Session Item

For locally advanced cervical cancer, earlier studies showed that a higher dose to the bone marrow leads to increased hematologic toxicity. Compared to photon therapy, proton therapy (PT) shows a large dose reduction to the bone marrow. A further reduction can be achieved with bone marrow sparing (BMS) technique. However, the dosimetric impact of BMS on other organs at risk (OAR) is currently unknown for PT. Therefore, we systematically studied the trade-off between BMS and dose to other OAR for intensity-modulated proton therapy (IMPT).

Twenty patients treated for FIGO 2018 stage IB3-IVA cervical cancer were included in this study. 2/20 patients had paraaortic nodal irradiation. With our in-house automated treatment planning system, 4-beam (90°, 150°, 210°, and 270°) Pareto-optimal IMPT plans were created. A library-of-plans online-adaptive strategy was assumed. A 5 mm set-up and 3% range robustness were used for robust optimization on 19 scenarios and evaluation on 27 scenarios. The whole pelvic bones were taken as a substitute for the bone marrow. For the Pareto fronts, first, an initial plan without BMS was created according to the EMBRACE-II planning constraints. Next, the bone marrow dose was reduced in each consecutive plan with steps of 1 Gy until maximum BMS was reached. Target coverage and conformality were kept equal to the initial plan during the optimization. The relation between the bone marrow dose and dose to the bladder, bowel, rectum, and sigmoid was evaluated.

Five to eight plans were created for each patient, resulting in a total of 127 IMPT plans. All plans fulfilled the EMBRACE-II constraints. The average [range] bone marrow Dmean for plans without BMS was 18.5 [16.1 – 22.3] Gy and for plans with maximum BMS 14.3 [12.6 – 17.6] Gy. For the plans with maximum sparing, the mean dose to the bladder, rectum, sigmoid, and bowel increased on average [range] by 2.5 [0.5 – 4.2] Gy, 3.1 [1.1 – 5.0] Gy, 2.3 [1.1 – 4.1] Gy, and 1.5 [0.2 – 3.0] Gy. Table 1 shows the average [range] OAR Dmean for 0, 2, 3 Gy BMS, and for the maximum BMS.

Figure 1 shows the average increase in bladder, rectum, bowel, and sigmoid Dmean for all patients as a result of decreasing the bone marrow dose. Every line corresponds to a patient and every point to an IMPT plan. For 20/20, 11/20, and 4/20 patients, BMS of respectively 2, 3, and 4 Gy could be realized without increasing the average dose to the four OAR with more than 1 Gy.

Although the bone marrow mean dose with IMPT is already lower than with photon irradiation, it can be further reduced by 2-4 Gy without a considerable increase in the OAR mean dose (<1 Gy). Further reduction of the bone marrow dose is at the expense of OAR dose increases of possibly >2 Gy. However, more BMS is possible for selected patients, which justifies an individualized BMS approach.