Impact Of Functional Bone Marrow On Acute Hematologic Toxicity In Patients With Gynecologic Cancer Receiving Pelvic Radiation Therapy: Does Vmat Make It Worse?

To investigate pelvic bone marrow (PBM) dosimetric parameters that correlate with hematological toxicity (HT) in patients treated with volumetric modulated arctherapy (VMAT) versus intensity-modulated radiotherapy (IMRT) radiation therapy for gynecologic cancer. A total of 171 patients with endometrial cancer and cervical cancer treated with VMAT or IMRT in our institution were retrospectively studied. Propensity Score Matching(PSM) were used to balance clinical bias. 51 patients treated with VMAT and 77 patients treated with IMRT radiotherapy were chosen to be equalized outcome by PSM.Dose volume histograms (DVH) for the organs at risk of PBM, lumbosacralspinal bone marrow(LSS), upper pelvic bone marrow(UP) and lower pelvic bone marrow(LP) were compared for two different techniques, and analyzed the correlation with HT events. VMAT group had significantly higher low dose volumes and lower high dose volumes. For low dose volumes, IMRT had better performance than VMAT in V5-V15 volumes (p=0.000). For high dose volumes, VMAT reduced the receiving doses of PBM slightly in V35-V35, however, no significant difference was demonstrated. Dmeans of VMAT plans were lower than IMRT plans in all three subsites (p<0.05). Dose receiving volume demonstrated difference between subsites. For LSS, VMAT had a better performance in V25-V45%. For UP and LP, IMRT had lower volume percentages in V5-V20 and V5-V10, respectively. During RT, 86 patients (86/128, 67.2%) and 67 patients (67/128, 52.3%) experienced leucopenia and neutropenia, respectively. Thrombocytopenia (12/128, 9.4%) occurred much less than WBC events. In univariated logistic regression analysis, only TP (concurrent chemotherapy with platinum and paclitaxel in the meantime of clinical parameters was associated with WBC G3+, ANC G2+ and ANC G3+ events. However, most volume parameters of PBM (V5-V20), UP (V5-V20) and LP (V5 and V10) were associated with WBC 2+ events. For PLT, the only predictive factor were dosimetric (PBM-V10,UP-V10 and LP-V5). There are no differences in each events of HT between VMAT and IMRT plans. In multivariate logistic regression analysis, for WBC and ANC events, both clinical and dosimetric factors were involved in the prediction, however dosimetric factor of PBM-V10 was the only predictive factor of PLT nadir. Cutoff value of PBM-V10 associated with WBC 2+ events and PBM-V15 associated with ANC 3+ were 85% and 80%, respectively. There are no differences in HT events of dosimetric contribution between VMAT and IMRT plans, however VMAT group had significantly higher low dose volumes of PBM. HT in radiation therapy for gynecologic cancer depends both on clinical and dosimetric factors. Sparing and dosimetric limitation of PBM will help the prevention of HT, and guarantee the proceeding of chemoradiation therapy.