White Cell Scan

The white cell scan (labelled leucocyte scintigraphy) finds sites of infection by following the patient's own white cells. A sample of blood is taken, the white cells are separated and labelled with a radioactive tracer, and they are returned to the patient.

Neutrophils migrate to sites of bacterial infection, so a focus of accumulating activity points to infection.

Two labels are used. Technetium-99m HMPAO (200 to 400 MBq) gives better images and a lower dose, with imaging at 30 minutes to 4 hours and again at 20 to 24 hours.

Indium-111 oxine (10 to 20 MBq) is stable and is imaged at 24 hours, with more tolerance for late images but a higher dose to the spleen.

About 40 to 50 mL of blood is drawn, and labelling takes 2 to 3 hours under sterile conditions. The effective dose is about 3 to 5 mSv for technetium and 10 to 15 mSv for indium.

Infection is a common and treatable cause of unexplained fever and bone pain, and osteomyelitis complicates up to 15% of diabetic foot ulcers.

Labelled leucocytes have a sensitivity and specificity of about 85 to 95% for peripheral bone infection, prosthesis infection and abdominal abscess, and they are more specific for infection than gallium or fluorodeoxyglucose.

Indications.

• Suspected osteomyelitis of the peripheral skeleton, including diabetic foot disease and post-traumatic infection.
• Infection of a joint prosthesis (combined with a marrow scan) or other orthopaedic hardware.
• Fever of unknown origin, particularly when an abdominal or pelvic source is suspected.
• Intra-abdominal abscess and post-operative sepsis.
• Activity and extent of inflammatory bowel disease (technetium label).
• Vascular graft infection and infected pacemaker or device leads.
• Infective endocarditis and cardiac device infection in selected cases.
• Infection in transplant recipients and in patients with neutrophilic inflammation.
• Sepsis in cystic fibrosis or chronic lung disease in selected cases.

Contraindications and precautions.

• Pregnancy: avoid unless essential. Breast feeding: a short interruption for technetium, and a long one for indium, as the local rule requires.
• Severe neutropenia (white count under about 2 x 10^9/L) gives poor labelling and a poor result.
• Reinfusion error is the main safety risk. The labelled blood is a blood product, and identity is checked at every step by two staff.
• Sterile technique is needed for the whole preparation, and the cells are injected without delay.
• Antibiotics may lower sensitivity, and corticosteroids, chemotherapy and recent surgery change cell function. The treatment history is recorded.
• High blood glucose does not matter for this test, unlike fluorodeoxyglucose imaging.
• Blood borne infection risk to staff is managed by universal precautions.
• A venous line is needed, and the patient must be able to return for late images.

Preparation and technique.

No fasting is needed. Blood is drawn into a syringe with an anticoagulant. Red cells are sedimented, the white cells are separated by centrifugation, washed and labelled, and the labelled cells are returned in 2 to 3 hours.

Technetium images are taken at 30 minutes to 1 hour, at 3 to 4 hours and at 20 to 24 hours, with the activity compared across time. Indium images are taken at 24 hours (and sometimes at 4 hours).

Whole body and spot views are taken, and SPECT-CT of the area of interest improves localisation.

For a prosthesis, a marrow scan with a colloid is done as well, so that normal marrow (activity in both scans) can be separated from infection (activity only in the white cell scan). Images are read with comparison to the earlier time points.

Normal appearances.

Labelled cells are seen in the spleen, liver and bone marrow, with the spleen being the most intense. Early images show lung activity, which clears by 4 hours.

With technetium, activity appears in the kidney, bladder and bowel at 3 to 4 hours as the label is excreted, and the gallbladder may be seen.

Normal marrow is seen in the axial skeleton and proximal long bones, with symmetrical distribution. The area of normal anatomy shows no focal uptake that increases with time.

Interpretation.

• Infection: a focus of uptake that increases in intensity with time (for technetium, compared at early and late images) and is more intense than the contralateral side or the background.
• Osteomyelitis: focal uptake in bone or in adjacent soft tissue. SPECT-CT separates bone from soft tissue, which is important in the diabetic foot.
• Prosthetic joint infection: uptake in the white cell scan without a matching increase in the marrow scan (a mismatch) indicates infection.
• Abdominal abscess: focal uptake that does not move on serial views. Bowel activity from the label itself in the technetium scan appears after 3 hours, so early images are used for bowel disease.
• Inflammatory bowel disease: segmental uptake that corresponds to active disease, with extent and activity assessed in the first hour.
• Vascular graft infection: uptake along the graft or around it.
• Endocarditis and device infection: uptake in the valve or lead with negative echocardiography.
• Photopenic area: a cold area in bone marrow is seen in spinal osteomyelitis, chronic infection or sterile inflammation, and the test is not reliable in the spine.
• Sterile inflammation and healing fractures may show mild uptake that does not increase with time.

Critical values and findings needing immediate action.

• Focal uptake in the abdomen or chest in a septic patient: urgent drainage or surgery, and antibiotics.
• Prosthetic joint infection: orthopaedic review for debridement or revision.
• Spinal infection with neurological signs (needs MRI): urgent spinal surgery review.
• Graft or device infection: vascular or cardiac surgical review.
• Labelling or reinfusion error: stop, report, and follow incident procedures.

Limitations and artefacts.

• Chronic infection with few neutrophils (tuberculosis, fungal disease, low grade osteomyelitis) can give false negative results.
• The test is not reliable in vertebral osteomyelitis, which often shows a cold defect. MRI is preferred.
• Bowel activity in later technetium images can mimic abdominal infection, and early images are used.
• Swallowed or inhaled labelled cells appear in the airway and stomach.
• Poor labelling gives high blood pool and liver activity. Free technetium appears in the kidney and bladder.
• Sterile inflammation, haematoma and tumours can accumulate cells, which causes false positive results.
• Antibiotic treatment, steroids and neutropenia lower sensitivity.
• Long preparation and the need for late images make it unsuitable in emergencies.
• Handling blood requires trained staff and a clean laboratory.
• Fluorodeoxyglucose PET-CT is faster and often replaces the test, but it has lower specificity in bone and joints.

Special groups.

• Children: technetium label with weight based dose, and a small blood volume.
• Pregnant patients: avoid.
• Patients with diabetes: foot imaging with SPECT-CT to separate soft tissue from bone infection.
• Patients with joint prostheses: combined white cell and marrow scans.
• Patients with neutropenia: test not suitable. Use fluorodeoxyglucose PET-CT or other imaging.
• Regions: tuberculosis, HIV-related infection and chronic osteomyelitis are common. The scan is available in very few centres, so ultrasound, radiographs, MRI, CT, aspiration and culture are used.

Follow up tests.

• MRI of the spine or the area of concern.
• CT or ultrasound with image directed aspiration for culture.
• Fluorodeoxyglucose PET-CT.
• Marrow scan with colloid for prosthesis.
• Blood cultures, C-reactive protein, erythrocyte sedimentation rate and full blood count.
• Bone biopsy with culture and histology.
• Echocardiography for endocarditis.
• Referral to infectious disease, orthopaedics, vascular surgery or gastroenterology.