Localization of infection site in femoral stabilization pin using SPECT/CT with radiolabeled leukocytes

Localization of infection site in femoral stabilization pin using SPECT/CT with radiolabeled leukocytes

By Partha Ghosh, MD, Siemens Healthineers, Hoffman Estates, Illinois, USA
Data and images courtesy of Queen Elizabeth Hospital Birmingham, Birmingham, United Kingdom


An approximately 70-year-old female with history of trauma with bilateral femoral fractures was treated with intramedullary rod insertions. The right femur was stabilized using an intramedullary rod with external fixation plate and screw, and the left femur had an intramedullary rod with a dynamic hip screw through the femoral neck and trochanter. The patient complained of severe pain in the right femoral fixation screws along with slight wound discharge. In view of the possibility of prosthetic infection, a 99mTc HMPAO-labeled leukocyte study was performed on Symbia Pro.specta[a] SPECT/CT. 

The patient was administered 8.91 mCi (330 MBq) of 99mTc HMPAO-labeled leukocytes and then the planar and SPECT/CT acquisition was performed after 3 hours. Initial CT scan performed with 110 kV and 150 reference mAs and 32 x 0.7 mm collimation. 1-mm slices were reconstructed with iterative metal artifact reduction (iMAR)[b] in order to ensure virtually artifact-free, thin-slice CT images. 

The SPECT study was acquired at 60 stops per detector with 20 seconds per stop. Data was reconstructed using OSEM3D with 5 iterations, 15 subsets, and a 128 x 128 matrix. Following CT attenuation correction (CTAC), the corrected SPECT data was fused with iMAR CT data for visual interpretation.

As seen in Figure 4, volume rendering of SPECT/CT data shows an accumulation of radiolabeled leukocytes along the entire length of the right lateral fixation plate as well as the femoral intramedullary nail with focal area of intense accumulation localized to the first fixation screw of the lateral plate (arrows). There is also a clear visualization of the leukocyte accumulation in the sinuses draining from the lower end of the lateral plate to the skin laterally as well as the lower third of the femoral intramedullary rod, which suggests a sinus tract in the lower third of the thigh as well.

As evident from the SPECT/CT images, the study shows increased uptake reflecting accumulation of radiolabeled leukocytes along the entire length of the lateral femoral plate. The uptake appears to be partly extruded from underlying bone with maximum focal uptake in the second fixation screw, which appears to be the site of maximum infective exudate. There is extrusion of the most distal fixation screws of the lateral plate with lucency surrounding the tip of the other remaining distal screws. This represents both a shear strain on the fixation plate due to malunion of the fractured femoral shaft as well as the osteolysis related to infection. There is also a soft-tissue collection at the lower end of the lateral plate with a sinus tract draining to the skin laterally. 

Furthermore, there is accumulation of radiolabeled leukocytes throughout the entire length of the femoral intramedullary nail representing infection throughout the entire prosthesis. The intramedullary nail also shows significant osteolysis surrounding its distal tip with fractures of the lowermost screw and the most proximal of the distal screws. The low-grade increased activity at the tip suggests active infection along with dependent collection of infective exudate. Linear increased accumulation of radiolabeled leukocytes along the proximal end of most of the distal screws represents focal infection at and around the screws and at the site of its attachment with the intramedullary plate, which extends along the subcutaneous sinus tract and appears to open onto the skin surface along the lateral right thigh. The left intramedullary nail in the femoral shaft as well as in the neck show complete union of the left subtrochanteric fracture without any evidence of active infection in the left femur or hip.

This case example demonstrates the clinical advantage of using SPECT/CT with radiolabeled leukocytes to exactly localize the site of abnormal accumulation of leukocytes and differentiate physiological leukocyte accumulation in the bone marrow from foci of infection. The CT component of the SPECT/CT is key for proper evaluation of the prosthesis, especially when evaluating the alignment, displacement, fracture or extrusion, periosteal reaction, bone sclerosis, bony cortical fractures, and the soft-tissue collections and sinus tracts associated with such infections. 

Symbia Pro.specta SPECT/CT enables a 32-slice acquisition with 0.7-mm collimation, which enables a reconstructed slice thickness of 1 mm with high image quality at a low dose. This enables high-quality coronal and sagittal reconstructions as well as volume rendering for optimal interpretation of the bony and prosthetic joint pathologies. 

In this case, the iMAR feature available on the CT of Symbia Pro.specta provided sharp, virtually artifact-free visualization of the intramedullary nail, lateral plate, and associated interlocking screws. This visualization was key in interpreting the displacement and extrusion, localizing the site, and determining the extent of abnormal accumulation of radiolabeled leukocytes, especially at the sites of the screw insertions. High-quality CT also enabled the localization of the sinus tract with retained radiolabeled leukocytes and the adjacent collections of infective exudate. 

Infection is the most serious complication after external or internal fixation with nails or plates for fracture or total- or partial-joint replacement with prosthesis. This occurs in 1.5-2.5% of primary procedures and up to 20% for revision surgeries.1 

Implanted devices and the presence of any foreign body is associated with a higher susceptibility for infection. Skin-derived bacteria are most often the ones to colonize an implant during surgery. Microorganisms form biofilms around the implant which protects the bacteria within the bio-film from phagocytes. The indolent nature of prosthetic joint infections may be due to the development of such protective biofilms on the surface of the infected implants, and very often, the infection cannot be controlled unless the implant is removed. The chemotactic factors secreted by the infective pathogens cause neutrophil recruitment at the site of infection, which is the basis of the effectiveness of infection imaging with radiolabeled leukocytes. In this case, the presence of radiolabeled leukocytes throughout the entire length of the lateral plate and intramedullary rod may reflect the presence of such biofilms harboring infective microorganisms. 

Although in this case there is minimal hematopoietic bone marrow within the femoral shaft due to the presence of the intramedullary rod, it is to be noted that radiolabeled leukocytes do accumulate in normal marrow and need to be differentiated from abnormal accumulations related to infection based on the location of such accumulation. Thus, the CT component of SPECT/CT is key for such localization. 

The CT performance of Symbia Pro.specta with thin-slice reconstruction with iMAR for metal-artifact-free visualization of the intramedullary plates and screws was instrumental in the correct assessment of the degree of extrusion as well as periprosthetic lucency associated with infection and sinus tracts, which was then accurately coregistered to the abnormal focal leukocyte accumulations.

The clinical evaluation of CT, SPECT, and SPECT/CT images suggests infection of the right lateral femoral plate throughout the length of the intramedullary femoral nail, especially at the proximal and distal ends, along with an infected sinus tract extending from the lower end of the femoral plate and intramedullary nail to the skin surface in the right distal thigh. 

The high-quality, thin-slice CT with iMAR made possible by the CT performance of Symbia Pro.specta was instrumental in the proper assessment of the degree of displacement, extrusion, and fracture of the lateral femoral plate, the interlocking screws and femoral intramedullary nail, as well as the localization of the focal accumulation of radiolabeled leukocytes, infective exudates, collections, and related sinus tracts. 

From the clinical impression of extensive infection within the femoral implants, along with the malunion of the fracture site and osteolysis associated with the periprosthetic bone, a total removal of the prosthesis with re-insertion of new stabilization rods with antibiotic beads would be a recommended approach to management.

Scanner: Symbia Pro.specta




Injected dose

8.91 mCi (330 MBq) 99mTc HMPAO-labeled leukocytes

Tube voltage

110 kV

Post-injection delay

3 hours

Tube current

150 ref mAs


60 stops per detector, 20 seconds per stop

Slice collimation

32 x 0.7 mm

Image reconstruction

128 x 128 matrix, OSEM3D 5i15s

Slice thickness

1 mm