The standard treatment for locally advanced cervical cancer is concurrent chemoradiotherapy. A common side effect in patients undergoing chemotherapy is HT. When receiving local radiotherapy alone, metabolic activity increased in the unirradiated bone marrow. The compensatory response is reduced during concurrent chemoradiotherapy owing to the superimposed toxicity, which increases the incidence of HT[11, 12]. The hematopoietic stem cells in the red bone marrow can self-renew and regenerate, which is the basis of maintaining normal hematopoietic function and repairing bone marrow injury. Wang et al. used ex vivo HRMAS 1 H NMRS and assessed microvascular perfusion status and changes occurring in the fat content composition in the bone marrow of rat femurs after total-body X-ray irradiation. Moreover, the bone marrow fat content gradually increased at days 4 and 7 post-irradiation, and the bone marrow microcirculation perfusion was correlated with fat content.
MRI IDEAL IQ FatFrct maps obtained by water-fat imaging can be used to measure the changes of bone marrow fat fraction quantitatively during the treatment course, providing quantitative information on marrow composition. IDEAL IQ is sensitive to marrow composition changes, quantitatively assessing bone marrow damage resulting from chemotherapy and radiation . Carmona et al.  used IDEAL IQ to assess the changes in vertebrae bone marrow fat fraction during chemoradiotherapy and detected that the bone marrow mean dose was associated with a 0.43% per Gy increase in PDFF (%). The changes in the latter are associated with peripheral blood cell counts. In this study, PDFF% increased by 58.5% during radiotherapy and had a positive dose-response relationship with dose accumulation. PDFF% is closely related to the changes in blood cell count, which is consistent with the study by Carmona .
Increased radiation dose to pelvic bone marrow enhances hematologic toxicity in patients undergoing chemoradiotherapy. The dose constraints of PBM could minimize the incidence of HT to some extent. For the implementation of individualized bone marrow protection, the effect of radiotherapy dose on the bone marrow in cervical cancer must be clarified. There are currently no standard criteria for optimal PBM dose limitation regimens. Some retrospective studies and NTCP models have demonstrated that a low dose of PBM was significantly associated with HT events [16, 17]. Kumar et al.  detected that G4 HT was related to PBM-V5 > 95% and V20 > 45%. Rogradientson et al.  demonstrated that active bone marrow V20 < 20 Gy was significantly correlated with WBC and ANC nadirs. Zhu et al.  concluded that with every 1 Gy increase in mean PBM dose, there was a reduction in ANC and WBC by 9.6/µL and 7.8/µL per week, respectively.
In this study, the analysis of PDFF% changes under different dose gradients demonstrated that the dose of radiotherapy caused a significant increase of 58.5% in PDFF% in the pelvic bone marrow. Moreover, there was a correlation between the V40 of the PBM and WBC nadirs. Significant positive correlations were observed between PDFF (%) changes and ANC nadirs at 5–10 Gy. Moreover, PDFF (%) changes are related to ALC nadirs at 5–40 Gy, which is consistent with previous studies [6, 20, 21]. It may be related to the irradiation range of low-dose area, suggesting that bone marrow sparing in radiotherapy for cervical cancer should reduce the range of low-dose bone marrow irradiation under the premise of ensuring the coverage of the target area.
From the perspective of radiobiology, bone marrow injury caused by high-dose radiation is more long-lasting and challenging to repair. In this study, WBC, ANC, and PLT nadirs occurred at RT mid-point, which decreased by 64.8%, 68.5%, and 80.1%, respectively; however, they only recovered to 56.4%, 70.2%, and 42.6%, one month after radiotherapy. PDFF% decreased in 6 months after the end of radiotherapy; however, it still increased by 55.95% compared with that at RT-Pre. Nevertheless, the fat content in bone marrow areas receiving high doses (> 30 Gy) continued to increase after radiotherapy, indicating that the bone marrow injury caused by high dose radiation is challenging to reverse or repair in a short time. This study found no correlation between PDFF% in the high-dose area (>30 Gy) and HT, which was primarily directly related to the small absolute volume of high-dose irradiation.
This study demonstrates that MRI IDEAL IQ FatFrac imaging can be used to quantify the changes in bone marrow composition during chemoradiotherapy. We confirmed the correlation between PDFF% and peripheral blood cell reduction, while PDFF% changes and dose accumulation demonstrated a significant dose-response relationship. By measurement of the bone marrow signal values, IDEAL IQ can distinguish red from yellow bone marrow. The combination of MRI and fat quantitative technology could accurately locate the position and range of active bone marrow; thereby, providing a visual basis for individualized bone marrow protection.
This study primarily assessed the changes of PDFF% under different dose gradients over the course of treatment and evaluated the correlation between changes in PBM fat contents and peripheral blood cells. However, the long-term change trend of bone marrow fat content still requires further assessment. Sini et al.  detected that the ALC was still at a low level one year after radiotherapy. Therefore, our team anticipates obtaining the fat fraction imaging and peripheral blood cell count of patients long after the end of radiotherapy through further follow-up, establishing a prediction model based on the bone marrow fat content dynamic changes and radiation dose to guide the individualized bone marrow sparing.
In conclusion, the present study used MRI IDEAL IQ fat fraction imaging to evaluate the fat content of bone marrow noninvasively, analyzed the changes of pelvic bone marrow fat content during and after radiotherapy, revealed the dose-effect of bone marrow fat content, and discussed the relationship between bone marrow changes and HT. A strong correlation between low-dose radiation and HT was detected. The individualized sparing of PBM low-dose irradiation during concurrent chemoradiotherapy for cervical cancer should be paid more attention.