Baseline characteristics

The age of 69 cervical cancer patients in this study ranged from 32 to 70 years old, and the average age was 53.4 years; BMI (kg/m2) values ranged from 17.7 to 29.2, with an average BMI value of 22.7; the tumour stages of the patients included stage IB-IIIC (67 cases), in addition to 2 cases of pelvic localized recurrence which can be treated by radical radiotherapy; there were 25 cases of radical radiotherapy and 44 cases of postoperative adjuvant radiotherapy; there were 61 cases of squamous cell carcinoma and 8 cases of nonsquamous cell carcinoma; the number of CCT cycles ranged from 3 to 6 cycles; 43 cases (62.3%) involved 3–4 cycles and 26 cases (37.7%) involved 5–6 cycles. The range of HGB values (g/L) before RT was 64–140 with a average of 118.6; the range of PLT (*109G/L) before RT was 98–542 with a average of 252.6; the range of NEU (*109G/L) before RT was 1.7–16.8 with a mean of 4.2; the range of LYM (*109G/L) before RT was 0.6–3.5 with a average of 1.6; and the range of NLR before RT was 0.7–30.6 with a average of 3.3. During radiotherapy, according to the RTOG acute HT classification, there were 54 cases (78.3%) with a grade 0–1 decrease in HGB and 15 cases (21.7%) with a grade 2–4 decrease. Regarding platelets, there were 52 (75.4%) patients with a grade 0–1 decrease and 17 (24.6%) patients with a grade 2–4 decrease in platelets. Regarding neutrophils, there were 34 cases (49.3%) with a grade 0–1 decrease and 35 cases (50.7%) with a grade 2–4 decrease in neutrophils. The above data are shown in Table 1.

Table 1 Patient characteristicsAnalysis of variables related to the nadir of haemoglobin

In analysing the variables influencing the nadir of HGB in 69 cervical cancer patients, the results of univariate analysis showed that the variables with P < 0.08 included weight, BMI, baseline hemoglobin (B-HGB), R-PBM-V10, et al., as shown in Table 2. Multivariate analysis showed that the nadir of HGB was positive correlated with B-HGB and negative correlated withR-LP-V10, R-LP-V25, R-LP-V50 and R-LP-mean. The cutoff values of these five variables were 96.150 g/l, 74.62%, 44.290%, 7.258% and 2020.850 Gy, respectively, as shown in Table 2. In the multivariate analysis of the absolute dose volume of BM irradiation, the nadir of HGB was positive correlated with B-HGB and negative correlated with A-LP-V15, A-LP-V20 and A-LP-V30. The cutoff values of these four variables were 96.150 g/l, 205.813 cm3, 151.947 cm3 and 96.740 cm3, respectively, as shown in Table 2.

Table 2 Multivariate analysis of the nadir of hemoglobin in 69 cervical cancer patientsAnalysis of variables related to the nadir of platelets

The nadir of platelets of 69 patients with cervical cancer was analysed by linear regression. In univariate analysis, the variables with P < 0.08 included BPLT, R-PBM-V50, et al., as shown in Table 3. After multivariate analysis of these variables, it was concluded that the nadir of platelets was positive correlated with baseline platelet (B-PLT) and negative correlated with R-LP-V50, and their cutoff values were 255.5*109 G/L and 17.356%, respectively, as shown in Table 3. The above operation was repeated for the volume of absolute dose of BM irradiation (including PBM, IL, LS, LP), and no factor with a P < 0.05 was found in the multivariate analysis as shown in Table 3.

Table 3 Multivariate analysis of the nadir of platelet in 69 cervical cancer patientsAnalysis of variables related to the nadir of neutrophils

The variables influencing the nadir neutrophil counts in 69 patients with cervical cancer were analysed by linear regression. In the univariate analysis, the only factor with P < 0.08 was age (P = 0.056). The absolute dose volume of BM (including PBM, IL, LS, LP) was recalculated, and the only factor with P < 0.08 was age (P = 0.056).

Considering the neutrophils falling apart and BM dose related, also associated with chemotherapy, while patients are treated first patients enrolled in the study, and not received neoadjuvant chemotherapy or induction chemotherapy, concurrent chemotherapy cycle number less than 3 cycles and is greater than 6 cycles also excluded, as far as possible to ensure the consistency of the baseline of cases was studied, but there are still differences concurrent cycles of chemotherapy. To explore the effect of the number of CCT cycles on myelosuppression, 69 patients were divided into two groups. The patients with 3–4 cycles of CCT were divided into group A (43 cases). Patients with 5–6 cycles of CCT were assigned to group B (26 cases). In group A, the nadir neutrophils occurred on days 8–52 of radiotherapy, and the median time was day 32. In group B, the nadir neutrophils occurred on days 7–45 of radiotherapy, with a median time of day 35. Single-factor linear regression analysis was performed to obtain data with a P < 0.08. Multivariate analysis was performed to obtain data with a P < 0.08. Two variables with P < 0.05 were obtained, namely, R-IL-V15 and R-IL-V50. The cutoff values were found to be 79.2% and 11.7%, as shown in Table 4. The above procedure was also repeated for the absolute dose volume of BM irradiation in group A, and the final multivariate analysis obtained a statistically significant factor, A-LP-V50, whose cutoff value was 30 cm3. The above procedure was also repeated in group B. Regardless of whether the relative dose volume or absolute dose volume of BM irradiation was analysed by univariate analysis, only one factor with a significant difference could not be further analysed by multivariate analysis. The only factor with a significant difference was age which was negative correlated with the nadir neutrophils (P = 0.026).

Table 4 Multivariate analysis of the nadir of neutrophil in 43 patients receiving 3–4 cycles of CCTAnalysis of variables related to the nadir of lymphocytes

The variables influencing the value of nadir LYM in 69 patients with cervical cancer were analysed by linear regression. Both in the univariate analysis of the relative and the absolute dose volume of BM, the only factors with P < 0.05 were baseline lymphocyte (B-LYM) (P < 0.001) and BMI (P = 0.036), while all the dosimetric parameters of BM irradiation were not significantly different.

Analysis of variables related to the vertice of neutrophil-to-lymphocyte ratio

The vertice of NLR for 69 cervical cancer patients was analysed by linear regression. In univariate analysis, the variables with P < 0.08 included age, baseline neutrophil-to-lymphocyte ratio (B-NLR), R-PBM-V5, et al., as shown in Table 5. After multivariate analysis of these variables and ROC curve analysis in which the value of the state variable was taken as the mean value of vertex NLR 15.5, it was concluded that the cutoff values of age (p < 0.001) and R-LP-V15 (p = 0.037) were 67.5 years and 30 cm3, respectively, as shown in Table 5. The above operation was repeated for the absolute dose volume of BM irradiation (including PBM, IL, LS, LP). Age (p = 0.001) and A-PBM-mean were identified in the multivariate analysis, the cutoff values of which were 67.5 years and 3040.6 Gy, respectively, as shown in Table 5.

Table 5 Multivariate analysis of the vertice of NLR in 69 cervical cancer patients