A volumetric analysis of GTVD and CTVHR as defined by the GEC ESTRO recommendations in FIGO stage IIB and IIIB cervical cancer patients treated with IGABT in a prospective multicentric trial (EMBRACE)

Purpose To quantify the gross tumor volume at diagnosis (GTV D ) and high-risk clinical target volume (CTV HR ) at brachytherapy (BT) and describe subgroups of patients with different patterns of response to chemoradiotherapy (CRT) in patients with FIGO stage IIB and IIIB cervical cancer treated with image-guided adaptive brachytherapy (IGABT). Additionally, to evaluate the feasibility of IGABT achieving adequate target coverage in these groups. Materials and methods Patients with FIGO stage IIB and IIIB cervical cancer enrolled in the EMBRACE study were analyzed. T2-weighted MRI scans were obtained at diagnosis and at BT. GTV D and CTV HR were defined as per the GEC ESTRO recommendations. Patients were classified taking into account that initial tumor volume and response to CRT represented by the volume of residual disease (CTV HR ) and extent of residual parametrial disease are all major factors determining local dose delivery by BT, local control, and overall disease outcome. These factors were quantified applying the following criteria: (1) volume of the GTV D relative to the median volume of the GTV D ; (2) the ratio (R) of CTV HR to GTV D for each patient; (3) the extent of residual parametrial disease at the time of BT. Accordingly, patients were classified into six groups (G1–G6): stage IB 1 -like tumors (G1), tumors with good response and any size (G2), small tumors with moderate response (G3), large tumors with moderate response (G4), tumors with poor response (G5) and those with progressive disease (G6). Tumor and treatment characteristics were then compared among the first five groups (only 3 patients were allocated to G6). Results A total of 481 patients were evaluated. The number of patients in the 6 groups were 55, 78, 123, 147, 75 and 3, respectively. The mean (SD) GTV D was 43.6 (32.8)cm 3 and the mean (SD) CTV HR was 31.6 (16.1)cm 3 . The mean GTV D and CTV HR were 12.6cm 3 and 23.7cm 3 in G1 ( R >1.1), 47.5cm 3 and 25.3cm 3 in G2 ( R <0.9), 23.9 cm 3 and 29.9cm 3 in G3 (R 0.9–1.1), 73.4cm 3 and 38.5cm 3 in G4 (R 0.9–1.1), 79.4cm 3 and 59.5cm 3 in G5 ( R >1.1), respectively. Parametrial disease extent at BT was as follows: no involvement in G1 and G2, proximal at most in G3 and G4, distal or to the pelvic wall in G5, progressive in G6. The use of interstitial needles was progressively higher among the groups (mean 0, 0, 2, 3, 6 in G1–5, P <0.001). The mean GTV BT D 100 in G1-5 was 103.1Gy, 91.8Gy, 93.5Gy, 88.3Gy and 87.1Gy. The mean CTV HR D 90 in G1-5 was 95.1Gy, 92.1Gy, 92.6Gy, 87.6Gy and 88.4Gy. Conclusions In patients with FIGO stage IIB and IIIB disease, intra-FIGO stage heterogeneity and overlap between the two stages exist with respect to tumor volume, treatment response and extent of parametrial disease at BT. Taking into account GTV D , parametrial disease at BT and the ratio of CTV HR /GTV D , five major groups exist. These enable prediction of GTV BT and CTV HR dose coverage through BT. IGABT, as performed in EMBRACE, accommodates to a considerable degree for the different variants of tumor regression in these groups through adaptation of the treatment technique including the use of needles. However, major variations remain at present with regard to dose to GTV BT and to CTV HR , which are most pronounced in G4 and G5. This new classification will be validated in future in regard to clinical outcome in EMBRACE.