A Study of Surgical Approaches to Retinal Vascular Occlusions
Total Page:16
File Type:pdf, Size:1020Kb
SURGICAL TECHNIQUE A Study of Surgical Approaches to Retinal Vascular Occlusions William M. Tang, MD; Dennis P. Han, MD Objective: To develop a surgical approach to retinal vas- nulations of central retinal arteries were successful in 0 cular occlusive diseases. of 2 procedures, and cannulations of central retinal veins were successful in 2 of 4 procedures. Arteriovenous Methods: Surgical manipulations were performed on the sheathotomies were successful in 4 of 7 procedures. In retinal vasculature to explore the feasibility of retinal vas- the in vivo model, surgical penetration of retinal blood cular surgery. In a human cadaver eye model (25 proce- vessels was accomplished in 5 of 6 eyes. Immediately post- dures, 21 eyes), we performed (1) cannulations of retinal operatively, thrombus formation with obstruction of the blood vessels with a flexible stylet and (2) arteriovenous retinal vasculature was observed. At 2 weeks postopera- sheathotomies. Histological findings were correlated with tively, the retinal vasculature was completely patent. surgical outcomes. In an in vivo model (6 eyes, 5 animals), we examined the technical feasibility and anatomical out- Conclusions: Multiple surgical techniques aimed at as- come of surgical penetration of retinal blood vessels. sisting recanalization of occluded retinal vasculature have been evaluated. Retinal vascular surgery has become more Results: Cannulations of branch retinal arterioles were feasible and deserves further investigation. successful in 7 of 9 procedures, cannulations of branch retinal venules were successful in 1 of 3 procedures, can- Arch Ophthalmol. 2000;118:138-143 ETINAL ARTERY and vein oc- endovascular therapy can lead to reversal clusions are among the of retinal vascular occlusions. most common causes of se- The technique of endovascular therapy vere vision loss. In the Blue involves 2 important steps: (1) mechani- Mountains Eye Study, the cal access of a catheter into the vascular lu- prevalence of retinal vein occlusion was men and (2) navigation of the catheter to R 10 found to be 1.6% in adults aged 49 years the desired site of action. We applied this or older.1 Standard treatment is generally technical concept to the retinal vascula- ineffective in restoring blood flow.2,3 ture. In a human cadaver eye model, we de- We are attempting to develop new veloped surgical techniques to (1) gain ac- treatments that might be effective in cess into the retinal vasculature and (2) restoring blood flow in retinal vascular cannulate retinal blood vessels with a flex- occlusions. We believe that significant ible stylet. In an in vivo model, we exam- therapeutic advances must be directed to ined surgical techniques for gaining ac- the obstructive process in the vascular cess into the retinal vasculature. The lumen.4-6 The unique opportunity to technical feasibility and anatomical out- visualize and access the retinal vascula- come of these procedures were evaluated. ture during vitreous surgery led us to Another portion of our study ad- From the Departments of consider developing an intraocular, dresses the recent interest in arteriove- Ophthalmology, Boston endovascular approach to retinal vascu- nous (AV) sheathotomy as a surgical treat- University School of Medicine, lar occlusions. Endovascular recanaliza- ment for branch retinal vein occlusion.11,12 Boston, Mass (Dr Tang), and tion procedures such as percutaneous Investigators have suggested that by me- Medical College of Wisconsin, transluminal angioplasty and regional chanical lysis of the AV sheath, compres- Milwaukee (Dr Han). The authors do not have any thrombolytic delivery have been effective sion on the retinal venule might be re- commercial or proprietary in restoring blood flow in various organ lieved, leading to reversal of vein occlusion. 7-10 interest in any product or systems. Timely intervention often In the present study, we examined the instrument discussed in the results in significant recovery of func- technical feasibility and anatomical out- article. tion. We postulate that intraocular, come of AV sheathotomy. ARCH OPHTHALMOL / VOL 118, JAN 2000 WWW.ARCHOPHTHALMOL.COM 138 ©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 MATERIALS AND METHODS mounted. In the case of the cannu- seconds via elevation of the infu- lated central retinal vein, tissue sec- sion bottle. Rabbit 1 underwent a bi- RETINAL VASCULAR tions were cut perpendicular to the lateral procedure and was killed im- CANNULATION: HUMAN axis of the optic nerve. Serial sec- mediately by intravenous injection CADAVER EYE MODEL tions were examined to (1) con- of pentobarbital/phenytoin solu- firm the entry of the nylon suture tion (1 mL per 5 kg). Two other rab- Human cadaver eyes were obtained into the vessel lumen, (2) identify bits underwent survival proce- from the Wisconsin Lions Eye Bank, any induced trauma to the endothe- dures. Fundus examinations by Milwaukee, and used within 48 lial lining due to cannulation, and indirect ophthalmoscopy were per- hours of expiration. The mean age (3) confirm the extent of vessel can- formed at weeks 1 and 2 after sur- was 80 years (age range, 63-91 nulation. gery. Fundus photographs were years). The anterior calotte was di- taken at week 2. The rabbits were vided from the posterior calotte at RETINAL VASCULAR then killed with pentobarbital/ the level of the equator. A total of 18 PENETRATION: ANIMAL phenytoin injection. The globes were cannulation procedures (17 eyes) EYE MODELS promptly enucleated. were performed. Adult hound dog eyes were op- Cannulations of branch reti- All animal research protocols were erated on during a nonsurvival ex- nal arterioles, branch retinal ven- reviewed and approved by the Ani- periment simultaneously conducted ules, central retinal arteries, and cen- mal Research Committee of the by cardiovascular physiologists study- tral retinal veins were performed. Medical College of Wisconsin, Mil- ing coronary blood flow. Anesthesia Under microscopic illumination, a waukee. In total, 6 eyes were oper- was maintained by intermittent doses bimanual technique with two 20- ated on: 4 eyes from 3 rabbits and 2 of intravenous pentobarbital (200 mg/ gauge microvitreoretinal (MVR) eyes from 2 dogs. The rabbit was se- kg) and barbital (26 mg/kg). Sys- blades was used to create an open- lected because of the presence of a temic medications relevant to our sur- ing in the vessel wall. Approxi- retinal vascular system. The dog was gical procedure included intravenous mately 3 to 4 disc diameters away selected because the caliber of its heparin. Surgical penetration of a reti- from the optic disc, an MVR blade retinal blood vessels more closely ap- nal venule was performed as de- was introduced to elevate a branch proximates that in humans. scribed earlier. Both eyes were enucle- retinal vessel from the surface of the Three-month-old New Zea- ated immediately after surgery. retina. Then, with the first MVR land white rabbits (2.4-3 kg) were After preservation, the poste- blade acting as a platform beneath obtained through the Animal Re- rior calotte was isolated. A surgical the retinal vessel, a second MVR search Facility at the Medical Col- blade was used to cut perpendicu- blade was used to make a small lon- lege of Wisconsin. Ketamine (44 mg/ lar to the direction of the retinal gitudinal arteriotomy/phlebotomy kg) and xylazine (5 mg/kg) were blood vessels to remove tissue 1⁄2 to incision. In the case of cannulation injected intramuscularly to pro- 1 disc diameter peripheral to the ves- of the central retinal artery/vein, the vide sedation and akinesia, respec- sel penetration site. The specimen same technique was used with the tively. Standard sterile surgical tech- was then embedded in paraffin, with arteriotomy/phlebotomy performed niques were employed. Illumination the cut edge marked for keratome on a first-order retinal arteriole/ was provided by the coaxial light sectioning (section thickness, 5 µm). venule immediately distal to the cen- source from an operating micro- For a distance of 1000 to 1500 µm tral bifurcation. A 10-0 black mono- scope. A 2.5-mm infusion cannula spanning the vessel penetration site, filament nylon suture held by a pair was placed through the pars plicata every fifth section was mounted. of smooth forceps was then used to at the inferotemporal quadrant. The Otherwise, every 10th to 100th sec- cannulate the retinal vessel through retina was visualized through an ir- tion was mounted. Serial sections the arteriotomy/phlebotomy open- rigating contact lens. An MVR blade were examined to determine the in- ing. Cannulation was considered suc- or a sharpened metal wire held by a tegrity of the retinal vasculature, as cessful if the nylon suture could be pair of intraocular forceps was in- well as the presence of associated advanced freely within the lumen of troduced through the pars plicata. retinal abnormalities. the blood vessel. A retinal vessel along the temporal After successful cannulation, myelin wing, at approximately a AV SHEATHOTOMY: HUMAN the specimen was preserved and third of a disc diameter from the op- CADAVER EYE MODEL photographed. Under a dissecting tic disc, was penetrated. (Because of microscope, a block of the poste- the similarity in color and size of the Human cadaver eyes were used as de- rior eye wall was isolated to in- retinal arterioles and venules, the scribedearlier.SevenAVsheathotomy clude the entire course of the can- retinal vessel selected for penetra- procedures (6 eyes) were performed. nulated retinal blood vessel. The tion was not specified as an arteri- Arteriovenous crossings were se- specimen was then embedded in par- ole or a venule.) Visualization of a lected from a first- or second-order affin. Tissue sections (5 µm thick) small burst of blood extravasating arteriole at a location where the were cut by a keratome in a direc- into the vitreous cavity indicated blood vessel was sufficiently filled tion roughly perpendicular to the successful penetration.