Change in FDG-PET uptake of bone marrow: Radiation dose-response and possible compensatory effect
Anders Beck Olin,
Denmark
PD-0157
Abstract
Change in FDG-PET uptake of bone marrow: Radiation dose-response and possible compensatory effect
Authors: Anders Beck Olin1, Ivan Richter Vogelius2, Liselotte Højgaard3, Flemming Littrup Andersen3, Barbara Malene Fischer3
1Rigshospitalet, Dept. of Oncology & Dept. of Clinical Physiology and Nuclear Medicine, Copenhagen, Denmark; 2Rigshospitalet, Dept. of Oncology, Copenhagen, Denmark; 3Rigshospitalet, Dept. of Clinical Physiology and Nuclear Medicine, Copenhagen, Denmark
Show Affiliations
Hide Affiliations
Purpose or Objective
Hematological
toxicity after chemoradiotherapy is partially caused by direct radiation to the
highly radiosensitive bone marrow, but the dose-response and the potential compensatory effect
in non-irradiated bone marrow remain to be better understood. For this
purpose, FDG-PET allows assessment of the metabolic activity within the bone marrow
and is correlated with higher cellularity and proliferative activity.
This study explored
how bone marrow within individual
vertebrae responds to radiotherapy
for NSCLC patients, by measuring changes in FDG-PET uptake from baseline to end
of radiotherapy and correlating with the dose to the individual vertebrae.
Material and Methods
We
identified and analyzed 15 NSCLC patients referred for radiotherapy with
curative intent (66 Gy in 33 fractions) between Jan 2019-Dec 2019, all
having an FDG-PET/CT before treatment and a follow-up PET/CT within 12 months.
Each
vertebra was segmented and labeled (cervical (C1-7); thoracic (T1-12); and
lumbar (L1-L5) vertebrae) using CT as input to the deep learning model winning MICCAI’s
VerSe challenge 2019.
The FDG-PET
mean standardized uptake values (SUVmean) within each vertebra at baseline
and follow-up were assessed and their relative differences were calculated
with baseline as reference (ΔSUVmean). The delivered mean dose (Dmean)
to each vertebra was obtained from the dose plan. We used Spearman's
correlation (ρ) to measure the association between the Dmean and ΔSUVmean
using data from all vertebrae across all patients.
Results
The SUVmean
for each vertebra at baseline and follow-up and the corresponding Dmean
are shown for an illustrative patient in Fig 1. A drop in SUVmean
for vertebrae receiving high doses is seen together with a potentially
compensating increased uptake for the vertebrae receiving only scatter dose.
Figure 1: SUVmean of each vertebra at baseline and follow-up (left y-axis) and the corresponding Dmean (right y-axis).
Across all vertebrae, there was a strong correlation between Dmean and ΔSUVmean
(ρ=-0.43, p-val<0.0001), with uptake decreasing with increasing dose. This
can also be seen from Fig 2, showing a scatterplot of the Dmean
and ΔSUVmean and a rolling mean line to indicate the trend. There is
a slightly increased SUVmean at low Dmean, supporting the
hypothesis of a compensatory response to radiation injury in the irradiated
vertebrae. However, the individual out-of-field responses are mixed so a
greater understanding is needed.
Figure 2: Scatterplot between and Dmean and ΔSUVmean. Each point denotes a vertebra and the solid line indicates the rolling mean.
Conclusion
This study
used longitudinal FDG-PET/CT scans as a unique method to demonstrate both a
dose-response of the hematopoietic activity in bone marrow as well as a
patient-dependent compensatory response in non-irradiated bone marrow. These
results are the first step towards an in-depth evaluation of using FDG-PET to
obtain a better understanding of the bone marrow’s dose-response, compensatory
response, and hematological toxicity.