Venlafaxine overdose treated with extracorporeal life support

Personalized venlafaxine dose prediction using artificial intelligence technology: a retrospective analysis based on real-world data

BACKGROUND

Venlafaxine dose regimens vary considerably between individuals, requiring personalized dosing.

AIM

This study aimed to identify dose-related influencing factors of venlafaxine through real-world data analysis and to construct a personalized dose model using advanced artificial intelligence techniques.

METHOD

We conducted a retrospective study on patients with depression treated with venlafaxine. Significant variables were selected through a univariate analysis. Subsequently, the predictive performance of seven models (XGBoost, LightGBM, CatBoost, GBDT, ANN, TabNet, and DT) was compared. The algorithm that demonstrated optimal performance was chosen to establish the dose prediction model. Model validation used confusion matrices and ROC analysis. Additionally, a dose subgroup analysis was conducted.

RESULTS

A total of 298 patients were included. TabNet was selected to establish the venlafaxine dose prediction model, which exhibited the highest performance with an accuracy of 0.80. The analysis identified seven crucial variables correlated with venlafaxine daily dose, including blood venlafaxine concentration, total protein, lymphocytes, age, globulin, cholinesterase, and blood platelet count. The area under the curve (AUC) for predicting venlafaxine doses of 75 mg, 150 mg, and 225 mg were 0.90, 0.85, and 0.90, respectively.

CONCLUSION

We successfully developed a TabNet model to predict venlafaxine doses using real-world data. This model demonstrated substantial predictive accuracy, offering a personalized dosing regimen for venlafaxine. These findings provide valuable guidance for the clinical use of the drug.

[Retrospective analysis of venlafaxine-induced hypoglycemia in patients with overdose]

INTRODUCTION

Recent publications relate the presence of hypoglycemia in venlafaxine (VLX) poisoning depending on the dose. Our aim was to analyze the clinical characteristics of patients who presented hypoglycemia induced by VLF overdose.

PATIENTS AND METHODS

Retrospective study carried out in the Balearic Islands (2020-2023).

INCLUSION CRITERIA

serum concentrations of VLX + O-desmethyl-venlafaxine (O-VLX)>800 ng/mL. The characteristics of patients with and without hypoglycemia were compared.

RESULTS

Twenty-one patients were included, 8 (38.1%) with hypoglycemia. No differences were found in the doses referred to in both groups. Peak concentrations of VLX + O-VLX (ng/mL) were 9,783 [4,459-17,976] in patients with hypoglycemia and 1,413 [930-1,719] in patients without hypoglycemia (p<0.0001). The presence of hypoglycemia was associated with: lower age and level of consciousness; and higher frequency of suicide attempts, seizures, mydriasis, tachycardia and serotonin syndrome, invasive respiratory support, fluid therapy and ICU admission (p<0.05).

CONCLUSIONS

The detection of hypoglycemia in a VLX overdose case is a readily available marker to suspect the severity of the patient. In any case, serum concentrations when available allow us to confirm intoxication.

Overdose and off-label psychotropic prescribing in patients with borderline personality disorder: A retrospective series

OBJECTIVE

Borderline personality disorder (BPD) is common and poses many clinical challenges. Despite limited evidence of effectiveness, psychotropic medications are often prescribed. We aimed to characterise overdose presentations in patients with BPD.

METHOD

This is a retrospective observational series of patients with BPD presenting to a tertiary hospital following an overdose from January 2019 to December 2020. Medical records were reviewed to determine baseline characteristics, overdose details, clinical features, treatment, and disposition.

RESULTS

There were 608 presentations in 370 people (76% female), median age 28 years (range 16-75 years). The majority (331[89%]) of patients were prescribed at least one psychotropic medication, with 129 (35%) being prescribed three or more different psychotropic agents. Of the total prescribed psychotropics, 520/1459 (36%) were for off-label indications. The majority of agents (860/1487[58%]) taken in overdose were prescribed. The commonest drug classes taken in overdose were benzodiazepines (241[16%]) and antipsychotics (229[15%]). Severe toxicity occurred in 99 (16%) cases with either coma (GCS<9) or hypotension (systolic BP <90 mmHg). The commonest agent associated with severe toxicity was quetiapine 39/99 (39%).

CONCLUSIONS

Psychotropic polypharmacy is common in BPD, often with off-label indications. Prescribed medications are commonly taken in overdose. Quetiapine is over-represented both in off-label prescribing and associated harm.

The Role of Angiotensin II in Poisoning-Induced Shock-a Review

Background

Shock in drug poisoning is a life-threatening condition and current management involves fluid resuscitation and vasopressor therapy. Management is limited by the toxicity of high-dose vasopressors such as catecholamines. Clinical trials have shown the efficacy of angiotensin II as an adjunct vasopressor in septic shock. The aim of this review is to assess the use of angiotensin II in patients with shock secondary to drug overdose.

Methods

Medline (from 1946), Embase (from 1947) and PubMed (from 1946) databases were searched until July 2021 via OVID. Included studies were those with shock due to drug poisoning and received angiotensin II as part of their treatment regimen. Of the 481 articles identified, 13 studies (case reports and scientific abstracts) were included in the final analysis with a total of 14 patients. Extracted data included demographics, overdose drug and dosage, angiotensin II dosage, time of angiotensin II administration, haemodynamic changes, length of hospital stay, mortality, complications, cardiac function and other treatment agents used.

Results

Thirteen studies were included consisting of 6 case reports, 6 scientific abstracts and 1 case series. Overdose drugs included antihypertensives (n = 8), psychotropics (n = 4), isopropanol (n = 1) and tamsulosin (n = 1). Out of a total of 14 patients, 3 patients died. Ten patients had their haemodynamic changes reported. In terms of MAP or SBP changes, three patients (30%) had an immediate response to angiotensin II, four patients (40%) had responses within 30 min, one patient (10%) within two hours and two patients (20%) did not have their time reported. Two patients were shown to have direct chronotropic effects within 30 min of angiotensin II administration. The median hospital stay for patients was 5 days (IQR = 4). The time from overdose until angiotensin II administration ranged from 5 to 56 h. Other vasopressors used included phenylephrine, noradrenaline, adrenaline, vasopressin, dobutamine, dopamine, methylene blue and ephedrine. A median of 3 vasopressors were used before initiation of angiotensin II. Twelve patients received angiotensin II as their final treatment.

Conclusions

Angiotensin II may be useful as an adjunct vasopressor in treating shock secondary to drug poisoning. However, the current literature consisted of only very low-quality studies. To truly assess the utility of angiotensin II use in drug-induced poisoned patients, further well-designed prospective studies are required.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13181-022-00885-4.