technologiesTechnologyStandard of care

Split-function perfusion scintigraphy before lung resection

aka Quantitative lung perfusion scintigraphy, Split lung function study, Differential perfusion scan

Before a surgeon removes part of a lung, the scan is used as an arithmetic tool rather than a picture: it measures what share of the breathing work that piece of lung is actually doing, so the team can predict the patient's lung function after the operation and decide whether they will tolerate it.

Predicting postoperative lung function is where perfusion scintigraphy earns its place in thoracic oncology. The ACCP physiologic evaluation guideline (Brunelli et al., Chest 2013) sets out the division of labour explicitly: for a pneumonectomy candidate the perfusion method is used, with predicted postoperative FEV1 taken as preoperative FEV1 times one minus the fraction of total perfusion for the resected lung, while for a lobectomy the anatomical segment-counting method is used. The reason is empirical, not dogmatic: before lobectomy, segment counting predicts postoperative FEV1 about as well as scintigraphy, so the scan is usually unnecessary; before pneumonectomy, the lung to be removed has to be measured.

The thresholds are what make the scan consequential. The ACCP guideline recommends no further testing when both predicted postoperative FEV1 and DLCO exceed 60% of predicted; a low-technology exercise test (stair climb or shuttle walk) when either falls below 60% but both stay above 30%; and formal cardiopulmonary exercise testing with VO2max measurement when either falls below 30%. The ERS/ESTS 2009 guideline routes any patient whose FEV1 or DLCO is below 80% of predicted to cardiopulmonary exercise testing, and regards predicted postoperative FEV1 or DLCO below 30% combined with a peak VO2 below 35% of predicted or 10 mL/kg/min as prohibitive risk for major resection. In its framing, segment counting is the default calculation and perfusion scintigraphy is reserved for patients whose function is borderline.

Whether tomographic quantification beats planar counting is genuinely unsettled. A 2024 series found SPECT/CT predicted postoperative FEV1 and DLCO more accurately than either planar scintigraphy or segment counting, and would have qualified 58 of 82 patients for resection against 47 by segment counting. A 2020 head-to-head study in the Journal of Nuclear Medicine found no significant difference between segment counting, planar perfusion and SPECT/CT, and concluded the extra effort may add nothing outside right lobectomy. What both agree on is the direction of the error in segment counting: it systematically underestimates postoperative function, which risks denying surgery to patients who would have tolerated it.

Generic schematic · not to scale · placeholder for the front

How it works

The same Tc-99m MAA perfusion scan used to look for embolism, counted rather than merely looked at. Activity is measured per lung (and per region on tomographic studies) to give the fraction of total perfusion each part contributes, and that fraction is used to predict what spirometry will look like after a resection: predicted postoperative FEV1 equals preoperative FEV1 multiplied by one minus the fraction of total perfusion belonging to the lung to be removed. For a lobectomy the anatomical version is used instead, multiplying by one minus the number of unobstructed segments to be removed over the total number of unobstructed segments, counting 19 segments in all.

Strengths

Turns a routine perfusion scan into a quantitative decision tool, using a formula stated in guidelines rather than judgement.
Measures what the lung to be resected actually contributes, which matters most when disease has already destroyed function in that lung; segment counting assumes every segment is equal.
Segment counting systematically underestimates predicted postoperative FEV1, so the scan can qualify patients for curative surgery who would otherwise be refused.

Limitations

Before a lobectomy it adds little: segment counting predicts postoperative FEV1 about as accurately, so guidelines reserve the scan for pneumonectomy candidates and borderline function.
Whether SPECT or SPECT/CT quantification beats planar counting is contested, with a 2024 series favouring SPECT/CT and a 2020 head-to-head study finding no difference.
It predicts spirometry, not outcome: the guidelines pair it with exercise testing precisely because a predicted value alone does not decide operability.

Products

1top

ApprovedIntravenous SPECT particle agent

Technetium Tc-99m albumin aggregated (MAA)

Tiny clumps of albumin, a normal blood protein, tagged with a radioactive marker and injected into a vein. They wedge briefly in the smallest blood vessels of the lung wherever blood is flowing, so the scan maps the lung's blood supply; the clumps then break up and clear.