Inflammation and Infection Scintigraphy

Gallium-67

Produced in cyclotron, electron capture decay with 93, 184 and 296 keV; HL 78.1h. Most is transferrin bound in plasma. Increased permeability and blood flow increases accumulation at inflammatory foci, also binds to lactoferrin at these foci, may be taken up directly by bacteria (siderophores). Accumulates despite absence of circulating leukocytes. Imaging at 18-72hrs after injection. Normal biodistribution in bone, marrow, liver, GU, GI, soft tissues, nasopharyngeal, lacrimal glands. Uptake in brests with hyperprolactinemic state (pregnancy, lactation, drugs). Increased uptake in kidneys, bladder, marrow, liver, colon after multiple transfusions (?iron receptor saturation) or gadolinium.

Indications:

  • Opportunistic infection – High negative predictive value in chest (esp if CXR normal). LN uptake in HIV usually mycobacteria or lymphoma. Focal/localised lung uptake in bacterial pneumonia; diffuse in PCP. Is not taken up by Kaposi sarcoma.
  • Interstitial lung disease – Uptake in any pulmonary inflammation including sarcoid, interstitial pneumonitis, pneumoconioses; generally correlates with severity. Activity compared to sternal or hepatic activity. Grade 0 similar to background; 1 increased; 2 less than liver; 3 equal to liver; 4 greater than liver. Higher grade may have photopenic cardiac area.
  • Interstitial nephritis – Takes up Ga more than spine, cf acute tubular necrosis has little or no renal uptake.
  • Fever/pyrexia of unknown origin (FUO/PUO) – >/=3/52 episodes of fever >38.3 deg without known origin. Infection in only 20-30%, neoplasm 15-25%, others collagen vascular disease, vasculitis, granulomatous disease, PE, CVA, drugs. Uptake of Ga-67 in infection, inflammation or tumour.
  • Spinal osteomyelitis – Used in conjuction with bone scan. Positive when Ga-67 uptake > bone scan or uptakes are spacially incongruent; equivocal when both are congruent; negative when G1-67 uptake < bone scan or is normal.

Radiolabeled Leukocytes

40-60mL blood taken, mixed with anticoagulant and leukocytes separated via sedimentation and centrifugation. Leukocytes labeled with indium-111 oxyquinoline or 99mTc-HMPAO (exametazine hyxamethyl propyleneamine oxime).

Techniques:

  • In111-WBC – Electron capture decay with 173 and 247keV, HL 67hrs. Immediate intense pulmonary activity which clears, abnormal if present at 24hrs. Imaged at 24hrs, uptake in liver, spleen, marrow.
  • Tc99m-WBC – More variable normal biodistribution; in lungs, liver, spleen, marrow, GU (shortly after injection), colon (within 4hrs), blood pool, occ GB. Generally imaged within a few hours, hence not as good for indolent infections that take longer to acumulate WBC.

WBC accumulation dependent on intact chemotaxis, type and number of WBC labelled (most are neutrophils), inflammatory response. Requires at least 2,000 WBC per microlitre. Best for bacterial infections (neutrophils); less useful for opportunistic infections, TB, sarcoid. Segmental/lobar lung uptake in bacterial pneumonia, CF/bronchiectasis (WBC in pooled secretions). Diffuse lung uptake in opportunistic infection, radiation pneumonitis, drug toxicity, ARDS, sepsis (neutrophils activated by cytokines unable to maneuvre through pulmonary capillaries). Unable to distinguish extent of bowel uptake, as peristalsis spreads activity over time.

Indications:

  • PUO – In-WBC more sensitive than Ga-67 in early disease, Ga-67 more sensitive in late disease.
  • Post-op infection – Differentiates infection from simple fluid collections, tumour, normal post-op changes. WBC don’t accumulate in healing surgical wounds, hence uptake indicates infection, except granulation in secondary intention (eg tracheostomy, ileostomy, gastrostomy, skin grafts). WBC rarely accumulate in non-infected tumours (cf Ga-67).
  • Myocardial abscess or prosthetic vascular graft infection, more sensitive than echo. False positives in perigraft haematoma, bleeding, thrombosis, pseudoaneurysm, graft endotheliasation.
  • CNS – Uptake suggests brain infection, may have faint uptake in tumours.
  • Osteomyelitis – Best imaged with bone scan, if there is no underlying disease. In-WBC scan improves sensitivity of complicating osteomyelitis (underling fracture/arthropathy etc). May be performed with Tc-99m sulfur colloid scan (taken up by haematopoietic marrow); hence areas of WBC uptake without sulfur-colloid is positive for infection (other patterns are negative/indeterminate). Not sensitive in spinal osteomyelitis as there may be decreased/absent activity in >50% (DDx tumour, infarction, fracture, Paget); hence better imaged with Ga-67.
  • Inflammatory bowel disease – Tc-WBC sensitive to colitis, may show skip areas, diffentiates active inflammation from scarring. Physiologic activity (?hepatobiliary excretion) appears after 3hrs in distal SB, is diffuse and mild, migrating to caecum by 4hrs. Sensitivity reduced by steroids.

Technetium-99m-Fanolesomab

Immunoglobulin, binds to CD15 receptors on leukocytes; in neutrophils that migrate to foci of infection as well as neutrophils and neutrophil debris already at infectious foci. No increased retention in lungs. Normal biodistribution in liver, spleen, marrow, GU, blood pool (reduces over time), large bowel (after 4hrs), occasionally small bowel. There is transient reduction in circulating WBC after injection, within 20min and recovery within 45min. High negative predictive value for appendicitis at 90min scan. Positive scans show activity in the ‘appendicitis zone’ (from pubic symphysis to lower pole right kidney) which persists over the 90min scan.

FDG-PET

Increased glocuse transporters in inflammation. Normal biodistribution in brain, myocardium, GU, thymus; variable in marrow, stomach, bowel; low grade liver, spleen (may be intense in infection). Imaging at ~1hr post injection. In infection/inflammation uptake is similar to Ga-67; being very sensitive, but nonspecific. Uptake in vasculitis, bacterial endocarditis, thromboembolic disease, sarcoid, chronic granulomatous disease, spinal osteomyelitis etc. Negative scan makes it very unlikely that a morphological origin of PUO will be identified.