Effective serum concentration of high-dose methotrexate therapy for osteosarcoma with dose adjustment for pharmacokinetic parameters: a retrospective study

Abstract

Background

In high-dose methotrexate (MTX) (HD-MTX) therapy, a dosage regimen utilizing the pharmacokinetic (PK) parameters of individual patients (PK method) has been developed to successfully reduce the intra- and inter-individual variability of serum MTX concentration. In this study, we evaluated the relationship between the therapeutic effect and serum MTX concentration in HD-MTX therapy performed under conditions in which the serum MTX concentration was precisely controlled using the PK method.

Methods

HD-MTX therapy consisted of 1 h of rapid dosing and 5 h of maintenance dosing, with a target serum concentration of 700–1000 µmol/L 1–6 h after the start of administration. The initial MTX dose was determined based on the body surface area, and subsequent courses were determined using PK parameters calculated from changes in MTX concentration at 1, 2, 4, 6, 7, 9, 12, 24, 48, and 72 h after the start of MTX administration. The relationship between the maximum serum MTX concentration (C_max) and mean concentration from 1 to 6 h after MTX administration [C_mean(1−6)] and efficacy was evaluated. The cut-off values for C_max and C_mean(1−6) were the values with the best sensitivity and specificity in terms of 5-year survival in the receiver operating characteristic curve.

Results

We included 33 patients with osteosarcoma without distant metastasis who underwent HD-MTX therapy. Those with a C_mean(1−6) > 768 µmol/L had significantly longer overall survival (OS) (p = 0.010) and tended to have longer event-free survival (EFS) (p = 0.061), with a 5-year survival rate as high as 92.9% (p = 0.077). In contrast, prolonged OS, EFS, and 5-year survival rates were observed in patients with C_max >936 µmol/L, but none of these differences were significant (p = 0.088, 0.200, and 0.103, respectively). Moreover, OS and EFS improved for C_mean(1−6) > 768 µmol/L regardless of C_max. In a subgroup analysis of 24 patients aged < 20 years, OS and EFS were significantly prolonged in patients with a C_mean(1−6) > 768 µmol/L (p = 0.010 and 0.025, respectively).

Conclusions

The usefulness of the serum MTX concentration was demonstrated by designing dosing using the PK method. Furthermore, C_mean(1−6) contributed more to prolonging OS and EFS than C_max.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-025-15437-6.

Background

Osteosarcoma is the most commonly diagnosed primary bone tumor in children and adolescents [1–5]. The treatment of osteosarcoma consists of surgery and systemic chemotherapy to prevent and treat distant metastases. With advances in combined systemic chemotherapy and surgical resection, the cure rate of osteosarcoma has exceeded 70% [6, 7]. One chemotherapy regimen reported to be highly effective for osteosarcoma is high-dose methotrexate (MTX) (HD-MTX) therapy.

MTX targets dihydrofolate reductase, an important metabolic component of nucleotide biosynthesis, and induces cell death by arresting the folate cycle. Treatment of osteosarcoma involves the inhibition of cancer cell growth by delivering MTX to cancer cells in a concentration-dependent (passive) manner by administering high doses of MTX and minimizing the effect on normal cells by administering leucovorin, an active folate that only normal cells have an uptake mechanism for (leucovorin rescue therapy), thereby minimizing the effect of MTX on normal cells [8–10]. Adverse events associated with HD-MTX therapy can be controlled by controlling the serum MTX concentration; therefore, it is recommended to monitor the serum concentration of MTX at the time of HD-MTX therapy and adjust the dose such that the serum MTX concentration is less than 10, 1, and 0.1 µmol/L at 24, 48, and 72 h after the initiation of MTX therapy, respectively [9, 11, 12]. However, because MTX is delivered to tumor cells in a concentration-dependent manner and its antitumor effect is time-dependent, controlling MTX at the highest possible serum concentration is also required. Several studies have reported the maximum serum MTX concentration (C_max) for high therapeutic effect and the relationship between C_max and adverse events [13–16]; in clinical practice, the target concentration range for C_max is 700–1,000 µmol/L for administration by continuous infusion for 6 h [15, 17–20]. However, there are many opposing and controversial views regarding the serum MTX concentration required to achieve the maximum therapeutic effect [13–16].

In 2010, Fujita et al. [21] developed a dosing design method that used the pharmacokinetic (PK) parameters of individual patients (PK method). By combining a 1-h loading dose with a 5-h maintenance dose, they could design dosing that reached and maintained the effective concentration range early. We previously reported that the PK method allows more precise control of serum MTX levels than the conventional method, which uses MTX doses calculated from body surface area (BSA) [22]. The PK method was also confirmed to provide more optimal control of serum MTX concentrations 72 h after administration and to increase the probability of controlling serum MTX concentrations within 6 h of administration to the target concentration range of 700–1,000 µmol/L [22]. Although our previous studies demonstrated that the PK method improves the attainment rate of target serum MTX concentrations, it has not been fully clarified whether maintaining MTX levels within the target range contributes to improved therapeutic outcomes. Therefore, this study was designed as a pilot investigation to explore the relationship between precisely controlled serum MTX concentrations, achieved through pharmacokinetically guided dosing, and clinical outcomes in patients with osteosarcoma.

Methods

Participants and HD-MTX regimen

All patients newly diagnosed with osteosarcoma who underwent HD-MTX therapy at the Department of Orthopedic Surgery, Gunma University Hospital, between April 2004 and March 2020 were enrolled in this study. Patients with recurrent or multiple lesions were excluded from the study. Patients with distant metastasis at the time of diagnosis were excluded. This study was approved by the Ethics Committees of Gunma University (Approval No.: HS2020-002) and Takasaki University of Health and Welfare (Approval No.: 2007) and was conducted in accordance with the Declaration of Helsinki. As this was a retrospective study using data collected since 2004, the ethics committees approved the use of an opt-out approach, and all study information was made publicly available to allow participants or their families to decline participation if they wished.

During HD-MTX therapy, the MTX dose during the first course was determined using the BSA method. Of the total dose (8–12 g/m²), approximately 45% (range, 40–50%) was administered as a 1-h loading dose, and the remaining 55% (range, 50–60%) was administered as a 5-h maintenance infusion. In the subsequent courses, loading and maintenance infusion doses were calculated by the PK method using the PK parameters of each patient, which were calculated based on their serum concentration profiles from the previous course, according to a report by Fujita et al. [21]. The target serum MTX concentration was defined as 700–1000 µmol/L after 1–6 h and less than 10, 1, and 0.1 µmol/L at 24, 48, and 72 h, respectively, from the start of administration. The loading and maintenance doses were slightly adjusted by approximately ± 10% from the calculated values. This adjustment range was determined in advance through discussions between our research group and the physicians in the Department of Orthopedic Surgery when the PK-based dosing method was introduced, and the same principle was applied consistently throughout the study. Leucovorin rescue was initiated 24 h after starting HD-MTX therapy. Leucovorin was started at a dose of 21 mg administered every 3 h and adjusted according to the serum concentration of MTX at 48 and 72 h. After 72 h of MTX treatment, leucovorin was continued until the serum MTX concentration reached 0.1 µmol/L. Sodium bicarbonate was administered to maintain a urine pH >7 and hydration and acetazolamide were administered to maintain urine volume.

Data collection

Electronic medical records from Gunma University Hospital were used to retrospectively survey the patient history, concomitant chemotherapeutic agents, and MTX-related laboratory data. The following characteristics were surveyed: age, sex, height, weight, BSA, diagnosis, site of onset, MTX dose, serum MTX concentration (a total of 10 points at 1, 2, 4, 6, 7, 9, 12, 24, 48, and 72 h after the start of MTX treatment, as C₁‒C₇₂), number of courses of HD-MTX therapy, laboratory data (hemoglobin [Hb], white blood cell [WBC], aspartate aminotransferase [AST], alanine aminotransferase [ALT], and serum creatinine levels), and treatment of patients with toxic MTX levels. Serum MTX concentrations were measured using fluorescence polarization immunoassay (TDxFLx Analyzer; Abbott Laboratories, Chicago, IL, USA).

Assessment of efficacy and safety

C_max is the highest serum concentration within C₁‒C₆, and C_mean(1−6) is the mean concentration from 1 to 6 h after MTX administration. Efficacy was evaluated based on the overall survival (OS) and event-free survival (EFS). OS was defined as the period from the date of diagnosis to death from any cause, whereas EFS was defined as the period from the date of diagnosis to the occurrence of recurrence, distant metastasis, or death due to any cause, whichever occurred first. The incidence of anemia, leukopenia, hepatic dysfunction, and renal dysfunction within 1 week of MTX administration was assessed to determine safety. Hb, WBC, AST, ALT, and creatinine clearance were used as indices of anemia, leukopenia, hepatic dysfunction, and renal dysfunction, respectively. Creatinine clearance was estimated using the Cockcroft–Gault equation. The Common Terminology Criteria for Adverse Events (CTCAE), version 5.0, was used to evaluate the grade of adverse events.

Calculation of dosage in the PK method

The PK parameters and dosages were calculated using the method described by Fujita et al. [21]. Briefly, serial blood samples were obtained at 1, 2, 4, 6, 7, 9, 12, 24, 48, and 72 h after the start of MTX infusion. To estimate the PK parameters, the serum concentration profiles of MTX were assumed to be a linear two-compartment model and analyzed by the non-linear least square method using the MULTI program [23]. The target concentration of C₁‒C₆ was set to 700 µmol/L.

Setting cut-off values

The sensitivity and specificity of C_max and C_mean(1−6) in terms of 5-year survival were assessed using a receiver operating characteristic (ROC) curve, and the optimal cut-off value was defined as the point at which the sum of sensitivity and 1 minus specificity was the greatest. Patients with C_max and C_mean(1−6) above the detected cut-off values were designated as C_max-H and C_mean(1−6)-H, respectively, and those below the cut-off values were designated as C_max-L and C_mean(1−6)-L, respectively. For subsequent analyses, the following designations were used to categorize patients into four groups according to the combinations of high or low values for C_max and C_mean(1–6): C_max-L/C_mean(1–6)-L, C_max-H/C_mean(1–6)-L, C_max-L/C_mean(1–6)-H, and C_max-H/C_mean(1–6)-H.

Statistical analysis

The Pearson’s chi-square test or Fisher’s exact test was used to compare the rates between the two groups. The Student’s t-test and Mann–Whitney’s U test were used to compare the means of the two groups. The log-rank test was used to determine the differences between the two groups in the survival curve. Adverse effects were categorized as CTCAE grade 2 or higher and less than 1. However, as there were no patients with an elevated serum clearance of grade 2 or higher, they were categorized as grade 1 or higher and grade 0. Statistical analysis was performed using the IBM SPSS software (version 26.0; IBM Corp., Armonk, NY, USA). Statistical significance was set at p < 0.05. In log-rank tests of survival curves for patients with C_max-H/C_mean(1−6)-L or C_max-L/C_mean(1−6)-H versus C_max-H/C_mean(1−6)-H, p < 0.025 was considered significantly different by Bonferroni correction.

Results

Characteristics of patients

The patient characteristics are shown in Table 1. Of the 43 patients who were newly diagnosed with osteosarcoma and treated during the study period, 10 with distant metastases at diagnosis were excluded. The remaining 33 patients without metastasis were included in the analysis (median age, 16.6 years). The male-to-female ratio was 3:2, and the site of onset was the lower extremity in 60% of patients. Among the 33 patients, 29 underwent wide or radical resection, and complete surgical resection was achieved in all of these patients. The remaining four patients had unresectable sacral osteosarcoma and were treated with carbon-ion radiotherapy. The median follow-up period from the date of osteosarcoma diagnosis was 91.1 months (interquartile range: 31.0–128.8 months). The 5-year survival rate for 28 patients was 75.0% (21/28), excluding 1 patient who was treated for less than 5 years and 4 who were lost to follow-up. Sex and site of onset (peripheral vs. central) were not associated with OS (p = 0.253 and 0.624, respectively; log-rank test; data not shown). In contrast, there was a trend toward shorter OS in younger patients, with 100% (8/8) of patients aged > 20 years and 65% (13/20) of patients aged < 20 years surviving for 5 years (p = 0.075; Fisher’s exact test).

Table 1.

Patients’ characteristics (n = 33)

Characteristic
Sex	Male	20	(60.6)
	Female	13	(39.4)
Age, y		16.6	(13‒27)
Body surface area, m2		1.55	(1.37‒1.66)
Total course of HD-MTX therapy		6	(4‒10)
Serum creatinine at diagnosis, mg/dL		0.57	(0.22‒1.67)
Creatinine clearance at diagnosis, mL/min		135.38	(61.07‒307.58)
Location	Upper limb	5	(15.2)
	Lower limb	20	(60.6)
	Pelvis	6	(18.2)
	Others	2	(6.1)
Loading dose of MTX, g/m2		6.64	6.00‒7.10
Maintenance dose of MTX, g/m2		8.52	7.75‒9.38
Total dose of MTX, g/m2		15.32	14.00‒16.20

Data are presented as median (interquartile range) or number (%)

HD-MTX high-dose methotrexate, MTX methotrexate

The median MTX clearance was 5.21 L/h (interquartile range: 4.33–5.82 L/h), with high interindividual variability; however, the median C_max and C_mean(1−6) were 955.2 µmol/L and 769.2 µmol/L, respectively, with an interquartile range of 700–1,000 µmol/L (Additional File 1).

All patients received HD-MTX therapy in combination with several chemotherapy regimens, which were administered before or after surgery or carbon-ion radiotherapy. The multi-drug regimens included doxorubicin (30 mg/m² on days 1, 2, and 3) + cisplatin (120 mg/m² on day 1); pirarubicin (30 mg/m² on days 1 and 2) + cisplatin (90 mg/m² on day 1); pirarubicin (30 mg/m² on day 1) + ifosfamide (3 g/m² on days 1–3) + vincristine sulfate (1 mg/m² on day 1); and bleomycin (15 mg/m² on days 1 and 2) + ifosfamide (3 g/m² on days 1 and 2) + actinomycin D (0.6 mg/m² on days 1 and 2). Additionally, some patients received doxorubicin monotherapy (30 mg/m² on days 1, 2, and 3) depending on their clinical condition. Patients with recurrent or metastatic disease were also treated with combinations of doxorubicin plus ifosfamide or gemcitabine plus docetaxel.

Setting cut-off values

Drawing an ROC curve using C_max and C_mean(1−6) as predictors for efficacy in 5-year survival resulted in a cut-off value of 936 µmol/L in C_max (0.667 sensitivity; 0.714 specificity) and 768 µmol/L in C_mean(1−6) (0.619 sensitivity; 0.857 specificity). The area under the curve (AUC) values were 0.667 for C_max and C_mean(1–6), showing no statistically significant difference (p = 0.147 and 0.105, respectively).

Relationship between serum MTX concentration and OS and EFS in all patients

The relationship between serum MTX concentration and efficacy is shown in Fig. 1A, B for all 33 patients, with no significant differences in OS or EFS between patients with C_max-H and those with C_max-L (p = 0.088 and 0.200, respectively; log-rank test). In contrast, there was a significant difference in OS between patients with C_mean(1−6)-H and those with C_mean(1−6)-L, and a significant trend was observed in the EFS (p = 0.010 and 0.061, respectively; log-rank test).

Fig. 1.

C_max (A) and C_mean(1-6) (B) in terms of efficacy in eligible patients (n = 33. The cut-off values for C_max and C_mean(1-6) were 936 µmol/L and 768 µmol/L, respectively. C_max, maximum concentration; C_max-H: C_max above the detected cut-off values; C_max-L: C_max below the detected cut-off values; C_mean(1-6), mean concentration during the maintenance dose; C_mean(1-6)-H: C_mean(1-6) above the detected cut-off values; C_mean(1-6)-L: C_mean(1-6) below the detected cut-off values

The 5-year survival rates for the 28 patients were 87.5% (14/16) for patients with C_max-H and 58.3% (7/12) for those with C_max-L, showing more favorable results in patients with C_max-H (p = 0.103; Fisher’s exact test). In addition, a significant difference in the 5-year survival trend was observed, with 92.9% (13/14) of patients with C_mean(1−6)-H and 57.1% (8/14) of those with C_mean(1−6)-L (p = 0.077; Fisher’s exact test).

Relationship between serum MTX concentration and OS and EFS in patients aged < 20 years

Among 24 patients aged < 20 years, those with C_max-H tended to have longer OS and EFS than those with C_max-L (p = 0.074 and 0.097, respectively; log-rank test; Fig. 2A). Moreover, OS and EFS were significantly prolonged in patients with C_mean(1−6)-H than in those with C_mean(1−6)-L (p = 0.010 and 0.025, respectively; log-rank test; Fig. 2B).

Fig. 2.

C_max (A) and C_mean(1-6) (B) were associated with efficacy in patients aged < 20 years (n = 24). The cut-off values for C_max and C_mean(1-6) were 936 µmol/L and 768 µmol/L, respectively. C_max, maximum concentration; C_max-H: C_max above the detected cut-off values; C_max-L: C_max below the detected cut-off values; C_mean(1-6), mean concentration during the maintenance dose; C_mean(1-6)-H: C_mean(1-6) above the detected cut-off values; C_mean(1-6)-L: C_mean(1-6) below the detected cut-off values

The 5-year survival rates for 20 patients, excluding 1 patient who was treated for less than 5 years and 3 patients who were lost to follow-up, were 83.3% (10/12) for patients with C_max-H and 37.5% (3/8) for those with C_max-L, showing a significant difference in trend (p = 0.062; Fisher’s exact test). In addition, a significant difference in the 5-year survival trend was found, with 90.0% (9/10) of patients with C_mean(1−6)-H and 40.0% (4/10) of those with C_mean(1−6)-L (p = 0.057; Fisher’s exact test).

Comparison of C_max and C_mean(1−6) in efficacy assessment

All patients and those aged <20 years were classified into four categories: C_max-H/C_mean(1-6)-H, C_max-H/C_mean(1-6)-L, C_max-L/C_mean(1-6)-H, and C_max-L/C_mean(1-6)-L, based on the C_max and C_mean(1-6) cut-off values. The OS and EFS of patients with C_max-H/C_mean(1-6)-H and those with C_max-H/C_mean(1-6)-L and C_max-L/C_mean(1-6)-H were evaluated (Fig. 3A, B). In the analysis of all patients, one patient with C_max-L/C_mean(1-6)-H had no relapse or death, and OS was similar to that of patients with C_max-H/C_mean(1-6)-H (p = 0.789; log-rank test). In contrast, four patients with C_max-H/C_mean(1-6)-L, two of whom relapsed and died, had poorer OS than those with C_max-H/C_mean(1-6)-H (p = 0.036; log-rank test). Furthermore, a trend toward poorer OS was also observed in patients aged <20 years with C_max-H/C_mean(1-6)-L (p = 0.094; log-rank test). Among all patients and those aged <20 years, patients with C_max-L/C_mean(1-6)-H had similar EFS to those with C_max-H/C_mean(1-6)-H (p = 0.506 and 0.567, respectively; log-rank test), whereas those with C_max-H/C_mean(1-6)-L had poorer EFS (p = 0.359 and 0.262, respectively; log-rank test).

Fig. 3.

Association of serum MTX concentration with OS and EFS in all patients (A, n = 33) and those aged <20 years (B, n = 24). The cut-off values for C_max and C_mean(1-6) were 936 µmol/L and 768 µmol/L, respectively. C_max, maximum concentration; C_max-H: C_max above the detected cut-off values; C_max-L: C_max below the detected cut-off values; C_mean(1-6), mean concentration during the maintenance dose; C_mean(1-6)-H: C_mean(1-6) above the detected cut-off values; C_mean(1-6)-L: C_mean(1-6) below the detected cut-off values

Adverse events

The relationship between the blood MTX concentration and adverse events in all 217 courses is shown in Table 2. Increased liver function values were observed in C_max-H and C_mean(1-6)-H, especially in C_mean(1-6)-H, with significant increases in AST and ALT values above grade 3 (p = 0.008 and 0.013, respectively; chi-square test). However, these elevations in liver function tests were transient and reversible, and no patients developed chronic hepatic dysfunction. C_max-H and C_mean(1-6)-H were not associated with anemia, leukopenia, or increased serum creatinine levels. Patients with elevated AST, ALT, and serum creatinine levels had a significantly higher AUC and lower MTX clearance than those without elevated levels (p < 0.05; Student’s t-test or Mann–Whitney U test; Additional File 2).

Table 2.

Adverse events of high-dose methotrexate therapy (217 courses)

	Cmax-L	Cmax-H		Cmean(1-6)-L	Cmean(1-6)-H
	(114)	(103)	p	(118)	(99)	p
Anemia
grade ≥2	44 (38.6)	39 (37.9)	0.9121	50 (42.4)	33 (33.3)	0.1721
grade ≥3	6 (5.3)	8 (7.8)	0.4531	8 (6.8)	6 (6.1)	0.8301
Leukopenia
grade ≥2	36 (31.6)	31 (30.1)	0.8131	39 (33.1)	28 (28.3)	0.4491
grade ≥3	17 (14.9)	18 (17.5)	0.6081	20 (16.9)	15 (15.2)	0.7201
AST increased
grade ≥2	42 (36.8)	35 (34.0)	0.6601	38 (32.2)	39 (39.4)	0.2701
grade ≥3	21 (18.4)	29 (28.2)	0.0891	19 (16.1)	31 (31.3)	0.0081
ALT increased
grade ≥2	47 (41.2)	44 (42.7)	0.8241	43 (36.4)	48 (48.5)	0.0731
grade ≥3	25 (21.9)	32 (31.1)	0.1271	23 (19.5)	34 (34.3)	0.0131
SCr increased
grade ≥1	0 (0.0)	2 (1.9)	0.1352	0 (0.0)	2 (2.0)	0.1212

Values are presented as number (%). ¹Chi-square test, ²Fisher’s exact test

ALT alanine aminotransferase, AST aspartate aminotransferase, C_max maximum concentration, C_max-H: C_max above 936 µmol/L; C_max-L: C_max below 936 µmol/L; C_mean(1-6), mean concentration during the maintenance dose; C_mean(1-6)-H: C_mean(1-6) above 768 µmol/L; C_mean(1-6)-L: C_mean(1-6) below 768 µmol/L; SCr, serum clearance

Discussion

The PK method used in this study estimated MTX clearance values from the concentration data of each individual patient. Consequently, the control rate within the target concentration range was improved, and intra-individual variability was reduced. This allowed for a more accurate evaluation of serum MTX concentrations, demonstrating that not only C_max but also C_mean(1–6) can be used as indicators of therapeutic efficacy. Furthermore, when comparing C_max and C_mean(1−6), we found that C_mean(1−6) had a stronger association with efficacy and could be an effective indicator of efficacy.

Since MTX exhibits time-dependent drug effects, it is important to evaluate serum concentrations, including temporal parameters. Most previous studies have used the AUC for this purpose. Indeed, AUC evaluation has been applied to many time-dependent anticancer drugs, including 5-fluorouracil [24, 25], and some reports have attempted to evaluate drug efficacy by AUC in HD-MTX therapy; however, a high AUC is strongly correlated with adverse events, such as high C_max. Hegyi et al. [26] evaluated C_max and the AUC versus OS and EFS in 105 pediatric patients with osteosarcoma treated with HD-MTX therapy and reported that C_max and the AUC were not associated with efficacy and were significantly correlated with adverse events, including hepatotoxicity. In addition, Comandone et al. [27] investigated the association between C_max, AUC, and efficacy in 25 patients with osteosarcoma treated with HD-MTX therapy and reported that a favorable clinical response was predicted by C_max and that severe toxicity was dependent on AUC.

Since AUC reflects not only early exposure (C_max) but also the late elimination phase (e.g., C₇₂), it can be strongly influenced by delayed MTX clearance; therefore, it may overestimate toxicity rather than efficacy. Consequently, AUC may not be an appropriate indicator for evaluating therapeutic efficacy in HD-MTX therapy, whereas early-phase parameters such as C_max and C_mean(1–6) may better reflect pharmacological effects. This study used a combination of loading and maintenance dosing to increase the C_max in a short period and maintain it at a high concentration for 6 h. Therefore, C_max and C_mean(1–6) showed weaker correlations with adverse events than did the AUC (Additional File 2). Nevertheless, AUC remains an important parameter for assessing overall drug exposure and ensuring safety, particularly to prevent toxicity related to delayed MTX elimination.

The usefulness of C_mean(1−6) for the effective concentration range of MTX in HD-MTX therapy has not yet been reported and, to the best of our knowledge, is mentioned for the first time in this study. Patients with C_mean(1−6)-H showed significantly prolonged OS (p = 0.010; log-rank test) and a trend toward longer EFS (p = 0.061; log-rank test), suggesting that C_mean(1−6) may be a useful indicator of treatment efficacy (Fig. 1). Furthermore, the effect of the serum MTX concentration was more pronounced in patients aged < 20 years than in all patients. In general, younger patients are reported to be more sensitive to chemotherapy than older patients. In this study, C_mean(1−6)-H was associated with significantly longer OS and EFS (p = 0.010 and 0.025, respectively, log-rank test; Fig. 2). It is possible that increasing the target concentration setting of MTX in the PK method, especially in younger patients, may help reduce the number of patients with C_mean(1−6)-L and lead to improved treatment efficacy, as long as chemotherapy is tolerated. However, because only 8 patients were aged 20 years or older, these age-related differences should be interpreted with caution. Thus, further validation in larger studies is warranted.

Patients were classified into four groups based on C_max and C_mean(1−6) cut-off values, and efficacy was compared. As expected, the prognosis was more favorable for C_mean(1−6)-H, regardless of C_max (Fig. 3). Since high C_max is negatively correlated with MTX clearance, and C_max ≥1,500 µmol/L is a poor prognostic factor [13, 16], it is important to design dosing based on C_mean(1−6) from a safety perspective. In contrast, C_mean(1−6)-H is associated with deteriorating hepatic functions. Although hepatic dysfunction with MTX is a frequently experienced adverse event [28–30], it is rarely fatal. As no association between C_mean(1−6)-H and adverse effects affecting prognosis, such as bone marrow suppression, has been confirmed, it is important to adjust the dosage to achieve C_mean(1−6)-H and continue treatment while monitoring liver function.

This study has several limitations. First, this was a retrospective, single-center study; therefore, confounders may have been overlooked. Surgical procedures and concomitant anticancer agents varied among patients, and magnetic resonance imaging and pathological data were not available for all cases, making uniform evaluation difficult. Moreover, pharmacogenetic data were not obtained. Second, the small number of samples made it difficult to calculate the optimal cut-off value and may have limited the statistical power of the ROC analysis, as reflected by the modest AUC values. However, despite this limitation, the log-rank test showed significant differences in OS and EFS between the high-concentration and low-concentration groups, suggesting a potential association that warrants further investigation and providing meaningful pilot data for future prospective studies. Finally, because the target maintenance concentration for the PK method was set at 700 µmol/L in this study, the usefulness of C_mean(1−6) at concentrations higher than 800 µmol/L could not be evaluated. Therefore, it is desirable to verify the efficacy and tolerability of higher concentrations of C_mean(1−6) in a larger number of patients to determine a more reliable cut-off value.

Conclusions

This study demonstrated the usefulness of C_max and C_mean(1−6) in designing dosing using a PK method with low intra-individual variability. Furthermore, C_mean(1−6) contributed more to the prolongation of OS and EFS than C_max, and there were no tolerability issues despite the presence of hepatic dysfunction; thus, to the best of our knowledge, the usefulness of C_mean(1−6) with MTX acting in a time-dependent manner was proven for the first time. As the present study was a pilot study, further large-scale prospective studies are needed to determine more accurate and clinically applicable cut-off values for serum MTX concentrations.

Abbreviations

ALT

alanine aminotransferase

AST

aspartate aminotransferase

AUC

area under the curve

BSA

body surface area

C_max

maximum serum methotrexate concentration

C_max-H

C_max above the detected cut-off values

C_max-L

C_max below the detected cut-off values

C_mean

mean concentration from 1 to 6 h after methotrexate administration

C_mean(1−6)-H

C_mean(1−6) above the detected cut-off values

C_mean(1−6)-L

C_mean(1−6) below the detected cut-off values

CTCAE

Common Terminology Criteria for Adverse Events

EFS

event-free survival

Hb

hemoglobin

HD-MTX

high-dose methotrexate

MTX

methotrexate

OS

overall survival

PK

pharmacokinetic

ROC

receiver operating characteristic

WBC

white blood cell

Authors’ contributions

A.N. and T.A. contributed to the study design and drafting of the manuscript. H.Y., A.K., and T.S. collected and analyzed the clinical data. T.Y., K.O., and K.Y. were involved in data interpretation and discussion. All authors read and approved the final manuscript.

Funding

No funding was received.

Data availability

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Gunma University (Approval No.: HS2020-002) and Takasaki University of Health and Welfare (Approval No.: 2007) and was conducted in accordance with the Declaration of Helsinki. This study was a retrospective study using data from the past 20 years, and it was difficult to obtain informed consent from all participants. Because this was a retrospective study using data collected since 2004, it was not feasible to obtain individual informed consent from all participants. Therefore, with approval from the institutional ethics committees, an opt-out approach was used, in which potential participants were considered to have agreed to participate unless they explicitly declined participation. In accordance with the committees’ instructions, information regarding the study purpose, subjects, and use of personal data was publicly disclosed on the hospital’s official website, and a contact point was provided to allow patients or their families to decline participation if they wished. This opt-out procedure is an ethically accepted and commonly used method for retrospective studies in Japan in which obtaining consent from all participants is impractical.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.