Pulmonary Scintigraphy

Ventilation-Perfusion (V/Q) Scan

When upright the ventilation/perfusion gradient of lungs increases inferiorly, with blood flow at the apex 1/3 of base. Greater basal intrapleural pressure, hence apical alveoli remain more open (greater oxygen tension) and basal alveoli greater respiratory cycle (hence 1.5-2x greater ventilation/gas exchange). Hence normal capillary perfusion and alveolar ventilation are matched to maximize gas exchange. When supine, ventilation/perfusion gradient shifts towards anterior-posterior. Autoregulation diverts blood away from hypoxaemic/diseased segments, but hypoperfusion rarely induces localised bronchoconstriction (V/Q mismatch); hence primary vascular disorders (PE) usually have normal ventilation unless associated with consolidation or infarction.

Ventilation radiopharmaceuticals require prevention of leakage into imaging room. Gas delivery systems have a shielded spirometer, oxygen delivery system, exhalational trap and good room ventilation.

  • Xenon-133 (Xe-133) – Fission product of Ur-125. HL 5.3d, beta and 81keV, 370-540Mbq. Should be performed before Tc99m perfusion due to Compton scattering interference with Xe-133 photopeak. First-breath image on breath-hold taken, then euqilibrium image after 5min of rebreathing. Serial 30sec washout images over 5min (HL <1min, complete 3min).
  • Xenon-127 – Cyclotron produced. HL 36.4d, 203/172/365keV, 296-555MBq. Performed as per Xe-133.
  • Krypton-81m (Kr-81m) – From rubidium-81-Kr-81m generator. HL 13s, 191keV, 370-740MBq. Ventilation scans taken immediately after perfusion on all six positions.
  • Tc-99m-diethylenetriaminepentaacetic acid (DTPA) aerosol – Nubuliser produced mist passed through a settling bag (traps larger particles), inhaled via non-rebreathing valve. Only is 2-10% deposited within lungs. Larger particles are heavier thus deposit centrally, >2 microns in trachea and pharynx. Microaerosols <0.5microns reach distal bronchila tree, but deposit more centrally if airways are narrowed (asthma, bronchitis, COPD) due to turbulance. Tc-99m-DTPA is absorbed across alveolar membrane with HL 60-90min (20min shorter in smokers due to increased alveolar permeability). Inhaled in supine position to avoid gradient.
  • Technegas – Created by heating Tc-99m-pertechnetate with graphite heating element (2550oC, just above Tc melting point), hence coating Tc with graphite vapour. Hydrophobic, hence repelled from aqueous surface and trapped by surfactant in alveoli with almost all activity reaching beyond 16th bronchial division.

Perfusion radiopharmaceutical particles are slightly larger than pulmonary capillaries (>8microns), lodging in precapillary arterioles. Tc-99m macroaggregated albumin (MAA) are irregularly shaped 20-40microns. >150microns can obstruct arterioles. Size and number of particles checked with light microscopy haemocytometer. MAA brakes into smaller particles passing through capillaries before removal by reticuloendothelial system with HL 2-9h. Patients with pulmonary HTN, right-to-left shunts and children should be given less particles. Contraindications include severe pulmonary HTN and allergy.

V/Q scans for PE, pulmonary function monitoring post-lung transplant, preop pneumonectomy estimate of lung function in lung cancer patients (split lung function study), right-to-left shunts and serial assesment of inflammatory lung disease. CXR should be done prior to any V/Q as alternative explanation for cause and interpretation of V/Q, ideally within 6-12h. Ventilation scan (V) performed to increase the low specificity of the perfusion scan (Q), where nonembolic disease should usually also cause V defect. V/Q should be performed over CTPA if probability is low and CXR normal, pregnancy or relative contraindication for iodinated contrast.

Normal V homogeneous distribution on all phases (initial breath, equilibrium, washout). Focal or diffuse retention/trapping indicates obstructive lung disease. Some activity is seen in trachea/bronchi and oesophagus/stomach. Restrictive changes/defects on single-breath/wash-in may dissappear as tracer bypasses via pores of Kohn or canals of Lambert. Normal Q should have well-defined margins on all views (posterior, anterior, LPO, LAO, RAO, RPO), sharp costophrenic angles, homogeneous. Smooth curviliear defect along left medial lung from heart; prominent focal triangular margin suggests perfusion defect abutting heart. Focal asymmetric hilar perfusion defects are abnormal (cardiomegaly, aortic tortuosity, lymphadenopathy).

V/Q findings are classified according to likelihood of emboli seen on angiography. Shape, location and size define segment/subsegment. <25% of a lung segment is small, 25-75% moderate, >75% large defect. 2 moderate of 4 small defects are equivalent to a full-segment defect. Usually underestimated. Mismatched defects have normal ventilation; matched defects have same size and location on V and Q; triple match defects also matched to CXR. Nonsegmental or non-wedge-shaped defects are not usually PE and need to cf CXR for possible cardiomegaly, effusions, adenopathy, masses, pacemenker, pneumonia, bullae, atelectasis, haemorrhage, aortic aneurysm or tortuosity. Defects still present after 3/12 of anticoagulation usually remain permanent. Follow-up scan may show new defects from occlusive thrombi fragmented from larger PE; recurent PE is more likely with multiple new large/moderete defects in previously normal areas. Mismatched defects DDx: new or old PE, extrinsic compression of vessels (mass or adenopathy), mediastinal fibrosis, intraluminal metastasis, sarcoma, lymphagitic carcinomatosis, vasculitides (radiation, Takayaso arteritis, SLE). False-negative V/Q if PE is only partially occlusive or very small.

Asthma causes focal segmental or subsegmental defects on first-breath which may wash in; mucous plugs cause persistent defects. Localised hypoxia induces vasoconstriction hence matched defects. COPD is similar with delayed wash-in and washout; matched perfusion defects due to hypoxia induced vasoconstriction or destruction of lung tissue (mottled if widespread), more pronounced in the apices (whereas a–antitripsin more pronounced in lower lobes). Masses displacing parenchyma produce matched defects. Consolidation with perfusion defects significantly larger than the ventilation/CXR has high probability for PE.

Stripe sign – Central perfusion defect with rim/stripe of increased activity, <10% probability of PE, but should be seen in different views to exclude pleural extension. PE have no overlying stripe of perfused lung and extend to pleural surface.

Fissure sign – Defects matching location and shape of fissures eg effusion. Can be confirmed with layering on lateral supine/decubitus view or plerual effusion on CXR.

Biello criteria include normal, low, intermediate or high probability scans. PIOPED (Prospective Investigation of Pulmonary Embolism Diagnosis) is modified Biello. Must be correlated with clinical suspicion. If intermediate V/Q and multiple risk factors or clinical DVT should have US (with CTPA if negative) or CTPA.

Ammended PIOPED Criteria

Category Criteria Likelihood of PE Prevalence of PE
High >/=2 mismatched large segments/equivalents without vent/CXR abnormality >/=80% 87%
Intermediate >/= 1 moderate segments/equivalents 20-80% 35%
Low Matched/triple matched defect; larger CXR abnormality; stripe sign; >3 small defects; nonsegmental defects <20% 12%
Very Low <3 small defects <20% 2.5%
Normal No defects; or exact match to shape of lungs on CXR <20% 0%

Other Lung Scans

  • Quantitative perfusion lung scan (preop estimate of post-op lung cancer function) measures counts over each lung ROI posterior view. Postop FEV1 (needs to be at least 800-1000mL) estimated by preop FEV1 x percent perfusion to remaining lung after pneumonectomy (same method via SPECT or lateral for lobectomy).
  • Tc99m aerosol lung clearance – Half-life normally ~60min, reduced with increaed permeability of inflammed epithelium (alveolitis, ARDS, smokers), increased with thickened alveolar membranes (fibrosis). Very sensitive, but nonspecific.
  • Xe-in-saline ventilation scan – Xe-133 under pressure dissolves in saline, given IV and remains in solution until it reaches the lungs where it is exhaled. Normal washout <2min, retention in areas of inhalational smoke injury (92% accurate).
  • Tc-99m MAA venogram – May be done by injection into dorsum of feet instead of arm. DVT shown as cutoff of activity and multiple collaterals. Acute DVT can be shown with antifibrin monoclonal antibodies and Tc-99m-labelled peptides with focal uptake.