Evaluation of serum and urine fetuin-A levels in children with acute lymphoblastic leukemia during and after high-dose methotrexate therapy: Relation to toxicity

Abstract

Background/objectives: Fetuin-A is a multifunctional protein with its urine level was considered as a marker of acute kidney injury. We investigated the serum and urine fetuin-A in acute lymphoblastic leukemia (ALL) children during and after high-dose methotrexate (HDMTX).

Methods: Twenty-two ALL children and 20 matched healthy controls were included. Liver transaminases, serum creatinine, estimated glomular filtration rate (eGFR), creatinine clearance (CrCl), serum β2 microglobulin (B2M), and serum and urine fetuin-A levels were assayed pre and 4 months after the consolidation. Among a subgroup of 15 patients, the investigations were performed 42 hours after the start of the second and the fourth HDMTX infusions.

Results: HDMTX was well tolerated. During HDMTX, there was significant decline in serum fetuin-A together with significant rise of urine fetuin-A and B2M levels compared to the control and to the pre-consolidation levels, changes that persisted 4 months after the consolidation despite recovery of the significantly altered renal functions. The second HDMTX-related serum fetuin-A level directly correlated with eGFR and CrCl (r = 0.86, P < 0.0001 and r = 0.67, P = 0.016, respectively). Four months after consolidation, urine fetuin-A directly correlated with serum creatinine (r = 0.54, P = 0.004) and inversely correlated with the eGFR (r = −0.66, P < 0.0001).

Conclusion: Significant disturbance in serum and urinary fetuin-A levels, which was related to renal functions, had occurred during HDMTX and persisted for at least 4 months after the consolidation. Serum and urine fetuin-A could be sensitive markers for subtle renal dysfunction in ALL children.

Keywords:

• Acute lymphoblastic leukemia
• Fetuin-A
• High-dose methotrexate
• Nephrotoxicity
• Renal biomarkers

Introduction

Childhood acute lymphoblastic leukemia (ALL) has a good prognosis, with about 80% of children achieving 5-year survival with the current intensive therapy.¹

Methotrexate, a folate analog, is an essential component of chemotherapy in nearly all pediatric ALL treatment protocols.² It acts as a competitive inhibitor of di-hydrofolate reductase enzyme.³

High-dose MTX (HDMTX) (defined as a dose ≥ 1 g/m²) is commonly given as ALL consolidation therapy.² This dose has been associated with severe toxicities with a mortality rate as high as 6%.⁴

Renal excretion is the primary route of MTX elimination with only about 10% of each dose is excreted unchanged in the bile.⁵ So, renal dysfunction is the main HDMTX toxicity and is caused fundamentally by tubular obstruction from crystal deposits of the drug (MTX) and its primary metabolite, 7-hydroxy MTX.⁶

HDMTX-induced acute kidney injury (AKI) and the resulting delay in MTX elimination can cause prolonged elevation of its systemic concentration with subsequent serious toxicity, including myelosuppression, mucositis, hepatic toxicity, and dermatitis.⁷

With the institution of standardized supportive care measures, including aggressive hydration, urine alkalinization, and leucovorin rescue, nowadays, HDMTX carries a much lower risk of severe toxicity. However, significant morbidity and mortality continue to be associated with HDMTX-induced renal dysfunction in approximately 1.8% of the patients (range, 2–12%).^6,8,9

AKI is diagnosed when there is progressive rise in serum creatinine over several days, which may or may not be accompanied with oliguria. Delayed diagnosis is a problem which may result in renal function deteriorations.¹⁰ Glomerular filtration rate (GFR) is difficult to measure in clinical practice.¹¹

Furthermore, serum creatinine could not detect renal failure until GFR decreases by more than 50%.¹² Hence, the detection of a reliable biomarker for early diagnosis of AKI would be very helpful in facilitating early intervention, evaluating the effectiveness of the therapeutic intervention and guiding pharmaceutical development.¹³

Fetuin-A (alpha-2-Heremans-Schmid glycoprotein, AHSG) is a 59-kDa glycoprotein consisting of two N-terminal cystatin domains and a smaller C-terminal domain. This liver synthesized molecule is found at high levels in the circulation and bone.¹⁴

It was classified as a negative acute phase protein (APP), where, decreased serum concentration occurs during acute inflammation¹⁵ as well as in end-stage renal disease.¹⁶

Fetuin-A urine level is found to be increased in intensive care unit (ICU) patients with AKI compared to those without AKI and healthy volunteers.¹⁷

Although the function of fetuin-A in AKI has not been clearly defined, it might have important role in tubular cell apoptosis.¹⁸

We investigated the hypothesis that serum and urine fetuin-A concentrations would be affected in ALL children under HDMTX during and after the consolidation phase in relation to HDMTX-induced toxicity.

Subjects and methods

Subjects

Twenty-two consecutive children with newly diagnosed ALL receiving HDMTX between August 2012 to October 2014 in the Pediatric Hematology and Oncology Unit, Menoufia University Hospital, Egypt were prospectively enrolled.

Diagnosis of ALL was performed according to standard clinical, morphological, cytochemical, and immune-phenotyping criteria. Patients were treated according to St Jude ALL Total Therapy Study XV treatment protocol.¹⁹ According to this protocol, the patients were classified in risk groups depending on pre-therapeutic factors and response to the induction therapy. The consolidation phase includes four HDMTX (2.5 g/m² for low risk and 5 g/m² for standard and high risk) 2 weeks apart. The first HDMXT starts at day 42 of induction after bone marrow (BM) and minimal residual disease (MRD) sampling.

The criteria for inclusion in the study were recently diagnosed ALL patients >1 year and <18 years who achieved complete remission (MRD < 0.01%) after induction therapy. The study did not include patients <1 year and >18 years old, those with Trisomy 21, patients on other therapy protocols, patients with impaired renal functions, liver functions, those with serological evidence of hepatitis C (HCV) or hepatitis B viral infection (HBV) at diagnosis or patients who did not complete the four HDMTX doses.

Subjects who had abnormal serum transaminases (ASAT/ALAT) (≥2 folds), serum creatinine (>1.4 mg/dl) values, or creatinine clearance <80 ml/minute/1.73 m² (who need MTX dose modification) prior to each HDMTX, as well as those with histories of oral mucositis, history of grade 3 or 4 gastrointestinal toxicity prior to MTX or history of typhlitis in previous chemotherapy, or oral herpes simplex virus I infection were also excluded. All included patients should be free from acute illness including infections with normal C-reactive protein (CRP) at all evaluation points.

Twenty normal age-, sex-, surface area-matched healthy children were involved as controls. They were clinically free from acute infection and inflammatory state with no previous history of any chronic illness. They had normal complete blood count (CBC) and normal CRP level. Controls were recruited from children presented to our general pediatric clinic for assessment and non-specific complaints.

Informed consent was obtained from the legal guardians of the included children and ethical clearance from Faculty of Medicine Menoufia University ethical committee was obtained before study beginning.

Methods

All included patients were subjected to detailed history taking and thorough clinical examination. Relevant medical data about the presenting manifestations, the risk stratification, renal and hepatic status, including functions at presentation, and the induction remission criteria were collected from the patients’ files.

For each participant, body weight and height were measured. Body surface area was calculated using the nomogram.

HDMTX delivery

According to the protocol, the consolidation phase started after the induction phase when absolute neutrophilic count (ANC) was ≥300/mm³, WBC was ≥1000/mm³, and platelet count was ≥50 × 10⁹/l. It included HDMTX (every other week for four doses). HDMTX was held if total bilirubin >2 mg/dl and direct bilirubin >1.4 mg/dl. Patients with low-risk ALL received 2.5 g/m², and those with standard-risk ALL received 5 g/m². The subsequent HDMTX was delayed if previous criteria are not fulfilled.

Pre-hydration: At least 2 hours before HDMTX, pre-hydration IV fluid [dextrose 5 water (D5W) + 40 mequiv. NaHCO₃/l] was administered. HDMTX treatment followed, provided that urinary pH was >6.5.

MTX was administered intravenously over 24 hours. During the MTX infusion, patients should receive hydration fluid with D5W + 40 mequiv./l NaHCO₃ at 100–150 ml/m²/hour. Urine pH is monitored with each void during infusion. During MTX infusion, the urine pH should be kept >6.5 by NaHCO₃.

The first plasma MTX level was assessed at 42 hours after the start of the infusion and then every 24 hours till the level be below 0.1 µM/l.

Leucovorin (LV) rescue: oral or iv leucovorin (10 mg/m² for low risk cases or 15 mg/m² for standard risk) was started at 42 hours after the start of MTX and repeated every 6 hours for a total of five doses. The dosage of leucovorin was increased in patients with high plasma MTX concentrations (>1.0 µM/l at 42 hours) and continued until the MTX concentration is less than 0.10 µM/l.

Concurrent triple intra-thecal therapy in age-adjusted doses was given for all children on the same day of the HDMTX administration.

Chlorhexidine mouthwash 0.12% was administered to all children two times daily during HDMTX therapy.

All included patients were investigated 24 hours before the first HDMTX and at 4 months after the fourth dose. Fifteen of the total 22 ALL children (patient subgroup) could be also evaluated at 42 hours after start of MTX infusion (with the first MTX leveling) during the second and the fourth doses. In such cases, blood samples were obtained simultaneously with serum MTX levels.

Clinical assessment was done daily specially for the presence of mucositis and gastrointestinal manifestations including vomiting and diarrhea with grading according to adverse event evaluation was assessed using the National Cancer Institute Common Terminology Criteria (NCI CTC) version 4.0.²⁰ The data at the points of evaluation were recorded.

Laboratory investigations

Both patients and controls were subjected to the following investigations:

1.	CBC using AC920 Autocounter after calibration.
2.	MTX level detection in plasma at 42 hours after the start of infusion and repeated every 24 hours until less than 0.1 µM/l (Cecil Instruments, England).
3.	Liver function tests, including alanine aminotransferase (ALT), aspartate aminotransferase (AST) by enzymatic colorimetric determination.
4.	Quantitative CRP was measured on Beckman Coulter AU480 full automated auto-analyzer using CRP kit for the Immunochemistry System (Beckman Coulter, Brea, CA, USA). According to the reagent manual, the cut-off value for abnormal serum CRP was 8 mg/l.
5.	Parameters of kidney functions include: a.  Serum creatinine concentration determined by the kinetic Jaffe’ method,21 using Beckman Coulter AU480, Japan. b.  The estimated glomerular filtration rate (eGFR) in ml/minute/1.73 m2 based on the Schwartz formula22 was calculated from serum creatinine and body length. c.  Creatinine clearance (CrCl) in ml/minute/1.73 m2: each participant provided a 24 hours urine collection, and CrCl was determined (ml/minute) as previously described.23 Using the following formula: urine Cr (mg/dl) × 24 hours urine volume (dl)/serum Cr (mg/dl) × 1440 (minute/day) in ml/minute. Then corrected CrCl for the surface area was measured; CrCl-corrected = CrCl × 1.73/surface area = in ml/minute/1.73 m2.

Sample collection and assay for other biochemical analyses

Blood samples

Blood samples were obtained from every investigated subject by sterile vein-puncture. The samples were transferred into plain tube, allowed to clot at 37°C and then were centrifuged for 10 minutes at 4000 rpm. The clear supernatant serum was separated from the clot and kept frozen at −80°C till analysis. Quantitative measurements of human fetuin-A (ng/ml) and β2 microglobulin (mg/l) were performed by ELISA techniques using ALPCO human fetuin-A (ALPCO, 26-G Keewaydin Drive, Salem NH 03079, USA) and β2 microglobulin (B2M) (BioVendor Research and Diagnostic Products Laboratorní medicína a.s. European Union, Czech Republic), respectively, according to the manufacturer's protocols.

Urine samples

An adequate and complete 24 hour urine collection was done. The volume was calculated for each participant. Five-milliliter sample was used for calculation of CrCl. Another 5 ml sample was kept at −20°C until measurement of urinary fetuin-A by ALPCO fetuin-A ELISA kits according to the manufacturer's instructions. From the total 24 hours urine volume, urinary fetuin-A in mg/day could be calculated.

Statistical method

The data were processed on an IBM-PC compatible computer using SPSS version 18 (SPSS, Inc., Chicago, IL, USA). Continuous parametric variables were presented as means ± SD, while for categorical variables, numbers (%) were used. In statistical analyses, compatibility with normal distribution was evaluated using Shapiro–Wilk normality test. Chi-square test was used for qualitative variables. The difference between two independent groups was performed by Student's t test and by Mann–Whitney (U) test for parametric continuous variables and non-parametric variables, respectively. For more than two groups, one-way analysis of variance test was used for parametric data, and the Kruskal–Wallis test was applied to discriminate differences in non-parametric variables. The least significant difference test (for parametric variables) and Tukey's honest significant difference test (for non-parametric variables) were applied for comparisons between individual groups when appropriate. For paired analysis, paired t test and Wilcoxon Signed Rank test were used for parametric and non-parametric variables, respectively. Pearson correlation (r) test was used to measure the association between two quantitative parametric variables, and Spearman correlation coefficient was applied for non-parametric data. P-value of <0.05 was considered statistically significant.

Results

Study population characteristics

ALL included children were 17 males (77.3%) and 5 females (22.7%) with a mean age of 5.85 ± 3.28 years (median of 4.5 years, range: 3.5–12.5 years). According to protocol risk stratification criteria, 16 patients (72.7%) were defined as standard risk while the remaining 6 (27.3%) were defined as low risk.

Regarding the ALL children who were investigated at the second and the fourth HDMTX doses, they were 10 males (66.7%) and 5 females (33.3%) with a mean age of 6.46 ± 3.4 years (median of 5 years, range: 3.5–11 years). This group involved six low-risk and nine standard-risk ALL children.

The controls were 11 males (73.3%) and 4 females (26.7%), with their ages ranging from 3.5 to 11 years (mean of 7.43 ± 2.49 years, median of 8 years).

There was no significant difference between all included ALL children or ALL subgroup and the controls regarding the age (P = 0.12 and P = 0.39 for the whole group and the subgroup, respectively) or the gender (P = 0.78 and P = 0.69 for the whole group and the subgroup, respectively).

Clinical assessment for HDMTX toxicities at 42 hours after the start of MTX infusion of the second and the fourth doses

According to the NCI CTC, none of the evaluated children had developed diarrhea.

Vomiting less than grade 2 was found in four patients (26.67%) and in three patients (20%) during the second and the fourth doses, respectively. No one had vomiting ≥grade 2.

None of the included patients except 1 (6.67%) who developed grade 1 mucositis at the time of evaluation during the fourth dose.

Regarding CBC findings, the mean ANC was 1.87 ± 0.98 × 10⁹/l and 1.84 ± 0.89 × 10⁹/l at 42 hours after the second and the fourth doses, respectively.

For the second dose, five patients (33.3%) developed grade 1 neutropenia and 2 (13.3%) had grade 2 neutropenia. While at the fourth dose, 1 (6.67%), 3 (20%), and 1 (6.67%) patients had developed grade 1, grade 2, and grade 3 neutropenia, respectively. All neutropenic patients were non-febrile. None of the evaluated patients had thrombocytopenia at the times of evaluation for both doses. No hepatotoxicity was detected in any of the tested ALL children for both doses.

Concerning renal toxicity, grade 1 renal toxicity had occurred in two of the tested patients (13.3%) while grade 2 was found in only one patient (6.67%) at the time of evaluation of the second dose. For the fourth dose, grade 1 renal toxicity had developed in only one patient (6.67%).

Comparison of the hepatic and renal functions between the ALL subgroup studied at different intervals and the controls

These comparisons were illustrated in Table 1. Both liver enzymes of the evaluated patients – at all points of evaluation – were significantly higher compared to the controls (Fig. 1A). At the point of evaluation during the second HDMTX, transaminases levels did not differ compared to their pre first HDMTX levels. While their fourth HDMTX-related levels were significantly higher in this regard being the highest detected levels (Fig. 2A).

Figure 1 (A) Comparison of AST (U/l), ALT (U/l), and urine fetuin-A (mg/day) in ALL subgroup at different points of evaluation between each others and the controls. (B) Comparison of eGFR (ml/minute/1.73 m²), creatinine clearance (ml/minute/1.73 m²), and serum fetuin-A (µg/ml) in ALL subgroup at different points of evaluation between each others and the controls. (C) Comparison of serum creatinine (mg/dl) and serum β2 M (mg/l) in ALL subgroup at different point of evaluations between each others and the controls.

Figure 2 The trend figures for the estimated parameters, *P = between the pre first HDMTX and at 4 months after consolidation values. (A) The trend of AST (U/l) and ALT (U/l) at different points of evaluation. (B) The trend of serum creatinine (mg/dl) at different points of evaluation. (C) The trend of eGFR (ml/minute/1.73 m²) and creatinine clearance (ml/minute/1.73 m²) at different points of evaluation. (D) The trend of serum fetuin-A (µg/ml) and urine fetuin-A (mg/day) at different points of evaluation. (E) The trend of serum β2 M (mg/l) at different points of evaluation.

Table 1 Comparison of liver and renal functions in ALL subgroup at different points of evaluation and the controls

	ALL children (15)
	Pre first MTX	42 hours post second dose infusion	42 hours post the fourth dose infusion	Four months after consolidation	Control (15)
AST (U/l)	31.1 ± 14.97*	34.5 ± 12.58*^$	52.1 ± 16.69*^&	51.6 ± 52.69*	18.07 ± 3.59
ALT (U/l)	32.3 ± 19.93*	36.6 ± 18.8*^$	51.9 ± 14.9*^&	44.7 ± 26.09*	19.73 ± 3.96
Serum creatinine (mg/dl)	0.49 ± 0.07	0.94 ± 0.3*^&$#	0.68 ± 0.18*^&#	0.46 ± 0.19	0.46 ± 0.12
eGFR (ml/minute/1.73 m^2)	129.44 ± 31.05	69.98 ± 15.14*^&$#	95.94 ± 33.18*^&#	153.86 ± 58.32	145.34 ± 27.66
Creatinine clearance (ml/minute/1.73 m^2)	126.58 ± 23.19	103.34 ± 18.36*^&#	108.95 ± 18.08*^&#	138.61 ± 30.01	142.55 ±  22.47

*Significant compared with the controls.

^&Significant compared with the pre first HDMTX dose.

^#Significant compared with at 4 months after consolidation.

^$Significant compared with 42 hours post the fourth HDMTX infusion.

At 4 months after the last HDMTX, both transaminases levels did not differ compared to their pre first MTX dose, the second or fourth HDMTX-related levels (Fig. 2A).

Regarding the renal functions, no significant difference was found comparing both pre first HDMTX serum creatinine (Fig. 1C), eGFR or CrCl (Fig. 1B) levels with the control levels. Both the second the fourth doses related serum creatinine were significantly higher, while the corresponding eGFR and CrCl values were significantly lower (Fig. 1B) compared to the controls and the pre first MTX dose (Fig. 1C and B, for serum creatinine and eGFR and CrCl, respectively). The second dose-related serum creatinine was significantly higher while its corresponding eGFR and CrCl were significantly lower compared to the counterpart fourth dose-related levels with these levels being the highest and the lowest for serum creatinine and eGFR and CrCl, respectively (Fig. 2B and C, respectively). At 4 months after the last MTX dose, the values of the three assessed parameters of renal functions returned to normal ranges being not significantly different compared to the control or the pre first MTX dose levels (Fig. 1B and C), but became significantly lower compared to the second and the fourth doses related levels (Fig. 2B and C).

Comparison of the serum B2M, serum fetuin-A, and urine fetuin-A between the ALL subgroup studied at different intervals and the controls

These comparisons were demonstrated in Table 2. No significant difference was found between the pre-MTX dose serum β2 M (Fig. 1C), serum, or urinary fetuin-A (Fig. 1B and A, respectively) levels and the controls. The second dose and the fourth dose-related serum fetuin-A levels were significantly lower while both B2M and urinary fetuin-A levels were significantly higher compared to the controls, to the pre first MTX dose levels (Fig. 1B, C, and A, respectively). Although B2M level was within the reference range, that of serum and urine fetuin-A were below and above the reference ranges, respectively. No significant difference was obtained between the two doses related levels in the three markers, although the lowest serum fetuin-A and the highest urinary fetuin-A (Fig. 2D) and B2M (Fig. 2E) levels were at the fourth dose-related evaluation point. At 4 months after the last MTX dose, serum fetuin-A persisted significantly lower while serum β2 M and urinary fetuin-A levels were significantly higher compared to the counterpart control and the pre first dose values (Fig. 1B, C, and A, respectively). At this point of evaluation, B2M became significantly lower compared to the second and the fourth doses relate levels, serum fetuin-A level did not differ while urinary fetuin-A was significantly lower than the fourth dose-related level and did not differ than the second dose-related level (Fig. 2E and D, respectively).

Table 2 Comparison of serum B2M, serum, and urine fetuin-A levels in ALL subgroup at different points of evaluation and the controls

	ALL children (15)
	Pre first MTX	42 hours post second dose infusion	42 hours post the fourth dose infusion	Four months after consolidation	Control (15)
Serum β2 microglobulin (mg/l)	0.99 ± 0.12	2.25 ± 0.18*^&#	2.40 ± 0.21*^&#	1.96 ± 0.11*^&	0.96 ± 0.12
Serum fetuin-A (µg/ml)	472.00 ± 22.88	346.60 ± 40.01*^&	333.40 ± 42.31*^&	374.60 ± 10.12*^&	476.07 ± 18.79
Urine fetuin-A (mg/day)	9.4 ± 1.9	55.4 ± 17.4*^&	62.4 ± 16*^&#	43.4 ± 7.2*^&	10.7 ± 2.1

*Significant compared with the controls.

^&Significant compared with the pre first HDMTX dose.

^#Significant compared with at 4 months after consolidation.

Comparison of the hepatic and renal functions between whole ALL studied group at different intervals and the controls

Comparisons of the laboratory parameters between the patients and the controls are demonstrated in Table 3. For all included ALL children, pre first HDMTX as well as at 4 months after the fourth HDMTX, they had significant higher liver enzymes (AST and ALT) without significant difference regarding renal functions (serum creatinine eGFR and creatinine clearance) compared to the controls. At 4 months after consolidation, AST level did not show any significant difference compared to the pre-consolidation level, while ALT became significantly higher. At this time of evaluation, all renal function parameters did not differ than the pre first MTX dose values.

Table 3 Comparison of the tested parameters in all included ALL children and the controls

	ALL children (22)
	At diagnosis	Pre the first HDMTX	Four months after the consolidation	Control (15)
AST (U/l)	29.9 ± 8.96	33.8 ± 13.14*	55.2 ± 50.93*	18.07 ± 3.59
ALT (U/l)	33.7 ± 10.52	31.75 ± 20.03*	49.45 ± 27.49*^&	19.73 ± 3.96
Serum creatinine (mg/dl)	0.75 ± 0.22	0.44 ± 0.09	0.45 ± 0.15	0.46 ± 0.12
eGFR (ml/minute/1.73 m^2)	82.93 ± 17.7	141.41 ± 40.3	144.69 ± 44.34	145.34 ± 27.66
Creatinine clearance (ml/minute/1.73 m^2)	112.4 ± 26.77	129.33 ± 22.32	141.58 ± 27.48	142.55 ± 22.47
Serum β2 microglobulin (mg/l)	No data	0.97 ± 0.12	1.97 ± 0.10*^&	0.96 ± 0.12
Serum fetuin-A (µg/ml)	No data	463.50 ± 29.39	370.40 ± 13.64*^&	476.07 ± 18.79
Urine fetuin-A (mg/day)	No data	9.9 ± 1.6	44.7 ± 7.2*^&	10.7 ± 2.1

*Significant compared with the controls.

^&Significant compared with the pre first HDMTX dose.

Comparison of the serum B2M, serum fetuin-A, and urine fetuin-A between whole ALL studied group at different intervals and the controls

These relations were demonstrated in Table 3. The pre-consolidation levels of B2M, serum fetuin-A, and urine fetuin-A did not differ compared to the control values. Four months after the last HDMTX, both serum B2M and urine fetuin-A became significantly higher while serum fetuin-A got significantly lower compared to the controls and to their pre first MTX dose levels.

The correlation analysis results of the tested parameters in the ALL subgroup

Correlations among these patients were illustrated in Table 4. Forty-two hours after the infusion of the second HDMTX, serum fetuin-A level had significant positive correlation with both eGFR and CrCl (r = 0.86, P < 0.0001 and r = 0.67, P = 0.016 for eGFR and CrCl, respectively). The 42 hours post the fourth dose infusion urinary fetuin-A level had significant positive correlation with the estimated MTX level at this point (r = 0.64, P = 0.024), while serum fetuin-A had a tendency toward negative correlation with MTX level that did not reach the significant level (r = −0.52, P = 0.08). At this time, serum and urine fetuin-A levels were correlated negatively to each other and B2M had correlated positively with serum creatinine (r = 0.87, P < 0.0001).

Table 4 Correlations of the tested parameters in ALL subgroup

	Post second HDMTX infusion parameters						Post fourth HDMTX infusion parameters
	Serum fetuin-A (µg/ml)		Urine fetuin-A (mg/day)		Serum β2-microglobulin (mg/l)		Serum fetuin A (µg/ml)		Urine fetuin A (mg/day)		Serum β2-microglobulin (mg/l)
	r	P	R	P	r	P	r	P	r	P	r	P
AST (U/l)	0.11	0.74	−0.25	0.44	0.35	0.26	0.08	0.8	−0.16	0.61	0.14	0.68
ALT (U/l)	0.09	0.78	0.11	0.74	0.42	0.18	−0.11	0.74	−0.11	0.73	0.09	0.79
Serum creatinine (mg/dl)	−0.06	0.84	−0.45	0.14	−0.07	0.83	−0.39	0.22	0.49	0.09	0.87	<0 . 0001
eGFR (ml/minute/1.73 m^2)	0.86	<0.0001	−0.1	0.76	0.06	0.85	0.02	0.95	0.02	0.95	−0.54	0.07
Creatinine clearance (ml/minute/1.73 m^2)	0.67	0.016	0.07	0.84	−0.14	0.67	0.38	0.23	−0.53	0.07	−0.34	0.28
Serum fetuin A (µg/ml)	–	–	0.06	0.87	0.08	0.82	–	–	−0.89	<0.0001	−0.35	0.27
Urine fetuin A (mg/day)	0.06	0.87	–	–	−0.37	0.23	−0.89	<0.0001	–	–	0.24	0.46
Serum β2 microglobulin (mg/l)	0.8	0.81	−0.37	0.23	–	–	−0.35	0.27	0.24	0.46	–	–
MTX level	−0.56	0.06	0.35	0.26	−0.9	0.79	−0.52	0.08	0.64	0.024	0.31	0.33

The bold numerical indicates significance.

The correlation analysis results of the tested parameters in the whole ALL studied group

Univariate analysis was performed to study the correlation between serum B2M, serum fetuin-A, and urine fetuin-A in one hand and the tested variables on the other hand at two different points of evaluation; pre the first HDMTX and 4 months after the last HDMTX (Table 5). The only significant correlations were found between the 4 months after the fourth HDMTX urine fetuin-A level and both the serum creatinine and the eGFR at this time; where it had significant positive correlation with serum creatinine (r = 0.54, P = 0.004) (Fig. 3A) and significant inverse correlation with the eGFR(r = −0.66, P < 0.0001) (Fig. 3C). At the same time, serum B2M had a tendency toward negative correlation with CrCl that did not reach a significant level (r = −0.42, P = 0.065). However, no significant correlation was found between these three markers with any of the other parameters pre the first HDMTX.

Figure 3 (A) Correlation between urine fetuin-A (mg/day) and serum creatinine (mg/dl) in all included ALL children at 4 months after the consolidation (r = 0.54, P = 0.004). (B) Correlation between urine fetuin-A (mg/day) and eGFR (ml/minute/1.73 m²) in all included ALL children at 4 months after the consolidation (r = −0.66, P < 0.0001).

Table 5 Correlations of the tested parameters in all included ALL children

	Pre the first HDMTX parameters						Four months after the last HDMTX parameters
	Serum fetuin A (µg/ml)		Urine fetuin A (mg/day)		Serum β2-microglobulin (mg/l)		Serum fetuin A (µg/ml)		Urine fetuin A (mg/day)		Serum β2-microglobulin (mg/l)
	r	P	r	P	r	P	r	P	r	P	R	P
AST (U/l)	0.05	0.8	0.33	0.16	−0.33	0.15	0.06	0.78	0.07	0.77	0.16	0.49
ALT (U/l)	−0.32	0.17	0.06	0.8	−0.28	0.22	0.01	0.95	0.03	0.89	0.27	0.25
Serum creatinine (mg/dl)	0.16	0.49	−0.34	0.14	0.17	0.47	0.26	0.2	0.54	0 . 004	0.09	0.68
eGFR (ml/minute/1.73 m^2)	−0.28	0.22	0.23	0.33	0.024	0.9	−0.09	0.63	−0.66	<0 . 0001	−0.09	0.69
Creatinine clearance (ml/minute/1.73 m^2)	−0.25	0.29	−0.17	0.48	−0.16	0.49	0.35	0.13	−0.29	0.2	−0.42	0.065
Serum β2 microglobulin (mg/l)	0.17	0.47	−0.05	0.84	–	–	0.22	0.35	0.0	0.99	–	–
Serum fetuin A (µg/ml)	–	–	−0.14	0.57	0.17	0.47	–	–	−0.41	0.07	0.22	0.35
Urine fetuin A (mg/day)	−0.14	0.57	–	–	−0.05	0.84	−0.41	0.07	–	–	0.0	0.99

The bold numerical indicates significance.

Discussion

As one of cancers with a high cure rate, the balance between efficacy and side effects in treating childhood ALL is one of the major clinical challenges.²

HDMTX application in the consolidation phase of ALL protocols can significantly increase cure rates and improve patients’ prognosis.²⁴

Unfortunately, this regimen may produce AKI, thereby impairing its own elimination and increasing its other toxicities.⁷

In this study, we could evaluate HDMTX toxicity during the second and the fourth doses in a subgroup of 15 ALL children of the total included 22 patients. So, we could evaluate 30 HDMTX courses; 12 infusions of 2.5 g/m² and 18 infusions of 5 g/m². The evaluation was done at different points; 24 hours before the first HDMTX dose infusion, at the 42 hours after the start of MTX infusion.

At the points of evaluation in this group namely at the first evaluation of MTX plasma level (42 hours after the start of the infusion), our studied patients did not exhibit frequent or significant gastrointestinal (GIT) complications including mucositis. Little evidence of HDMTX-related myelo-suppression is existed with the use of LV rescue. Myelo-suppression may become evident if rescue is delayed or in the setting of prolonged elevated serum levels.²⁵

Significant non-febrile neutropenia in the form of NCI CTC grade 3 (ANC <1 and >0.5 × 10⁹/l) was observed in only one patient (6.67%) during the fourth dose. This is in accordance with previous reports.²⁶

HDMTX can cause an acute transient elevation in the serum transaminases from 2- to 20-fold normal levels. This occurs in as many as 60–80% of patients and typically resolves spontaneously within 1–2 weeks.²⁷

In this regard, although there was significant rise of both transaminases during the course of the second and the fourth HDMTX infusions compared to the controls and the pre-consolidation levels, hepatic toxicity – according to NCI CTC – was not found in any of the included ALL children at the points of evaluation. Actually, no one of our patients had 2-fold or more rise in the transaminases.

This comes in accordance with what was reported that HDMTX use in the treatment of childhood ALL is not usually associated with major evidence of direct acute hepatotoxicity, that, if occurred, is transient and peaked during the first week after the infusion.²⁸

So, collectively, in this study significant degrees of GIT, hepatic, and BM toxicities were infrequent among our studied ALL children at the points of evaluations. Indeed, apart from renal toxicity, other HDMTX-related toxicities do not occur usually as early as these points of evaluation. They usually present following a delay of several days (5–14 days after the dose).²⁶

HDMTX-induced nephrotoxicity has to be counteracted as medical emergency as it may be lethal.⁹ The cornerstone of successful treatment for patients with HDMTX-induced renal dysfunction includes early recognition and prompt institution of increased LV doses.^6,8

An abrupt rise in serum creatinine during or shortly after MTX infusion indicates the development of renal dysfunction and can result in significantly elevated plasma MTX concentration.²⁹

Based on the serum creatinine, using the NCI CTC criteria, grade 1 nephrotoxicity had developed in 10% of the tested infusions (3/30), two were related to the second dose and one to the fourth dose. Grade 2 toxicity was diagnosed in only one second dose infusion (3.33%). So collectively, grade 2 renal toxicity was found among 3 out of the tested 15 ALL children (20%) who did not manifest any sign of systemic toxicity. Moreover, nephrotoxicity more than grade 2 was not detected in this study.

In the previous studies, HDMTX-induced renal toxicity was reported in different frequencies up to 40% of the tested populations of different malignancies under different dose regimens.^8,30–33

Of these studies, few were performed upon pediatric ALL. While AKI defined as a rise in serum creatinine to >1.5 times the patient's baseline serum creatinine value was reported in only 1.3% (10 of 741) ALL children treated per the total XV and total XVI in one study.³⁰ Mikkelsen et al.³³ documented the highest prevalence for induced renal toxicity in 18.5% of all HDMTX 5 g/m² infusions and in 40.0% of all HDMTX 8 g/m² infusions in the study which included 43 and 4 children with ALL and lymphoma, respectively.

The disparity in the estimated frequencies of renal toxicity in different studies may be related to the difference in the included sample sizes, the applied protocols using different doses, the used assessment scale of toxicity and finally in the time points of toxicity assessment.³⁴

The tendency toward renal toxicity was also evident in this study in the significant rise in the mean serum creatinine coupled with significant decline in eGFR and creatinine clearance values assisted at the 42 hours after the start of the second and the fourth HDMTX infusions compared to controls and the pre-MTX levels.

In agreement with this, Abelson et al.³⁵ found transient decreases in GFR in nine children who were receiving HDMTX and had a nontoxic course.

Nevertheless, Ylinen et al.³⁶ did not find any significant changes in plasma creatinine level during or after HDMTX treatment.

Also, Krawczuk-Rybak et al.³² had documented that the values of serum creatinine were in normal range without significant change 24–48 hours after the last HDMTX, with transient decline in creatinine clearance in ALL children compared to the control values.

Early recognition of drug-induced nephrotoxicity is so valuable. Therefore, the need for more sensitive biomarkers for nephrotoxicity has been increased.¹³

Latent renal toxicity, without significant changes in serum creatinine, was documented before, during and after HDMTX infusion in pediatric ALL patients using plasma cystatin C as sensitive marker.³⁶

For the first time, in this study, serum and urine fetuin-A levels were studied in relation to HDMTX-induced toxicities. The salient findings of this work were, the significant decline in serum fetuin-A that was accompanied by significant rise in its urinary level measured at 42 hours after the start of the second and the fourth MTX doses infusions compared to the controls and their pre first MTX dose levels, findings that persisted 4 months after the last MTX dose among the patients subgroup as well as in the whole group. At these points of evaluation, serum fetuin-A levels were below the normal range.

Fetuin-A is a multifunctional protein, which has been recognized as a potential diagnostic tool and/or therapeutic agent. Its synthesis is divergently regulated in response to injury or infection, classifying it as a positive or negative APP that counter-regulates both injury- and infection-elicited inflammatory responses, respectively.³⁷ Beyond the neonatal period, serum fetuin-A levels are independent of age and gender³⁸ ranging from 450 to 600 μg/ml.³⁹

Since it is produced exclusively by hepatocytes, serum concentrations of fetuin-A have been linked to human metabolic alterations and can serve as an indicator of liver cell function.⁴⁰

There are conflicting data about fetuin level disturbance in different liver diseases.

Significantly lower fetuin-A levels were found in patients with primary biliary cirrhosis than in healthy controls.⁴¹ Among patients with acute A, B, and Epstein–Barr virus hepatitis, alcoholic cirrhosis, and hepatocellular cancer, significantly lower fetuin-A levels were found in patients with the latter two conditions. The authors attributed this to hepatocelluar dysfunction rather than the acute phase reaction, since other acute phase reactants did not differ significantly.⁴²

On the other hand, serum fetuin-A levels were significantly increased in patients with non-alcoholic fatty liver disease and are associated with insulin resistance.⁴⁰ Serum fetuin-A level was positively associated with elevated ALT and the metabolic syndrome in the general population.⁴³

Regarding its relation to renal dysfunction, serum concentrations of fetuin-A were found to be depressed in patients with end-stage renal disease.¹⁶ Its level decrease followed the decrease in eGFR.⁴⁴

Serum fetuin-A was reduced in children on dialysis and in those with chronic kidney disease (CKD) stage 2 while those with moderate CKD (stages 3 and 4) and transplanted children presented with apparently normal fetuin-A concentrations.⁴⁵

Returning to our results, serum fetuin-A level decline during HDMTX may be likely due, at least in part, to renal dysfunction rather than defective hepatic production remembering that hepatic toxicity of any grade was not detected in any of the studied ALL children at any point of evaluations. Moreover, there was no significant correlation between serum fetuin-A and liver transaminases. Indeed, the influence of inflammation is not likely to be present due to absence of clinical evidence of infection or inflammation (fever or mucositis) and the CRP negativity at the time of evaluations. The presence of direct relation between serum fetuin-A with eGFR (r = 0.86, P < 0.0001) and creatinine clearance (r = 0.67, P = 0.016) 42 hours post to the second HDMTX infusion, could strengthen this assumption.

Published data about serum fetuin-A level in ALL are scarce. One previous study reported decrease of serum fetuin-A levels in certain hematological malignancies including ALL. Its level was further reduced by concomitant infection and cytostatic treatment during a follow-up duration of 18 months. The authors related this to decreased production caused by infiltration of the liver, a hepatotoxic effect of cytostatic treatment and to a lesser degree, to increased consumption. In a result that contradicts our finding, serum fetuin-A showed statistically significant negative correlations with AST.⁴⁶

There is a growing body of evidence that urinary fetuin-A could be considered as an important biomarker of AKI through tubule cell apoptosis.¹⁸

Urinary exosomal fetuin-A level was found to be markedly increased in rats following cisplatin injection. It increased 50–52.5 fold at day 2 (1 day before serum creatinine increase and tubule damage), and remained elevated 51.5-fold at day 5 (peak renal injury). Urinary exosomal fetuin-A also increased in three ICU patients with AKI compared to those without AKI, denoting that, urinary fetuin-A might be a predictive biomarker of structural renal injury.¹⁷

In addition, urinary fetuin-A was demonstrated as a risk factor for reduction of GFR in diabetic nephropathy.⁴⁷

In this context, the parallel increase of urinary fetuin-A level that accompanied serum fetuin-A decline and the good negative relation between them (r = −0.89, P < 0.0001) 42 hours post the fourth HDMTX could support the relation between these markers change and HDMTX-related renal toxicity rather any cause in this work.

Urinary concentration of fetuin-A may depend on hepatic production, alterations in permeability through glomerular basement membrane by capillary damage and changes in tubular re-absorption.⁴⁷

The documented rise in urinary fetuin-A in absence of increased – but rather – decline in its serum level gives the possibility that these changes could be due to the tubular injury. Supporting this, our results revealed that urinary fetuin-A measured at 42 hours post the fourth HDMTX infusion had positive correlation with plasma MTX level (r = 0.64, P < 0.024).

This study included B2M evaluation as an another renal marker at the points of evaluation.

B2M – a low molecular weight protein (11.8 kDa) – is a light chain of major histo-compatibility class (MHC) I expressed on the cell surface of every nucleated cell.⁴⁸ Circulating B2M is cleared almost exclusively by the kidney, where it is freely filtered by the glomeruli, and reabsorbed and metabolized in the proximal tubules.¹² As such, its blood concentration has been proposed as a measure of GFR.⁴⁹

B2M had been reported to have a better diagnostic sensitivity for detection of impaired GFR with the advantages of age and muscle mass independence.⁵⁰

In our study, B2M level was significantly increased during the second and the fourth HDMTX compared to the controls and to the pre-MTX levels.

These data, coupled with the knowledge that these changes came in parallel to serum and urinary fetuin-A levels mentioned before, give another possible argument that the detected changes in these two markers were more likely to be related to renal toxicity especially that B2M level was directly correlated with serum creatinine at 42 hours post the fourth HDMTX infusion (r = 0.87, P < 0.0001).

In disagreement with our results, Krawczuk-Rybak et al.³² did not find any change in serum B2M levels, that remained within the normal rage after HDMTX among ALL children.

The possible influence of malignancy or inflammation, as documented causes of B2M rise,⁵¹ should not be a confounding factor because all our included children were in remission and the possibility for presence inflammation is weak as discussed before.

It is noteworthy that, studying the trend of the tested parameters from the second through the fourth HDMTX, the results of this work revealed significant rise in eGFR with significant decline in serum creatinine at the time of evaluation related to the fourth dose in comparison to that of the second dose. Nevertheless, non-significant changes were detected regarding creatinine clearance, serum, and urinary fetuin-A or B2M levels.

In accordance with this, progressive renal dysfunction, defined as steady increase in baseline serum creatinine, with each subsequent cycle of HDMTX has not been reported in the literature.³¹

Much attention has been focused on the acute renal toxicity of HDMTX, but there is a paucity of data evaluating renal dysfunction encountered after therapy.⁷

In 23 consecutive adults treated with greater than four consecutive cycles of HDMTX for the treatment of primary central nervous system lymphoma, the authors reported persistence of renal dysfunction in terms of serum creatinine at least 4 months following the conclusion of therapy in nearly 30% of patients.³¹

In contrast to the previous study, 4 months after the consolidation phase, in this study, there was recovery of the ordinary renal functions, namely serum creatinine, eGFR, and creatinine clearance that became comparable to the controls and the pre-consolidation levels among the subgroup as well as in the whole group. This could be related to the difference in the doses and the cycle numbers.

However at this time, there was persistent significant decline of serum fetuin-A (that remained below the normal range), with significant rise in one or both liver transaminases, urinary fetuin-A and B2M levels in comparison to the controls and the pre-consolidation levels. This may denote the possible presence of persistent but subtle renal dysfunction at this time that could not be detected by the ordinary renal function tests but by more sensitive markers. The possibility of hepatic affection contribution in serum fetuin-A decline as previously suggested by another authors⁴⁷ could not be ignored. But the presence of increased urinary fetuin-A that had significant positive correlation with serum creatinine (r = 0.54, P = 0.004) and negative correlation with eGFR (r = −0.66, P < 0.0001) at that time augment the role of renal dysfunction. However, it is of note that, during the 4 months duration after consolidation, ALL children were under the maintenance phase of chemotherapy. The possible contribution of such chemotherapy on these findings could not be totally ruled out, unless, the investigations were made at regular intervals after each phase of therapy which was not possible in our study.

The possible role of other nephrotoxic agents used by these patients like nephrotoxic antibiotics, anti-inflammatory drugs, and analgesics, in this non-overt renal dysfunction at this time may be a possible cause. However, the possible contribution of nephrotoxic antibiotics was excluded in two previous studies.^32,52

Limitations to this study include the relatively small number of the included children and the lack of 24 hourly follow-up of the tested parameters.

The availability of new easy ELISA tests for both serum and urine fetuin-A that are more or less equal in expense to other sensitive markers like cystatin C and urinary B2M may open the door for the wide use of serum and urine fetuin-A, and for other studies that combine theses markers together so, the researchers can compare the sensitivity of each marker specially if done on a large sample sizes. The simultaneous evaluation of several renal markers will be able to answer the raised question about the possibility of fetuin-A to replace the ordinary renal markers (especially serum creatinine with its known disadvantage) due to its peculiarity to detect the non-overt abnormalities.

In conclusion, this study demonstrated that, HDMTX therapy was well tolerated in ALL children and was not associated with significant clinical toxicities at least at the points of evaluation. However, this regimen was accompanied by significant but, rather, transient disturbance of ordinary renal functions. In addition, for the first time, this study results revealed significant disturbance in serum and urinary fetuin-A levels which persisted for at least 4 months after HDMTX course. These disturbances were significantly associated with renal functions. The findings may lend credence to the idea that serum and urine fetuin-A could be sensitive markers for subtle renal disturbance in ALL children. This also affirms the prognostic significance of serum fetuin-A level estimation in ALL children. Long-term follow-up studies are highly recommended.

Disclaimer statements

Contributors S.M.R. is responsible for the study concept and design, overseeing data collection, data interpretation, literature search, drafting the manuscript, figures and tables, and writing/revising the manuscript. E.A.B. is responsible for performing the biochemical analyses of the studied parameters and sharing in writing/revising the manuscript.

Funding Personal.

Conflicts of interest The authors have no conflict of interest to declare.

Ethics approval Ethical clearance from Faculty of Medicine Menoufia University ethical committee was obtained before study beginning.