As a vitreoretinal surgeon working in a tertiary care setting, I consider treatment of proliferative vitreoretinopathy (PVR) to be the most demanding and difficult surgical procedure that I perform. PVR is the most common reason for failure of surgery for rhegmatogenous retinal detachment, and it’s often responsible for multiple failures. Detachment-related proliferation of RPE cells leads to formation of both epiretinal and subretinal membranes that contract the retina, producing new breaks and recurrent detachments. With no way to prevent the cell proliferation, and no pharmacologic treatment proven to be effective, surgery to dissect and peel the membranes and flatten the retina is currently the best way to reattach it and prevent further vision loss.

Visual Prognosis

The visual prognosis for eyes in which PVR has caused recurrent detachment of the retina is poor but dependent on several factors. Surgical reattachment of the retina should be attempted in most cases. Reattachment surgery should be visually beneficial except in those cases that have developed afferent pupillary defect (APD), which will have a more guarded visual prognosis. For eyes with no APD, I aim to perform repair or re-attachment as soon as possible, before the PVR becomes more aggressive. That said, if PVR involves only a small quadrant of the retina and the rest remains attached, the timing of surgery is less urgent.

In my experience, the most important prognostic indicator of vision following PVR surgery is preoperative visual acuity and whether the preceding detachment(s) was macula-on or macula-off. Typically, eyes that come to me with 20/50 to 20/70 visual acuity most likely hadn’t had macular involvement. I expect surgery can improve vision by at least two lines in that scenario. Eyes that present to me with 20/200 or worse visual acuity have likely had macular involvement, and if not APD, probably irreversible RPE changes in the macula. In that scenario, I expect 20/80-20/100 to be the best vision we can achieve with surgery.

I’m careful to ensure patients understand the difference between anatomic and visual surgical outcomes. I want to avoid at all costs giving them false hope regarding their vision potential, especially given that they’ve already undergone one or more difficult surgeries. I provide my honest assessment of their prognosis and let them decide whether to have another surgery. If the decision is not to proceed, we focus on keeping the eye pain-free and monitoring it to make sure IOP stays controlled and inflammation doesn’t develop.

In the OR

Although OCT can show whether and to what extent the retina and macula are detached prior to another surgery for PVR, only clinical exams can indicate the extent of the PVR itself. As I examine PVR eyes prior to surgery, I note the area that’s most severely affected, which is an area I can address first during surgery to begin making the contracted retina more mobile.

Supporting the peripheral retina in PVR is crucial, which is why, in my opinion, a scleral buckle is a necessity. Even in the setting of a failed previous surgery, the buckle will have at least served to reduce the retinal foreshortening that the PVR would have caused had the buckle not been in place. If no buckle is in the eye when a patient reaches me for surgery, I always put one in. I usually use a Type 240 band, placed low in order to support as much of the vitreous base and peripheral retina as possible.

Placing the scleral buckle is my first step during surgery, followed by the core vitrectomy, which I perform at 7,500 cuts per minute with the Constellation Vision System (Alcon). Because the proliferative membranes sometimes can be barely visible, one of the biggest challenges in surgery for PVR is knowing how much to peel. Yet it’s of the utmost importance to remove as much vitreous as possible. Therefore, it’s rare that I don’t inject diluted triamcinolone to stain the vitreous to help ensure I don’t leave any behind. Membranes left behind will only re-accumulate.

The more severe the PVR, the more likely I am to use a perfluorocarbon (PFO) liquid to flatten as much of the retina as possible. PFO will flatten what needs to be flattened centrally, and any area in the periphery it fails to flatten is the most severely diseased and, therefore, an ideal location to begin peeling. In the periphery, I tend to do quite a bit of vitreous shaving at the edges of the retinal breaks or detachment because that’s where PVR originates. The preset shave mode available on the Constellation is very helpful for this. Consisting of low vacuum yet a high cut rate, shave mode allows me to safely work very close to the retinal surface. The safety of this close work is also enhanced because the location of the cutter port is close to the cutter’s end, which has a stabilizing effect on the intraocular environment. I utilize other preset Constellation modes as I proceed with surgery. Scissor mode is very useful for cutting and peeling membranes and is also an effective substitute for various types of scissors that in the past would have to be moved in and out of the eye. Most of the necessary peeling and cutting can be accomplished with the vitrectomy cutter alone. All of the preset modes raise or lower the cut rate and other parameters in such a way as to maximize the safety of the chosen maneuver. However, the surgeon can override a chosen preset at any time during surgery. Both 25- and 27-gauge instrumentation is available for the Constellation. I use the 25-gauge for PVR surgeries most frequently, although an advantage of the 27-gauge is that it enables working between membranes.

Wide-angle viewing systems are also beneficial in surgery for PVR. They improve visualization in the periphery to the extent that a contact lens or scleral depression isn’t necessarily needed. In addition, it’s not unusual for a PVR case to require a bimanual approach for grasping and peeling membranes, in which case the chandelier lighting systems available today are a significant help. They don’t require the surgeon’s hand to hold them in place.

If I encounter an area of PVR that’s not amenable to peeling because it is already integrated into the retina, I may perform a relaxing retinectomy, sealing off the vessels around the PVR with cautery and removing that section of tissue. I’ve found a retinectomy is preferable to leaving that part of the retina intact as it would only impede retinal flattening. The retinectomy also creates a small hole through which I can suction fluid to further assist flattening. The last two steps of the PVR procedure are applying laser and injecting the tamponade. I tend to use silicone oil rather than gas in severe cases because, in my experience, it’s more stable. These eyes require approximately a month of tamponade against the retina to ensure it remains flat, and in the past, I had seen detachments start to recur as soon as C3F8 began to dissipate.

I take what might be considered a conservative approach to face-down positioning following PVR surgery for failed detachment procedure(s). I ask my patients to remain face down for as much time as they can during the first 3 to 5 days postop and to sleep on their side at night. I find this to be very helpful in safeguarding the outcome, especially during the first week. After that, I don’t believe positioning makes much of a difference, if any.

Avoid Making a Difficult Situation Worse

While technology is helpful in executing a safe and effective procedure, the surgeon must remain vigilant of several factors intraoperatively to maximize safety. One key aspect of safety is to avoid penetrating into the eye with the scleral buckle suture. In addition, as previously mentioned, clear visualization of the surgical field is crucial, so the surgeon should be prepared to remove a clouded crystalline lens during the procedure if necessary. Although mild cataract is rarely an issue, 2-3+ nuclear sclerosis can become very cloudy, particularly during PFO/fluid exchange or PFO/air exchange. Patients should be made aware of the possibility of lensectomy, without immediate IOL placement, prior to surgery.

Intraoperative IOP also must be closely monitored as it fluctuates frequently with all the different maneuvers this surgery requires. The surgeon must recognize that each fluctuation is occurring and ensure the pressure doesn’t dip too low, which could cause suprachoroidal hemorrhage, or rise too high, which could create a host of problems, not the least of which could be damage to the optic nerve. The Constellation Vision System assists in this regard because it constantly monitors infusion pressure, and its IOP Compensation feature automatically makes adjustments to keep pressure stable.

Further Improvements to Come

Proliferative vitreoretinopathy remains the leading cause of failure in rhegmatogenous retinal detachment surgery. In the past decade, surgical instrumentation, technology, and viewing systems have all improved, but the rate-limiting step in PVR is still the fact that we have no way to control cellular proliferation immediately after retinal detachment and after surgery. Once that occurs, we’ll have a higher success rate for these cases and will better serve our patients. ■

Dr. Lakhanpal, a pioneer in the field of micro-incisional vitrectomy surgery, is managing partner with Eye Consultants of Maryland.