Introduction

INTACS intrastromal ring segment implantation was first developed to improve mild myopia. In current practice, the procedure is more commonly used to improve irregular corneal astigmatism in patients with keratoconus or post-LASIK kerectasia. INTACS implantation is considered a safe and effective treatment modality for progressive keratoconus [1]. However, the fragile nature of the ectatic cornea, alongside a demonstrated potential for INTACS migration within the stroma, presents a risk of anterior and posterior corneal surface perforation. Of particular concern are the risks of posterior corneal damage and loss of fluid equilibrium that may necessitate penetrating keratoplasty (PKP). Such occurrences are exceedingly rare, and thus far have only been reported in the intraoperative or immediate postoperative period. In this report, we describe the first known case of late-onset INTACS segment intrusion into the anterior chamber causing corneal hydrops 7 years after implantation.

Case Report

A 44-year-old man was referred to our clinic following sudden onset irritation, blurry vision, and loosening of the single, inferiorly placed INTACS device in his left eye. His past ocular history is significant for LASIK surgery in 1996 and INTACS implantation in 2009 for LASIK-induced ectasia. At the time of presentation, his best corrected visual acuity (BCVA) was 20/500 in the left eye and 20/25 in the right eye. Slit-lamp examination revealed intrusion of the INTACS corneal ring segment into the anterior chamber that resulted in pronounced corneal edema, stromal opacification, and aqueous accumulation in the INTACS tunnel. These clinical signs were confirmed by elevated pachymetry (1062 µm) and significant inferior topographic steepening (K max = 63.4 D) (Fig. 1). Anterior segment optical coherence tomography (OCT) indicated a rupture through Descemet’s membrane that compromised the endothelium with epithelial microcystic edema overlying the arc of the ring segment (Figs. 2, 3).

Fig. 1
figure 1

Inferior corneal steepening (left) and pachymetry showing corneal thickening (right) at the time of presentation

Fig. 2
figure 2

Transverse anterior segment OCT image taken at the time of presentation (plane denoted by white arrow in frontal view). Fluid accumulation is clearly visible in the stromal tunnel surrounding the INTACS ring segment. Microcystic edema is present in the epithelium overlying the INTACS ring segment

Fig. 3
figure 3

Sagittal anterior segment OCT image taken at the time of presentation (plane denoted by white arrow in frontal view). Fluid accumulation is clearly visible in the stromal tunnel surrounding the INTACS ring segment. Microcystic edema is present in the epithelium overlying the INTACS ring segment

Progressive corneal thinning was determined to be the precipitating cause exacerbated by mechanical tissue trauma related to habitual eye rubbing. Corneal edema was refractory to 3 weeks of treatment with 5% sodium chloride solution bid and to combination brimonidine/timolol bid. BCVA further worsened to counting fingers at six feet at 1 week from presentation. The persistent edema was determined to be due to perforation of the endothelium by the intrastromal ring segment. We elected to explant the INTACS segment through a 1.1 mm incision placed over the original implantation incision site with a Sinskey hook. Successful liberation of the INTACS device resulted in an egress of fluid from the anterior chamber onto the ocular surface. The anterior chamber became mildly shallow, but maintained fluid. This supported our original hypothesis of the presence of a fistula caused by acute perforation leading to the development of hydrops. A single suture was placed over the incision, and a bandage contact lens was applied under a pressure patch. At the 3-week post-operative visit, BCVA improved to 20/60 in the treatment eye, with significant decrease in corneal edema. Maximum thickness was reduced on pachymetry from 1062 to 814 µm, but topography indicated persistent inferior steepening above the point of perforation (K max = 63.1 D) (Fig. 4). Anterior segment OCT showed full resolution of epithelial microcystic edema (Fig. 5).

Fig. 4
figure 4

Persistent inferior corneal steepening (left) and pachymetry showing reduced corneal thickening (right) after INTACS segment explant

Fig. 5
figure 5

Transverse (above) and sagittal (below) anterior segment OCT images taken 6 weeks after removal of the INTACS ring segment. Note significant reduction in stromal edema and resolution of epithelial microcystic edema. A subtle endothelial break can be appreciated

The patient involved was provided a written informed consent in accordance with the tenets of the Declaration of Helsinki to having their data used for research purposes.

Discussion

Acute corneal hydrops describes stromal edema secondary to disruption of Descemet’s membrane and endothelium [2]. Although the exact etiology of corneal hydrops is unclear, it is often associated with progressive ectatic disease, severe allergic eye disease, and habitual eye rubbing [2]. Rare cases can cause perforation and fistula formation [3]. In keratoconus patients, the onset of hydrops is generally spontaneous and proportional to the degree of corneal steepening [2]. As such, controlling or arresting the progression of ectasia with modalities such as collagen cross-linking (CXL), INTACS implantation, or newer allogenic intrastromal ring segments such as CAIRS is likely to reduce the probability of acute hydrops.

However, INTACS implantation may itself prompt corneal hydrops in rare instances. Güell et al. report an experience with acute corneal hydrops related to an intraoperative break in Descemet’s membrane while positioning an INTACS device in a keratoconic patient [4]. Corneal edema within 1 week of INTACS implantation has been reported and hypothesized to be related to the induction of endothelial dysfunction by the femtosecond laser used to create the INTACS stromal tunnel [5, 6]. Antonios et al. report the only instance of significantly delayed corneal hydrops or edema with the concurrent presence of an INTACS device was in a patient with severe recurring allergic keratoconjunctivitis and keratoconus who underwent CXL and INTACS implantation 3 years prior [7]. There was no evidence of INTACS migration or corneal perforation, and the presence of the INTACS ring appears coincidental rather than causal. Our observed case of corneal hydrops and edema 7 years post-implantation has a clear causal link to the INTACS device due to a perforation in Descemet’s membrane and endothelium.

Multiple instances of INTACS penetration into the anterior chamber during or shortly after implantation have also appeared in the literature [8,9,10]. Similar to the cases involving corneal edema, these instances are hypothesized to be due to an intraoperative insult of Descemet’s membrane and underlying endothelium. This case is distinct because the penetration seems to be due to posterior corneal thinning and habitual eye rubbing. Moreover, although superficial migration and ring segment extrusion are well established, but rare complications of INTACS, the apparent posterior migration and intrusion into the anterior chamber seen in this patient has not been reported [11].

In order to avoid complications associated with shallow segment implantation, aiming for an implantation depth of 70% total corneal thickness is common practice [12, 13]. Using three points above and below the implantation site, we calculated the average depth of our patient’s INTACS segment to be 75.08% total corneal thickness [13]. It is unclear if this depth was part of the surgical plan or if it was due to posterior migration of the implant. While deep implantation helps to prevent ring segment extrusion and epithelial breakdown, posterior migration of the intrastromal ring segment could result in Descemet’s rupture, endothelial compromise, and fistula formation as seen in our case. This is more likely in patients with ectatic conditions where corneal thinning can occur. Additionally, continuous mechanical push from habitual eye rubbing could also have contributed to posterior ring segment migration, especially since the segment appeared non-planar to the endothelial surface.

Conclusion

This case of intrastromal ring intrusion into the anterior chamber 7 years post-implantation indicates the importance of longitudinal follow up of INTACS patients to monitor the change in proximity of the intrastromal ring to the posterior cornea. Treatment of implant-associated corneal hydrops should involve timely removal of the implant. Subsequent migration of endothelial cells to damaged areas could allow for the restoration of pump function and result in improved visual outcomes.