Abstract
Background/Aim: High-dose methotrexate (HD-MTX) is widely used to treat osteosarcoma. However, some patients develop hepatic toxicity, leading to dose modification and delays in the scheduled chemotherapy. The present study aimed to identify the risk factors of hepatotoxicity in patients with osteosarcoma. Patients and Methods: We conducted a retrospective study of patients with osteosarcoma treated with HD-MTX between January 2014 and June 2020 at the National Cancer Center Hospital, Japan. The risk factors for MTX-induced hepatotoxicity (≥grade 3) were identified using multivariate logistic regression analysis. Results: The final analysis included 88 courses of 36 patients. Hepatotoxicity occurred in 51 (58.0%) of the 88 courses. Female sex, MTX dose (>10.2 g/m2), and serum calcium concentration (>9.3 mg/dl) were identified as risk factors for HD-MTX-induced hepatotoxicity. Conclusion: Identifying the risk factors of HD-MTX-induced hepatotoxicity may contribute to improvements in the safety and management of HD-MTX therapy.
Osteosarcoma is the most common primary malignancy of the bone in children and adolescents (1). This is a very rare form of cancer; it has an incidence of 2-3 per million people per year in the general population. However, it has a higher incidence in adolescence, peaking annually at 8-11 per million per year between the ages of 15 and 19 years (2). The current treatments for osteosarcoma comprise systemic chemotherapy and surgery (3). The introduction of multi-agent chemotherapy has dramatically improved the rate of successful of osteosarcoma. Recently, approximately two-thirds of children with osteosarcoma have been cured (4). Standard regimens now include preoperative (neoadjuvant) and postoperative (adjuvant) chemotherapy (5). Multiagent methotrexate (MTX), doxorubicin (DXR), and cisplatin (CDDP) chemotherapy are used to treat osteosarcoma. Among them, intravenous high-dose methotrexate (HD-MTX) is considered essential (6).
The dose of HD-MTX for patients with osteosarcoma is 8-12 g/m2 (7, 8). HD-MTX has various side effects, such as kidney injury, myelosuppression, mucositis, neurotoxicity, and hepatotoxicity, which lead to morbidity, occasional mortality, and discontinuation of cancer treatment, leading to inferior anticancer outcomes. The serum methotrexate concentration is monitored, and hydration, urinary alkalization, and leucovorin rescue are used to prevent and mitigate these toxicities (9).
Hepatic toxicity related to HD-MTX occurs frequently, and the concentrations of hepatic transaminases, such as alanine aminotransferase (ALT) and aspartate aminotransferase (AST), are elevated (10, 11). These liver enzymes are transiently elevated and rarely cause chronic liver diseases (12). HD-MTX hepatic damage in osteosarcoma is rare. However, hepatic toxicity results in modifications, such as reduction or discontinuation of the dose and delays in scheduled chemotherapy, which have been major obstacles to the treatment schedules (13). The risk factors for hepatotoxicity, including alcohol and hepatitis B and C infections, are listed for long-term oral low-dose MTX dosing that is used in patients with rheumatoid arthritis (14-17).
However, few studies have investigated the risk factors of HD-MTX-induced hepatotoxicity in patients with osteosarcoma. A better understanding of the potential risk factors of hepatotoxicity is required to improve the safety of HD-MTX administration. Early detection of hepatotoxicity is an important issue. Therefore, we performed a retrospective study to identify the risk factors of hepatotoxicity in patients with osteosarcoma treated with HD-MTX therapy at the National Cancer Center Hospital, Japan.
Patients and Methods
Study population. This was a single-center retrospective observational study that evaluated the risk factors for HD-MTX-induced hepatotoxicity in the treatment of osteosarcoma. We retrospectively reviewed the electronic medical records of patients with osteosarcoma treated with HD-MTX between January 2014 and June 2020 at the National Cancer Center Hospital, Japan. We analyzed the data from the first to the fourth course of HD-MTX therapy during the preoperative period. Courses with missing liver function test data before and after HD-MTX therapy, and those with grade 3 or worse hepatic dysfunction, which was graded according to the Common Terminology Criteria for Adverse Events (CTCAE) version 4.0, before HD-MTX administration, were excluded.
Chemotherapy. The osteosarcoma patients were treated with the MAP regimen according to the JCOG0905 protocol (18). The standard dose of MTX was 12 g/m2 in patients younger than 19 years, 10 g/m2 in patients aged 20-39 years, and 8 g/m2 in patients aged 40 years and older. The dose of MTX was infused for 4-6 h. Leucovorin rescue was initiated 24 h after the initiation of MTX infusions at 15 mg every 6 h (in principle). The standard dose of cisplatin was 120 mg/m2 in patients younger than 29 years and 100 mg/m2 in patients older than 30 years. The standard dose of doxorubicin was 30 mg/m2 (≤39 years)×2 and 25 mg/m2×2 (>40 years). All patients received adequate hydration, and urine alkalinization using 7% sodium bicarbonate was added when the urine pH was 7 or less.
Assessment of hepatic dysfunction. For each course, blood tests were conducted from day 2 to day 6. AST, ALT and total bilirubin (T-Bil) data were collected during each course of HD-MTX therapy, and the most severe were evaluated. Hepatic dysfunction was graded according to CTCAE version 4.0. The CTCAE defines grade 3 and grade 4 toxicity levels of AST and ALT as 5 to 20 times, and more than 20 times the upper limit of normal, respectively, and grade 3, and grade 4 toxicity levels of T-Bil as 3 to 10 times, and more than 10 times the upper limit of normal, respectively. In each course, we defined the course with grade 3 or worse hepatic dysfunction that occurred after administration of HD-MTX as the “hepatic dysfunction group,” and the other course was defined as the “non-hepatic dysfunction group.”
Data collection. The following baseline clinical data were collected: sex; age; body mass index (BMI); tumor size; stage; doses of MTX, cisplatin, and doxorubicin; body surface area; serum chloride, sodium, potassium, and calcium concentrations; alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, total bilirubin, blood urea nitrogen (BUN), and serum creatinine concentrations; creatinine clearance; white blood cell, platelet, monocyte, neutrophil, and eosinophil counts, hemoglobin concentration, and urine volumes on the day of HD-MTX administration; and administration of liver-supporting therapy such as glycyrrhizin preparations until the day of MTX administration. The Cockcroft-Gault and Schwartz formulas were used to calculate creatinine clearance for patients aged 12 or more and those aged 11 years or less, respectively.
Ethical statement. This study was conducted according to the World Medical Association Declaration of Helsinki and the ethical principles for clinical research. The protocol was approved by the ethics committee of the National Cancer Center Hospital, Japan (No. 2019-115), and the Tokyo University of Science (No. 19018).
Statistical analysis. Univariate analysis and multivariate logistic regression analyses were used to identify the risk factors for HD-MTX-associated hepatic dysfunction. Univariate analysis was performed to compare the patient characteristics of the hepatic dysfunction and non-hepatic dysfunction groups. Continuous variables were analyzed using the Mann-Whitney U-test or t-test. Pearson’s chi-squared test was used to analyze the categorical data. To avoid multiple collinearity, Spearman’s correlation coefficients were estimated for the relationships between the extracted factors with a p-value of <0.05. When there was a strong correlation between factors (Spearman’s correlation coefficient: |ρ| >0.7), one factor was selected for clinical importance. These variables were entered into the multivariate logistic regression analysis to identify any independent risk factors associated with hepatic dysfunction. Receiver operating characteristic (ROC) curve analysis was performed on factors showing significance in the multivariate logistic regression analysis to determine the cutoff values for continuous variables. We also analyzed the relationship between the number of risk factors and the occurrence of hepatic dysfunction. Pearson’s chi-squared test was used to analyze these relationships. All statistical analyses were performed using R software version 3.6.3 (The R Foundation for Statistical Computing, Vienna, Austria), and the mice package was used to complement the missing values in the multiple imputation method (19). The ROCR package was used to create the ROC curve (20). Statistical significance was set at p<0.05.
Results
Patient characteristics. A total of 119 courses of 36 osteosarcoma patients were treated with HD-MTX therapy during the study period. Thirty-one courses were excluded because they had missing liver function data before or after HD-MTX therapy (n=5) and grade 3 or worse hepatic dysfunction before HD-MTX administration (n=26). Therefore, 88 courses in 36 patients were included. Table I shows the baseline characteristics before HD-MTX therapy; the median age was 18 years, and 57 men and 31 women was observed in 88 courses. Table II shows the occurrence of hepatic dysfunction after the administration of HD-MTX. The number of courses with grade 3 or worse hepatic dysfunction with elevated ALT, AST, and T-Bil were 51 (58.0%), 34 (38.6%), and 0 (0%), respectively. Therefore, hepatic dysfunction was associated with elevated transaminase concentrations and was often judged by elevated ALT that exceeded grade 3.
Baseline characteristics of the patients.
Hepatic dysfunction after high-dose methotraxate administration.
Risk factors for HD-MTX associated hepatic dysfunction. Table III shows the comparison of the characteristics of the hepatic dysfunction and non-hepatic dysfunction groups before the administration of HD-MTX.
Comparison of the characteristics of the hepatic dysfunction and non-hepatic dysfunction groups after high-dose methotraxate therapy in univariate analysis.
The univariate analysis showed that age, body surface area, and serum creatinine concentrations were significantly lower in the hepatic dysfunction group than in the non-hepatic dysfunction group. There were significantly more women in the hepatic dysfunction group than in the non-hepatic dysfunction group (49.0% and 13.5%, respectively). Tumor size; cisplatin dose; MTX dose; serum potassium and calcium concentrations, lactate dehydrogenase concentration; and the platelet and eosinophil counts were significantly higher in the hepatic dysfunction group than in the non-hepatic dysfunction group. To prevent multiple collinearity, the correlations between these 12 potential factors, for which both groups showed significant differences in the univariate analysis, were assessed. Age, body surface area, and serum creatinine concentration were correlated with each other (age and body surface area: ρ=0.743, age and serum creatinine: ρ=0.713, body surface area and serum creatinine: ρ=0.754). Age and serum creatinine concentration have been reported to be correlated in children (21). Therefore, we selected age based on these results. There was a negative correlation between age and MTX dose (ρ=0.723). However, both factors were selected based on their clinical importance. Finally, age, sex, tumor size, cisplatin dose, MTX dose, serum potassium and calcium concentrations, lactate dehydrogenase concentration, and eosinophil counts were entered into the multivariate logistic regression analysis.
Table IV shows the results of the multivariate logistic regression analysis, which revealed that female sex, higher MTX doses, and higher serum calcium concentrations were independent risk factors for HD-MTX-induced hepatic dysfunction.
Multivariate logistic regression analysis for high-dose methotraxate-associated hepatic dysfunction risk factor identification.
Relationship between the number of risk factors and the incidence of hepatic dysfunction. ROC curve analysis was performed to determine the cutoff value for serum calcium and methotrexate doses. As shown in Figure 1, the optimal cutoff values of the serum calcium concentration and methotrexate dose were 9.3 mg/dl and 10.2 g/m2, respectively. The serum calcium (Ca) concentration of >9.3 mg/dl and MTX dose of >10.2 g/m2 were considered as independent risk factors for hepatic dysfunction. Figure 2 shows the relationship between the number of risk factors and the incidence of hepatic dysfunction. The incidence of hepatic dysfunction was 0% in patients without risk factors (Ca≤9.3 mg/dl, males with MTX dose of ≤10.2 g/m2) and 54.8% in patients with one risk factor (Ca>9.3 mg/dl, males with MTX dose ≤10.2 g/m2 or Ca of ≤9.3 mg/dl, males with MTX dose of >10.2 g/m2 or Ca of ≤9.3 mg/dl, females with MTX dose of ≤10.2 g/m2). In patients with two risk factors, it was 82.9%, and in patients with three risk factors, it was 100%. Patients with three risk factors had a significantly higher incidence of hepatic dysfunction than those with no or one risk factor.
Receiver operating characteristic curve of serum calcium concentrations (A) and methotrexate dose (B) for HD-MTX induced hepatic dysfunction. The optimal cutoffs for the serum calcium concentration and methotrexate dose were 9.3 mg/dl and 10.2 g/m2, respectively. AUC: Area under the curve; HD-MTX: high dose-methotrexate.
Correlation between the incidence of hepatic dysfunction and the number of risk factors. Ca ≤9.3 mg/dl: male patients with MTX dose of ≤10.2 g/m2 have no risk. Ca >9.3 mg/dl: male patients with MTX dose of ≤10.2 g/m2 or Ca of ≤9.3 mg/dl, male patients with MTX dose of >10.2 g/m2 or Ca of ≤9.3 mg/dl, and female patients with MTX dose of ≤10.2 g/m2 had one risk factor. Ca >9.3 mg/dl: female patients with MTX dose of ≤10.2 g/m2 have three risk factors. **p<0.01, *p<0.05 (Pearson’s chi-squared test).
Discussion
This retrospective observational study showed that female sex, higher MTX dose, and higher serum calcium concentration were independent risk factors for HD-MTX-associated hepatic dysfunction in patients with osteosarcoma. Our study also showed that an increase in the number of risk factors, which were serum calcium concentration of >9.3 mg/dl, MTX dose of >10.2 g/m2, and female sex, was associated with an increase in the incidence of hepatic toxicity. To our knowledge, this is the first study to identify the risk factors for developing hepatic toxicity after HD-MTX administration in patients with osteosarcoma. These results suggest that female sex, higher MTX dose (>10.2 g/m2), and higher serum calcium concentration (>9.3 mg/dl) at the beginning of HD-MTX therapy were associated with an increased risk of hepatic toxicity related to the treatment of osteosarcoma. This finding may indicate that it is necessary to confirm whether or not patients have these risk factors and monitor hepatic dysfunction in osteosarcoma patients who have them before starting HD-MTX treatment. These findings may contribute to the improvement of the safety and management of HD-MTX therapy.
In the present study, hepatotoxicity of grade 3 or worse occurred in 51 (58.0%) of the 88 courses. In a similar study, in which the dose of HD-MTX was the same for patients with osteosarcoma, the incidence of grade 3 and 4 hepatotoxicity was 14.3% for 882 courses of the 113 patients (22). Regarding these differences in incidence, our study investigated only the patients during the preoperative chemotherapy period, but this study also included patients who received postoperative chemotherapy. Tsurusawa et al. reported that hepatic toxicity was higher after the first drug administration during HD-MTX therapy (23), and appropriate supportive care, such as repeated administration of leucovorin, was provided subsequently, and also it is thought that liver damage may have been reduced (24). From these studies, it should be noted that hepatic dysfunction is more likely to occur during the preoperative period in HD-MTX therapy.
Multivariate logistic regression analysis showed that female sex was an independent risk factor for hepatic dysfunction in patients with osteosarcoma, which is consistent with previous studies showing that female patients are affected by the development of MTX-induced hepatic toxicity after the administration of a high dose of MTX for osteosarcoma (25) and a low dose for juvenile idiopathic arthritis (26). The effects of female hormones and differences in pharmacokinetics related to sex have been implicated (26). The incidence of osteosarcoma differs with sex; osteosarcoma is more incident in men (27). It is important that women were selected as risk factors for hepatic dysfunction after HD-MTX therapy for a limited number of cases.
The dose of HD-MTX was extracted as an independent risk factor for hepatic dysfunction in patients with osteosarcoma receiving HD-MTX therapy. These findings are consistent with previous reports that the occurrence of hepatic dysfunction reflects MTX exposure (22, 28). Our results showed that there was a significant positive correlation between MTX dose and plasma MTX concentration at 24 h, 48 h, and 72 h after the administration of HD-MTX (data not shown). Previous studies reported that the hepatoxicity group had higher MTX peak concentrations at 24 h and 48 h than the non-hepatotoxicity group after administration, which can also be explained by the causation of MTX hepatic toxicity by a direct cytotoxic effect on hepatocytes proportional to MTX exposure (22). On the other hand, methotrexate is predominantly eliminated by the kidneys (29, 30). Thus, the plasma MTX concentration is affected by renal function (31). In this study, there was no significant difference in creatinine clearance between the two groups. These results suggest that renal function did not affect the plasma concentration of MTX in this study. In addition, our previous study reported that the co-administration of HD-MTX and glycyrrhizin causes pharmacokinetic interactions, resulting in elevated liver enzymes in rats (32). In the present study, we investigated the administration of liver-supporting therapy, such as glycyrrhizin administration, until the day of MTX administration, but no difference in the supporting therapy was found between the two groups. Our findings also showed that the optimal cutoff value of the methotrexate dose was 10.2 g/m2, and an MTX dose of ≤10.2 g/m2 was an independent risk factor. This corresponds to a dose of 19 or less years in this regimen. Since it has been reported that patients with a higher blood concentration of MTX have a better prognosis (22, 33), it is necessary to pay close attention to the occurrence of hepatic dysfunction.
We also found that the serum calcium concentration was an independent risk factor for HD-MTX-associated hepatic dysfunction in patients with osteosarcoma. These were unexpected results. The cut-off value of the serum calcium concentration was 9.3 mg/dl, which was within the normal range, but it was significantly higher in the hepatic dysfunction group than in the non-hepatic dysfunction group. Patients with osteosarcoma have been reported to have higher serum calcium concentrations than patients with other tumors, although they are within the normal range (34). It has been reported that the over-expression of the parathyroid hormone-related protein (PTHrP) and its receptor (PTHR1) lead to tumor growth, progression, and metastasis in osteosarcoma, which can also occur in breast and lung cancers, chondrosarcoma, squamous cell carcinoma, and melanoma (35). In addition, PTHrP has been reported to increase serum calcium concentration and excrete urinary calcium to enhance bone resorption (36). Corrected serum calcium concentrations are positively associated with advanced tumor stages in melanoma studies (37). Regarding the high corrected calcium concentrations, it has been reported that only a few patients had concentrations above the reference (approximately 2%), and the rest had elevated concentrations within the normal range (13%) (37). From the above, the serum calcium concentration increases with the progression of cancer, and it may be high but within the normal range. In this study, the tumor size was significantly greater in the hepatic dysfunction group than in the non-hepatic dysfunction group in the univariate analysis. From these findings, there may have been more patients with advanced cancer in the hepatic dysfunction group, and the serum concentrations were high but within the normal range.
The present study has limitations. First, the data were limited to retrospective observational studies. Therefore, there may be other unknown factors. Second, our findings were limited to a single center. Further large-scale studies are required to confirm our results. Third, this study focused on preoperative chemotherapy, and it is necessary to investigate postoperative chemotherapy. Finally, some reports have shown that single nucleotide polymorphisms, such as in SLCO1B1, affect the pharmacokinetics of MTX (38). Further studies on the relationship between single nucleotide polymorphisms and hepatic dysfunction in HD-MTX therapy are needed.
In conclusion, we revealed that female sex, MTX dose (>10.2 g/m2), and serum calcium concentration (>9.3 mg/dl) are risk factors for hepatic dysfunction in patients with osteosarcoma receiving HD-MTX therapy. Examination of these factors may help predict hepatic dysfunction, and patients with these factors should be monitored to reduce hepatotoxicity in osteosarcoma patients receiving HD-MTX therapy. We provide suggestions for preventing hepatic dysfunction and reducing the severity of hepatotoxicity in osteosarcoma patients receiving HD-MTX therapy. These findings may contribute to the improvement of the safety and management of HD-MTX therapy. Further studies are needed to confirm these results because our patient population is limited.
Acknowledgements
The Authors would like to thank the pharmacists and doctors who were involved in this study.
Footnotes
Authors’ Contributions
KA, AM, and YM conceived the study. KA, AM, and TI analyzed the data and performed the statistical analysis. KA, AM, and YM drafted the manuscript. KA, AM, SI, EK, TS, YK, HH, MY, TF, SM, and YM contributed to the discussion and reviewed the final manuscript. All Authors approved the final manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
- Received November 9, 2021.
- Revision received November 30, 2021.
- Accepted December 2, 2021.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.