Sage Journals: Discover world-class research

Abstract

BACKGROUND AND AIM:

There is a limited data at the moment regarding the clinical value of inflammatory indices and malnutrition markers in children with acute leukemias. We have examined the usefulness of prognostic nutritional index (PNI), Glasgow prognostic score (GPS), Prognostic Index (PI), monocyte to lymphocyte (MLR), neutrophil to lymphocyte (NLR), and platelet to lymphocyte (PLR) ratios to stratify patients as regards the response to induction therapy correlating them to different prognostic factors.

PATIENTS AND METHODS:

Children with acute leukemia and without microbial-induced inflammation at the time of diagnosis were prospectively recruited. Preliminary total and differential CBC, c-reactive protein (CRP), serum albumin (ALB) were used to calculate different inflammatory indicators including NLR, MLR, PLR, PNI, GPS, and PI.

RESULTS:

Higher PNI was significantly more associated to children who achieved remission as compared to those without remission ( $p<$ 0.0001). Patients without remission had GPS 1 or 2 compared to GPS 0 or 1 in those who entered remission ( $p=$ 0.001). NLR was significantly lower in patients in remission than in those without remission ( $p=$ 0.005). Similarly, complete remission was significantly associated to MLR $\leqslant$ 0.45 as compared to MLR $>$ 0.45 ( $p<$ 0.0001).

CONCLUSION:

Pretreatment PNI, GPS, CRP, serum albumin, NLR, MLR, and PLR are remission promising prognostic markers in pediatric acute leukemias, which deserve to be further investigated in large-scale studies.

Keywords

Prognostic nutritional index Glasgow prognostic score prognostic index monocyte to lymphocyte neutrophil to lymphocyte platelet to lymphocyte ratios AML ALL

1. Introduction

In the pediatric population, acute leukemia is the most common cancer representing 25% of all malignancies, with acute lymphoblastic leukemia comprising 20% and acute myeloid leukemia 5% of all cases; the estimated US and worldwide incidence is 3.7–4.9 cases per 100,000 children aged 0–14 years [1]. In the past few decades, there has been a dramatic improvement in 5-year survival rates that now reach 90% for ALL, and 65–70% for AML [2].

An inflammatory environment for malignant tumors is assumed to modulate the growth of malignant cells, thus contributing to angiogenesis, mutation, cellular migration, and metastasis acting through the release of different cytokines and other inflammatory mediators [3].

Inflammation can result from trauma, ischemia-reperfusion injury, chemically-induced injury, and other causes unrelated to infectious organisms. Sterile inflammation can occur in acute pictures and is similar to microbial-induced inflammation in that it is marked by the recruitment of neutrophils and macrophages, in addition to the production of pro-inflammatory cytokines and chemokines, such as tumor necrosis factor- $\alpha$ (TNF- $\alpha$ ) and interleukin-1 (IL-1). The pathophysiology of this sterile inflammation is not completely clarified [4].

Different cohort studies had examined C-reactive protein (CRP) and the risk of malignancy, including a meta-analyses of the risk of incidence of colorectal [5], lung [6], ovarian [7], and breast cancers [8]. Although the results didn’t find significant associations, other case-control studies have reported low positive predictive values for inflammatory markers, including CRP, in a variety of malignancies, such as bladder and renal cancers, Hodgkin’s and non-Hodgkin’s lymphoma, and multiple myeloma [9, 10, 11, 12].

Systemic inflammation is increasingly being recognized as an important outcome determinant in cancer patients [13, 14, 15, 16], as cancer progression and prognosis are affected by the inflammatory response of the patient in the surrounding tumor microenvironment. Neutrophils, lymphocytes, and platelets are recognized to have an important role in carcinogenesis and hence in tumor progression [17].

In a prospective cohort study, the cancer incidence increased with rising of inflammatory markers in a dose-response relationship [18]. Out of 506 people with ESR $\geqslant$ 100, 13.6% developed malignancies within 1 year; for those 1983 individuals with CRP $\geqslant$ 100, 6.81% had cancer.

Recently, several peripheral blood-derived, inflam- mation-based scores and indices such as neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), Glasgow and modified Glasgow prognostic Scores (GPS), and prognostic nutritional index (PNI) have been assumed to be prognostic markers in cancer [19].

Many studies showed that preoperative inflammatory markers, including prognostic index (PI), were found in univariate analysis to be associated with overall and disease-free survival and were considered to be independent predictors of survival in gastric cancer [20, 21].

In a population-based study, the concentrations of different cytokines and inflammatory markers from 178 children aged 1–9 years with B-ALL were measured and compared to 178 matched controls [22]. The results supported the hypothesis that children with B-ALL were born with dysregulated immune functions.

A growing body of evidence indicates that cancer-related malnutrition, as expressed in GPS and PNI, has adverse effects on treatment and survival outcomes for many solid malignancies [23, 24, 25].

There is a limited data at the moment regarding the clinical value of inflammatory indices and malnutrition markers in children with acute leukemias. We have examined the usefulness of prognostic nutritional index (PNI), Glasgow prognostic score (GPS), Prognostic index (PI), monocyte to lymphocyte (MLR), neutrophil to lymphocyte (NLR), and platelet to lymphocyte (PLR) ratios to stratify patients as regards the response to induction therapy correlating them to different prognostic factors.

2. Patients and methods

This was a prospective study conducted at The South Egypt Cancer Institute and Assiut University Hospital, which was approved by the Assiut University Ethical Committee. The study involved a cohort 73 children diagnosed with acute leukemias. Patients with microbial-induced inflammation as detected by cultures were excluded.

The diagnosis of acute leukemia depends on peripheral hemogram assessment, morphologic bone marrow (BM) examination, cytochemical studies, and flow cytometry to identify the cell lineages and subsets.

CD56, a marker of natural killer (NK) cell differentiation, defines a rare subgroup representing about 3% of ALL which often displays other early T-cell antigens such as CD2 CD5, CD7 and sometimes cCD3 [26] as measured by flow cytometry.

At diagnosis, preliminary total and differential CBC, c-reactive protein (CRP), serum albumin (ALB) were used to calculate different inflammatory indicators including: neutrophil to lymphocyte ratio, calculated by dividing absolute neutrophilic count by absolute lymphocytic count; monocyte to lymphocyte ratio, calculated by dividing absolute monocytic count by absolute lymphocytic count; platelet to lymphocyte ratio, calculated by dividing absolute platelet count by absolute lymphocytic count; Glasgow prognostic score (GPS); defined by CRP (g/L) and serum ALB (g/L) as 0, 1, and 2. GPS score 0 is given for those with CRP $<$ 10 g/L and ALB $\geqslant$ 35 g/L; GPS score 1 for those with either CRP $>$ 10 g/L, or ALB $<$ 35 g/L; and GPS score 2 for those with CRP $>$ 10 g/L and ALB $<$ 35 g/L. The prognostic nutritional index (PNI) was calculated by adding serum ALB (g/L) plus total lymphocytic count per $\mu$ L after being multiplied by 0.005 (ALB $+$ TLC $\times$ 0.005); PNI value $\geqslant$ 50 is considered normal; PNI value $<$ 50– $\geqslant$ 45 indicates mild malnutrition; PNI value $<$ 45– $\geqslant$ 40 indicates moderate malnutrition; and PNI value $<$ 40 is considered indicator of serious malnutrition. Prognostic index (PI) was determined as follows: PI 2, both CRP $>$ 1.0 mg/dl and white cell count $>$ 11 $\times$ 10 ${}^{9}$ /l; PI 1, either CRP $>$ 1.0 mg/dl or white cell count $>$ 11 $\times$ 10 ${}^{9}$ /l, but not both; and PI 0, normal values for both CRP and white cell count.

The treatment for newly diagnosed ALL has 3 major phases, reflecting a reliance on multidrug regimens to avoid the development of resistance, with variable intensities according to each patient’s risk. Remission induction is the first phase of chemotherapy, lasting 4 to 6 weeks, that includes agents such as vincristine, corticosteroids, asparginase, and anthracycline. It is followed by consolidation, which consists of 8 weeks of multidrug chemotherapy given to eradicate the microscopic residual disease that persists after induction of remission, with higher risk patients receiving longer and more intensified regimens. The final and long-lasting phase is maintenance chemotherapy in which less intensified usage of cornerstone drugs such as methotrexate, and 6-mercaptopurine, proceeds for 2–3 years. In parallel, prophylaxis and treatment of CNS disease is achieved through intrathecal chemotherapy, administration of systemic chemotherapy able to penetrate the blood-brain barrier, and cranial irradiation.

Treatment of AML depends on induction with anthracycline and cytarabine for all patients in course 1 and for non-high risk (HR) patients in course 2, and mitoxantrone combined with cytarabine in high risk patients in course 2, different regimens are used in induction of remission including FLAG-Ida, ADE, this is followed by consolidation chemotherapy given in courses 3, 4 consisting of two courses of high dose Ara-C plus intrathecal chemotherapy for standard risk AML, while FLAG-Ida is given in course 3 followed by high dose Ara-C and L-Asparginase for high risk patients if stem cell transplantation is not available. HR patients are sent for stem cell transplantation at specialized centers if feasible.

ALL remission has been defined by morphology ( $<$ 5% lymphoblasts in bone marrow with hematologic count recovery). Remission status after induction chemotherapy has critical implications because it can affect patient eligibility for salvage regimens or clinical trials in experimental agents.

According to the revised recommendations of the International Working Group (IWG), the criteria used to define complete remission are: $<$ 5% blasts in the bone marrow, no blasts with Auer rods, normal maturation of all cellular components in the bone marrow, absence of extramedullary disease (e.g., CNS, soft tissue disease), neutrophils $\geqslant$ 1,000/ $\mu$ L, platelets $\geqslant$ 100,000/ $\mu$ L, and transfusion independent [27].

2.1 Statistics

Descriptive statistics including percentages, mean, median, standard error were used in this study. Inferential statistics utilized to determine data significance included independent sample $t$ -test, one way Anova, and Chi-square. A Roc curve was used to find a cutoff value for MLR. Pearson correlation was employed to determine the degree of association between scale variables, while Phi test was used to find the degree of associations between nominal variables. Cramer’s V was chosen to determine the degree and strength of the association between nominal and ordinal variables. Logistic regression was used for multivariate analysis. All data were analyzed using SPSS version 20, and were considered significant at $p$ -value $<$ 0.05.

3. Results

Seventy-three children with acute leukemia were recruited including 48 patients with ALL and 25 patients with AML; the median age was 9 years, male predominance was observed, Table 1. Different laboratory parameters were illustrated in Table 2.

Table 1
Demographic data of 73 children with acute leukemia

Data	Descriptive
Age mean $\pm$ SE	8.74 $\pm$ 0.6 ys
Median	9 ys
Min-max	1–18 ys
Male/female	42/31
Leukemia
ALL	48 (65.8%)
AML	25 (34.2%)

Table 2

Laboratory parameters of 73 studied patients with acute leukemia

CBC parameter	Number (%)
Leucocytic count (mean $\pm$ SE)	63.4 $\pm$ 11.95 $\times$ 10 ${}^{9}$ /L
Normal	25 (34.2%)
Leucopenia	15 (20.5%)
Leukocytosis	33 (45.2%)
Lymphocytic count	5.5 $\pm$ 1.1 $\times$ 10 ${}^{9}$ /L
Normal	17 (23.3%)
Lymphopenia	39 (53.4%)
Lymphocytosis	17 (23.3%)
HB	9.6 $\pm$ 0.22 g/ $\mu$ L
$<$ 11 g/ $\mu$ l	49 (62.1%)
$>$ 11 g/ $\mu$ l	24 (32.9%)
Platelet count	67.7 $\pm$ 10.3 $\times$ 10 ${}^{9}$ /L
Normal	9 (12.3%)
Thrombocytopenia	63 (86.3%)
Thrombocytosis	1 (1.4%)
Monocytes	2.41 $\pm$ 2.03 $\times$ 10 ${}^{9}$ /L
Basophils	0.39 $\pm$ 1.2 $\times$ 10 ${}^{9}$ /L
Esinophils	0.99 $\pm$ 0.36 $\times$ 10 ${}^{9}$ /L
Neutrophils	16.01 $\pm$ 2.4 $\times$ 10 ${}^{9}$ /L
CRP	61.12 $\pm$ 3.6
ALB	35.98 $\pm$ 0.68
BM blasts mean $\pm$ SE	56.78 $\pm$ 4.1
Peripheral blasts mean $\pm$ SE	40.53 $\pm$ 4
CD56	19.35 $\pm$ 3.5%
CD34	16.45 $\pm$ 3.1%
ESR 1 ${}^{\text{st}}$ hour	59.2 $\pm$ 4.6 seconds
2 ${}^{\text{nd}}$ hour	89.88 $\pm$ 6.2 seconds

HB; hemoglobin, CRP; c-reactive protein, ALB; albumin, BM; bone marrow, ESR; erythrocyte sedimentation rate.

The mean BM blasts was 56.78 $\pm$ 4.1, while the mean peripheral blasts was 40.5 $\pm$ 4. The mean value $\pm$ SE of CRP, ALB, CD34, and CD56 were provided in Table 3.

Eighteen out of forty-eight ALL patients achieved complete remission to induction therapy, while 9/25 patients with AML achieved complete remission, Table 3.

3.1 Prognostic nutritional index (PNI) and remission

Significantly, all patients without remission were shown to have PNI $<$ 40 and PNI 40– $<$ 45, compared to 9 patients in remission having PNI 45– $<$ 50 and PNI $\geqslant$ 50 ( $p<$ 0.0001), Fig. 1.

3.2 Glasgow prognostic score (GPS) and remission

By the same token, all patients without remission were found to have GPS 1 and 2, while most patients in remission (21/27) were found to have GPS 0 and 1 ( $p=$ 0.001), Fig. 2.

3.3 Relation of PI to induction results

Although both CRP and white blood cells were elevated in acute leukemias, PI had no relation to response to induction chemotherapy, Fig. 3.

3.4 Relations of NLR, MLR, and PLR and induction results

The mean NLR for patients who achieved complete remission was 3.5 $\pm$ 0.82, while for those without remission it was 7.0 $\pm$ 0.92, $p=$ 0.005, Fig. 4. Furthermore, the mean MLR for the former patients was 1.8 $\pm$ 1.2, while for the later patients it was 2.6 $\pm$ 0.57, $p=$ 0.5. Additionally, the mean PLR for those in remission was 87.9 $\pm$ 32.9, and 99.1 $\pm$ 36.3, $p=$ 0.8 in absence of remission, Fig. 4.

Table 3
Response to induction therapy

Diagnosis	Induction results		Total
	Remission	Not remission
ALL
Count	18	30	48
% within diagnosis	37.5%	62.5%	100%
AML
Count	9	16	25
% within diagnosis	36.0%	64.0%	100%
Total
Count	27	46	73
% within diagnosis	37.0%	63.0%	100%

Figure 1.

Induction results according to different PNI categories, ( $p<$ 0.0001), Chi ${}^{2}$ test for significance.

Figure 2.

Induction results according to different GPS categories, $p=$ 0.001, Chi ${}^{2}$ test for significance.

Figure 3.

Relation between PI categories and induction results, Cramer’s V value $=$ 0.21, $p=$ 0. 2.

Figure 4.

Difference of mean NLR according to induction results.

Figure 5.

The ROC curve for MLR and PLR showed AUC for MLR $=$ 0.272 $\pm$ 0.1 (95% CI $=$ 0.146–0.399) that was significant, $p=$ 0.001. The cutoff value of MLR of highest detected sensitivity and specificity was 0.45. Conversely, we didn’t find a cutoff value for PLR because AUC $=$ 0.533 $\pm$ 0.1 (95% CI $=$ 0.397–0.669), $p=$ 0.6.

3.5 ROC curve of MLR and PLR

A ROC curve was built to determine the cutoff values for MLR and PLR allowing adequate discrimination between patients who achieved complete remission and those without complete remission, Fig. 5.

Table 4
Relation between MLR categories and treatment response

	No remission	Remission	Total
MLR $\leqslant$ 0.45	8 (29.6%)	19 (70.4%)	27 (100%)
MLR $>$ 0.45	38 (82.6%)	8 (17.4%)	46 (100%)
Phi value	$+$ 0.53
$p$ -value	$<$ 0.0001

Data analyzed by Phi test.

Table 5

Differences in ALL and AML regarding inflammatory indices

Inflammatory markers	ALL (N $=$ 48)	AML (N $=$ 25)
ESR at 1 ${}^{\text{st}}$ hour ESR at 2 ${}^{\text{nd}}$ hour	53.6 $\pm$ 5.5 89.8 $\pm$ 7.4	55.4 $\pm$ 8.5 90.1 $\pm$ 11.3
ALB	36.9 $\pm$ 0.76	34.2 $\pm$ 1.33
CRP	6.1 $\pm$ 0.5	6.1 $\pm$ 0.6
NLR	5.3 $\pm$ 0.76	6.4 $\pm$ 1.4
MLR	2.8 $\pm$ 0.8	1.17 $\pm$ 0.4
PLR	89.5 $\pm$ 34.3	115.0 $\pm$ 37.2
PI $=$ 0 PI $=$ 1 PI $=$ 2	1 (2.1%) 22 (45.8%) 25 (52.1%)	– 13 (52%) 12 (48%)
GPS 0 1 2	03 (6.3%) 23 (47.9%) 22 (45.8%)	01 (4%) 12 (48%) 12 (48%)
PNI Normal nutrition Mild malnutrition Moderate malnutrition Serious malnutrition	02 (4.2%) 4 (8.3%) 14 (29.2%) 28 (58.3%)	01 (4%) 2 (8%) 4 (16%) 18 (72%)
MLR categories MLR $\leqslant$ 0.45 MLR $>$ 0.45	017 (35.4%) 31 (69.6%)	010 (40%) 15 (60%)

Data expressed as mean $\pm$ SE, number and percentages.

Table 6

Univariate analysis of inflammatory indices in treatment results of ALL patients

	ALL
	Remission (18/48)	No remission (30/48)	$p$ -value
Age (mean $\pm$ SE)	6.4 $\pm$ 1.1	8.1 $\pm$ 0.8	0.21
WBCs	60.9 $\pm$ 26.6	103.3 $\pm$ 22.1	0.23
CRP (mean $\pm$ SE)	3.1 $\pm$ 0.6	7.9 $\pm$ 0.4	0.0001
ALB (mean $\pm$ SE)	36.8 $\pm$ 1.6	37.0 $\pm$ 0.8	0.9
NLR (mean $\pm$ SE)	3.15 $\pm$ 1.1	6.6 $\pm$ 1.0	0.024
MLR (mean $\pm$ SE)	2.5 $\pm$ 1.8	3.03 $\pm$ 0.8	0.8
PLR (mean $\pm$ SE)	32.6 $\pm$ 12.02	115.7 $\pm$ 53.9	0.2
PI $=$ 0 $=$ 1 $=$ 2	01 (5.6%) 8 (44.4%) 9 (50%)	00% 14 (46.7%) 18 (53.3%)	0.4
GPS 0 1 2	03 (16.7%) 11 (61.1%) 4 (22.2%)	00% 12 (40%) 18 (60%)	0.008
PNI Normal Mild Moderate Serious	02 (11.1%) 4 (22.2%) 4 (22.2%) 8 (44.4%)	00% 0% 10 (33.3%) 20 (66.7%)	0.01
MLR MLR $\leqslant$ 0.45 MLR $>$ 0.45	011 (61.1%) 7 (38.9%)	06 (20%) 24 (80%)	0.004

Data analyzed by using independent sample $t$ -test, Ch ${}^{2}$ test.

3.6 Relation between MLR categories and treatment response

Patients without remission had a significantly higher MLR than those in remission as illustrated in Table 4.

Figure 6.

Mean CRP value $\pm$ SE according to treatment results in 73 patients with acute leukemia, independent sample $t$ -test, $p<$ 0.0001.

3.7 Relation of serum CRP, albumin and ESR to treatment results

Significant differences were observed in the mean CRP values obtained from patients who attained complete remission (CRP $=$ 34.7 $\pm$ 5.1), and those without remission who had CRP $=$ 76.7 $\pm$ 3.1, $p<$ 0.0001, Fig. 6.

No significant differences were found in the mean $\pm$ SE of serum albumin, and 1 ${}^{\text{st}}$ hour and 2 ${}^{\text{nd}}$ hour ESR between those in complete remission and those with no remission ( $p=$ 0.12, $p=$ 0.09, and $p=$ 0.1, respectively).

Table 7
Univariate analysis of inflammatory indices in the treatment results of AML patients

	AML
	Remission (9/25)	No remission (16/25)	$p$ -value
Age (mean $\pm$ SE)	9.4 $\pm$ 2.03	12.1 $\pm$ 1.0	0.2
WBCs	8.2 $\pm$ 2.2	22.3 $\pm$ 5.8	0.036
CRP (mean $\pm$ SE)	4.1 $\pm$ 0.94	7.2 $\pm$ 0.6	0.007
ALB (mean $\pm$ SE)	38.6 $\pm$ 1.7	31.8 $\pm$ 1.5	0.001
NLR (mean $\pm$ SE)	4.1 $\pm$ 1.1	7.8 $\pm$ 2.0	0.1
MLR (mean $\pm$ SE)	0.22 $\pm$ 0.1	1.71 $\pm$ 0.6	0.026
PLR (mean $\pm$ SE)	198.4 $\pm$ 87.5	68.03 $\pm$ 27.4	0.2
PI $=$ 0 $=$ 1 $=$ 2	00 7 (77.8%) 2 (22.2%)	00 6 (37.5%) 10 (62.5%)	0.05
GPS 0 1 2	01 (11.1%) 6 (66.7%) 2 (22.2%)	00% 6 (37.5%) 10 (62.5%)	0.09
PNI Normal Mild Moderate Serious	01 (11.1%) 2 (22.2%) 4 (44.4%) 2 (22.2%)	00% 0% 0% 16 (100%)	0.001
MLR MLR $\leqslant$ 0.45 MLR $>$ 0.45	08 (88.9%) 1 (11.1%)	02 (12.5%) 14 (87.5%)	0.0001

Data analyzed by using independent sample $t$ -test, Chi ${}^{2}$ test.

Table 8

Logistic regression of inflammatory indices in ALL patients

Logistic regression of inflammatory indices in AML patients ${}^{\text{a}}$
Variable	B	Wald	Odds ratio	$p$ -value
GPS	1.7	7.6	5.52	0.006
PNI	4.5	3.8	89.8	0.05
MLR groups	5.6	3	253.6	0.09
CRP	1.3	2.7	3.7	0.1
NLR	0.9	2.3	2.5	0.1
MLR groups	4.03	9.6	56.0	0.002
CRP	0.49	5.4	1.63	0.021
ALB	$-$ 0.22	4.5	0.8	0.034

Data were analyzed by using logistic regression with the method forward stepwise (likelihood ratio) method. ${}^{\text{a}}$ : Estimation of PNI, MLR, WBCs, PI were omitted because the parameters’ estimates changed by less than 0.001, Wald $=$ 0, and Nagelkerke’s R ${}^{2}$ $=$ 1 (i.e. the variation in treatment response was completely explained by the variation in PNI, MLR, WBCs, and PI or complete fit, so Wald attained a value of 0, i.e. not important).

We found similarity between both ALL and AML as regards the distribution of inflammatory markers and indices in ALL and AML with the majority of cases showing higher scores of PI, GPS, PNI, and MLR $>$ 0.45, Table 5.

Patients in complete remission had significantly lower NLR ( $p=$ 0.024), lower GPS ( $p=$ 0.008), better PNI ( $p=$ 0.01), and lower MLR group ( $p=$ 0.004), There was not any significant difference in MLR ( $p=$ 0.8), PLR ( $p=$ 0.2), and PI ( $p=$ 0.4) compared to those without complete remission, Table 6.

Differently from ALL, patients with AML achieving remission were found to have significantly lower MLR mean $\pm$ SE ( $p=$ 0.026), and better PI score ( $p=$ 0.05) than those without complete remission, Table 7.

Table 9

Correlations between different prognostic features of acute leukemia and inflammatory indices

		BM blasts	CD34	CD56	CRP	ALB	NLR	MLR	PLR	pb. blasts
BM blasts	$r$	NR	0.122	$-$ 0.122	0.213	$-$ 0.017	0.134	$-$ 0.096	0.057	0.340
	$P$		0.304	0.303	0.071	0.888	0.257	0.418	0.630	0.003
CD34	$r$	0.122	NR	0.199	0.229	0.270	$-$ 0.158	$-$ 0.022	$-$ 0.222	0.178
	$P$	0.304		0.091	0.050	0.021	0.181	0.855	0.049	0.131
CD56	$r$	$-$ 0.122	0.199	NR	0.057	0.115	$-$ 0.089	$-$ 0.092	$-$ 0.060	0.113
	$P$	0.303	0.091		0.634	0.334	0.455	0.438	0.613	0.343
CRP	$r$	0.213	0.229	0.057	NR	$-$ 0.101	$-$ 0.035	0.127	$-$ 0.108	0.256
	$P$	0.071	0.050	0.634		0.394	0.772	0.286	0.362	0.029
ALB	$r$	$-$ 0.017	0.270	0.115	$-$ 0.101	NR	$-$ 0.080	0.034	$-$ 0.091	0.195
	$P$	0.888	0.021	0.334	0.394		0.504	0.774	0.446	0.099
NLR	$r$	0.134	$-$ 0.158	$-$ 0.089	$-$ 0.035	$-$ 0.080	NR	$-$ 0.022	0.318	$-$ 0.317
	$P$	0.257	0.181	0.455	0.772	0.504		0.850	0.006	0.006
MLR	$r$	$-$ 0.096	$-$ 0.022	$-$ 0.092	0.127	0.034	$-$ 0.022	NR	$-$ 0.073	0.005
	$P$	0.418	0.855	0.438	0.286	0.774	0.850		0.539	0.965
PLR	$r$	0.057	$-$ 0.222	$-$ 0.060	$-$ 0.108	$-$ 0.091	0.318	$-$ 0.073	NR	$-$ 0.190
	$P$	0.630	0.049	0.613	0.362	0.446	0.006	0.539		0.108
pb. blast	$r$	0.340	0.178	0.113	0.256	0.195	$-$ 0.317	0.005	$-$ 0.190	NR
	$P$	0.003	0.131	0.343	0.029	0.099	0.006	0.965	0.108

Data analyzed by Pearson correlation, $r$ ; correlation coefficient, pb; peripheral blood, BM; bone marrow.

Figure 7.

Differences in the mean $\pm$ SE of prognostic factors of acute leukemia in PNI groups, one way Anova, $P<$ 0.0001.

Figure 8.

Different prognostic features of acute leukemia according to MLR categories, independent sample $t$ -test, $p=$ 0.023.

The independent predictors of treatment response for ALL were GPS (Wald $=$ 7.6), and PNI (Wald $=$ 3.8), Table 8. In the case of AML, MLR categories, CRP, and albumin were considered independent predictors for treatment response, Table 8.

3.8 Relation of different prognostic features to inflammatory indices

Significant positive correlations were found between CD34 and CRP ( $r=$ 0.229, $p=$ 0.05) and CD34 and ALB ( $r=$ 0.270, $p=$ 0.021), while CD34 exhibited significant negative association with PLR ( $r=-$ 0.222, $p=$ 0.049). CRP correlated significantly with peripheral blasts ( $r=$ 0.256, $p=$ 0.029), Table 9.

3.9 Relations of PNI, GPS, and MLR categories to adverse prognostic factors of 73 children with acute leukemias

Significant differences were found in the mean values of CD34 among different PNI groups, with higher mean CD34 in PNI $<$ 45– $\geqslant$ 40, $p=$ 0.008; additionally, significant progressive increase was detected in the serum level of CRP and peripheral blasts with lowering of PNI ( $p=$ 0.009, and $p<$ 0.0001, respectively), as shown in Fig. 7.

Patients with MLR $>$ 0.45 were found to have significantly higher BM blasts, CD34, CRP, and peripheral blasts ( $p=$ 0.005, $p=$ 0.014, $p<$ 0.0001, and $p=$ 0.023, respectively) compared to those with MLR $\leqslant$ 0.45, while there was no significant differences with CD56 and serum ALB level ( $p=$ 0.3, and $p=$ 0.9 respectively), as illustrated in Fig. 8.

4. Discussion

Pediatric acute leukemia has moved from being an incurable fatal disease to a disease with a cure rate exceeding 75% [28], and 5-year survival rates approaching 90% in ALL and 70% in AML.

Management of childhood acute leukemia in Egypt suffers from multiple challenges, including limited health resources, inadequate financial support, treatment abandonment, decreased adherence to standardized protocols, and limited diagnostic measures for adequate risk stratifications.

In the current study, we focused on simple and easily applicable measurements that might correlate with prognosis, including GPS, MLR, NLR, and CRP, which were found to be significantly higher among those children unable to achieve complete remission, as well as PNI whose increase was associated to remission.

PNI was originally proposed to assess the perioperative surgical risk and immunological status in gastrointestinal malignancies, with values $<$ 40 being associated with increased postoperative mortality [29]. Thereafter, low PNI was found to be a prognostic marker associated with poor overall survival (OS) in many solid tumors, and high TNM stages in gastric and colorectal malignancies [30].

On the other hand, little is known about the role of PNI in hematologic malignancies. Thus, Go et al. reported that PNI $<$ 40 associated with lower complete response rate, high treatment toxicities, and frequent treatment discontinuation [31]. As a consequence, it was considered as an independent prognostic factor for OS in multivariate Cox regression analysis of patients with diffuse large B-cell lymphoma (DLBCL). In agreement to the previous study, our results indicated that low PNI was associated with decreased complete remission rate in pediatric acute leukemia, supporting the role of malnutrition in disease progression and poor survival. However, additional good quality prospective studies are needed to draw a precise conclusion regarding its impact on acute leukemias.

Systemic inflammation promotes angiogenesis, proliferation, and survival of tumor cells as well as tumor progression and metastases. Inflammation-based prognostic scores, including platelet-to-lymphocyte ratio, neutrophil-lymphocyte ratio, and Glasgow Prognostic Score- which combines both C-reactive protein as a marker of inflammation and albumin as a marker of nutrition-, have been attributed prognostic value for patients with cancer [32, 33]. Furthermore, the role of GPS remains to be defined in hematologic malignancies, although a retrospective evaluation of 252 patients with DLBCL demonstrated that high GPS, high NLR, high PLR and low PNI in univariate analysis, were all associated with poor overall survival ( $p<$ 0.05) and event-free survival ( $p<$ 0.05), while in multivariate analysis indicated that only GPS remained an independent prognostic predictor in DLBCL (HR $=$ 1.781, 95% CI $=$ 1.065–2.979, $p=$ 0.028).

Interleukin-6 is implicated in quite a number of acute and chronic inflammation-related diseases through promotion of positive acute phase reactants (e.g., CRP) and negative reactants (e.g., ALB), which participate in tumor growth and invasion. Previous studies reported that IL-6 was associated with poor survival in DLBCL [34], and therefore, the shutting-off of inflammatory cytokines greatly interfere with tumor progression. Likewise, patients with higher CRP had lower remission rates in both ALL and AML ( $p=$ 0.0001, and $p=$ 0.007, respectively), while only AML patients with high albumin had higher remission rate ( $P=$ 0.01).

Although GPS is a good index of inflammation and malnutrition and has suggested prognostic value in many solid malignancies [35, 36, 37, 38], studies discussing its role in hematologic malignancies are lacking. To our knowledge, the current report is the only one to address its prognostic significance in pediatric acute leukemias.

Moreover, systemic inflammation induces the release of inhibitory cytokines, including transforming growth factor- $\beta$ , and interleukin-10, resulting in impaired lymphocyte cytolytic activity and lymphocytopenia, an effect correlated with adverse outcomes in hematologic malignancies, especially in non-Hodgkin’s lymphoma [39, 40]. Neutrophils promote tumor progression through induction of DNA damage, inhibition of apoptosis, and promotion of angiogenesis. NLR has been used to compare the neutrophil-based inflammatory response to cancer immune response by lymphocytes in DLBCL. Patients with NLR $<$ 2.32 were found to achieve better OS and progression-free survival (PFS) than those with NLR $\geqslant$ 2.32 [41].

Patients with multiple myeloma ineligible for autologous BMT and treated with melphalan, bortezomib and prednisone with NLR $\geqslant$ 1.51 experienced lower complete response rate, and significantly lower 2-year OS. Hence, high NLR was an independent poor prognostic factor for OS in these cases (HR $=$ 2.217, $p=$ 0.0458) [42]. Consistent with previous studies, our results demonstrated that patients entered in complete remission after induction therapy had a significantly lower NLR than those without remission. However, when we analyzed this ratio on ALL and AML separately, we didn’t find a prognostic value owing to small sample size.

Inflammatory cytokines such as IL-3, IL-6, and IL-10 produced in different cancers that cause thrombosis and enhance tumor growth and motility, contributing to metastasis. It has been previously shown that patients with PLR $<$ 150 experienced improved OS and PFS as compared to those with PLR $\geqslant$ 150 in DLBCL [41]. Our results, however, didn’t agree with the latter findings possibly due to small sample size and heterogeneity of the studied acute leukemias.

Because of low cost, easy reproducibility, and wide availability of these inflammatory markers and indices, they were extensively studied in many cancers to evaluate their prognostic impact and their capacity to provide an easy and rapid outcome prediction. However, the best marker and the optimal cutoff value for each type of cancer are still debatable.

The relation between severity of symptomatic burden and several inflammatory markers was analyzed in a prospective study involving 95 patients with AML before and during the induction phase of the treatment. This study concluded that the change in severity of symptoms during therapy was not translated significantly into a change in levels of inflammatory markers suggesting that the absolute levels of inflammatory markers during induction treatment was more important than their relative change over time [43].

Serum albumin was considered an independent prognostic factor in AML, given that patients with levels $\geqslant$ 30 g/L were more associated with remission and favorable survival than those with albumin $<$ 30 g/L [44]. Although, we identified a positive significant correlation between albumin and CD34 ( $r=+$ 0.270, $p=$ 0.021), our results failed to find the association between albumin and remission/survival, what might be attributed to the small number of children with acute leukemia included in our study.

The current study is limited by its single arm nature with relatively small number of analyzed patients, and also by of the acute leukemias heterogeneity. Pretreatment measurement of inflammatory indicators is not currently a standard practice in acute leukemia and should not be overestimated until a properly designed prospective trial with the correct sample size proves the potential role of these markers.

5. Conclusion

Pretreatment CRP, ALB, PNI, GPS, NLR, MLR, and PLR are promising prognostic markers of remission in pediatric acute leukemias that need to be further investigated in large scale studies.

Footnotes

Conflict of interest

None.

References

Pui

C.H.

Mullighan

C.G.

Evans

W.E.

and Relling

M.V.

, Pediatric acute lymphoblastic leukemia: Where are we going and how do we get there? Blood 120(6) (2012), 1165–1174.

Siegel

R.L.

Miller

K.D.

and Jemal

, Cancer statistics, 2018, CA: A Cancer Journal for Clinicians 68(1) (2018), 7–30.

Balkwill

, Cancer and the chemokine network, Nature Reviews Cancer 4(7) (2004), 540–550.

Chen

G.Y.

and Nunez

, Sterile inflammation: Sensing and reacting to damage, Nature Reviews Immunology 10(12) (2010), 826–837.

Zhang

Liu

and Zhou

, Circulating levels of C-reactive protein, interleukin-6 and tumor necrosis factor-alpha and risk of colorectal adenomas: A meta-analysis, Oncotarget 7(39) (2016), 64371–64379.

Zhou

Liu

Wang

Z.M.

and Xi

, C-reactive protein, interleukin 6 and lung cancer risk: A meta-analysis, PloS One 7(8) (2012), e43075.

Zeng

Wei

Yeoh

Zhang

Ren

Z.-F.

Colditz

G.A.

et al., Inflammatory Markers of CRP, IL6, TNFα, and soluble TNFR2 and the risk of ovarian cancer: A meta-analysis of prospective studies, Cancer Epidemiology Biomarkers & Prevention 25(8) (2016), 1231–1239.

Guo

Liu

Zhang

Chen

Zhang

Quan

et al., C-reactive protein and risk of breast cancer: A systematic review and meta-analysis, Scientific Reports 5(1) (2015), 10508.

Shephard

Neal

Rose

Walter

Litt

and Hamilton

, Quantifying the risk of myeloma from symptoms reported in primary care patients: A large case-control study using electronic records, Br J Gen Pract 65 (2014), 631.

10.

Shephard

E.A.

Stapley

Neal

R.D.

Rose

Walter

F.F.M.

and Hamilton

W.T.

, Clinical features of bladder cancer in primary care, Br J Gen Pract 62(602) (2012), e598–e604.

11.

Shephard

E.A.

Neal

R.D.

Rose

P.W.

Walter

F.M.

and Hamilton

W.T.

, Quantifying the risk of Hodgkin lymphoma in symptomatic primary care patients aged ⩾ 40 years: A case-control study using electronic records, Br J Gen Pract 65(634) (2015), e289–e94.

12.

Shephard

E.A.

Neal

R.D.

Rose

P.W.

Walter

F.M.

and Hamilton

W.T.

, Quantifying the risk of non-Hodgkin lymphoma in symptomatic primary care patients aged ⩾ 40 years: A large case-control study using electronic records, Br J Gen Pract 65(634) (2015), e281–e8.

13.

Hanahan

and Weinberg

R.A.

, The hallmarks of cancer, Cell 100(1) (2000), 57–70.

14.

Hanahan

and Weinberg

R.A.

, Hallmarks of cancer: The next generation, Cell 144(5) (2011), 646–674.

15.

McMillan

D.C.

, The systemic inflammation-based Glasgow Prognostic Score: A decade of experience in patients with cancer, Cancer Treatment Reviews 39(5) (2013), 534–540.

16.

Roxburgh

C.S.

and McMillan

D.C.

, Role of systemic inflammatory response in predicting survival in patients with primary operable cancer, Future Oncology (London, England) 6(1) (2010), 149–163.

17.

Mollinedo

, Neutrophil degranulation, plasticity, and cancer metastasis, Trends in Immunology 40(3) (2019), 228–242.

18.

Watson

Salisbury

Banks

Whiting

and Hamilton

, Predictive value of inflammatory markers for cancer diagnosis in primary care: A prospective cohort study using electronic health records, British Journal of Cancer 120(11) (2019), 1045–1051.

19.

Guthrie

G.J.

Charles

K.A.

Roxburgh

C.S.

Horgan

P.G.

McMillan

D.C.

and Clarke

S.J.

, The systemic inflammation-based neutrophil-lymphocyte ratio: Experience in patients with cancer, Critical Reviews in Oncology/Hematology 88(1) (2013), 218–230.

20.

Pan

Q.X.

Z.J.

Zhang

J.H.

Wang

C.R.

and Ke

S.Y.

, A comparison of the prognostic value of preoperative inflammation-based scores and TNM stage in patients with gastric cancer, OncoTargets and Therapy 8 (2015), 1375–1385.

21.

Toyokawa

Muguruma

Tamura

Sakurai

Amano

Kubo

et al., Comparison of the prognostic impact and combination of preoperative inflammation-based and/or nutritional markers in patients with stage II gastric cancer, Oncotarget 9(50) (2018), 29351–29364.

22.

Soegaard

S.H.

Rostgaard

Skogstrand

Wiemels

J.L.

Schmiegelow

and Hjalgrim

, Neonatal inflammatory markers are associated with childhood b-cell precursor acute lymphoblastic leukemia, Cancer Research 78(18) (2018), 5458–5463.

23.

Morgan

T.M.

Tang

Stratton

K.L.

Barocas

D.A.

Anderson

C.B.

Gregg

J.R.

et al., Preoperative nutritional status is an important predictor of survival in patients undergoing surgery for renal cell carcinoma, European Urology 59(6) (2011), 923–928.

24.

Esfahani

Ghoreishi

Abedi Miran

Sanaat

Ostadrahimi

Eivazi Ziaei

et al., Nutritional assessment of patients with acute leukemia during induction chemotherapy: Association with hospital outcomes, Leukemia & Lymphoma 55(8) (2014), 1743–1750.

25.

Xue

Tan

Yang

and Wei

, Impact of the preoperative prognostic nutritional index on survival outcomes in upper tract urothelial carcinomas, Cancer Medicine 8(6) (2019), 2971–2978.

26.

Vardiman

J.W.

Thiele

Arber

D.A.

Brunning

R.D.

Borowitz

M.J.

Porwit

et al., The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: Rationale and important changes, Blood 114(5) (2009), 937–951.

27.

Cheson

B.D.

Bennett

J.M.

Kopecky

K.J.

Buchner

Willman

C.L.

Estey

E.H.

et al., Revised recommendations of the international working group for diagnosis, standardization of response criteria, treatment outcomes, and reporting standards for therapeutic trials in acute myeloid leukemia, Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology 21(24) (2003), 4642–4649.

28.

Pui

C.H.

Cheng

Leung

Rai

S.N.

Rivera

G.K.

Sandlund

J.T.

et al., Extended follow-up of long-term survivors of childhood acute lymphoblastic leukemia, The New England Journal of Medicine 349(7) (2003), 640–649.

29.

Onodera

Goseki

and Kosaki

, Prognostic nutritional index in gastrointestinal surgery of malnourished cancer patients, Nihon Geka Gakkai Zasshi 85(9) (1984), 1001–1005.

30.

Sun

Chen

and He

, The prognostic significance of the prognostic nutritional index in cancer: A systematic review and meta-analysis, Journal of Cancer Research and Clinical Oncology 140(9) (2014), 1537–1549.

31.

S.I.

Park

Kang

M.H.

Kim

H.G.

Kim

H.R.

and Lee

G.W.

, Clinical impact of prognostic nutritional index in diffuse large B cell lymphoma, Annals of Hematology 98(2) (2019), 401–411.

32.

Sanchez-Salcedo

, de-Torres

Martinez-Urbistondo

Gonzalez-Gutierrez

Berto

Campo

et al., The neutrophil to lymphocyte and platelet to lymphocyte ratios as biomarkers for lung cancer development, Lung Cancer (Amsterdam, Netherlands) 97 (2016), 28–34.

33.

Chen

Chang

Kang

Han

et al., Modified Glasgow prognostic score might be a prognostic factor for hepatocellular carcinoma: A meta-analysis, Panminerva Med 59(4) (2017), 302–307.

34.

Uskudar Teke

Gunduz

Akay

O.M.

Bal

and Gulbas

, Are the high serum interleukin-6 and vascular endothelial growth factor levels useful prognostic markers in aggressive non-hodgkin lymphoma patients? Turkish Journal of Haematology: Official Journal of Turkish Society of Haematology 32(1) (2015), 21–28.

35.

Yotsukura

Ohtsuka

Kaseda

Kamiyama

Hayashi

and Asamura

, Value of the glasgow prognostic score as a prognostic factor in resectable non-small cell lung cancer, Journal of Thoracic Oncology: Official Publication of the International Association for the Study of Lung Cancer 11(8) (2016), 1311–1318.

36.

Nishida

Nakamura

Haraga

Ogawa

Kusumoto

Seki

et al., The glasgow prognostic score determined during concurrent chemoradiotherapy is an independent predictor of survival for cervical cancer, International Journal of Gynecological Cancer: Official Journal of the International Gynecological Cancer Society 25(7) (2015), 1306–1314.

37.

Shiba

Misawa

Fujiwara

Futagawa

Furukawa

Haruki

et al., Glasgow prognostic score predicts outcome after surgical resection of gallbladder cancer, World Journal of Surgery 39(3) (2015), 753–758.

38.

Moriwaki

Ishige

Araki

Yoshida

Nishi

Sato

et al., Glasgow prognostic score predicts poor prognosis among advanced biliary tract cancer patients with good performance status, Medical Oncology (Northwood, London, England) 31(11) (2014), 287.

39.

Oki

Yamamoto

Kato

Kuwatsuka

Taji

Kagami

et al., Low absolute lymphocyte count is a poor prognostic marker in patients with diffuse large B-cell lymphoma and suggests patients’ survival benefit from rituximab, European Journal of Haematology 81(6) (2008), 448–453.

40.

Salazar-Onfray

López

M.N.

and Mendoza-Naranjo

, Paradoxical effects of cytokines in tumor immune surveillance and tumor immune escape, Cytokine & Growth Factor Reviews 18(1) (2007), 171–182.

41.

Wang

Sun

Shi

Jiang

et al., Prognostic significance of pretreatment neutrophil/lymphocyte ratio and platelet/lymphocyte ratio in patients with diffuse large B-cell lymphoma, BioMed Research International 2018 (2018), 9651254.

42.

Lee

G.W.

Park

S.W.

S.I.

Kim

H.G.

Kim

M.K.

Min

C.K.

et al., The derived neutrophil-to-lymphocyte ratio is an independent prognostic factor in transplantation ineligible patients with multiple myeloma, Acta Haematologica 140(3) (2018), 146–156.

43.

Lacourt

T.E.

Kavelaars

Galloway-Pena

J.R.

Sahasrabhojane

P.V.

Shah

N.D.

Futreal

et al., Associations of inflammation with symptom burden in patients with acute myeloid leukemia, Psychoneuroendocrinology 89 (2018), 203–208.

44.

Filliatre-Clement

Broseus

Muller

Hosseini

Rotonda

Schirmer

et al., Serum albumin or body mass index: Which prognostic factor for survival in patients with acute myeloblastic leukaemia? Hematological Oncology 37(1) (2019), 80–84.

Role of pretreatment inflammatory indicators in pediatric acute leukemias;where do we stand? A prospective cohort study

Abstract

BACKGROUND AND AIM:

PATIENTS AND METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Patients and methods

2.1 Statistics

3. Results

Table 1 Demographic data of 73 children with acute leukemia

3.2 Glasgow prognostic score (GPS) and remission

3.3 Relation of PI to induction results

3.4 Relations of NLR, MLR, and PLR and induction results

Table 3 Response to induction therapy

Table 4 Relation between MLR categories and treatment response

Table 7 Univariate analysis of inflammatory indices in the treatment results of AML patients

3.9 Relations of PNI, GPS, and MLR categories to adverse prognostic factors of 73 children with acute leukemias

4. Discussion

5. Conclusion

Footnotes

Conflict of interest

References

Table 1
Demographic data of 73 children with acute leukemia

Table 3
Response to induction therapy

Table 4
Relation between MLR categories and treatment response

Table 7
Univariate analysis of inflammatory indices in the treatment results of AML patients