Intraoperative haemorrhagic choroidal detachment during MicroIncision Vitrectomy Surgery (MIVS): a unique mechanism

Abstract

Haemorrhagic choroidal detachment (HCD) is a dreaded intraoperative complication of ophthalmic surgery, despite being rare. Multiple systemic and ocular risk factors of HCD have been reported. Acute hypotony during surgery is considered one of the most important precipitating factors. We herein describe a series of events during pars plana vitrectomy surgery for management of rhegmatogeneous retinal detachment which lead to localised HCD. We believe that the pathogenesis of localised HCD is related to compressive decompressive forces during the final tie of the encircling element after fluid air exchange.

Background

Haemorrhagic choroidal detachment (HCD) is a rare, but serious intraoperative complication which can develop intraoperatively.¹ It has been reported during various intraocular procedures and is usually associated with marked fluctuations in intraocular pressure (IOP).^{2 3} The reported overall incidence of intraoperative HCD in ocular surgeries is 0.19%.³ The incidence varies with different surgical procedures, with reports of 0.15% for glaucoma surgeries, 0.16% for lens related procedures, 0.41% for retinal and vitreous procedures and as high as 0.56% for keratoplasty procedures.³ During pars plana vitrectomy (PPV), the incidence of HCD for 23G PPV has been estimated to be higher (1.18%) than 20 GPPV (0.4%–0.5%).^{4 5} However, studies have shown that small gauge vitrectomy probably carries similar risk profile as 20 gauge vitrectomy.⁶ Reported ocular risk factors associated with development of intraoperative HCD include high preoperative IOP, high myopia, aphakia, pseudophakia and previous history of glaucoma.^{3 7} Systemic risk factors comprise advancing age, atherosclerosis, hypertension, diabetes and intraoperative tachycardia.^{3 8 9} Acute intraocular hypotony is believed to precipitate the development of serous choroidal detachment (CD) and subsequently HCD in some cases.⁴ We herein report a case which developed a sudden intraoperative localised HCD during 25 gauge PPV for repair of a rhegmatogeneous retinal detachment in a pseudophakic patient. We hypothesise that the compression and decompression forces during the tying of the 240 silicone encircling element induced the occurrence of this phenomenon.

Case presentation

A 56-year-old man presented to our retina clinic with sudden onset reduction in visual acuity in his left eye (OS) which had started 5 days before. Best-corrected visual acuity (BCVA) in the right eye (OD) was 6/6 and hand motion in OS. Anterior segment evaluation in OD revealed nuclear sclerosis grade 1 and in OS showed the presence of a posterior chamber intraocular lens in the capsular bag. Fundus examination in OD was unremarkable and in OS revealed the presence of a total rhegmatogeneous retinal detachment with multiple nasal horseshoe tears. A PPV along with a 240 encircling element procedure was planned. Axial length in OD and OS was 22.5 mm and 22.2 mm, respectively. The patient had no known systemic illnesses such as hypertension, diabetes mellitus or coronary artery disease. Surgical steps involved the passage of an encircling element using 5–0 polyester scleral sutures followed by PPV and thorough vitreous base dissection. A drainage retinotomy was made nasal to the optic disc and fluid air exchange was performed. Once the retina was attached, laser retinopexy was performed around the tears and the retinotomy site. The surgery was uneventful until the final tie of the encircling element. When the assistant was asked to pull both the ends of the encircling element, the assistant pulled both the ends with excessive force causing marked compression of the globe. The assistant was asked to immediately release both the ends in order to relieve the globe compression and avoid a very high buckle indent and then the encircling element was tied with careful titration. Intraocular examination at the end of the surgery showed development of a localised CD inferiorly reaching up to the inferior arcade along with development of a subretinal bleed near the retinotomy site (figure 1A). The infusion cannula was checked immediately and was found to be inside the vitreous cavity without evidence of any retraction or blockage. Finally, oil tamponade was performed, followed by a strict prone position.

Figure 1.

(A) Intraoperative image showing the presence of large inferior choroidal mound with associated subretinal bleed near retinotomy site. (B) Postoperative fundus colour photograph and USG at day 1 suggestive of inferior choroidal mound (left panel, red arrow) and haemorrhagic component within the choroidal detachment (right panel, yellow arrow). (C) Postoperative fundus colour photograph and USG at day 7 suggestive of decrease in height of inferior choroidal detachment (left panel, red arrow) and resolution of haemorrhagic component within the choroidal detachment (right panel, yellow arrow). USG, ultrasonography.

Investigations

Postoperative examination at day 1 revealed the presence of a large choroidal elevation inferiorly with attached retina under oil (figure 1B, left panel). IOP was 14 mm Hg and anterior chamber depth was normal. Ultrasonography (USG) confirmed the presence of haemorrhagic CD as against a serous CD (figure 1B, right panel).

Treatment

Oral steroids (tablet prednisolone 50 mg) along with topical steroids (betamethasone 0.1% 6 t/d), cycloplegics (atropine 1% three times a day) and topical antibiotics (moxifloxacin 0.5% 4 t/d) were started and the patient was instructed to maintain a strict prone position.

Outcome and follow-up

At 1 week postoperatively, there was reduction in height of HCD (figure 1C), which was resolved completely at 2 weeks. At 1 month of follow-up, the retina was still attached and the BCVA of OS improved to 6/36. Postoperative IOP was 14 mm Hg at 1 week, 12 mmHg at 2 weeks and 16 mmHg at 1 month follow-up.

Discussion

Suprachoroidal space is a potential space between the choroid and the sclera. The presence of fluid in this suprachoroidal space can develop into CD. The presence of clear fluid into the suprachoroidal space is classified as choroidal effusion or serous CD, whereas the presence of blood in this potential space is classified as HCD. The extent of HCD can range from a localised detachment to a vision threatening expulsive choroidal haemorrhage. Acute choroidal effusion and expulsive haemorrhage in fact may represent two stages of the same process.¹⁰ Expulsive haemorrhage may often begin as choroidal effusion that turns to haemorrhage as the choroidal and ciliary vessels stretch and rupture. It may be difficult to determine intraoperatively whether the CD is serous or haemorrhagic.¹⁰

The most frequently proposed mechanism for development of HCD is the presence of initial hypotony, with or without a pre-existing serous detachment causing stretching and rupture of ciliary or choroidal vessels.^{7 11 12} The extent of HCD may be limited by a subsequent rise in IOP, intraluminal pressure or by the clotting of the blood in suprachoroidal space.

In the era of MicroIncision Vitrectomy Surgery, prolonged acute hypotony is not frequently found during PPV, despite marked fluctuations in IOP being possible especially with non-valved cannulas. There are other factors like excessive cryotherapy or choroidal perforations during passage of scleral sutures which might contribute to the development of CD intraoperatively.¹³ Tarantola et al have shown retraction of infusion cannula as the most common cause of both serous and HCD during 23G PPV.⁴ Both retraction and blockade of infusion cannula lead to a transient hypotony causing intraoperative choroidal effusion. All cases of HCD reported by Tarantola et al were localised and resolved spontaneously during the post-operative follow-up period.

We propose a new mechanism involved in pathogenesis of intraoperative HCD. First, our case neither had acute hypotony nor retraction of cannula during the surgery prior to development localised HCD. Second, HCD due to cannula retraction is usually localised in the quadrant where the infusion cannula is placed,⁴ and localised HCD in our case was away from the site of infusion cannula.

We believe that pulling both the ends of the encircling element very tightly by the assistant caused an acute compressive force on the air filled globe causing compression of the large choroidal vessels along with an outflow obstruction in the vortex veins leading to sudden increase in choroidal congestion. This was followed by release of both the ends of the encircling element, which lead to the outwards recoiling of the choroidal vessels with subsequent stress and rupture. As this was a very acute event lasting for a very short time, the resultant bleeding into the suprachoroidal space from these large choroidal vessels led to a localised HCD. Figure 2 depicts the proposed mechanism for HCD in our case. USG during the postoperative period showed inferior CD with moderate to high echogenicity suggestive of haemorrhagic nature of CD. Acute compressive and decompressive forces probably also induced subretinal bleeding near the retinotomy site due to rupture of retinal vessels.

Figure 2.

(A) Proposed mechanism depicting acute compressive (B) and decompressive forces (C) during final tying of encircling element in air filled vitreous cavity causing a localised HCD (D). HCD, haemorrhagic choroidal detachment.

Conclusion

This report presents a new mechanism for development of intraoperative HCD. Tying of the encircling element during retinal detachment surgery can cause acute compressive and decompressive forces on the globe. This warrants the need for careful titration during the final tie of the encircling element to prevent any dreaded complication intraoperatively. Performing the final tie before fluid air exchange is another option one can consider to avoid such a complication, as compression and decompression in a fluid filled cavity will be less than in an air filled vitreous cavity.

Learning points.

• This case highlights a new mechanism for the development of haemorrhagic choroidal detachment (HCD) in the era of MicroIncision Vitrectomy Surgery. Sudden compression followed by decompression during titration of the encircling 240 element at the time of the final tie in an air filled eye can cause development of HCD.

• Careful titration of encircling elements and avoiding acute compression and decompression of the globe at the time of final tie can prevent such dreaded intraoperative complications.

• One can also consider performing the final tie before fluid air exchange as the impact of compression and decompression forces tends to be higher in air filled eyes.

Ethics statements

Patient consent for publication

Obtained.