nature.com

Nomogram for the prediction of valproic acid induced platelet decline: a nested case–control study

Abstract

Platelet decline is a frequent side effect of valproic acid, a medication commonly prescribed to prevent seizures in neurosurgical patients. However, the risk factors for valproic acid-associated platelet decline remain poorly understood, and it remains unknown whether linezolid or levetiracetam in combination with valproic acid is associated with thrombocytopenia, as both drugs could lead to the decrease in platelet count. This three center, retrospective nested case–control study aimed to develop a predictive model for the prediction of valproic acid-induced platelet decline in a cohort of 356 participants. Multivariate analyses identified advanced age (OR: 1.05; 95% CI 1.03–1.08; P = 0.030) and combination therapy between valproic acid and levetiracetam (OR: 3.03; 95% CI 1.43–6.65; P = 0.005) as independent risk factors, while a trough concentration of valproic acid below 100 μg/mL (OR: 0.41; 95% CI 0.24–0.69; P = 0.010) was an independent protective factor for platelet decline. A nomogram was developed based on these factors, demonstrating robust performance with an area under the curve value of 0.85 in the training cohort and 0.81 in the validation cohort. Calibration plots showed strong agreement between predicted and observed outcomes. This model provides a valuable tool for assessing the risk of platelet decline in valproic acid-treated neurosurgical patients.

Introduction

Seizures are a frequent complication associated with neurosurgical procedures1. For individuals undergoing surgical resection of cerebral neoplasms, seizures are reported in 8–10% of cases, while procedures requiring substantial brain retraction may result in post-surgical convulsions in up to 25% of instances2. Though uncommon, post-surgical convulsions can be dangerous, requiring the administration of antiseizure medications throughout the perioperative period for neurosurgical patients3.

Valproic acid (VPA) has been approved for the treatment of generalized and focal seizures, migraine headache, and bipolar disorder and is widely used for the prevention of agitation and seizures following neurological surgeries. Although numerous new antiseizure drugs have been approved by the FDA in recent years, the critical role of VPA in the treatment of seizures remains irreplaceable, particularly in developing countries. However, the widespread use of VPA is limited by its adverse reactions, e.g., hepatotoxicity and thrombocytopenia4.

Thrombocytopenia refers to an abnormally low number of platelets in the blood. It is a common adverse effect of VPA with an incidence of 12–18%5. Moreover, a study conducted in China reported that the incidence of VPA-associated thrombocytopenia reached up to 37%6. Unlike other adverse reactions, VPA-associated thrombocytopenia often presents as an asymptomatic laboratory finding, making it less noticeable to doctors and patients, yet it carries the potential risk of leading to hemorrhagic stroke5,7.

Levetiracetam, a novel antiseizure drug, has demonstrated favorable therapeutic effects in generalized tonic–clonic seizures, myoclonic seizures, refractory epilepsy, and status epilepticus, as supported by numerous clinical studies8,9,10. Levetiracetam is preferred due to its reliable pharmacokinetics, lack of association with hypotension during administration, and the absence of a requirement for serum monitoring. Furthermore, levetiracetam is recommended for preventing postoperative seizures following craniotomy11. However, several case reports and clinical studies have documented that levetiracetam may lead to a reduction in platelet count in both adult and pediatric patients12,13,14,15. In patients with refractory or complex epilepsy undergoing craniotomy, combination therapy with multiple antiseizure drugs is often necessary. While both VPA and levetiracetam have been independently associated with platelet decline, it remains unclear whether the concurrent use of these two drugs significantly increases the risk of platelet decline. Further research is required to address this critical question.

Patients undergoing neurosurgery are prone to intracranial infections, which often require anti-infective treatment. Linezolid, an oxazolidinone antibiotic capable of crossing the blood–brain barrier, is commonly used to treat intracranial infections caused by Gram-positive cocci. Platelet decline is a well-documented adverse reaction associated with linezolid therapy16. However, the current knowledge does not clearly elucidate whether the co-administration of VPA with linezolid increases the risk of platelet decline.

To date, no predictive model for valproic acid-induced platelet decline has been identified in the PubMed or Web of Science databases. To fill this knowledge gap, a three-center, retrospective, nested case–control study was conducted to develop a predictive model and investigate the association between selected risk factors and VPA-associated platelet count decline.

Materials and methods

A retrospective nested case–control study was conducted in a cohort of patients who underwent neurosurgery and received VPA intravenous therapy. The study was performed at three tertiary hospitals: Shanxi Provincial People’s Hospital, Shanxi Provincial Cancer Hospital, and Beijing Friendship Hospital. Ethical approval was obtained from the Ethics Committee of Shanxi Provincial People’s Hospital (Approval No. 2019-36). Written informed consent was waived by the Ethics Committee due to the retrospective nature of the study.

Population of the study

Eligible participants were adults (aged 18 years or older) who underwent neurosurgical procedures, such as the removal of specific brain tumors or treatment for intracranial or subarachnoid hemorrhage, and had no history of hematological disorders. Exclusion criteria included treatment with VPA intravenous for less than 5 days, abnormal platelet count at admission, prior VPA treatment within one month, incomplete clinical information, current sepsis or disseminated intravascular coagulation, concomitant use of both linezolid and levetiracetam, or concurrent use of drugs known to affect platelet count or VPA concentration, e.g., glucocorticoids, heparin, and meropenem (Fig. 1). The initial VPA dose was 400 mg, administered three times daily, with adjustments based on VPA trough concentrations (reference range: 50–100 μg/mL). Levetiracetam was administered at a dose of 500 mg twice daily, and linezolid was administered at a dose of 600 mg twice daily. Combination therapy was defined as the concurrent use of VPA and at least one other drug for at least 5 days. All eligible patients were randomly divided into training and validation cohorts in a 7:3 ratio.

Fig. 1

figure 1

Study flowchart.

Full size image

Definition for cases and controls

The normal platelet count range was 150–450 × 109/L, with thrombocytopenia defined as a platelet count of < 150 × 109/L17. Patients exhibiting thrombocytopenia or a platelet count reduction of more than 40% during the course of therapy were categorized into the case group. A reduction in platelet count of more than 40% from baseline was chosen as a clinically significant threshold, based on evidence that such a decline increases the risk of bleeding18. This dual criterion was chosen to capture both overt thrombocytopenia and subclinical platelet decline, ensuring a comprehensive assessment of VPA-induced hematological toxicity. In this study, the case group was defined as the platelet decline group. Individuals meeting the eligibility criteria but without a significant decline in platelet count were assigned to the control group. Subsequently, the association between VPA-related platelet decline and various factors, including patient demographics, medication regimens, VPA trough concentrations, and baseline laboratory test results was analyzed.

Data collection

Relevant data were extracted from the medical facilities’ electronic databases. The collected medical information included demographic details, underlying diseases, treatment dosage and duration, concomitant medications, Glasgow Coma Scale scores, VPA trough concentrations, prothrombin time (PT), activated partial thromboplastin time (APTT), fibrinogen levels, white blood cell count, platelet count, hemoglobin levels, total bilirubin, transaminase levels, albumin, and creatinine levels. Baseline laboratory values were obtained within 48 h prior to the initiation of medication, and at least two laboratory tests were conducted during the treatment period. The worst laboratory values observed during treatment with VPA or VPA combined with levetiracetam/linezolid were defined as the post-treatment laboratory results.

Statistical analysis

Sample size estimation was conducted prior to study initiation. The calculations were based on an assumed alpha level of 0.05, a statistical power (1 − β) of 0.9, and a projected dropout rate of 10% to account for potential sample loss due to data anomalies. Linezolid treatment was designed as the exposure factor, with an assumed exposure rate of 40% in the case group and 15% in the control group. Using these parameters, the minimum required sample size for the case group was determined to be 72 participants. Given a case-to-control ratio of 1:3, the total cohort size was calculated to include at least 288 participants.

Data normality was assessed using the Shapiro–Wilk test. Descriptive statistics for quantitative variables are reported as medians with interquartile ranges (nonnormally distributed data) or as means with standard deviations (normally distributed data). Categorical data are expressed as number (percentages). In univariate comparisons, chi-square analysis is used for qualitative variables, and Student’s t-test or Kruskal–Wallis H-test is used for continuous variables, as appropriate. Univariate and multivariate logistic regression models were employed to explore the correlation between variables and platelet decline. Variables with a P-value < 0.05 in the univariate analysis were selected for inclusion in the multivariate logistic regression model. A stepwise regression approach was employed to sequentially incorporate these variables into the multivariate model. To evaluate the associations between quantitative variables, Pearson or Kendall τ correlation coefficients were calculated. In the model development phase, a forward stepwise selection approach was employed with a significance threshold of P < 0.05 to construct a simplified model. A nomogram was then created for the training cohort. Receiver operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was computed along with 95% confidence intervals (CIs) through 1000 bootstrap resamplings to assess the nomogram’s discriminatory performance. Calibration curves were also used to evaluate model performance. Statistical analysis was performed using the Extreme Smart Analysis platform (available at https://www.xsmartanalysis.com).

Results

Baseline information

Following the application of inclusion and exclusion criteria, a total of 356 patients were enrolled in the study between January 2020 and June 2023, with a mean age of 58.10 ± 12.40 years. The indications for neurosurgery among the enrolled patients were as follows: brain tumor (23.31%), subarachnoid hemorrhage (26.97%), intracranial hemorrhage (42.42%), Parkinson’s disease (4.21%), and brain abscess (3.09%) (Table 1). No bleeding events or mortality occurred during the hospital stay.

Table 1 Demographic and clinical patient characteristics.

Full size table

Characteristics of the case and control groups

Platelet count decline was observed in 37.08% (132/356) of the overall cohort and in 31.12% (47/151) of patients receiving VPA monotherapy. The mean age of patients in the platelet decline group was significantly higher compared to the control group. The proportions of patients in the case group receiving VPA monotherapy, VPA combined with linezolid, and VPA combined with levetiracetam were 35.61%, 43.18%, and 21.21%, respectively, while the corresponding proportions in the control group were 47.77%, 30.36%, and 21.87% (P = 0.034). The VPA trough concentration in the case group (93.16 [76.02, 108.27]) was significantly higher than that in the control group (77.45 [53.91, 86.34]). However, no statistically significant differences were found between the two groups in terms of sex, baseline platelet count, or duration of VPA treatment (Table 2).

Table 2 Comparison between case and control group.

Full size table

Risk factors for VPA-associated platelet decline

Multivariate logistic regression analyses revealed that advanced age (OR: 1.05; 95% CI 1.03–1.08; P = 0.030), combination therapy with VPA and levetiracetam (OR: 3.03; 95% CI 1.43–6.65; P = 0.005) significantly increased the risks of VPA-induced platelet decline. On the other hand, a trough concentration of VPA below 100 μg/mL (OR: 0.41; 95% CI 0.24–0.69; P = 0.010) is an independent protect factor for platelet decline (Table 3).

Table 3 Univariate and multivariate logistic regression models.

Full size table

Development and validation of a VPA-associated nomogram

The nomogram model was developed to predict the risk of platelet decline based on age, treatment regimen, and VPA trough concentration in the training cohort. Individual values for each variable are plotted along their respective axes, and corresponding scores are assigned based on these values. The scores are then summed and mapped onto the total score scale. Finally, the total score is vertically projected onto the bottom axis to estimate the individual’s risk of platelet decline (Fig. 2).

Fig. 2

figure 2

Nomogram for the prediction of valproic acid induced platelet decline.

Full size image

Predictive accuracy

The AUC values for the nomogram were 0.85 (95% CI 0.79–0.91) in the training cohort and 0.81 (95% CI 0.66–0.95) in the validation cohort (Fig. 3). Internal validation was performed using 1000 bootstrap resamplings. Calibration plots demonstrated robust agreement between the model’s predicted probabilities and the observed probabilities (Fig. 4).

Fig. 3

figure 3

(A) ROC curve for valproic acid induced platelet decline prediction in the training cohort; (B) validation Cohort ROC Curve for valproic acid induced platelet decline prediction.

Full size image

Fig. 4

figure 4

(A) Calibration curves for predicting valproic acid induced platelet decline in the training cohorts. (B) Calibration curve for valproic acid induced platelet decline prediction in the validation cohorts.

Full size image

Subgroup analysis in patients with linezolid

The median duration of linezolid therapy in the case and control groups were 86,10 days and 65,8 days, respectively. Subgroup analysis based on treatment duration revealed that patients receiving linezolid for more than 7 days had a significantly higher risk of platelet decline compared to those receiving short-term therapy (≤ 7 days) (42.31% vs. 28.77%, OR: 1.85, 95% CI 1.12–3.06, P = 0.017) (Table 4).

Table 4 Association between treatment duration of linezolid and platelet decline in subgroup analyses.

Full size table

Discussion

This study developed a predictive model to accurately identify the risk of platelet count reduction in patients receiving VPA. To the best of our knowledge, this is the largest study to date investigating the impact of VPA on platelet counts and the first to establish a comprehensive predictive model for this adverse reaction.

In this study, the case group included not only patients with thrombocytopenia but those whose platelet counts decreased by more than 40% relative to baseline values. This approach enabled a broader and more precise assessment of the impact of VPA on platelet counts. Since severe infections and certain medications, such as glucocorticoids and heparin, can influence platelet levels, these factors were excluded to minimize their potential confounding effects on the results18.

Our findings indicated an incidence of platelet decline of 31.12%, which is significantly higher than reported in previous studies19,20. Notably, these earlier studies were conducted in predominantly Caucasian populations6, whereas the incidence in the Chinese population was found to be as high as 37%. This discrepancy implies that VPA-associated platelet decline may be influenced by racial or ethnic differences.

The mechanism underlying VPA-induced platelet decline remains unclear. Bone marrow examinations in patients receiving VPA have revealed an increased number of megakaryocytes, suggesting enhanced peripheral platelet destruction4,21. Additionally, VPA has been shown to reduce the production of platelet malonyl dialdehyde, potentially impairing platelet function22. Another plausible explanation is the direct toxic effect of VPA on the bone marrow, resulting in reduced platelet production23.

Linezolid-induced platelet decline has been confirmed in multiple clinical studies16. Surprisingly, the combination of linezolid and VPA did not appear to increase the risk of platelet decline in our study. Linezolid-induced platelet decline is typically associated with prolonged treatment duration and renal dysfunction24,25. Therefore, we also conducted a subgroup analysis in patients with linezolid. The results demonstrated that a longer treatment duration (> 7 days) was associated with a higher risk of linezolid-induced platelet decline. This finding underscores the importance of monitoring platelet counts during prolonged linezolid therapy, particularly in neurosurgical patients who may already be at risk for thrombocytopenia due to concurrent medications, such as VPA.

The combination of VPA and Levetiracetam is a commonly used antiseizure treatment regimen. Our study indicates that the concurrent use of these two drugs increases the risk of platelet decline. Levetiracetam shares a similar chemical structure with piracetam, which has been shown to inhibit platelet aggregation and exert antithrombotic effects26. Previous case reports suggest that Levetiracetam-induced platelet decline may be associated with immune deficiency14. However, most patients in this study were not screened for immune function, and therefore, the potential impact of Levetiracetam on platelet levels should be interpreted with caution.

The relationship between the concentration of VPA and hematological toxicity remains a topic of debate. Some studies have found that VPA can cause adverse hematological effects even within its normal therapeutic concentration range27,28. Another study reported no severe hematological adverse reactions when VPA concentrations were between 90 and 160 µg/mL29. A negative correlation has been observed between VPA concentration and platelet count, with platelet levels being significantly lower in patients whose VPA concentrations exceed the therapeutic window30. For example, the incidence of platelet decline is 0.2% when VPA concentrations are below 80 µg/mL, compared to 30.2% when concentrations exceed 120 µg/mL20. In this study, only 33.33% of patients in the platelet decline group had trough concentrations below 100 μg/mL, whereas the corresponding proportion in the control group was 56.25%. Multivariate analysis indicates that a VPA trough concentration below 100 µg/mL is an independent protect factor for platelet decline. Our study confirms that VPA-induced platelet decline is associated with elevated trough concentrations. Additionally, VPA-related hepatotoxicity is also associated with its trough concentration31. Therefore, routine monitoring of VPA trough concentrations and maintaining them below 100 µg/mL is recommended.

A double-blind, multicenter clinical trial indicated that women and elderly patients are more prone to developing platelet decline, with women exhibiting twice the risk compared to men20. Our study corroborates these findings, confirming that advanced age is an independent risk factor for VPA-induced platelet decline. This increased susceptibility may be linked to a decline in hematopoietic capacity associated with aging.

This study has several limitations, the most prominent being its retrospective design. Despite the implementation of strict inclusion criteria, selection bias and confounding factors could not be fully avoided. Second, while our nomogram demonstrated robust discriminatory performance in internal validation, external validation in diverse populations is essential to confirm its generalizability. Future studies should prioritize prospective, multicenter cohorts to evaluate the predictive accuracy of the model in real-world clinical settings. Additionally, patient outcomes were not further evaluated in this analysis.

Conclusion

In summary, this study established a simple, user-friendly, and accurate prediction model to assess the risk of VPA-related platelet decline, which may serve as a valuable tool for informing clinical decision-making. Our model identified advanced age, the combination of VPA with levetiracetam, and a VPA concentration < 100 µg/mL as significant factors associated with platelet decline.

Data availability

The data supporting the research findings can be obtained from the corresponding author, Professor Jinlin Guo, upon reasonable request.

References

Sarwal, A. Neurologic complications in the postoperative neurosurgery patient. Continuum (Minneap Minn). 27(5), 1382–1404 (2021).

PubMedMATHGoogle Scholar

Wu, A. S. et al. A prospective randomized trial of perioperative seizure prophylaxis in patients with intraparenchymal brain tumors. J. Neurosurg. 118(4), 873–883 (2013).

ArticleCASPubMedPubMed CentralGoogle Scholar

Anbarlouei, M., Emamikhah, M., Basiri, K. & Farzanegan, G. Perioperative routine EEG can reliably predict de novo seizures after supratentorial craniotomy. Neurol. Sci. 43(4), 2717–2722 (2022).

ArticlePubMedGoogle Scholar

Morris, N., Barr, R. D., Pai, K. R. & Kelton, J. G. Valproic acid and thrombocytopenia. Can. Med. Assoc. J. 125(1), 63–64 (1981).

CASPubMedPubMed CentralGoogle Scholar

Buoli, M., Serati, M., Botturi, A. & Altamura, A. C. The risk of thrombocytopenia during valproic acid therapy: A critical summary of available clinical data. Drugs R&D 18(1), 1–5 (2018).

ArticleCASGoogle Scholar

Tang, J. L., Lin, Q., Yan, J., Suying. Analysis of the factors related the severity of valproic acid-related hematologic adverse drug reactions. Pract. Pharm. Clin. Remedies 23(5), 466–470 (2020).

Kim, D. W., Kim, W., Lee, C. H. & Chun, Y. I. Thrombocytopenia during intravenous valproic acid therapy in the neurological intensive care unit. J. Clin. Pharm. Ther. 45(5), 1014–1020 (2020).

ArticleCASPubMedMATHGoogle Scholar

Nevitt, S. J., Sudell, M., Weston, J., Tudur Smith, C. & Marson, A. G. Antiepileptic drug monotherapy for epilepsy: A network meta-analysis of individual participant data. Cochrane Database Syst. Rev. 12(12), CD011412 (2017).

PubMedMATHGoogle Scholar

Trinka, E. & Kalviainen, R. 25 years of advances in the definition, classification and treatment of status epilepticus. Seizure 44, 65–73 (2017).

ArticlePubMedGoogle Scholar

Shorvon, S. D., Bermejo, P. E., Gibbs, A. A., Huberfeld, G. & Kalviainen, R. Antiepileptic drug treatment of generalized tonic-clonic seizures: An evaluation of regulatory data and five criteria for drug selection. Epilepsy Behav. 82, 91–103 (2018).

ArticlePubMedGoogle Scholar

Iuchi, T. et al. Levetiracetam versus phenytoin for seizure prophylaxis during and early after craniotomy for brain tumours: A phase II prospective, randomised study. J. Neurol. Neurosurg. Psychiatry 86(10), 1158–1162 (2015).

ArticlePubMedGoogle Scholar

Hedges, A. et al. Levetiracetam dosing for seizure prophylaxis in neurocritical care patients. Brain Inj. 37(10), 1167–1172 (2023).

ArticlePubMedGoogle Scholar

Kim, J. & Shin, J. W. Levetiracetam-induced thrombocytopenia in a patient with status epilepticus. Epileptic Disord. 19(1), 104–108 (2017).

ArticlePubMedMATHGoogle Scholar

Nwaedozie, S. T., Aatrai, S., Najjar Mojarrab, J. & Al, T. A. Acute levetiracetam-induced thrombocytopenia in an anticoagulated immunocompetent patient. Cureus 13(3), e13817 (2021).

PubMedPubMed CentralGoogle Scholar

Ammad Ud Din, M., Ather Hussain, S. & Bodrog, A. Delayed Recognition of Levetiracetam-induced Pancytopenia. Eur. J. Case Rep. Intern. Med. 8(3), 002449 (2021).

PubMedPubMed CentralMATHGoogle Scholar

Zhao, H. et al. A nomogram incorporating linezolid and metabolite concentrations for predicting linezolid induced thrombocytopenia in patients with renal impairment. Sci. Rep. 14(1), 26064 (2024).

ArticleCASPubMedPubMed CentralGoogle Scholar

Post, D. S. et al. Assessment of need for hemostatic evaluation in patients taking valproic acid: A retrospective cross-sectional study. PLoS One 17(2), e0264351 (2022).

ArticleCASPubMedPubMed CentralGoogle Scholar

Vanderschueren, S. et al. Thrombocytopenia and prognosis in intensive care. Crit. Care Med. 28(6), 1871–1876 (2000).

ArticleCASPubMedGoogle Scholar

Allarakhia, I. N., Garofalo, E. A., Komarynski, M. A. & Robertson, P. L. Valproic acid and thrombocytopenia in children: A case-controlled retrospective study. Pediatr. Neurol. 14(4), 303–307 (1996).

ArticleCASPubMedGoogle Scholar

Nasreddine, W. & Beydoun, A. Valproate-induced thrombocytopenia: A prospective monotherapy study. Epilepsia 49(3), 438–445 (2008).

ArticleCASPubMedMATHGoogle Scholar

Neophytides, A. N., Nutt, J. G. & Lodish, J. R. Thrombocytopenia associated with sodium valproate treatment. Ann. Neurol. 5(4), 389–390 (1979).

ArticleCASPubMedGoogle Scholar

von Voss, H., Schulte-Berbuhl, R. & Gobel, U. Side effects of sodium valproate (sodium di-N-propylacetate) on malonyldialdehyde-production and other platelet function tests. Minerva Pediatr. 30(17), 1397–1404 (1978).

Google Scholar

Verrotti, A., Scaparrotta, A., Grosso, S., Chiarelli, F. & Coppola, G. Anticonvulsant drugs and hematological disease. Neurol. Sci. 35(7), 983–993 (2014).

ArticleCASPubMedGoogle Scholar

Al Qamariat, Z. et al. Rapid onset and recovery linezolid-induced thrombocytopenia: A large-sample, single-center retrospective cohort study. Drug Healthc. Patient Saf. 16, 43–49 (2024).

ArticlePubMedPubMed CentralGoogle Scholar

Thabit, A. K. et al. Linezolid-associated thrombocytopenia: Assessment of risk factors in patients without hemato-oncologic diseases. J. Clin. Med. 13(8), 2380 (2024).

ArticleCASPubMedPubMed CentralGoogle Scholar

Olaizola, I., Brodde, M. F., Kehrel, B. E. & Evers, S. The impact of levetiracetam and valproate on platelet functions-A double-blind, placebo-controlled crossover study. J. Clin. Med. 12(3), 933 (2023).

ArticleCASPubMedPubMed CentralGoogle Scholar

Hongyan, M. Q. G. Literature analysis on 322 cases of fibrinogen reduction induced by valproate. China Pharm. 22(1), 144–147 (2019).

MATHGoogle Scholar

Dong Yan, D. K. & Aixiang, H. Effect of valproate sodium on fibrinogen in children with epilepsy. Guangdong Med. J. 38(2), 305–7 (2017).

MATHGoogle Scholar

Zeng Yan, L.L.-I. & Yan, S. Analysis of liver function and blood routine of the patients with sodium valproate serum concentration above. Chin. J. Drug Appl. Monit. 10(2), 66–70 (2013).

MATHGoogle Scholar

Nasreddine, W. et al. Predicting the occurrence of thrombocytopenia from free valproate levels: A prospective study. Seizure 94, 33–38 (2022).

ArticlePubMedMATHGoogle Scholar

Guo, J. et al. Valproic acid after neurosurgery induces elevated risk of liver injury: A prospective nested case-control study. Ann. Pharmacother. 56(8), 888–897 (2022).

ArticleCASPubMedMATHGoogle Scholar

Download references

Funding

This research was funded by Shanxi Provincial Social Development Project (ID: 201903D321033).

Author information

Author notes

Gang Cheng, Xinfeng Cai and Tianning Zhang contributed equally to this paper.

Authors and Affiliations

Department of Neurosurgery, Shanxi Provincial People’s Hospital, Taiyuan, Shanxi, People’s Republic of China

Gang Cheng, Jiuhong Ma & Xiufeng Zhang

Department of Pharmacy, Shanxi Province Cancer Hospital/Shanxi Hospital Affiliated to Cancer Hospital, Chinese Academy of Medical Sciences/Cancer Hospital Affiliated to Shanxi Medical University, Taiyuan, Shanxi, People’s Republic of China

Xinfeng Cai

Department of Pharmacy, Yunnan Infectious Diseases Hospital, Kunming, Yunnan, People’s Republic of China

Tianning Zhang

Department of Pharmacy, Shanxi Provincial People’s Hospital, Shuangtasi Street 59#, Taiyuan, Shanxi, 030012, People’s Republic of China

Jinlin Guo

Department of Pharmacy, Beijing Friendship Hospital, Capital Medical University, Beijing, People’s Republic of China

Xingang Li

Authors

Gang Cheng

View author publications