Introduction
Micro-invasive glaucoma surgery (MIGS) has rapidly expanded the surgical armamentarium for glaucoma management over the past decade, offering a favorable safety profile compared with conventional filtration procedures.1 Among MIGS procedures, ab-interno goniotomy and related angle procedures, including procedures performed with the Kahook Dual Blade (KDB; New World Medical), gonioscopy-assisted transluminal trabeculotomy (GATT), and microhook-style knives, reduce IOP by opening the trabecular meshwork and inner wall of Schlemm's canal, thereby enhancing conventional aqueous outflow.2, 3
Despite its relatively favorable safety record, goniotomy is not without complications. Transient hyphema is the most frequently reported early postoperative event, with incidence varying according to the extent of angle incision, anticoagulant exposure, and concurrent lens surgery.4 Retrograde bleeding can occur when IOP falls below episcleral venous pressure (EVP), allowing blood to enter the anterior chamber through Schlemm's canal and collector channels.5 Delayed or intermittent bleeding after angle surgery is uncommon but may have important clinical consequences. Sozen-Delil et al. reported intermittent bleeding from the iridocorneal angle after GATT, accompanied by inferior vitreous hemorrhage and followed by ghost cell glaucoma.6 This report provides an important clinical precedent linking postoperative angle bleeding with subsequent ghost cell glaucoma after angle surgery.
Ghost cell glaucoma, first described by Campbell et al., in 1976, is a form of secondary open-angle glaucoma in which degenerated, less deformable erythrocytes obstruct aqueous outflow.7 Classically, the vitreous cavity serves as a low-turnover reservoir after vitreous hemorrhage, allowing erythrocytes to degenerate before they enter the anterior chamber. However, ghost cells or ghost-cell features may also be encountered in secondary hemorrhagic contexts, including hyphema, trauma, and intraocular surgery.
We report a case of delayed blood reflux-related secondary open-angle glaucoma with ghost cells more than 3 years after ab-interno goniotomy. Paracentesis revealed pressure-dependent reflux, and aqueous pathology supported a ghost-cell component.
Case description
A 73-year-old man with a history of primary open-angle glaucoma (POAG) was referred for evaluation of blurred vision and refractory elevated IOP in the left eye. Approximately 45 months before the current presentation, he had undergone left ab-interno goniotomy. The angle procedure was performed under gonioscopic visualization using a Tanito microhook (TMH; Inami & Co., Ltd., Tokyo, Japan), with incision of the trabecular meshwork over 120°.
The operative course was uneventful, and on postoperative day 1 the left-eye IOP was 13.0 mmHg with a documented 1-mm anterior chamber hyphema. His medical history included hypertension; no coagulopathy or anticoagulant use was identified in the available records.
After an early postoperative IOP fluctuation to 27.0 mmHg on postoperative day 8, the left-eye IOP was 17.0 mmHg at 3 months and 14.0 mmHg at 1 year. Available late follow-up records showed controlled left-eye IOP of 18.0 mmHg at postoperative month 36 and 15.3 mmHg at postoperative month 44. Approximately 2 weeks before admission, the left-eye IOP increased to 45.3 mmHg despite topical therapy.
At admission, the left-eye IOP was 45.1 mmHg. Slit-lamp biomicroscopy showed corneal edema and a small inferior hyphema without active iris or angle neovascularization. Gonioscopy demonstrated an open angle with blood staining near the prior nasal goniotomy site and layered blood at the 6-o'clock angle in the left eye (Fig.
1A and C). Khaki-colored anterior chamber cells were not specifically documented in the contemporaneous clinical record. B-scan showed no clinically evident vitreous hemorrhage.
IOP measurements were obtained using noncontact tonometry (TX-20; Canon, Tokyo, Japan).Fig. 1(A) Gonioscopic image showing blood staining near the prior goniotomy site (arrow). (B) Intraoperative view during anterior chamber paracentesis/decompression showing blood-tinged reflux from the nasal angle corresponding to the prior goniotomy site (arrowhead).
(C) Gonioscopy showing a small inferior hyphema with layered blood at the 6-o'clock angle. (D) Representative routine hematoxylin-eosin-stained direct smear of the aqueous specimen; the pathology report described a small number of degenerated erythrocytes and ghost cells.
A diagnostic and therapeutic anterior chamber paracentesis with anterior chamber washout was performed on the day of admission. A paracentesis was created at the 10-o'clock clear cornea, 0.5 mm inside the limbus, using a 20-gauge scleral puncture knife. During the early phase of paracentesis/decompression, 0.1 mL of aqueous humor was aspirated with a 1-mL syringe and submitted for pathologic examination.
During decompression and washout, blood-tinged reflux was observed from the nasal angle corresponding to the prior goniotomy site, rather than circumferentially around the angle, and was recorded on intraoperative video (Fig. 1B; Supplementary Video 1).
The following is the supplementary data related to this article:Supplementary Video 1Anterior chamber paracentesis and ocular decompression. During decompression, blood-tinged reflux was observed from the nasal angle corresponding to the prior goniotomy site. Multimedia component 1
The aqueous specimen was smeared directly onto glass slides and processed using the pathology laboratory's routine hematoxylin-eosin staining protocol. The pathology report described a small number of degenerated erythrocytes and ghost cells, and a representative direct-smear field is shown in Fig. 1D.
No cell block, special stain, or quantitative cell differential was documented. These findings supported a ghost-cell component but did not establish the exact site or timing of erythrocyte degeneration.
Because anterior chamber paracentesis and washout did not achieve sustained IOP control, the patient underwent left trabeculectomy with anterior chamber reformation approximately 1 week after admission. Mitomycin C 0.4 mg/mL was applied under the conjunctival and scleral flaps for 5 minutes, followed by irrigation with 100 mL of balanced salt solution. The scleral flap was closed with two adjustable 10-0 nylon sutures to permit controlled postoperative outflow titration.
On postoperative day 1 after trabeculectomy, the left-eye IOP was 57.5 mmHg and decreased to 18 mmHg after ocular massage. Subsequent follow-up showed durable IOP control, with left-eye IOP of 12.0 mmHg at postoperative month 6 and 10.0 mmHg at postoperative month 11. At the last available follow-up, the anterior chamber was clear and no recurrent hyphema was documented.
Discussion
We interpret this case as delayed blood reflux-related secondary open-angle glaucoma with a pathology-supported ghost-cell component after ab-interno goniotomy. The key findings were late IOP decompensation, pressure-dependent blood reflux localized to the nasal angle corresponding to the prior goniotomy site, and an aqueous smear showing degenerated erythrocytes and ghost cells.
Ab-interno goniotomy and related angle procedures open the trabecular meshwork and inner wall of Schlemm's canal to improve conventional outflow.2, 3 Retrograde blood entry may occur when IOP falls below episcleral venous pressure.5 In this case, decompression during paracentesis made reflux visible at the nasal angle corresponding to the prior goniotomy site rather than circumferentially around the angle. Although blood reflux can also occur through intact trabecular meshwork when IOP is sufficiently reduced, the localized finding supports the prior goniotomy site as a plausible route of blood entry rather than proving it to be the sole source. Importantly, Sozen-Delil et al. reported intermittent bleeding from the iridocorneal angle after GATT, accompanied by inferior vitreous hemorrhage and followed by ghost cell glaucoma.6 Their report provides a relevant clinical precedent for delayed hemorrhagic complications after angle surgery and subsequent ghost-cell–related IOP elevation.
However, the presence of vitreous hemorrhage in their case represents an important distinction from the present case, in which no clinically evident vitreous hemorrhage was identified.
This distinction makes the mechanism of ghost-cell formation in our case less straightforward. Although this case does not represent the classic presentation of ghost cell glaucoma, the presence of pathology-confirmed ghost cells suggests that these cells may have contributed to the uncontrolled IOP. Classic ghost cell glaucoma usually follows longstanding vitreous hemorrhage, with the vitreous acting as a low-turnover reservoir in which erythrocytes degenerate before entering the anterior chamber.7 In this case, khaki-colored anterior chamber cells were not documented, and B-scan showed no clinically evident vitreous hemorrhage or hyaloid disruption.
Therefore, the precise site and timing of erythrocyte degeneration cannot be established. Intermittent blood reflux may have contributed to the presence of degenerated erythrocytes, but occult or previously resolved vitreous microhemorrhage cannot be excluded.
Interpretation of the aqueous cytology is also limited by the specimen characteristics. Only 0.1 mL of aqueous humor was obtained, and the sample was collected during the early phase of decompression. The specimen was examined qualitatively as a direct smear using routine hematoxylin-eosin staining, without a cell block, special stain, cytospin quantification, or quantitative cell differential.
Therefore, the cytologic findings support a ghost-cell component but do not establish the complete pathogenesis of classic ghost cell glaucoma.
The intraoperative video provides complementary hemodynamic information. It documented blood-tinged reflux during ocular decompression, localized to the nasal angle corresponding to the goniotomy site. This finding supports a pressure-dependent route of blood entry but does not establish where the erythrocytes subsequently degenerated.
Thus, the video and aqueous cytology provide complementary information: the video documents reflux during decompression, whereas the cytology demonstrates the presence of degenerated erythrocytes and ghost cells.
Because anterior chamber washout did not provide sustained IOP control, further glaucoma surgery was required. Pars plana vitrectomy was not favored because no clinically significant vitreous hemorrhage reservoir was identified. Repeat angle surgery or laser treatment would not directly address the prior goniotomy site.
Glaucoma drainage device implantation was the main alternative external filtration procedure, but early overfiltration and hypotony can still occur even with valved or flow-restricting devices.8 In this eye, avoiding uncontrolled early hypotony was a particular concern because decompression had demonstrated pressure-dependent blood reflux. Trabeculectomy with two adjustable scleral-flap sutures was therefore selected to provide durable IOP reduction while allowing controlled early filtration. Postoperative outflow could subsequently be titrated with ocular massage and staged suture adjustment or removal as bleb function and wound healing evolved.9 This represents a case-specific surgical rationale rather than evidence that trabeculectomy is superior to glaucoma drainage device implantation.
In conclusion, delayed blood reflux-related secondary open-angle glaucoma may occur years after angle surgery. In this case, localized pressure-dependent reflux and pathology-confirmed ghost cells suggest that degenerated erythrocytes may have contributed to the late IOP elevation in a non-classical hemorrhagic setting. By creating a direct communication between the anterior chamber and Schlemm's canal, ab-interno goniotomy may produce a long-term hemodynamic vulnerability in which transient reversal of the IOP–episcleral venous pressure gradient permits retrograde blood entry through the incised segment.
CRediT authorship contribution statement
Miao Zhang: Writing – original draft, Formal analysis. Jinfeng Jia: Writing – original draft, Formal analysis. Junwei Tan: Data curation.
Xinbo Gao: Writing – original draft, Supervision, Formal analysis, Data curation.
Ethics declaration
This study was performed in compliance with relevant laws, regulatory frameworks and guidelines where the research took place. The Zhongshan Ophthalmic Center, Sun Yat-sen University granted an exemption. The authors provided the following explanation: Formal ethical approval was exempt by the IRB for this study, as it is a single-case report that involves no experimental intervention or alteration to standard medical management.
This research follows the CARE guidelines and the CARE checklist.
Patient consent
Written informed consent to publish this case report, including clinical images and the intraoperative video, has been obtained from the patient. This report does not contain any personal identifying information beyond what is standard for a de-identified case report.
Authorship
All authors attest that they meet the current ICMJE criteria for Authorship.
Funding
No funding or grant support.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.