# Limbal versus Pars plana extraction of posterior segment IOFBs using 23-gauge vitrectomy: anatomical and visual outcomes, and predictive factors for postoperative retinal detachment and poor visual prognosis

## TL;DR

This study compares two surgical methods for removing foreign objects from the eye and finds that both methods have similar outcomes, but certain factors increase the risk of complications.

## Contribution

The study provides evidence that the route of extraction does not significantly affect outcomes, while identifying new predictive factors for retinal detachment and poor vision.

## Key findings

- Both limbal and pars plana extraction methods showed comparable anatomical and visual outcomes.
- Impacted IOFBs and longer foreign body diameter were independent predictors of postoperative retinal detachment.
- Poor visual outcome was associated with longer wound length and worse preoperative visual acuity.

## Abstract

Retained posterior segment intraocular foreign bodies (IOFBs) present a surgical challenge with potential for serious complications, including retinal detachment (RD) and vision loss. While pars plana vitrectomy (PPV) is the standard technique, there is no consensus regarding the optimal route for IOFB extraction, whether through a limbal incision or via the pars plana. This study aims to compare anatomical and visual outcomes between the two surgical approaches and to identify risk factors for postoperative RD and poor visual outcome.

This prospective comparative study included 51 eyes with retained posterior segment IOFBs and an attached retina at presentation. Patients were randomly assigned to IOFB extraction via either a limbal incision (n = 26) or a pars plana incision (n = 25), following 23-gauge PPV. All patients were followed for a median of 12 months. Primary outcomes included the incidence of postoperative RD recorded over the entire first postoperative year and best-corrected visual acuity (BCVA) at one year. Univariate and multivariate logistic regression analyses were performed to identify predictors of RD and poor visual outcome, defined as BCVA worse than 0.8 logMAR.

Postoperative RD developed in 8 eyes (15.7%): 2 eyes (7.7%) in the limbal group and 6 eyes (24%) in the pars plana group (P = 0.14). Impacted IOFB (P = 0.042) and longer foreign body diameter (P = 0.032) were independent predictors of RD. Both groups showed significant improvement in BCVA from baseline (P < 0.001), with no significant difference between groups at any follow-up point. Poor visual outcome was independently associated with longer wound length (P = 0.011) and preoperative VA ≤ 1.6 logMAR (P = 0.005). The route of extraction was not a significant predictor of anatomical or functional outcome.

IOFB extraction via the limbus or pars plana using 23-gauge PPV provides comparable anatomical and visual outcomes. Impacted IOFBs and larger foreign bodies increase the risk of postoperative RD, while wound length and poor baseline visual acuity are strong predictors of poor final visual outcome. Early recognition of these predictors is important for optimizing surgical planning and patient counseling.