Real-World Experience Using Sodium Thiosulfate Pentahydrate Off-Label for Cisplatin Otoprotection in Children, Adolescents, and Young Adults

BACKGROUND

Cisplatin is used to treat solid tumors but causes irreversible hearing loss. Pedmark, a formulation of sodium thiosulfate (STS), is approved to prevent cisplatin-induced hearing loss (CIHL). Prior to approval, non-Pedmark formulations of STS pentahydrate (STS-P) were prescribed off-label for otoprotection and continue to be used in the absence of data.

PROCEDURE

This multicenter retrospective study examined tolerability, toxicity, and hearing outcomes of STS-P used off-label for otoprotection. Exploratory analyses compared toxicity and hearing data in patients receiving STS-P versus the pre-Pedmark investigational formulation (STS-inv) tested in trials.

RESULTS

Fifty-nine patients received STS-P (16 or 20 g/m2). Infusion-related reactions (IRR) occurred in 14% (8/59), more commonly in patients receiving 20 g/m2. No severe adverse events occurred. One patient (2%) discontinued STS-P for IRR. The prevalence of CIHL (International Society of Paediatric Oncology [SIOP] Grade ≥2) at the end of therapy and at the most recent hearing assessment was 30% at both timepoints (12/40 and 8/27, respectively). In exploratory analyses comparing STS-P with STS-inv (n = 14), there was no difference in tolerance or toxicity. In multivariable analysis, a lower risk for CIHL at the end of therapy was found for age ≥5 years, higher dosing of 20 g/m2, and received STS-inv (odds ratio 0.02, 95% confidence interval: 0.0003-0.691, p < 0.01). No difference was present at the most recent exam.

CONCLUSIONS

STS-P off-label for otoprotection following cisplatin was tolerable in a real-world setting across age groups and cancer types. Formal testing in larger studies of different STS formulations is needed to explore possible differences in toxicity and CIHL prevention.

Building a machine learning-assisted echocardiography prediction tool for children at risk for cancer therapy-related cardiomyopathy

BACKGROUND

Despite routine echocardiographic surveillance for childhood cancer survivors, the ability to predict cardiomyopathy risk in individual patients is limited. We explored the feasibility and optimal processes for machine learning-enhanced cardiomyopathy prediction in survivors using serial echocardiograms from five centers.

METHODS

We designed a series of deep convolutional neural networks (DCNNs) for prediction of cardiomyopathy (shortening fraction ≤ 28% or ejection fraction ≤ 50% on two occasions) for at-risk survivors ≥ 1-year post initial cancer therapy. We built DCNNs with four subsets of echocardiographic data differing in timing relative to case (survivor who developed cardiomyopathy) index diagnosis and two input formats (montages) with differing image selections. We used holdout subsets in a 10-fold cross-validation framework and standard metrics to assess model performance (e.g., F1-score, area under the precision-recall curve [AUPRC]). Performance of the input formats was compared using a combined 5 × 2 cross-validation F-test.

RESULTS

The dataset included 542 pairs of montages: 171 montage pairs from 45 cases at time of cardiomyopathy diagnosis or pre-diagnosis and 371 pairs from 70 at-risk survivors who didn't develop cardiomyopathy during follow-up (non-case). The DCNN trained to distinguish between non-case and time of cardiomyopathy diagnosis or pre-diagnosis case montages achieved an AUROC of 0.89 ± 0.02, AUPRC 0.83 ± 0.03, and F1-score: 0.76 ± 0.04. When limited to smaller subsets of case data (e.g., ≥ 1 or 2 years pre-diagnosis), performance worsened. Model input format did not impact performance accuracy across models.

CONCLUSIONS

This methodology is a promising first step toward development of a DCNN capable of accurately differentiating pre-diagnosis versus non-case echocardiograms to predict survivors more likely to develop cardiomyopathy.