Over the past 2 decades, thanks in part to the development of optical coherence tomography, clinicians have gained increasing understanding of the critical role played by the vitreous in macular and retinal vascular disease. Age-related posterior vitreous detachment (PVD), once thought to be an acute event with precipitous onset and rapid progression, is now known to progress slowly over many months or years before its culmination at the time of vitreopapillary separation. In most individuals, the early stages of PVD are physiologic—uncomplicated and asymptomatic until vitreous separation from the optic disc margin produces symptoms and signs of a Weiss ring. However, in a minority of patients, early-stage PVD is pathologic and may cause a variety of macular complications. These complications result from the tractional effects of perifoveal PVD with vitreomacular adhesion, from the process of vitreoschisis, in which cortical vitreous remnants on the retinal surface after vitreous detachment become scaffolds for fibrocellular proliferation, or both.

It is now apparent that a significant number of the disorders commonly treated by vitreoretinal specialists are caused or exacerbated by these 2 pathologic mechanisms associated with early-stage PVD. Fibrocellular organization of vitreous remnants left on the retinal surface during vitreoretinal separation is considered by most authorities to be the likely cause of idiopathic epiretinal membrane. Idiopathic lamellar and full-thickness macular hole, tractional cystoid macular edema, and the vitreomacular traction syndrome are caused directly by the tractional effects of early-stage (perifoveal) PVD with vitreomacular adhesion. Other conditions, such as myopic traction maculopathy, diabetic macular edema, and possibly neovascular AMD, are not caused primarily, but may be exacerbated, by pathologic features of PVD. Partial PVD also plays an important pathogenic role in proliferative retinopathies and in the vitreopapillary traction syndrome.

Until now, vitreoretinal surgery has been the only treatment option for eyes with substantial vision loss caused by vitreomacular disorders. Although usually effective at improving vision, surgery is costly and has inherent risk. Vitreous surgery, performed with or without peeling of the internal limiting membrane, also is limited by its inability to restore a normal physiologic state at the retinal surface. For example, clinical and histologic studies have shown that thin patches of vitreous cortex commonly are found on the retinal surface after vitrectomy and may organize into fibrocellular epiretinal membranes. Attempts to eliminate such remnants by internal limiting membrane peeling may increase the risk of surgical complications and visual morbidity. Given the inherent limitations of vitrectomy, there is longstanding interest in developing pharmacologic methods for the nonsurgical induction of PVD, a technique known as pharmacologic vitreolysis .

In theory, pharmacologic vitreolysis represents an attractive alternative to surgery, given its potential for inducing a clean and complete PVD, thereby releasing traction while avoiding cortical vitreous remnants that may lead to fibrocellular membrane formation or recurrence. Compared with surgical vitrectomy, PVD induction by a safe and effective pharmacologic agent would be expected to be safer and easier, cheaper, and possibly more effective at providing fast visual rehabilitation with optimal and stable visual outcomes. These advantages could well allow earlier intervention in disease progression, before visual function has dropped to the level that would justify surgical risk.

Pharmacologic agents are candidates for vitreolysis if they have the ability to induce vitreous liquefaction, weaken the vitreoretinal adhesion, or both. A variety of agents have been studied to date, including collagenase, chondroitinase, dispase, hyaluronidase, nattokinase, plasmin, tissue plasminogen activator, Vitreosolve (Vitreoretinal Technologies Inc, Irvine, California, USA), arginine-glycine-aspartate peptides, and ocriplasmin (formerly microplasmin). Development of most of these agents has been hindered or abandoned by concerns about toxicity, lack of efficacy, or both. However, 2 phase 3 clinical trials of ocriplasmin in patients with symptomatic vitreomacular adhesion recently were completed. After a single intravitreal injection, vitreomacular separation occurred more often in eyes treated with ocriplasmin than in placebo-injected eyes (26.5% vs 10.1%). Based on these results, the Food and Drug Administration Dermatologic and Ophthalmic Drugs Advisory Committee has recommended that ocriplasmin be granted approval for symptomatic vitreomacular adhesion. However, the less-than-robust results of the ocriplasmin trials point to the complexity of pharmacologic vitreolysis and suggest that the ideal vitreolytic agent (or combination of agents) has yet to be identified.

It has long been known that intravitreal gas injection can induce PVD and potentially can treat vitreomacular disorders such as idiopathic macular hole. In this issue of the Journal , Rodrigues and associates report the results of a small retrospective interventional case series involving 15 eyes with symptomatic vitreomacular traction that underwent a single intravitreal injection of perfluoropropane gas as primary treatment. Complete resolution of vitreomacular traction was achieved in 40% of eyes by 1 month after treatment and in 60% of eyes within 6 months of treatment, typically with restoration of a normal foveal contour. Responder analysis showed that anatomic success rates were higher in eyes with smaller vitreomacular adhesion sizes, less maximal foveal thickness, and low vitreous face reflectivity. Disappointingly, the mean final visual acuity in this small cohort was unchanged from baseline, which may be attributable in part to the fact that nearly half the patients had traction diabetic macular edema. No retinal tears or other serious complications developed as a consequence of either gas injection or PVD development. Although the study is limited by its retrospective design, small size, and relatively short follow-up, it provides interesting pilot data suggesting that pneumatic PVD induction, which may be termed pneumatic vitreolysis , may be a reasonable alternative to vitrectomy for select patients with vitreomacular traction in a variety of clinical contexts.

When treating a patient with vitreomacular traction causing significant visual loss, clinicians likely soon will have 3 management alternatives in addition to observation: vitreous surgery, pharmacologic vitreolysis with ocriplasmin, or pneumatic vitreolysis. In the absence of comparative clinical trials, we currently have little firm scientific data to guide our decision about which method to use. However, several perspectives can be offered based on what we know at this time. First, vitreous surgery currently remains the gold standard for treating significant vitreomacular disorders and likely will continue for some time to be the preferred treatment in eyes with large vitreomacular adhesion sizes, an epiretinal membrane component, or both. Epiretinal membrane anchors the vitreous to the retina, and thereby inhibits both pharmacologic and pneumatic vitreolysis.

Second, it is reasonable to consider a less invasive initial step, either ocriplasmin or gas injection, in patients with a small vitreomacular adhesion size and no significant epiretinal membrane, with or without a small macular hole. Data from several small pilot studies suggest that PVD can be induced more consistently in such eyes with an intravitreal injection of low-cost perfluoropropane gas than with a single ocriplasmin injection. Nevertheless, many clinicians likely will choose ocriplasmin as initial treatment because of the randomized clinical trial data supporting its use and the lack of level 1 evidence for the safety and efficacy of gas injection. That said, it is not difficult to predict that there will be some level of disappointment in ocriplasmin, given its relatively low success rate, and studies of combined or sequential pharmacologic and pneumatic vitreolysis would be an interesting next step.

Finally, the perfect vitreolytic drug—capable of inducing PVD consistently with a clean retinal surface and no toxicity concerns—would be the preferred treatment, if it were available, for virtually all cases involving significant vitreomacular traction. Such an agent almost certainly would be superior in efficacy and safety to both vitreous surgery and pneumatic vitreolysis and would enjoy cost advantages over vitreous surgery and the cataract surgery that typically follows. Until this silver bullet is developed, a thoughtful and tailored approach to vitreomacular disorders—one that considers efficacy, safety, cost, and individual patient characteristics—is needed.