Lymphocyte recovery from radiation-induced lymphopenia in locally advanced esophageal squamous cell carcinoma: correlations with prognosis and lymphocyte-related organs

Patient characteristics

Two hundred thirty-two eligible patients were enrolled for this study, including 185 (79.7%) males and 47 (20.3%) females. The median age of all patients was 62 years. Most of them (69.8%) were in normal health conditions with ECOG-PS-0. Fifty-eight (25.0%) patients at stage II and 137 (59.1%) patients with tumors at cervical and upper esophagus. The median tumor length was 5.0 cm. Of 232 patients, 94.8% finished the 61.2 Gy radiotherapy course. The number of patients treated with PF, TF, TP and TC was 64 (27.6%), 102 (44.0%), 30 (12.9%) and 36 (15.5%), respectively. After concurrent chemoradiotherapy, 178 (76.7%) patients received 2 cycles of consolidation chemotherapy and 54 (23.2%) patients received 0–1 cycle. Clinical characteristics in details were listed in Table 1.

Table 1 Patient characteristicsALC data and patient subgroups

The ALC data of 232 patients were obtained. During treatment, ALCs declined every week and generally reached a plateau at week 5, continuing till the end of treatment, then gradually elevated to near-normal levels (Fig. 1). The median of ALCs at baseline was (× 109/L) 1.68, and 0.99, 0.80, 0.61, 0.50, 0.41, 0.43, 0.43 for week1–7 during dCCRT, respectively. The cumulative incidence of G4 ALC nadir and G1–3 nadir was 29.7% (N = 69), and 70.3% (N = 163).

Fig. 1figure 1

The dynamic changes of ALCs from baseline (before dCCRT) through week 1–7 during dCCRT and 6 months after the end of dCCRT represented by Tukey box-and-whisker plots. The red dashed line represented the lower limit of normal lymphocyte counts (1.1 × 109/L). ALC absolute lymphocyte count, dCCRT definitive concurrent chemoradiotherapy

Six months after the end of dCCRT, there were 104 (44.8%) patients whose ALCs returned to normal level (≥ 1.1 × 109/L), while more than half of them (55.2%) were still accompanied by varying degrees of lymphopenia: 77 (33.2%) with grade 1 lymphopenia, 47 (20.3%) with grade 2 and 4 (1.7%) with grade 3. The median of ALC at 6 months after dCCRT was 1.06 × 109/L (range 0.41–3.98).

With regard to the status of lymphocyte recovery the median of LRI was 67.8% (range 19.8–193.1). According to the cut-off point of LRI (60%) by maxstat based on OS (Additional file 1: Figure S2) and the development of G4 and G1–3 ALC nadir during dCCRT, the population were classified into 4 groups: group A (G4 → unrecovered) included patients with G4 ALC nadir during dCCRT and LRI < 60% (N = 27, 11.6%), group B (G4 → recovered) with G4 ALC nadir during dCCRT and LRI ≥ 60% (N = 42, 18.1%). For patients with G1–3 ALC nadir during dCCRT, group C (G1–3 → unrecovered) included those with LRI < 60% (N = 67, 28.9%), while group D (G1–3 → recovered) with LRI ≥ 60% (N = 96, 41.4%). Distributions of clinical characteristics in these 4 groups could be found in Additional file 2: Table S1.

Prognosis of lymphocyte recovery combined with ALC nadir during dCCRT

At analysis, 117 (50.4%) patients died with 69.9 months of median OS time. The overall PFS, LRFS and DMFS median time was 27.4 months, 46.2 months and 49.2 months, respectively. As displayed in Fig. 2, among patients with G4 ALC nadir during dCCRT, significantly poorer 5-year OS rate (18.5% vs 53.8% p < 0.001) and 5-year PFS rate (3.7% vs 31.7%, p < 0.001) were observed in group A vs group B. Also, prominent differences in the 5-year LRFS rate (14.8% vs 45.9%, p < 0.001) and 5-year DMFS rate (7.4% vs 49.0%, p < 0.001) existed in group A vs group B. Among patients with G1–3 ALC nadir during dCCRT, survival curve analysis indicated poorer 5-year OS (46.8% vs 62.1%, p = 0.005) and 5-year PFS (38.0% vs 52.2%, p = 0.017) in group C vs group D. The 5-year LRFS rate (58.4% vs 39.9%, p = 0.009) and 5-year DMFS rate (59.9% vs 42.3%, p = 0.008) in group D was respectively significantly higher than that in group C.

Fig. 2figure 2

Kaplan Meier curves of A overall survival, B progression-free survival; C local recurrence-free survival, and D distant metastasis-free survival between 4 groups. Group A (G4 → unrecovered) included patients with G4 ALC nadir during dCCRT and LRI < 60% (N = 26), group B (G4 → recovered) with G4 ALC nadir during dCCRT and LRI ≥ 60% (N = 42). For patients with G1–3 ALC nadir during dCCRT, group C (G1–3 → unrecovered) included those with LRI < 60% (N = 67), while group D (G1–3 → recovered) with LRI ≥ 60% (N = 96). ALC, absolute lymphocyte count dCCRT, definitive concurrent chemoradiotherapy; DMFS, distant metastasis-free survival; LRFS, local recurrence-free survival; LRI, lymphocyte recovery index; OS, overall survival; PFS, progression-free survival

Comparing group A with group C, OS and PFS were both worse in patients with G4 ALC nadir than those with G1–3 ALC nadir, with a 5-year OS rate 18.5% vs 46.8% months (p = 0.005) and 5-year PFS rate 3.7% vs 38.0% (p < 0.001), respectively. Besides, poorer 5-year LRFS rate (14.8% vs 39.9%, p = 0.003) and 5-year DMFS rate (7.4% vs 42.3%, p < 0.001) were observed in group A vs group C.

Based on univariable Cox analysis (Additional file 2: Tables S2, S3), following multi-factor adjustment in Table 2 revealed that lymphocyte unrecovered from G4 ALC nadir at post 6 months was independently related to shorter OS (HR, 2.80; 95% CI 1.47–5.34; p = 0.002) and PFS (HR, 3.67; 95% CI 2.09–6.46; p < 0.001), as well as poorer LRFS (HR, 2.82; 95% CI 1.51–5.26; p = 0.001) and DMFS (HR, 3.65; 95% CI 1.96–6.78; p < 0.001). Among patients with G1–3 ALC nadir during dCCRT, inadequate lymphocyte recovery was still found to be an independent factor associated with poorer OS (HR, 1.70; 95% CI 1.07–2.72; p = 0.025) and LRFS (HR, 1.60; 95% CI 1.01–2.52; p = 0.040). However, no statistically prognostic significance was observed in PFS and DMFS (Additional file 2: Table S4, Fig. 2).

Table 2 Multivariable Cox analysis for survival outcomes between group A and BPrognosis of lymphocyte recovery within the same tumor stage

Based on the tumor stage, the population was re-stratified for further investigation. In the subset of patients with stage II (N = 58), no significant correlations were observed between lymphocyte recovery status and clinical outcomes (data not shown). However, within the stage III + IV group (N = 174, including 137 with stage III and 37 with stage IV), patients who did not experience sufficient lymphocyte recovery at 6 months post-therapy had inferior OS, PFS, LRFS and DMFS (Additional file 1: Figure S3, Additional file 2: Table S5). Furthermore, the multivariable Cox analysis in Additional file 2: Table S6 demonstrated an independent association between the inadequate lymphocyte recovery and shorter OS (HR, 1.77; 95% CI 1.19–2.64; p = 0.005), shorter PFS (HR, 1.64; 95% CI 1.15–2.35; p = 0.007), as well as inferior LRFS (HR, 1.65; 95% CI 1.13–2.42; p = 0.010) and inferior DMFS (HR, 1.80; 95% CI 1.23–2.65; p = 0.003).

Dose-volume parameters

In 232 patients, the median of V5, V10, V20, V30 and V50 of the BM was 42.2% (range 12.6–69.2), 33.5% (range 9.3–54.2), 26.1% (range 5.5–41.8), 18.8% (range 3.3–34.7), 5.3% (range 0.5–16.2), respectively. The max of the spleen V5, V10, V20, V30 and V50 was 96.0%, 87.7%, 80.3%, 55.2% and 6.7%, respectively. For EDIC, the median dose was 9.0 Gy (range 2.0–14.6).

In group A vs B, the median of V5, V10, V20, V30 and V50 of the BM was 46.0% vs 40.7%, 37.2% vs 33.8%, 27.9% vs 25.7%, 20.3% vs 19.9% and 5.3% vs 5.4%, and the max values of spleen V5, V10, V20, V30 and V50 was 77.4% vs 96.0%, 60.0% vs 87.7%, 35.5% vs 80.3%, 7.9% vs 55.2% and 0 vs 1.4%, respectively. In group C vs D, the median of related BM dose-volume parameters mentioned above was respectively 44.7% vs 40.8%, 34.0% vs 32.2%, 26.3% vs 25.5%, 19.0% vs 17.6% and 5.5% vs 4.8%; and max values of spleen V5, V10, V20, V30 and V50 were 70.0% vs 87.0%, 60.5% vs 64.7%, 50.7% vs 53.8%, 32.4% vs 45.8% and 6.5% vs 6.7%. The median of EDIC in group A vs B and group C vs D was 14.1 Gy vs 13.6 Gy and 8.59 Gy vs 8.50 Gy, respectively.

Predictors of lymphocyte recovery at 6 months after dCCRT

In univariable logistic regression analysis, tumor stage (II vs. III + IVa), dose volume parameters of BM gave p-values of less than 0.05 whether in comparison with group A and group B (Additional file 2: Table S7) or group C and group D (Additional file 2: Table S8). Between group A and group B, BM V5 ≥ 40.7% (OR 5.75; 95% CI 1.69–19.52; p = 0.005) was positively associated with lymphocyte unrecovery, while, BM V5 ≥ 46.0% (OR 2.91; 95% CI, 1.50–5.66; p = 0.002) for patients with G1–3 ALC nadir and lymphocyte unrecovery. Subsequently, multivariable adjustment indicated that BM V5 ≥ 40.7% (OR 4.24; 95% CI 1.18–15.20; p = 0.027) and ≥ 46.0% (OR 2.29; 95% CI 1.11–4.73; p = 0.025) was independently correlated with lymphocyte unrecovery 6 months after dCCRT respectively in patients with G4 and G1–3 ALC nadir (Table 3).

Table 3 Multivariable Logistic regression analysis of factors related to lymphocyte unrecovered from radiation-induced lymphopenia

留言 (0)

沒有登入
gif