that robotic assistance does not increase infection risk but improves screw accuracy. In our case, we believe the infection was related to host factors and biomechanical stress rather than the surgical technique itself.

Several factors may explain the reason for this rare complication in this 43-years-old man. First, pelvic fractures are high-energy injuries, and severe tile C-type fractures are often accompanied with marked soft-tissue damage and disruption of blood supply. This compromises local resistance and significantly increases the infection risk. Second, this patient had congenital fusion of the contralateral (left) sacroiliac joint, which completely eliminated its mobility. This placed increased biomechanical stress on the injured right side where stability of the posterior pelvic ring depended entirely on the two iliosacral screws. Excessive friction between the cannulated screws and the cortical bone may have caused persistent micromotion and a foreign-body reaction. This in turn could have weakened local resistance, promoted low-grade biofilm formation, and ultimately led to delayed infection. Third, although hematogenous spread is the most common cause of delayed infection following fracture fixation, this seemed less likely in this case. The patient denied any history of postoperative infections such as periodontitis, skin infection, pneumonia, and urinary tract infection. Intraoperative cultures of necrotic tissue were negative for bacteria, fungi, and anaerobes. Therefore, although the possibility of hematogenous spread cannot be entirely excluded, it is considered improbable.

Establishment of a clinical diagnosis in this patient was challenging for several reasons, which contributed to some delay. First, the clinical symptoms were atypical. The main signs were recurrent low-grade fever, which improved temporarily with oral anti-inflammatory drugs, and diffuse pain in the right lower back. These features could easily be mistaken for postoperative recovery, lumbosacral strain, or irritation due to INFIX, leading to misdiagnosis or delayed recognition. It was particularly difficult to consider delayed sacroiliac joint infection after robot-assisted percutaneous iliosacral screw fixation. Furthermore, laboratory findings were nonspecific. The white blood cell count was normal, while the erythrocyte sedimentation rate was elevated (64 mm/h), and the CRP level was increased (101.46 mg/L). A white blood cell count of 10.44 × 10⁹/L with 76.7% neutrophils showed only slight elevation, which did not help distinguish this infection from others. Early imaging, including plain pelvic radiography and CT, were also normal. Osteolysis around the screw was detected only at 10 months postoperatively, and MRI showed high signal intensity on T2-weighted images around the two right iliac screws, with edema and inflammation of the iliopsoas muscle. These findings could not reliably differentiate between sterile loosening and infectious loosening.

We considered several alternative explanations for the clinical presentation. Mechanical loosening was considered but deemed less likely because although peri-screw osteolysis was present, the presence of systemic symptoms such as fever and chills, along with elevated inflammatory markers (CRP level, 101.46 mg/L and ESR, 64 mm/h), made isolated mechanical loosening improbable. Intraoperatively, purulent and necrotic tissue was observed, confirming infection. Inflammatory foreign-body reaction was also considered; however, histopathology showed acute and chronic inflammatory infiltrates including neutrophils, lymphocytes, and plasma cells. However, no giant cells, granulomas, or polarizable foreign material were observed, ruling out a pure foreign-body reaction. Pseudoarthrosis was excluded as radiological union was confirmed at 10 months postoperatively (Figure 4(a) to (c)), and intraoperatively, the fracture site was healed.

Blood cultures during febrile episodes were negative. Intraoperative cultures of inflammatory and necrotic tissue from the right sacroiliac joint were also negative for bacteria, anaerobes, and fungi. To ensure detection of slow-growing organisms such as Cutibacterium acnes, anaerobic cultures were maintained for a minimum of 14 days. Despite these measures, no growth was observed. Thus, a definitive diagnosis could not be established based on cultures alone. Culture-negative orthopedic infections are not uncommon, with reported rates ranging from 7% to 43% in periprosthetic joint infections. ²¹ Possible explanations include prior antibiotic use, biofilm formation reducing bacterial metabolic activity, inadequate sampling technique or delayed transport, fastidious or slow-growing organisms such as Cutibacterium acnes and Corynebacterium species, and low bacterial load. In such cases, a combination of clinical presentation, laboratory markers, imaging, histopathology, and response to treatment becomes essential for diagnosis. Intraoperative frozen section analysis in our case showed ≥5 neutrophils per high-power field in at least 5 separate microscopic fields based on the criteria described by Mirra et al. ¹² and Athanasou et al. ¹³ This supported the diagnosis of infection, and pathology confirmed acute and chronic inflammatory cell infiltrations.

Notably, the patient had received empirical antibiotic therapy (intravenous ceftriaxone 2 g daily) for 5 days prior to surgical debridement for persistent high fever and chills. This preoperative antibiotic exposure likely contributed to the negative culture results, despite appropriate sampling and extended anaerobic incubation. Therefore, although a specific pathogen could not be isolated, the diagnosis of culture-negative deep sacroiliac joint infection was established based on the combination of clinical presentation (delayed-onset fever and pain), radiographic findings (progressive peri-screw osteolysis and bone marrow edema on MRI), elevated inflammatory markers (CRP level, 101.46 mg/L and ESR, 64 mm/h), and compelling histopathological evidence of acute inflammation on both frozen section and permanent pathology.

The management of the delayed sacroiliac joint infection involved a single-stage procedure consisting of debridement, INFIX removal, and implantation of calcium sulfate antibiotic beads. This was followed by intravenous ceftriaxone (2 g daily) for 4 days, intravenous vancomycin (1 g daily) for 10 days, and finally oral levofloxacin (0.5 g daily) for 4 days. Intraoperative debridement of the infected bone tract and sacroiliac joint, along with complete removal of INFIX, were critical for radiological union of the fracture at 10 months postoperatively. The patient recovered well after surgery, with no pain or fever in the right hip during a 2-years follow-up period.

The limitations of this case should be acknowledged. First, no routine, preoperative single photon emission computed tomography (SPECT) imaging was performed. Second, no tuberculosis-related examinations were conducted. Third, the lesion tissue was not subjected to high-throughput genetic testing to identify the causative pathogen and guide treatment.

Conclusion

Robot-assisted sacroiliac screw fixation is an effective, minimally invasive technique for unstable pelvic fractures. However, surgeons should be aware of the potential risk of delayed sacroiliac joint infection, although this complication is extremely rare. In patients presenting with unexplained low-grade fever and sacroiliac pain, systemic anti-infective treatment alone may not be sufficient. Delayed infection after percutaneous INFIX should not be excluded as the causative pathogen may arise from a blood-borne source elsewhere in the body. Early targeted treatment can be challenging in such cases. Open surgical intervention using a posterior sacroiliac joint approach for complete debridement and drainage remains curative for sacroiliac joint infection.