Main Session
Sep 28
PQA 02 - Lung Cancer/Thoracic Malignancies, Patient Reported Outcomes/QoL/Survivorship, Pediatric Cancer

2439 - A Novel Imaging Biomarker for Radiation Induced Cardiotoxicity

04:45pm - 06:00pm PT
Hall F
Screen: 28
POSTER

Presenter(s)

Arezoo Modiri, PhD - University of Maryland School of Medicine, Baltimore, Baltimore, MD

A. Modiri1, I. Vogelius2, C. Terrones Campos3, D. Kutnar4, A. Sawant1, and J. Petersen2; 1University of Maryland, School of Medicine, Baltimore, MD, 2Department of Oncology, Rigshospitalet Copenhagen University Hospital, Copenhagen, Denmark, 3Department of Infectious Diseases, Rigshospitalet, Copenhagen, Denmark, 4Dept. Of Oncology Copenhagen University Hospital – Rigshospitalet, Copenhagen, Denmark

Purpose/Objective(s): Cardiotoxicity is the second major cause of death after the cancer itself among thoracic cancer patients receiving radiotherapy (RT). Unfortunately, there are no reliable non-invasive biomarkers to predict RT-induced cardiotoxicity. In cardiology, the pericardium sac is considered a fast responder to cardiac injury. Any decline in pericardium health can potentially be an early biomarker of a later cardiac event. We investigated if RT-induced radiographic pericardium changes might serve as early imaging biomarkers for late cardiotoxicity.

Materials/Methods: We performed a retrospective study of 521 patients, 234 males & 287 females, age: 28 to 93 (median 68) yrs, treated with chemo-RT for small/non-small-cell lung cancer at 1 institution within 2009-2020. Median for mean heart doses was 2.8 Gy (max: 36.4 Gy). The heart and its 4-millimeter outmost layer (pericardium sac) were contoured on standard-of-care (Pre-RT) baseline/plan CT scans. The 5-to-8-month post-RT follow-up CTs were deformably registered on the baseline CTs using Plastimatch. The overall HU change histogram was studied and voxel-based mass change (VCM) in each dose range and tissue composition HU range was reported as average normalized value:

VCMDose & HU range = mean{S? ? Dose & HU range(HUbaseline(?)-HUfollowup(?)) × Volume? } / mean{S? ? Dose range(HUbaseline(?) + 1000) × Volume? }

where ? is a pericardium voxel.

The CTs had mixed contrast status. To approximate a contrast calibration factor for each tissue composition HU range (CCFHU range), we used contrast-enhanced baseline and follow-up CTs from 281 patients. For each HU range, we assumed that in the absence of dose (i.e., in voxels with <1Gy) VCM was the effect of contrast: CCFHU range.

Results: The table below shows that voxels within calcification and fat HU ranges had baseline-to-followup mass changes consistent with the dose increase (p values < 0.02). When studying the 281 cases with contrast enhanced baseline and follow-up CTs, there was a dose-dependent baseline-to-followup (max: 2.6 gr) increase in overall pericardium mass (p=0.006). Additionally, the histograms of HUbaseline-HUfollowup per patient had skewed distributions with mean & skewness that were correlated (p values<0.04) with mean dose to pericardium sac (Spearman correlation factor [SCF] -0.7 & 0.7, respectively) & mean dose to the whole heart (SCF -0.6 and 0.6, respectively).

Conclusion: Pericardium HU histogram and composition distribution have dose-dependent changes at around 6 months post-RT. These changes are detectable on the standard-of-care CTs and can potentially serve as early markers of late cardiotoxicity.

Table 1

Dose response

 (VCM – CCF) %

Per pericardium voxel dose range
Tissue (HU range)

 <1 (Gy)

 1-5 (Gy)

 5-10 (Gy)

 10-20 (Gy)

 >20 (Gy)
Fat (-500<HU<-10)

 0.30

 0.09

 0.06

 -0.11

 -0.16
Effusion (-5<HU< 30)

 0.25

 0.11

 0.20

 0.10

 0.16
Normal tissue (31<HU<60)

 0.19

 0.13

 0.25

 0.18

 0.22
Fibrosis (65<HU<120)

 -0.02

 0.03

 0.12

 0.23

 0.19
Calcification (500>HU>130)

 0.11

 -0.02

 -0.14

 -0.36

 -0.90