Encrypted login | home

Program Information

PET Image-Guided Dose Escalation Study for Cervical Cancer Patients Receiving HDR Brachytherapy

no image available
J Anderson

J Anderson, K Kiel, R Yao, Y Liao, D Bernard, N Biswal, J Turian, J Chu, Rush University Medical Center, Chicago, IL

SU-E-T-453 Sunday 3:00PM - 6:00PM Room: Exhibit Hall

PURPOSE: To study the feasibility and benefit of boosting dose to active regions of cervical cancer identified by positron emission tomography (PET) imaging during brachytherapy.

METHODS: 5 patients imaged with MRI and PET/CT were retrospectively studied. An experienced physician contoured the GTV and HR-CTV using the MRI scan alone, and then contoured the region of PET-hot activity using a PET/CT image. The MRI and PET/CT images where then fused, and a PET-HR-CTV volume was constructed from the union of the HR-CTV and PET-GTV.

Each patient had been treated using a Syed applicator with 4-20 needles. Two new plans were optimized using Oncentra's IPSA package with the following objectives: 1) a "PET-blindA plan in which the HR-CTV-D90 received the prescription dose (Rx), and 2) a plan where the PET-HR-CTV D90 received Rx and the PET-GTV D90 was boosted to 1.5 times Rx. In both cases the EQD2 D2cc dose to the bladder and rectum were kept below the GEC-ESTRO guidelines of 90 and 75 Gy, including the contribution from external beam therapy. Dose-volume parameters and TCP biological predictors were calculated for the PET-target volumes in both plans.

RESULTS: Significant differences were seen between the non-PET and PET-GTV (average size difference 21.8cc, average overlap 50%). In the unboosted plans, the average percent D90 to the PET-GTV was 171%, whereas in the boosted plans it increased to 223%. Average TCP, calculated using risk-scaled parameters for IR-, HR-CTV and GTV, increased from 76% to 84% with the boost to the PET positive region.

CONCLUSION: PET imaging modality has the potential to identify residual cervical cancer otherwise not visible on MRI and CT studies. This work has found that MRI target contouring regularly does not include regions of PET activity, and PET-based planning can lead to improved outcome predictions over plans using MRI alone.


Funding Support, Disclosures, and Conflict of Interest: Funding support partly provided by a Varian research grant.

Contact Email: