Presenter(s)
Y. Xu1,2, Y. Song3, Y. Li1,2, and D. Yan3; 1Lung Cancer Center, West China Hospital, Sichuan University, Chengdu, Sichuan, China, 2Department of Radiation Oncology, Cancer Center, West China Hospital, Sichuan University, Chengdu, Sichuan, China, 3Radiotherapy Physics and Technology Center, Cancer Center, West China Hospital, Sichuan University, Chengdu, Sichuan, China
Purpose/Objective(s): Tumor voxel dose response was assessed using 2-point FDG-PET/CT imaging feedback during the treatment course in locally advanced NSCLC treated with concurrent chemoradiotherapy (CCRT) versus induction chemoimmunotherapy (ICI). The potential response-guided treatment adaptation was evaluated.
Materials/Methods: Baseline tumor voxel metabolic activity (SUV0) and dose response matrix (DRM) were quantified using a pre-treatment FDG PET/CT and a second scan obtained during the 3rd treatment week in the CCRT group (n=11) or after ICI in the ICI group (n=32). The ICI group included patients who received 2 (n=9), 3 (n=12), or 4 cycles (n=11). Equivalent doses (EQD2) of the chemotherapy in CCRT and the chemoimmunotherapy in ICI treatment were determined, respectively, by assuming the treatment achieved the same mean DRM or the mean tumor voxel metabolic reduction. The inter-/intra-tumoral variations of the tumor voxel (SUV0, DRM) were calculated within each treatment group. The required treatment doses, achieving expected tumor control probabilities (TCP) of 0.95 and 0.99, were derived from tumor voxel (SUV0, DRM) distributions and a TVCP-lookup Table.
Results: EQD2 dose for 1 cycle of chemotherapy was about 8.5 Gy. The EQD2 doses for the 2, 3, and 4 ICI treatments were 46Gy (23Gy/cycle), 48.9Gy (16.3Gy/cycle), and 56.5Gy (14.1Gy/cycle), respectively. The CVs of tumor voxel SUV0 and DRM were 60% and 68% in the CCRT group, and 69% and 52% in the ICI group. In both CCRT and ICI groups, intra-tumoral variation significantly exceeded inter-tumoral variation for tumor voxel SUV0 (58%, 60% vs. 14%, 32%) and DRM (63%, 44% vs. 28%, 29%). The mean total treatment doses (EQD2) required to achieve TCP=0.95/0.99 were 95/110Gy (CCRT group) and 86/100Gy (ICI group). Based on all patients’ data, to achieve the TCP=99%, the radiation dose after 2-cycle ICI would be largely different (58±27 Gy EQD2). 9 of 43 patients required the radiation dose >75Gy EQD2 (86-112Gy). Thus, local hypofractionation could be applied effectively for highly resistant tumor regions.
Conclusion: Large spatial heterogeneity in the pre-treatment tumor metabolic activity and dose response was observed in the multi-modality treatment of locally advanced NSCLC. Incorporating immunotherapy into induction chemotherapy can significantly increase the tumor cell killing effect in the first 2 cycles. However, the effect was saturated afterward with a small increase with more cycles, suggesting 2 ICI cycles as an effective window to initiate radiotherapy in the response-guided adaptive therapy. To achieve a high tumor local control, radiotherapy doses required for the individuals after the 2 ICI cycles were very different, with about 21% of the tumors needing to perform hypofractionation to their locally resistant regions.
| Characteristics | CCRT (n = 11) | ICI (n = 32) | p | |
| Age, years | Mean (range) | 65(49-72) | 66(39-76) | 0.4 |
| Clinical stage | IIIA | 4 | 11 | 0.75 |
| IIIB | 6 | 16 | ||
| IIIC | 1 | 5 | ||
| GTV volume (cc) | Mean (range) | 75.3(19.7-209.4) | 84.7(8.7-255.5) | 0.32 |