The ratio of absolute lymphocyte count at interim of therapy to absolute lymphocyte count at diagnosis predicts survival in childhood B-lineage acute lymphoblastic leukemia

Leukemia Research 39 (2015) 144–150

Contents lists available at ScienceDirect

Leukemia Research journal homepage: www.elsevier.com/locate/leukres

The ratio of absolute lymphocyte count at interim of therapy to absolute lymphocyte count at diagnosis predicts survival in childhood B-lineage acute lymphoblastic leukemia Yuping Cheng, Zebin Luo, Shilong Yang, Ming Jia, Haizhao Zhao, Weiqun Xu, Yongmin Tang ∗ Division of Hematology-Oncology, Children’s Hospital, Zhejiang University School of Medicine, #57 Zhuganxiang Road, Yan-an Street, Hangzhou 310003, PR China

a r t i c l e

i n f o

Article history: Received 26 August 2014 Received in revised form 1 November 2014 Accepted 22 November 2014 Available online 9 December 2014 Keywords: Childhood Acute lymphoblastic leukemia Absolute lymphocyte count Ratio Lymphocyte subsets Prognosis

a b s t r a c t Absolute lymphocyte count (ALC) after therapy has been reported to be an independent prognostic factor for clinical outcome in leukemia. This study mainly analyzed ALC at interim of therapy on day 22 (ALC22) and the ratio of ALC-22 to ALC at diagnosis (ALC-0) on the impact of survival and the relation of ALC to lymphocyte subsets in 119 pediatric B-lineage acute lymphoblastic leukemia (B-ALL) patients. Univariate analysis revealed that ALC-22/ALC-0 ratio <10% was significantly associated with inferior overall survival (OS) (hazard ratio (HR) = 12.24, P = 0.0014) and event-free survival (EFS) (HR = 3.3, P = 0.0046). In multivariate analysis, ALC-22/ALC-0 ratio remained an independent prognostic factor for OS (HR = 6.92, P = 0.0181) and EFS (HR = 2.78, P = 0.0329) after adjusting for age, white blood cell (WBC) count and minimal residual disease (MRD) status. A Spearman correlation test showed that CD3+ T cells had a negative correlation with ALC-0 (r = −0.7204, P < 0.0001) and a positive correlation with ALC-22 (r = 0.5061, P = 0.0071). These data suggest that ALC-22/ALC-0 ratio may serve as a more effective biomarker to predict survival in pediatric B-ALL and ALC is mainly associated with CD3+ T cells. © 2014 Elsevier Ltd. All rights reserved.

1. Introduction Initial studies have shown that early absolute lymphocyte count (ALC) recovery after hematopoietic stem cell transplantation (HSCT) is associated with survival in both children and adults with hematological malignancies, such as acute lymphoblastic leukemia [1,2], acute myeloid leukemia [3,4] and lymphoma [5,6]. These results suggest that ALC is associated with an increased graftversus-tumor effect and might serve as an indicator of bone marrow recovery. In the non-transplant setting, ALC, absolute neutrophil count (ANC) and absolute monocyte count (AMC) at diagnosis have been reported to be the prognostic factors in non-leukemia hematological diseases [7–9] and other solid tumors [10–12]. These findings further confirm the role of immune cells in cancer therapy. While in leukemia frequently with abnormal white blood cell (WBC) count at diagnosis, ALC after induction treatment has been reported to be a prognostic factor for clinical outcome. In the pediatric setting, several studies have shown the relation of

∗ Corresponding author. Tel.: +86 571 88873450; fax: +86 571 87033296. E-mail address: Y M [email protected] (Y. Tang). http://dx.doi.org/10.1016/j.leukres.2014.11.013 0145-2126/© 2014 Elsevier Ltd. All rights reserved.

ALC to survival with different cut-off values (350 cells/␮l [13,14], 500 cells/␮l [15], 1000 cells/␮l [14,16] and 1500 cells/␮l [14,17,18]) and different time points during induction therapy on day 15 [13,16], day 28 [13,16], day 29 [14,17], day 36 [18] and day 43 [15] in either ALL or AML. Also, MRD measurement by using flow cytometry or molecular techniques and other traditional prognostic factors such as age, WBC and cytogenetics have been used to predict outcome in leukemia [19,20]. Despite the predictive role of ALC with survival after chemotherapy in pediatric ALL has been evaluated by several studies, some issues remain to be addressed: (1) previous studies always have demonstrated the impact of ALC on survival in ALL (both B-ALL and T-ALL) as a whole cohort, which has not given a clear idea for either B-ALL or T-ALL; (2) although studies have suggested that ALC after therapy may reflect the degree of bone marrow recovery in ALL [15,21], it may be partially affected by ALC at diagnosis due to abnormal initial WBC count in leukemia. So whether ALC after therapy reflects the actual hematopoietic recovery remains unclear; (3) various ALC cut-off values and time points have caused difficulties in clinical practice, thus it is necessary to find more appropriate and simpler prognostic indicators to overcome the diversity; (4) natural killer (NK) cells and T cells have been shown to target hematopoietic

Y. Cheng et al. / Leukemia Research 39 (2015) 144–150

malignancies and play an important role in maintaining remission [22–24], however, it is unclear whether ALC is associated with T cells, NK cells or even B cells responsible for its effect. Notably, in our previous study, we showed that higher ALC at interim of the induction therapy is significantly correlated with lower MRD level at interim of therapy in childhood B-ALL [25]. Based on the above observations, we evaluated the prognostic value of ALC at the midcourse of induction therapy (ALC-22) and the ratio of ALC-22 to ALC at diagnosis on day 0 (ALC-0) with survival in 119 consecutive, newly diagnosed childhood B-ALL patients. Blood cell recovery according to ALC-22 and ALC-22/ALC-0 ratio and the correlation of ALC and lymphocyte subsets were also investigated. 2. Methods 2.1. Patients From August 2006 to June 2009, a total of 119 children (age ≤ 18 years) with newly diagnosed B-ALL at the Children’s Hospital of Zhejiang University School of Medicine were enrolled in this study. Patients with expression of other lineage molecules such as myeloid or T lineage markers were excluded to avoid the possible impact of abnormal lineage cells on the level of ALC. Treatment regimens were described elsewhere [26]. In brief, the treatment regimen included 7 days of prophase treatment with prednisone and followed by 4 weeks of induction therapy consisting of four drugs (vincristine, daunorubicin, l-asparaginase and dexamethason) according to the protocol NPCAC97. This study was approved by the Ethics Committee of Children’s Hospital, Zhejiang University School of Medicine, and informed consents were obtained from all patients’ parents or guardians in accordance with the declaration of Helsinki. 2.2. Analysis General medical records were reviewed to determine gender, age, cytogenetics, WBC, ANC, red blood cell (RBC) count, hemoglobin (Hb) and platelets (Plts). ALC values calculated from the complete blood cell count (CBC) were obtained at the time of diagnosis on day 0, at the mid-course of induction therapy on day 22 and at the end of induction therapy on day 36. MRD values were determined at interim of the induction therapy on day 22 and were assayed with a combination of CD19/CD10/CD34/CD45 antibodies, which has been described elsewhere [26]. Lymphocyte subset analyses were obtained at diagnosis with CD3/CD4/CD8 and CD3/CD56/CD16/CD19 antibody combinations by flow cytometry. The lymphocytes were mainly gated as T cells (CD3+) which were further gated as CD4+ T cells (CD3+CD4+) and CD8+ T cells (CD3+CD8+), B cells (CD3−CD19+) and NK cells (CD3−CD16+CD56+). 2.3. Statistical methods Receiver operating characteristic (ROC) curve analysis was used to determine the cut-off values of the ALC-0, ALC-22, and ALC-22/ALC-0 ratio. The value with maximum sensitivity and specificity was selected as the best cut-off value. OS was measured as the time between the first day of diagnosis and the date of death from any cause or the last follow-up and EFS was calculated from the first day of diagnosis to the date of first event (disease progression, second malignancy, relapse or death of any causes) occurrence or last follow-up. OS and EFS rates were estimated by Kaplan–Meier analysis and compared by using the log-rank test. Potential risk factors for OS and EFS outcome were evaluated in univariate and multivariate analyses with the Cox proportional hazards regression model. Categorical variables between groups were compared by using Chi-square tests. Continuous variables between

145

groups were evaluated by using Mann–Whitney U test or Wilcoxon signed-rank test. Spearman’s rank correlation coefficients were used to evaluate the associations for continuous variables. Statistical analysis of the data was performed using the SPSS 19.0 software package (SPSS, Chicago, IL, USA). A two-sided P-value less than 0.05 was considered statistically significant.

3. Results 3.1. Cut-off values for the ALC-0, ALC-22 and ALC-22/ALC-0 ratio The cut-off values of ALC-0, ALC-22 and ALC-22/ALC-0 ratio were selected from the ROC curve analysis based on its utility as a marker for the survival status of death/survival. The ALC-0 value of 15.7 × 109 /L had an AUC of 0.79 (95%CI, 0.65–0.93, P = 0.001) with a sensitivity of 67% and a specificity of 86% (Fig. 1A). The ALC-22 value of 0.9 × 109 /L had an AUC of 0.71 (95%CI, 0.57–0.86, P = 0.0198) with a sensitivity of 69% and a specificity of 73% (Fig. 1B). The most discriminative cut-off value of ALC-22/ALC-0 ratio was 10% with an AUC value of 0.83 (95%CI, 0.74–0.93, P = 0.0003) on the ROC curve (75% sensitivity and 83% specificity) (Fig. 1C). 3.2. Patient characteristics The median age of the study group at diagnosis was 4.4 years (range: 1.1–14.3 years). The 5-year OS and EFS rates estimated for the entire cohort were 89.91% and 79.83%, respectively. Distributions of baseline characteristics for these patients were presented in Table 1 according to the ALC-22/ALC-0 ratio (≥10% vs. <10%). Patients with ALC-22/ALC-0 ratio <10% was more likely to have higher initial WBC count (≥50 × 109 /L) (P < 0.0001), higher initial LDH level (≥normal) (P < 0.0001), and higher MRD-positive status (≥0.01%) (P = 0.0015) on day 22. Although ALC-22/ALC-0 ratio <10% was more correlated with older age (≥10 years), male and unfavorable cytogenetics as compared with ALC-22/ALC-0 ratio ≥10%, no statistical significance was observed (P > 0.05). 3.3. Prognostic significance of ALC-22/ALC-0 ratio Twelve (10.1%) of 119 patients had died in this cohort. Recurrence (8/12) and progression (4/12) of the disease were the main causes of death. To determine which time point was the best to predict survival, we also analyzed ALC at the end of therapy on day 36 (ALC-36) with survival. However, ALC-36 was not significantly associated with OS or EFS when analyzed as a continuous variable (P > 0.05). Otherwise, patients with ALC-22 ≥ 0.9 × 109 /L had an inferior 5-year OS rate (80% vs. 94.94%, P = 0.0276, Fig. 2A) and 5-year EFS rate (70% vs. 84.81%, P = 0.1295, Fig. 2B) as compared with ALC-22 < 0.9 × 109 /L. ALC-22/ALC-0 ratio <10% was significantly associated with a lower 5-year OS rate (72.97% vs. 95.56%,

Fig. 1. Receiver operating characteristic (ROC) curve and area under the curve (AUC) for ALC-0 (A), ALC-22 (B), and ALC-22/ALC-0 ratio (C). ALC, absolute lymphocyte count; CI, confidence interval; ALC-0, measured at diagnosis (day 0); ALC-22, measured at interim of induction therapy on day 22.

146

Y. Cheng et al. / Leukemia Research 39 (2015) 144–150

Table 1 Patients’ characteristics according to ALC-22/ALC-0 ratio. Characteristics

All cases (n = 119)

ALC-22/ALC-0 ratio <10% (n = 37)

ALC-22/ALC-0 ratio ≥10% (n = 82)

Age (years) 1–10 ≥10

101 18

28 (75.68%) 9 (24.32%)

73 (89.02%) 9 (10.98%)

0.06

63 56

21(56.76%) 16(43.24%)

42 (51.22%) 40 (48.78%)

0.5754

WBC-0 (×109 /L) <50 ≥50

105 14

23 (62.16%) 14 (37.84%)

82 (100%) 0 (0%)

<0.0001

WBC-0 (×109 /L), median (range) ALC-0 (×109 /L), median (range) ALC-22 (×109 /L), median (range)

6.45 (1.2–450) 4.8 (0.8–83.58) 1.23 (0.16–6.27)

32.5 (5.1–450) 17.07 (4.72–83.58) 0.8 (0.16–2.01)

4.23 (1.2–31.1) 2.79 (0.8–23.01) 1.47 (0.31–6.27)

<0.0001 <0.0001

Sex Male Female

P-valuea

LDH at diagnosis ≤Normal >Normal

39 80

2 (5.41%) 35 (94.59%)

37 (45.12%) 45 (54.88%)

<0.0001

MRD level at day 22 of therapy Negative (<0.01%) Positive (≥0.01%) No application

60 51 8

13 (36.11%) 23 (63.89%) 0 (0%)

47 (56.63%) 28 (33.73%) 8 (9.64%)

0.0015

Cytogeneticsb Favorable Intermediate Unfavorable No application

25 36 11 47

10 (27.03%) 10 (27.03%) 6 (16.22%) 11 (29.73%)

15 (18.29%) 26 (31.71%) 5 (6.1%) 36 (43.9%)

0.257

WBC, white blood cell; ALC, absolute lymphocyte count; Hb, hemoglobin; Plts, platelets; LDH, lactate dehydrogenase; MRD, minimal residual disease; WBC-0 and ALC-0, measured at diagnosis on day 0; ALC-22, measured at interim of induction therapy on day 22. a Mann–Whitney U test for continuous variables; Chi-square test by two-sided linear-by-linear association for cytogenetics group while Chi-square test by two-sided Pearson’s exact test for other categorical variables. b Favorable: hyperdiploidy, t(12; 21); unfavorable: hypodiploidy, t(9;22), t(4;11), t(1;19); intermediate: abnormalities other than those in categories favorable or unfavorable.

P < 0.0001, Fig. 2C) and 5-year EFS rate (62.16% vs. 87.8%, P = 0.0026, Fig. 2D) when compared with ALC-22/ALC-0 ratio ≥10%. The results of the univariate and multivariate analyses for factors associated with OS and EFS were presented in Tables 2 and 3, respectively. In univariate analysis, patients with ALC-22/ALC0 ratio <10% were more likely to have inferior OS (HR = 12.24, P = 0.0014) and EFS (HR = 3.3, P = 0.0046) compared with ALC22/ALC-0 ratio ≥10%, while ALC-22 < 0.9 × 109 /L was associated with inferior OS (HR = 3.63, P = 0.0396) but not EFS (HR = 1.86, P = 0.1357). MRD-22 ≥ 0.01% was correlated with shorter OS (HR = 7.58, P = 0.0089) and EFS (HR = 3.97, P = 0.004) compared with MRD-22 < 0.01% (Table 2). In addition, when ALC-22 and ALC-22/ALC-0 ratio were analyzed as continuous variables, they were also significantly associated with survival (data not shown). In multivariate analysis, ALC-22/ALC-0 ratio retained prognostic significance both for OS (HR = 6.92, P = 0.0181) and EFS (HR = 2.78, P = 0.0329) as well as MRD-22 (OS, HR = 5.01, P = 0.0456; EFS, HR = 2.78, P = 0.0329) (Table 3), while ALC-22 was not found to be an independent prognostic factor for survival (Table S1).

3.4. Hematologic recovery comparisons at the interim and the end of therapy according to ALC-22 and ALC-22/ALC-0 ratio To assess whether ALC-22/ALC-0 ratio could preferentially reflect the bone marrow recovery, the levels of blood cell count at interim of therapy (day 22) and at the end of therapy (day 36) were analyzed by Wilcoxon signed-rank test (Table 4). The median values of ANC, Hb and Plts on day 22 and on day 36 were all higher in the group with ALC-22/ALC-0 ratio ≥10% than in the group with ALC-22 ≥ 0.9 × 109 /L, while lower in the group with ALC-22/ALC-0 ratio <10% than in the group with ALC-22 < 0.9 × 109 /L. However, significant difference was only observed in Plts recovery on day 22 (median 67 × 109 /L vs. 96 × 109 /L, P = 0.0348) and day 36 (median 195 × 109 /L vs. 277.5 × 109 /L, P = 0.0462) between the group with ALC-22/ALC-0 ratio <10% and the group with ALC-22 < 0.9 × 109 /L. 3.5. Correlation between ALC and lymphocyte subsets In order to further analyze which cell subset might be accounted for ALC, we performed a correlation study between ALC and

Table 2 Univariate analysis for overall and event-free survival. Prognostic factors

Age (≥10 years vs. 1–10 years) WBC-0 (≥50 × 109 /L vs. <50 × 109 /L) ALC-0 (>15.7 × 109 /L vs. ≤15.7 × 109 /L) ALC-22 (<0.9 × 109 /L vs. ≥0.9 × 109 /L) ALC-22/ALC-0 ratio (<10% vs. ≥10%) ALC-36 (continuous) MRD-22 (≥0.01% vs. <0.01%) Cytogenetics (unfavorable vs. others)

Overall survival

Event-free survival

HR

95% CI

P-value

HR

95% CI

P-value

3.39 3.42 6.01 3.63 12.24 0.92 7.58 2.17

1.02–11.26 2.34–9.52 3.01–13.33 1.06–12.42 2.64–56.75 0.66–1.3 1.66–34.64 0.59–8.05

0.0465 0.0127 0.0022 0.0396 0.0014 0.6552 0.0089 0.2446

2.09 3.14 3.31 1.86 3.3 0.81 3.97 2.01

0.83–5.26 1.71–9.05 1.51–7.7 0.82–4.23 1.45–7.54 0.59–1.11 1.55–10.16 0.65–6.18

0.1183 0.0317 0.0131 0.1357 0.0046 0.197 0.004 0.2224

ALC, absolute lymphocyte count; WBC, white blood cell count; MRD, minimal residual disease; CI, confidence interval; HR, hazard ratio; WBC-0 and ALC-0, measured at the time of diagnosis; ALC-22 and MRD-22, measured at interim of the induction therapy on day 22; ALC-36, measured at end of the induction therapy on day 36.

Y. Cheng et al. / Leukemia Research 39 (2015) 144–150

147

Fig. 2. Kaplan–Meier estimates of overall survival (OS) and event-free survival (EFS) in childhood B-ALL. (A) OS for absolute lymphocyte count on day 22 (ALC-22). (B) EFS for ALC-22. (C) OS for ALC-22/ALC-0 ratio. (D) EFS for ALC-22/ALC-0 ratio.

T/B/NK cells (Table 5). The results showed that ALC-22 was positively associated with both the level of CD3+CD4+ T cells (r = 0.4659, P = 0.0135) and CD3+CD8+ T cells (r = 0.533, P = 0.0042) at diagnosis. On the contrary, ALC-0 was negatively associated with the level of CD3+CD4+ T cells (r = −0.8823, P < 0.0001) and CD3+CD8+ T cells (r = −0.5537, P = 0.0027) at diagnosis. No significant correlation of B cells with ALC-0 or ALC-22 was observed (P > 0.05). NK cells showed a weaker correlation with ALC-0 (r = −0.4153, P = 0.0436) but not with ALC-22 (r = 0.3536, P = 0.0901).

4. Discussion In the present study, the impact of ALC on OS and EFS in pediatric B-ALL has been evaluated. We have found that ALC at the interim of therapy but not at the end of therapy is associated with survival in univariate analysis. The result is similar to the study by De Angulo et al. [13], who has reported the importance of ALC-15 with the cut-off value of 350 cells/␮l rather than ALC at other time points (i.e. days 21 and 28) on survival using the similar treatment (3–4 drugs) in pediatric patients. However, the result with ALC-36 is in

Table 3 Multivariate analysis for overall and event-free survival with ALC-22/ALC-0 ratio. Prognostic factors

Overall survival HR

Age (≥10 years vs. <1–10 years) WBC-0 (≥50 × 109 /L vs. <50 × 109 /L) ALC-22/ALC-0 ratio (<10% vs. ≥10%) MRD-22 (≥0.01% vs. <0.01%)

3.56 1.8 6.92 5.01

Event-free survival 95% CI 0.98–13 0.45–7.26 1.39–34.43 1.03–24.35

P-value

HR

0.0546 0.4096 0.0181 0.0456

2.58 1.92 2.78 3.2

95% CI 0.96–6.94 0.61–6.07 1.09–7.13 1.2–8.54

P-value 0.0615 0.2663 0.0329 0.0201

ALC, absolute lymphocyte count; WBC, white blood cell count; MRD, minimal residual disease; CI, confidence interval; HR, hazard ratio; WBC-0 and ALC-0, measured at the time of diagnosis; ALC-22 and MRD-22, measured at interim of the induction therapy on day 22.

148

Y. Cheng et al. / Leukemia Research 39 (2015) 144–150

Table 4 Blood cell recovery comparisons after induction therapy according to ALC-22 and ALC-22/ALC-0 ratio. Characteristics

P-valuea

Low group

ANC-22 (×109 /L) Hb-22 (g/L) Plts-22 (×109 /L) ANC-36 (×109 /L) Hb-36 (g/L) Plts-36 (×109 /L)

ALC-22

Ratio

0.98 (0.01–6.11) 87 (67–117) 96 (9–346) 4.07 (0.15–58.42) 94 (62–123) 277.5 (3.8–626)

0.5 (0.01–6.62) 84.5 (67–115) 67 (9–271) 2.49 (0.2–25.24) 93 (62–112) 195 (33–543)

0.2423 0.667 0.0348 0.1035 0.6267 0.0462

P-valuea

High group ALC-22

Ratio

1.15 (0.01–19.05) 92 (67–125) 101 (12–533) 2.3 (0.01–24.43) 92.5 (73–118) 262 (10–543)

1.33 (0.01–19.05) 92 (67–125) 119 (12–533) 2.59 (0.01–58.42) 93.5 (75–123) 283 (3.8–626)

0.807 0.7698 0.4695 0.0533 0.93 0.3906

ANC, absolute neutrophil count; Hb, hemoglobin; Plts, platelets. ANC-22, Hb-22 and Plts-22, measured at interim of induction therapy on day 22; ANC-36, Hb-36 and Plts-36, measured at the end of induction therapy on day 36. Low group means the group with ALC-22 <0.9 × 109 /L or ALC-22/ALC-0 ratio <10%; High group means the group with ALC-22 ≥0.9 × 109 /L or ALC-22/ALC-0 ratio ≥10%. a Wilcoxon signed-rank test.

and therefore overcome the diversity of ALC cut-off values in different studies caused by the selection of samples (age, ethnicity etc.). However, it cannot overcome the diversity of cut-off values caused by different therapeutic strategies and time point selections. NK cells have been reported to be correlated with continued remission in AML patients [22] and associated with ALC recovery post-transplant [30]. CD8+T cells may mediate leukemia-associated antigens to exert anti-leukemia effects [31,32]. Moreover, patients with slow recovery of CD8+ T cells and NK cells after HSCT had higher relapse rate and poorer outcome [33,34]. CD4+ T cells were also previously implicated in anti-tumor growth [35,36] and mediated graft-versus-tumor effect [37]. In our study, the median ALC-0 is 4.8 × 109 /L (range: 0.8–83.58 × 109 /L) in the entire cohort and is higher than in healthy controls. Furthermore, ALC-0 is negatively associated with the levels of CD3+ T cells and NK cells but not B cells. So we speculate that a higher level of ALC-0 may contain a large number of abnormal B lymphocytes or non-functional lymphocytes caused by hematopoietic dysfunction in leukemia. With treatment and normal hematologic recovery, non-functional lymphocytes would decrease and normal functional lymphocyte would become the major component of lymphocytes. To some extent, our results that CD3+ T cells are positively associated with ALC-22 have confirmed the above speculations. McIver et al. [38] also have shown that higher T lymphocyte counts in donors are associated with a rapid lymphocyte recovery, a reduced risk of extensive chronic graft-versus-host disease, and superior survival in the recipient following allogeneic HSCT. So, the ALC responding to leukemia during therapy may be CD3+ T cells. However, the association of ALC-22 with lymphocyte subsets at the same time point (day 22) after therapy is not analyzed due to the missing data. Future investigations are needed to characterize the association of ALC with CD3+ T lymphocytes responsible for survival in B-ALL. To our knowledge, this is the first report to demonstrate the association between ALC-22/ALC-0 ratio and survival in pediatric B-ALL. Firstly, ALC at interim of therapy on day 22 rather than ALC at the end of therapy on day 36 is associated with survival in our setting. Secondly, we believe that ALC-22/ALC-0 ratio rather than ALC at certain time point reflects the actual host immunity,

contrast to Jeffrey et al’s finding that ALC with the cut-off value of 500 cells/␮l or as a continuous variable at the end of remission induction (day 43) is significantly associated with EFS and OS in univariate analysis [15]. This difference of observations may be due to the treatment regimens (4 drugs for 4-week period in our setting versus 8 drugs for 6-week period in the Total Therapy XV trial). Although ALC-22 shows significant difference with survival in univariate analysis, it fails to show any prognostic effects in multivariate analysis. The cut-off value of ALC-22 (900 cells/␮l) is also different from De Angulo et al.’s study (350 cells/␮l) [13] and other studies (1000 cells/␮l) [16,17] at the similar time point. So, a lack of consensus on critical ALC cut-off values may be a problem in clinical application. In our study, we have presented a new immune-related prognostic indicator, ALC-22/ALC-0 ratio, which is superior to ALC22 to predict survival in following aspects. (1) ALC-22/ALC-0 ratio <10% is significantly more prevalent among patients with unfavorable presenting characteristics such as higher WBC count, higher LDH level, and positive MRD status. (2) Rubnitz et al. [15] have suggested that MRD is superior to ALC at the end of induction therapy as an independent predictor of OS and EFS. In our study, both ALC-22/ALC-0 ratio and MRD continue to show strong prognostic significance on OS and EFS in multivariate analysis but not ALC-22. ALC-22/ALC-0 ratio represents comparable or even better performance on OS prediction compared with MRD assessment. In addition, ALC-22/ALC-0 ratio can be easily obtained from the CBC testing which is simple, cost effective and readily available. The use of peripheral blood rather than bone marrow is less invasive. So, ALC-22/ALC-0 ratio may represent a supplementary or even an alternative parameter to MRD value as a promising predictor in pediatric B-ALL, especially in low income countries. (3) A better recovery of blood cells especially Plts is more likely to be found in patients with ALC-22/ALC-0 ratio ≥10%. A rapid recovery of blood cells after chemotherapy may reflect the regeneration of normal hematopoietic recovery, thus it would be a favorable prognostic indicator [27–29]. In fact, ALC-22/ALC-0 ratio rather than ALC value at certain time point can be a relatively dynamic and suitable marker to reflect the variation of blood cells and to predict bone marrow recovery in patients. (4) More importantly, the application of ALC-22/ALC-0 ratio may minimize individual differences,

Table 5 Correlation between ALC and T/B/NK cells. ALC-0 (n = 27)

CD3+ T cells CD4+ T cells CD8+ T cells NK cells B cells

ALC-22 (n = 27)

ALC-22/ALC-0 ratio (n = 27)

r

P-value

r

P-value

r

P-value

−0.7204 −0.8223 −0.5537 −0.4153 0.3557

<0.0001 <0.0001 0.0027 0.0436 0.0881

0.5061 0.4695 0.533 0.3536 −0.2548

0.0071 0.0135 0.0042 0.0901 0.2296

0.7747 0.848 0.6642 0.441 −0.3774

<0.0001 <0.0001 0.0002 0.031 0.0691

NK cells, natural killer cells; ALC-22, ALC at interim of the induction therapy on day 22; ALC-0, ALC at the time of diagnosis on day 0. r, Spearman rank correlation coefficient.

Y. Cheng et al. / Leukemia Research 39 (2015) 144–150

bone marrow recovery and therapeutic response. Moreover, ALC22/ALC-0 ratio may minimize the diversity of ALC cut-off values in various studies. Thirdly, ALC-22/ALC-0 ratio can be easily obtained from the routine clinical test, i.e. complete blood cell count. However, some limitations remain. As this is a retrospective study with a relatively small sample size, it is necessary to be explored by a multi-center study with a larger sample size. Furthermore, the association of CD3+ T cells and NK cells with clinical outcome may supplement and validate the relation of ALC to survival in pediatric B-ALL. In conclusion, ALC-22/ALC-0 ratio <10% is a simple, inexpensive, and independent unfavorable prognostic factor for survival in childhood B-ALL. Conflict of interest The authors declare no conflict of interest. Funding source This study was supported in part by grants from the National Natural Science Foundation of China (Nos. 30971283, 81470304 and 81170502), Zhejiang Provincial Natural Science Foundation of China (No. LZ12H08001), Leukemia Research Innovative Team of Zhejiang Province (No. 2011R50015), Zhejiang Provincial Fund of Health Bureau (No. 2007B122) and Zhejiang Provincial Fund of Education Bureau (No. 20061407). Acknowledgements We thank Mrs. Baiqin Qian, Mr. Hongqiang Shen, Mr. Jianhua Feng, Mrs. Sisi Li and Mr. Ning Zhao at Hematology–Oncology Laboratory for their excellent technical support. We also would like to thank all the physicians from the Hematology-Oncology Division for their careful care of patients enrolled in this study. Contributions. Y.-M.T. was the principal investigator and takes the primary responsibility for the paper. Y.-P.C., Z.-B.L., M.J., H.-Z.Z., S.-L.Y. and W.-Q.X. collected the data. Y.-P.C. and Y.-M.T. analyzed the data. Y.-P.C. wrote the manuscript draft which was amended by Y.-M.T. All the authors approved the final version submitted. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.leukres. 2014.11.013. References [1] Kumar S, Chen MG, Gastineau DA, et al. Lymphocyte recovery after allogeneic bone marrow transplantation predicts risk of relapse in acute lymphoblastic leukemia. Leukemia 2003;17:1865–70. [2] Ishaqi MK, Afzal S, Dupuis A, et al. Early lymphocyte recovery postallogeneic hematopoietic stem cell transplantation is associated with significant graft-versus-leukemia effect without increase in graft-versus-host disease in pediatric acute lymphoblastic leukemia. Bone Marrow Transpl 2008;41:245–52. [3] Le Blanc K, Barrett AJ, Schaffer M, et al. Lymphocyte recovery is a major determinant of outcome after matched unrelated myeloablative transplantation for myelogenous malignancies. Biol Blood Marrow Transpl 2009;15:1108–15. [4] Porrata LF, Litzow MR, Tefferi A, et al. Early lymphocyte recovery is a predictive factor for prolonged survival after autologous hematopoietic stem cell transplantation for acute myelogenous leukemia. Leukemia 2002;16:1311–8. [5] Porrata LF, Gertz MA, Inwards DJ, et al. Early lymphocyte recovery predicts superior survival after autologous hematopoietic stem cell transplantation in multiple myeloma or non-Hodgkin lymphoma. Blood 2001;98:579–85. [6] Porrata LF, Inwards DJ, Micallef IN, et al. Early lymphocyte recovery postautologous haematopoietic stem cell transplantation is associated with better survival in Hodgkin’s disease. Br J Haematol 2002;117:629–33. [7] Deel MD, Kong M, Cross KP, et al. Absolute lymphocyte counts as prognostic indicators for immune thrombocytopenia outcomes in children. Pediatr Blood Cancer 2013;60:1967–74.

149

[8] Behl D, Ristow K, Markovic SN, et al. Absolute lymphocyte count predicts therapeutic efficacy of rituximab therapy in follicular lymphomas. Br J Haematol 2007;137:409–15. [9] Tadmor T, Fell R, Polliack A, et al. Absolute monocytosis at diagnosis correlates with survival in diffuse large B-cell lymphoma-possible link with monocytic myeloid-derived suppressor cells. Hematol Oncol 2013;31:65–71. [10] Sadeghi N, Badalato GM, Hruby G, et al. Does absolute neutrophil count predict high tumor grade in African-American men with prostate cancer. Prostate 2012;72:386–91. [11] Chew V, Chen J, Lee D, et al. Chemokine-driven lymphocyte infiltration: an early intratumoural event determining long-term survival in resectable hepatocellular carcinoma. Gut 2012;61:427–38. [12] Li J, Jiang R, Liu WS, et al. A large cohort study reveals the association of elevated peripheral blood lymphocyte-to-monocyte ratio with favorable prognosis in nasopharyngeal carcinoma. PLoS ONE 2013;8:e83069. [13] De Angulo G, Yuen C, Palla SL, et al. Absolute lymphocyte count is a novel prognostic indicator in ALL and AML: implications for risk stratification and future studies. Cancer 2008;112:407–15. [14] Hatzipantelis E, Pana ZD, Vlachou M, et al. Peripheral blood lymphocyte recovery and overall survival in pediatric acute lymphoblastic leukemia. Pediatr Blood Cancer 2014;61:181–3. [15] Rubnitz JE, Campbell P, Zhou Y, et al. Prognostic impact of absolute lymphocyte counts at the end of remission induction in childhood acute lymphoblastic leukemia. Cancer 2013;119:2061–6. [16] Anoceto MA, Gonzalez OA, Guerchicoff DSE, et al. Absolute lymphocyte count as a prognostic factor in children with acute lymphoblastic leukemia. An Pediatr (Barc) 2012;76:10–1. [17] Rabin KR, Gramatges MM, Borowitz MJ, et al. Absolute lymphocyte counts refine minimal residual disease-based risk stratification in childhood acute lymphoblastic leukemia. Pediatr Blood Cancer 2012;59:468–74. [18] Alkayed K, Halalsheh H, Khattab E, et al. Lack of prognostic significance of absolute lymphocyte count after intensive induction therapy in childhood acute lymphoblastic leukemia. Pediatr Blood Cancer 2012;59:351. [19] Borowitz MJ, Devidas M, Hunger SP, et al. Clinical significance of minimal residual disease in childhood acute lymphoblastic leukemia and its relationship to other prognostic factors: a Children’s Oncology Group study. Blood 2008;111:5477–85. [20] Conter V, Bartram CR, Valsecchi MG, et al. Molecular response to treatment redefines all prognostic factors in children and adolescents with B-cell precursor acute lymphoblastic leukemia: results in 3184 patients of the AIEOP-BFM ALL 2000 study. Blood 2010;115:3206–14. [21] Behl D, Porrata LF, Markovic SN, et al. Absolute lymphocyte count recovery after induction chemotherapy predicts superior survival in acute myelogenous leukemia. Leukemia 2006;20:29–34. [22] Lowdell MW, Craston R, Samuel D, et al. Evidence that continued remission in patients treated for acute leukaemia is dependent upon autologous natural killer cells. Br J Haematol 2002;117:821–7. [23] Ohnishi K, Yamanishi H, Naito K, et al. Reconstitution of peripheral blood lymphocyte subsets in the long-term disease-free survivors of patients with acute myeloblastic leukemia. Leukemia 1998;12:52–8. [24] Rakhra K, Bachireddy P, Zabuawala T, et al. CD4(+) T cells contribute to the remodeling of the microenvironment required for sustained tumor regression upon oncogene inactivation. Cancer Cell 2010;18:485–98. [25] Shen HQ, Feng JH, Tang YM, et al. Absolute lymphocyte count is associated with minimal residual disease level in childhood B-cell precursor acute lymphoblastic leukemia. Leuk Res 2013;37:671–4. [26] Xu XJ, Tang YM, Shen HQ, et al. Day 22 of induction therapy is important for minimal residual disease assessment by flow cytometry in childhood acute lymphoblastic leukemia. Leuk Res 2012;36:1022–7. [27] Zeidler L, Zimmermann M, Moricke A, et al. Low platelet counts after induction therapy for childhood acute lymphoblastic leukemia are strongly associated with poor early response to treatment as measured by minimal residual disease and are prognostic for treatment outcome. Haematologica 2012;97:402–9. [28] Yanada M, Borthakur G, Garcia-Manero G, et al. Blood counts at time of complete remission provide additional independent prognostic information in acute myeloid leukemia. Leuk Res 2008;32:1505–9. [29] Faderl S, Estrov Z. Hematopoietic recovery following induction therapy of acute leukemias: prognostic implications and a new look at the definition of remission. Leuk Lymphoma 2004;45:67–71. [30] Porrata LF, Gastineau DA, Padley D, et al. Re-infused autologous graft natural killer cells correlates with absolute lymphocyte count recovery after autologous stem cell transplantation. Leuk Lymphoma 2003;44:997–1000. [31] Molldrem JJ, Lee PP, Wang C, et al. Evidence that specific T lymphocytes may participate in the elimination of chronic myelogenous leukemia. Nat Med 2000;6:1018–23. [32] Greiner J, Schmitt M, Li L, et al. Expression of tumor-associated antigens in acute myeloid leukemia: implications for specific immunotherapeutic approaches. Blood 2006;108:4109–17. [33] Koehl U, Bochennek K, Zimmermann SY, et al. Immune recovery in children undergoing allogeneic stem cell transplantation: absolute CD8+ CD3+ count reconstitution is associated with survival. Bone Marrow Transpl 2007;39:269–78. [34] Giannelli R, Bulleri M, Menconi M, et al. Reconstitution rate of absolute CD8+ T lymphocyte counts affects overall survival after pediatric allogeneic hematopoietic stem cell transplantation. J Pediatr Hematol Oncol 2012;34: 29–34.

150

Y. Cheng et al. / Leukemia Research 39 (2015) 144–150

[35] Corthay A, Skovseth DK, Lundin KU, et al. Primary antitumor immune response mediated by CD4+ T cells. Immunity 2005;22:371–83. [36] Stevanovic S, Griffioen M, Nijmeijer BA, et al. Human allo-reactive CD4+ T cells as strong mediators of anti-tumor immunity in NOD/scid mice engrafted with human acute lymphoblastic leukemia. Leukemia 2012;26:312–22.

[37] Jiang YZ, Barrett J. The allogeneic CD4+ T-cell-mediated graft-versus-leukemia effect. Leuk Lymphoma 1997;28:33–42. [38] Mciver Z, Melenhorst JJ, Wu C, et al. Donor lymphocyte count and thymic activity predict lymphocyte recovery and outcomes after matched-sibling hematopoietic stem cell transplant. Haematologica 2013;98:346–52.