Room: Karl Dean Ballroom A1
Purpose: Calculating lung mean dose (LMD) from Y90-microsphere therapy requires estimating lung shunt fraction (LSF) and mass of the lungs. The standard model estimates LSF with planar imaging following Tc99m-MAA injection to the target site(s) and assumes 1 kg lung mass for all patients. We propose a new LMD calculation by the deriving the LSF from Tc99m-MAA SPECT/CT (spectCT) and the patient-specific lung mass from diagnostic CT (diagCT).
Methods: Our novel LMD (LMDnew) calculation consists of the following: 1) Estimate liver counts from liver contour drawn on spectCT; 2) Estimate left lung count density [cts/g] from spectCT and multiply it by patient-specific lung mass [g] estimated from diagCT (using lung segmentation) to calculate total lung counts; 3) Compute LMDnew using LSF calculated from liver and lung counts (LSFnew) and the total lung mass from diagCT. We compared LMDnew, LSFnew, and lung mass estimates to standard model values (LMDstd, LSFstd, 1 kg) in 44 patients with HCC undergoing Y90 therapy.
Results: On average, LSFnew and LMDnew were 41% and 50% of the standard model estimates, respectively. The novel method proposed resulted in median [0%/25%/75%/100% percentile] decrease of 0.043 [0.009/0.029/0.062/0.101] in LSF and 4.0 [0.2/1.6/6.6/23.2] Gy in LMD compared to standard model. CT-based lung masses ranged 544-1182 g (median 816 g) and were lower than the nominal 1 kg in 91% of cases. The maximum dose to target volumes (150 Gy) was limited by LMDstd exceeding 30 Gy in 19% of cases, whereas LMDnew did not exceed 30 Gy for any cases.
Conclusion: The proposed method to calculate lung mass, LSF, and LMD using diagnostic chest CT and Tc99m-MAA SPECT/CT indicates that the current standard model: 1) substantially overestimates LSF and LMD; and 2) unnecessarily restricts treatment dose options in ~20% of cases due to 30 Gy LMD limit.