Segmentation of Whole-Body Images into Two Compartments in Model for Bone Marrow Dosimetry Increases the Correlation with Hematological Response in 177Lu-DOTATATE Treatments

Linn Hagmarker,Johanna Svensson,Tobias Rydén,Peter Gjertsson,Peter Bernhardt
DOI: https://doi.org/10.1089/cbr.2017.2317
Abstract:Background: In 177Lu-DOTATATE treatments, bone marrow (BM) is one of the most important organs at risk. The authors previously developed an image-based two-compartment method for BM dosimetry, showing a significant correlation between absorbed dose to BM and hematological toxicity in 177Lu-DOTATATE treatments. In the present study, they aimed to further evaluate this BM dosimetry method by finding optimal settings for dividing the whole body into two compartments; in terms of minimizing the coefficient of variation (CV) for the individual absorbed doses and studying its correlation to the BM response. The authors have also added specific absorbed fractions for male and female. Finally, they compare this two-compartment method with whole-body dosimetry. Methods: This study included 46 patients with advanced neuroendocrine tumors treated with 177Lu-DOTATATE on two to five occasions at Sahlgrenska University Hospital. Planar gamma camera images were collected at four time points postinjection, and a segmentation tool using a normalized number of uptake foci (nNUF) to divide the whole body into high- and low-uptake compartments was used. The authors characterized the two-compartment model and compared it with whole-body dosimetry. Results and conclusion: The dosimetry method was robust, with an optimal nNUF value of 0.1-0.2. Using an nNUF value of 0.15, the absorbed BM dose was estimated as 0.20 Gy/7.4 GBq, and the CV as 8.4%. Compared to whole-body dosimetry, stronger correlation was found between absorbed dose to BM and hematological response using the two-compartment method. The two-compartment method has potential as a valuable image-based alternative to blood-based BM dosimetry.
What problem does this paper attempt to address?