Evaluation of plan quality and robustness of IMPT and helical IMRT for cervical cancer

Both plan quality and robustness were investigated through comparing some dosimetric metrics between intensity modulated proton therapy (IMPT) and helical tomotherapy based intensity modulated radiotherapy (IMRT) for cervical cancer.

Both a spot-scanning robust (SRO) IMPT plan and a helical tomotherapy robust (TRO) IMRT plan were generated for each of 18 patients. In order to evaluate the quality of nominal plans without dose perturbations, planning scores (PS) on clinical target volume (CTV) and five organs at risk (OARs) based on clinical experience, and normal tissue complication probabilities (NTCP) of rectum and sigmoid were calculated based on Lyman-Kutcher-Burman (LKB) model. Dose volume histogram bands width (DVHBW) were calculated in 28 perturbed scenarios to evaluate plan robustness.

Compared with TRO, the average scores of SRO nominal plans were higher in target metrics [V_46.8Gy, V_50Gy, Conformity and Homogeneity](16.5 vs. 15.1), and in OARs metrics (60.9 vs. 53.3), including bladder [V₃₅,V₄₅, D_mean,D_2cc], rectum [V₄₀,V₄₅,D_2cc,D_max], bowel [V₃₅,V₄₀,V₄₅, D_max], sigmoid [V₄₀,D_max] and femoral heads [V₃₀,D_max]. Meanwhile, NTCP calculation showed that the toxicities of rectum and sigmoid in SRO were lower than those in TRO (rectum: 2.8% vs. 4.8%, p < 0.05; sigmoid: 5.2% vs. 5.7%, p < 0.05). DVHBW in target coverage for the SRO plan was smaller than that for the TRO plan (0.6% vs. 2.1%), which means that the SRO plan generated a more robust plan in target.

Better CTV coverage and OAR Sparing were obtained in SRO nominal plan. Based on NTCP calculation, SRO was expected to allow a small reduction in rectal toxicity. Furthermore, SRO generated a more robust plan in CTV target coverage.