Nanothin Descemet stripping automated endothelial keratoplasty performed on a lacerated cornea after globe rupture: a case report

Abstract

Background

When performing corneal endothelial transplantation, the presence of low intraocular pressure, abnormal host corneal shape, and septal defects between the anterior and posterior chamber may cause poor graft adhesion. We report our experience with Descemet stripping automated endothelial keratoplasty (DSAEK) using a nanothin graft in a case of bullous keratopathy (BK) in a lacerated cornea after globe rupture—a situation that encompasses these risk factors.

Case presentation

An 81-year-old man was admitted to our department after he had ruptured his right eyeball following a fall. On the same day, he underwent sclerocorneal suturing, vitrectomy, and intraocular injection of silicone oil. Because the sclerocorneal laceration was healed and there was no recurrence of vitreous hemorrhage or retinal detachment, the silicone oil was removed, and intrascleral fixation of the intraocular lens was performed 2 months later. However, as BK progressed, DSAEK was performed. We observed low intraocular pressure (approximately 5 mmHg), an abnormal corneal shape, and a septum defect between the anterior and posterior chamber, which could cause poor graft adhesion. The nanothin graft (diameter, 7.75 mm; thickness, 47 μm) gradually developed poor adhesion near the corneal laceration on the temporal side postoperatively. However, during reoperation, the graft was eccentrically placed on the nasal side, and air injection into the anterior chamber was added, which ultimately made it possible to attach the graft without sutures. Visual acuity was 20/2000 preoperatively but was 20/200 3 months postoperatively.

Conclusions

We experienced a severe case of BK in lacerated cornea after globe rupture, wherein nanothin DSAEK was performed with graft reposition, and the graft adhered well to the host cornea, resulting in improved visual function.

Background

When performing corneal endothelial transplantation, the presence of (1) low intraocular pressure, (2) abnormal host corneal shape, and (3) septal defect between the anterior and posterior chamber may cause poor graft adhesion. In such cases, suturing the graft to the host cornea or penetrating keratoplasty (PKP) may be indicated; however, there are concerns that this may increase the risk of irregular astigmatism, infection, and rejection.

We herein report a case of bullous keratopathy (BK) that developed in an eye with a corneal laceration after globe rupture, in which there was concern that graft adhesion would be poor owing to the aforementioned risk factors; however, Descemet stripping automated endothelial keratoplasty (DSAEK) was performed without suturing the graft and host cornea.

Case presentation

An 81-year-old man suffered a right-eye injury and globe rupture owing to a fall and was referred to our department. Examination revealed a sclerocorneal laceration in the temporal half, extending from the superior rectus muscle attachment point to the inferior corneal periphery. Additionally, there was prolapse of the lens and iris, vitreous hemorrhage, and total retinal detachment due to a circumferential retinal tear. Emergency surgery was performed on the same day under general anesthesia. Surgery included sclerocorneal suturing, vitrectomy, and intraocular silicone oil insertion, and the patient was left aphakic. As there was no recurrence of vitreous hemorrhage, and the retina was attached, the silicone oil was removed 2 months later, and intrascleral fixation of the IOL was performed simultaneously.

Corneal edema gradually developed, resulting in BK, and a corneal endothelial transplantation (DSAEK) was performed 4 months later. Figure 1 shows the findings before corneal endothelial transplantation. There was a cleft scar on the temporal half of the limbus from 6 to 12 o'clock (Fig. 1A). Anterior segment optical coherence tomography (AS-OCT) (CASIA2; TOMEY Corporation, Aichi, Japan) examination revealed a septum defect between the anterior and posterior chamber due to traumatic aniridia, although the IOL was well fixed (Fig. 1B). Corneal topography showed that the central corneal thickness had increased to 960 μm (Fig. 1C left). An increase in irregular astigmatism due to abnormal corneal shape was also observed, and the high-order aberrations (HOAs) in the 4-mm pupillary area were 1.45 μm (Fig. 1C right). Best-corrected visual acuity was 20/2000, and intraocular pressure was 6 mmHg, which was low due to decreased ciliary function.

Fig. 1

Findings before corneal endothelial transplantation. A Slit lamp photography shows a cleft scar on the temporal half of the limbus from 6 o'clock to 12 o'clock. B Anterior segment OCT (AS-OCT) (upper: Horizontal, lower: vertical section) shows the intraocular lens (IOL) fixed and a septum defect between the anterior and posterior chamber. C Corneal topography shows central corneal thickness had increased to 960 µm (C left) and 1.45-µm high-order aberrations (HOAs) in the 4-mm pupillary area (C right)

Figure 2 shows the intraoperative findings of DSAEK. To prevent the graft from falling into the vitreous cavity, a single 10–0 nylon needle was sutured to the graft as a lifeline (Fig. 2A). After Descemet's membrane of the host cornea was removed, a nanothin graft (diameter, 7.75 mm; thickness, 47 μm) was inserted into the anterior chamber through a 4.1-mm corneal incision in the 5 o'clock direction, avoiding the corneal laceration site in the temporal half (Fig. 2B). This operation was performed using the retraction method via the BUSIN Glide (Moria, Bourbon-l'Archambault, France) spatula [1]. The graft was deployed with the posterior lifeline in place, and the lifeline was removed after injecting air into the anterior chamber (Fig. 2C). The graft was moved slightly to the nasal side, avoiding the temporal corneal tear area, and massage was performed from the host corneal side to drain the anterior aqueous humor between the graft and host cornea (Fig. 2D). Graft adhesion was confirmed using intraoperative AS-OCT (RESCAN700; Carl Zeiss Meditec, Jena, Germany) (Fig. 2D).

Fig. 2

Intraoperative findings of DSAEK. A A single 10–0 nylon needle was sutured to the graft as a lifeline. B A nanothin graft was inserted into the anterior chamber at the 5 o'clock direction. C The graft deployed with the posterior lifeline (highlighted with a black line) in place (C left), and the lifeline was removed after injecting air into the anterior chamber (C right). D The graft moved slightly to the nasal side (D left), and adhesion to the host cornea was confirmed using intraoperative AS-OCT (D right)

Figure 3 shows the findings after DSAEK surgery. On postoperative day (POD) 1, the air in the anterior chamber had migrated into the vitreous cavity and disappeared (Fig. 3A), and the intraocular pressure was low (4 mmHg). The end of the graft on the temporal side was away from the corneal laceration, but slight adhesion failure was observed when confirmed by AS-OCT (Fig. 3B). Therefore, we performed air reinjection into the anterior chamber, and graft reattachment was confirmed using intraoperative AS-OCT. However, postoperatively, the air bubble easily migrated again to the posterior cavity and did not remain in the anterior chamber. By POD7, the poor adhesion of the graft had progressed further (Fig. 3C). Since the intraocular pressure remained low (3 mmHg), and air tamponade was difficult to maintain, it was considered necessary to suture the graft to the host cornea. Moreover, AS-OCT revealed that the graft was interfering with an uneven area on the endothelial surface that was thought to be residual tissue after the iris tear (Fig. 3C). Therefore, before suturing the graft, we first attempted to shift the graft further toward the nasal side. With the anterior chamber filled with air, the periphery of the graft was hooked with the tip of a 32-G needle to shift it (Fig. 3D). After massaging from the host cornea side to drain the anterior aqueous humor between the graft and host cornea, intraoperative AS-OCT showed that the graft was well-adhered; therefore, the graft was not sutured (Fig. 3E).

Fig. 3

Findings after DSAEK surgery. A On postoperative day 1 (POD1), the air in the anterior chamber had disappeared. B Arrows in AS-OCT (horizontal section) show slight adhesion failure in the end of the graft on the temporal side. C On POD7 after air reinjection, AS-OCT (horizontal section) shows that the poor adhesion of the graft had progressed further, and the graft was interfering with an uneven area showed by an arrowhead on the endothelial surface. D The graft was shifted further toward the nasal side showed by arrows using a 32-G needle. E Intraoperative AS-OCT shows that the graft was well adhered after massaging

Figure 4 shows progress after the graft replacement. One week later, intraocular pressure was still low (4 mmHg), but the adhesion of the graft was maintained (Fig. 4A and B). The transparency of the cornea had increased, and the right corrected visual acuity had improved to 80/2000. Three months postoperatively, the central corneal endothelial cell density was approximately 1200 cells/mm² (data not shown), and with improvement in the corneal edema (Fig. 4C and D), the central corneal thickness had thinned to approximately 540 μm (Fig. 4E left); the corrected visual acuity in the right eye had improved to 20/200. Moreover, several sutures for corneal laceration were removed to adjust the astigmatism, and HOAs were reduced to 0.46 μm (Fig. 4E right).

Fig. 4

Progress after the graft displaced. A One week later, slit lamp photography shows that the transparency of the cornea had increased. B AS-OCT (horizontal section) shows that the adhesion of the graft was maintained. C and D Three months after DSAEK, slit lamp photography (C) and AS-OCT (upper: Horizontal, lower: vertical section) (D) shows improvement in corneal edema. E Corneal topography shows the central corneal thickness had thinned to approximately 540 μm (E left), and the HOAs were reduced to 0.46 um (E right)

Discussion and conclusions

In recent years, applications for corneal endothelial transplantation have been expanding from DSAEK (graft thickness, 100–150 μm)—which transplants the endothelial layer including the posterior stroma—to Descemet membrane endothelial keratoplasty (DMEK; graft thickness, 15–20 μm)—which transplants the endothelial layer only and improves postoperative visual acuity to a higher degree [2]. However, with DSAEK, it is possible to create a graft with a thinner posterior parenchyma (ultrathin, 50–100 μm; nanothin, < 50 μm), and there are reports that nanothin DSAEK can improve postoperative visual acuity close to that of DMEK [3].

In DSAEK for eyes with aniridia, including after a globe rupture, owing to the presence of a septal defect between the anterior and posterior chamber, we must be careful not to allow the graft to fall into the vitreous cavity when inserted. In this case, referring to Yokogawa et al.'s "Lifeline Suture," 10–0 nylon was sutured to the posterior side around the graft as a lifeline, and it was inserted into the eye using the retraction method using a BUSIN glide spatula [4]. At this time, if the graft is grasped and controlled until it is deployed just below the host cornea, the risk of it falling into the vitreous cavity is mitigated, and subsequent position adjustments become easier.

Moreover, when we perform DSAEK for eyes with aniridia, the air in the anterior chamber tends to migrate into the vitreous cavity postoperatively, and in cases where there is coexistence of a low intraocular pressure state due to decreased ciliary function after globe rupture, these may cause poor graft adhesion. Normally, sulfur hexafluoride gas remains in the anterior chamber longer than air. Therefore, in low intraocular pressure conditions, it is thought to be effective for graft adhesion. However, when there is a septal defect between the anterior and posterior chamber, as in this case, it is expected that the gas will escape into the vitreous cavity, making it less effective [5]. Accordingly, nanothin DSAEK, which has a thinner and lighter graft, is considered more advantageous for graft adhesion compared to regular DSAEK [3].

In contrast, regarding irregularities in the host corneal endothelial surface, even nanothin DSAEK was the cause for poor graft adhesion, as in this case. There are also reports that the effect of graft deviation on visual function is unclear [6], and if poor graft adhesion is observed due to endothelial surface irregularity, there is a possibility of adhering the graft without suturing it, even in a state of low intraocular pressure and inadequate air tamponade, by adjusting the position of the graft.

In Japan, there are many complicated cases with concomitant problems in addition to BK that required corneal endothelial transplantation [7], as in this case, but nanothin DSAEK has made it possible to perform corneal endothelial transplantation even in cases where PKP was previously indicated. Furthermore, in recent years, the indication for DMEK has been expanded to include complicated cases, and a surgical procedure using the Safety-net suture method has been reported for traumatic aniridia [8]. However, in this case, the eye also had a corneal laceration, which was accompanied by abnormal corneal shape, irregular endothelial surface, and low intraocular pressure, and it is possible that suturing of the graft edge was required when DMEK was performed. Compared to conventional DSAEK, nanothin DSAEK has a thinner graft and requires more delicate handling as it tends to wrinkle during graft manipulation, which can cause endothelial loss. Nevertheless, the deployment dynamics after the graft is inserted into the anterior chamber are not significantly different from conventional DSAEK, and for complex cases, surgeons who are familiar with DSAEK may choose nanothin DSAEK instead of DMEK. We believe that good postoperative results can be obtained even with the former.

In this case report, we performed nanothin DSAEK for BK in lacerated cornea after globe rupture. Nanothin DSAEK can be considered as a surgical option in complex BK cases with factors associated with poor graft adhesion.