Intra-arterial chemotherapy for retinoblastoma: a performance analysis

To conduct a performance analysis of intra-arterial chemotherapy (IAC).

Comprehensive data search in MEDLINE from inception until December 2023. Peer-reviewed, English-language, clinical studies (retrospective, prospective, trials, non-randomised) and case series. Two authors abstracted data independently using a standardised form. Study heterogeneity calculation and influence statistics with a leave-one-out analysis were carried out. Globe salvage rates (GSR) (overall, primary and secondary), per international classification of retinoblastoma groups and adjuvant intravitreal chemotherapy were carried out. Performance was assessed using learning curve, experience and study variable effects (site, size and year).

416 identified studies met the selection criteria. 43 studies (3174 eyes) were included for random-effects meta-analysis. The statistical test for heterogeneity was significant (p value<0.001). Overall GSR was 73% (95% CI 68 to 77%) and comparable when used as primary (69%, 95% CI 63 to 75%) or as secondary treatment (66%, 95% CI 58 to 74%). GSR was higher (66%, 95% CI 60 to 71%), for group D eyes than for group E eyes (49%, 95% CI 39 to 59%). There was insufficient evidence to conclude the modifying effect of adjuvant intravitreal chemotherapy. Analysis for learning curve effect revealed no improvement in GSR from a unique centre with time. Analysis for the experience effect was precluded due to lack of sufficient data. GSR was not impacted by the national income level of the study site (p value 0.07), study size (<50 eyes vs >=50 eyes, p value 0.449) or year (<=2020 versus >2020, p value <0.0.075).

Improved selection criteria are needed to identify eyes suitable for IAC.