Vitreomacular traction (VMT) can cause significant loss of central vision with scotoma and metamorphopsia and can lead to a macular hole. Until recently, retina specialists have had only two options for treating symptomatic vitreomacular adhesion or traction: monitor the condition until the traction spontaneously resolves with improvement in vision or perform a vitrectomy with membrane peel and fluid gas exchange. The latest option to treat VMT is intravitreal injections of ocriplasmin (Jetrea; ThromboGenics, Iselin, NJ).1–7 Ocriplasmin is a proteolytic enzyme indicated for the treatment of symptomatic vitreomacular adhesion. This enzyme cleaves fibronectin and laminin, which are thought to be responsible for the vitreous adherence to the retina.8–9 Adverse effects of ocriplasmin include floaters, conjunctival hemorrhage, eye pain, photopsia, dyschromatopsias, blurred vision, macular hole, reduced visual acuity, visual impairment, and retinal edema. Recent reports have shown a transient disruption in the ellipsoid layer on OCT.9,10 There are rare reports of electroretinogram (ERG) abnormalities after ocriplasmin injection.1,9
Patient 1 had developed VMT that, over 4 months, eventually evolved into a small full-thickness macular hole. Her subjective vision loss was disproportionate to the observed VMT. For this reason a pattern VEP was performed, which documented a significant “functional component” to her vision loss prior to any intervention. It became necessary to test her visual acuity using pattern VEPs subsequently. After an intravitreal injection of ocriplasmin, successful VMT release and closure of the macular hole were achieved. Three days after the intravitreal injection, marked ERG changes were observed under all conditions of testing (Figures 1A and 2A; Table 1): scotopic (rod-isolated), mesopic (maximal), photopic (cone-dominated), and flicker (cone-isolated). Twenty days after injection, scotopic as well as maximal responses continued to be significantly depressed, whereas the photopic and flicker responses began to partially recover but never returned to baseline. Two months after injection, the scotopic responses had worsened even further and were no longer recordable, and the maximal responses continued to be depressed. The photopic and flicker responses continued to improve. The implicit times were unaffected under maximal, photopic, and flicker conditions after injection. Her visual acuity had to be measured using VEP because of the functional component to her vision loss. Her visual acuity by VEP improved from 20/200 to 20/30 along with consistent OCT findings (Figure 2A). After 15 months, she recovered most of her photoreceptor function but did not return entirely back to baseline. Paralleling the ERG, the patient’s ellipsoid layer became indistinct 1 week after ocriplasmin injection. However, the ellipsoid layer became more visible at 6 months and significantly recovered at 15 months. The patient initially reported “darker vision” and photopsias but later noted her vision had improved substantially. The loss of the inner segment/outer segment layer directly correlated to the closure of the macular hole. However, the partial recovery of the ellipsoid layer at months 6 and 15 paralleled the improvement in cone and rod function as well as visual acuity.
(A) Electroretinogram data and images of patients 1 through 3 before and after treatment with ocriplasmin. Demonstrates a successful macular hole closure of patients 1 and 2 as well as rod and some cone function loss. Patient 3 did not experience successful macular hole closure but lost cone photoreceptor function. (B) Results obtained before and after ocriplasmin injection of six patients. No significant changes were seen in patients 4 through 9.
(A) Spectral-domain optical coherence tomography (SD-OCT) of three patients before and after ocriplasmin injection. Patients 1 and 2 experienced successful macular hole closure, whereas patient 3 did not. All three patients experienced disruption of their ellipsoid layer after ocriplasmin injection. (B) SD-OCT of six patients before and after ocriplasmin injection. Patients 4 through 9 displayed no significant changes in photoreceptor function, and closure of the macular hole was unsuccessful. All post-injection OCT images were taken 1 week after ocriplasmin injection.
Electroretinogram Amplitudes of Three Patients Before and After Ocriplasmin Injection
Patient 2 also experienced successful closure of her macular hole after an intravitreal injection of ocriplasmin. Two days after the injection, the patient experienced significant visual acuity improvement. However, she had marked ERG depression noted under all conditions of testing (Figures 1A and 2A; Table 1) along with a loss of the ellipsoid layer on OCT. Implicit times were normal. OCT images documented the closure of the macular hole (Figure 2A). Patient 2 also experienced a partial disappearance of the ellipsoid layer 1 week after the injection. The ellipsoid layer partially recovered at month 6 concomitant with the improvement in the patient’s visual acuity along with recovery of the ERG.
Patient 3 did not experience successful closure of her macular hole but exhibited a moderate reduction in cone and rod function, which recovered after 2 months. The disappearance in the ellipsoid layer 1 week after injection paralleled the reduction in the ERG. However, after 2 months, the ellipsoid layer partially reappeared and the ERG recovered (Figures 1A and 2A; Table 1). Visual acuity recovered after successful macular hole closure with vitrectomy.
Patients 4 through 9
The six other patients had no response to intravitreal ocriplasmin injection either in vitreous release of the macula, visual acuity, or ERG findings. These six patients had VMT with early small macular holes. Although minor changes are visible in Tables 1 and 2 for patients 4 through 9, the changes were not significant and no change in photoreceptor functional was observed. One month after failure of the intravitreal ocriplasmin, they underwent successful vitrectomy with closure of the hole and visual acuity improvement. Patient 4 had a relatively broad VMT, which did not respond to ocriplasmin, and refused surgery. There were no significant changes in the ellipsoid layer when the macular hole did not close in patients 4 through 9 (Figure 2A).
Electroretinogram Amplitudes of Six Patients Before and After Ocriplasmin Injection
Post-market experiences with intravitreal injections of ocriplasmin for VMT report successful closure of macular holes from 20% to 30%.4,5 Our study showed a 25% success rate of macular hole closure. Most studies report a slow improvement in visual acuity after successful use of ocriplasmin. Typically a pocket of subretinal fluid under the closed hole slowly absorbs during a period of months. There are rare reports of permanent vison loss, such as in the case reported by Fahim et al.1 Their single patient showed no macular hole closure and worsening of vision perhaps as the hole enlarged. There was no long-term follow-up information on this case. None of our patients experienced permanent vision loss.
Our report has significant differences compared to others. There are few single case reports of ERG abnormalities with ocriplasmin.2–3,9,11 These other cases have very brief follow-up periods, whereas our study consists of long-term data for nine patients for at least 15 months. We also have pre- and post-ERG data on all patients. Our study showed a 30% occurrence of early and significant ERG changes. If we find this degree of change in a relatively small study, we suspect this is much more prevalent than previously thought. Based on our experience, we suspect that there are likely significant ERG depressions in all patients who experience dyschromoatopsia or “darkening of the vision” and afferent pupillary defects.
We observed that these ERGs changes were most profound in those patients who were successfully treated and achieved hole closure. This suggests that the better this enzyme works, the more likely the macular hole will close and the more likely it is to cause ERG changes and symptoms of “darkened vision,” photopsia, and dyschromatopsia. OCT changes and blurring of the ellipsoid layer seem to parallel the ERG abnormalities. We believe these ERG changes are not due to the mere mechanical aspect of the vitreous interface peeling away from the retina, because we did not find any such changes in the patient who proceeded to undergo vitrectomy and membrane peel (data not shown). Therefore, it appears that these findings are likely due to the enzymatic effect of ocriplasmin.
There are extensive preclinical data of intravitreal plasmin dating back more than 20 years showing possible toxicity as evidenced by ERG. Verstraeten et al. demonstrated that in plasmin-treated rabbit eyes, the ERGs showed a transient (3 days), decreased b-wave amplitude, which recovered quickly.12 The investigators speculated, without much supporting data, that this was due to the osmolality of plasmin. Another rabbit study by Sakuma et al. showed that ERG depression was dose-dependent for microplasmin.13 Their study showed that at 90 days, b-wave amplitude had incompletely recovered in the animals injected with 250 µg microplasmin. The a-wave remained depressed throughout the study.12,13
Plasmin is a member of the trypsin protease family and cuts proteins at the same basic amino acids (lysine and arginine) as this digestive enzyme.11 The BRENDA database shows numerous known and tested substrates such as fibrin, laminin, fibronectin, and plasminogen, and even more might exist.11 Laminins, which ocriplasmin enzymatically digests, are known to be expressed extensively throughout the adult and developing retina. They showed that laminin 14 (composed of the alpha 4, beta 2, and gamma 3 chains) and laminin 15 (composed of the alpha 5, beta 2, and gamma 3 chains) along with laminin 5 (the alpha 3, beta 3, and gamma 2 chains) are found within the interphotoreceptor matrix and in the outer plexiform layer.16,17 Laminins may play a role in photoreceptor production, stability, and synaptic organization.16,17 Beebe has expressed concern about likely adverse effects of injecting an enzyme with such broad substrate specificity into the eye.11 The enzymatic cleavage of fibronectin and laminin proteins may have a direct and detrimental effect on retinal function. After injection of ocriplasmin, the ellipsoid layer becomes indistinct, causing some to speculate that there is a loss of that anatomical layer altogether.2,3
Human data are neither extensive nor detailed, particularly with regard to electrophysiological findings. In the original ocriplasmin clinical trial of 976 participants, 141 reportedly had ERG data collected.5 Of these, 11 showed decreased a- and b-waves after 1 month, but no follow-up data beyond that were disclosed. In the 10% of participants who showed ERG depression, the depression was reportedly “resolved” in seven, whereas four “remained persistently depressed.”4–5 No other details are available, making it difficult to determine the significance of these findings.
Since ocriplasmin’s availability, there have been only two recent case reports of ERG changes. A single case study by Tibbets et al. of a 71-year-old woman with symptomatic VMT and visual acuity of 20/60 improved to 20/40 4 months after ocriplasmin but complained of darkness in vision.9 She was found to have a depressed full-field and multifocal ERG. They suggested that rods might be more susceptible than cones, and the discontinuity of the ellipsoid layer may be caused by a diffuse enzymatic effect on photoreceptors or the retinal pigment epithelium and not limited to areas of vitreomacular adhesion.9 The fact that the full-field ERG was depressed in Tibbets et al. as well as in our cases demonstrates a panretinal effect not limited to the focal vitreomacular adhesion.
Fahim et al. also reported a single case where the subject experienced vision loss (from 20/40 to 20/125) showing full-field ERG depression, but no follow-up was reported.1 The “vision loss” may have been due to the fact that the macular hole did not close, which usually means it has opened further (data not available). They also suggested that the effect of ocriplasmin is not limited to the macular region alone but involves the entire retina. Fahim et al. expressed concern that an enzymatic cleavage of intraretinal laminin may be a biologically plausible mechanism for acute ocriplasmin retinal toxicity.1 Our data support this hypothesis and that ocriplasmin is causing a panretinal dysfunction of deeper layers of the retina, such as the photoreceptors, and not the vitreoretinal interface alone. Outer segment defects, visual acuity dysfunction, photopsia, and dyschromatopsia, as well as rod photoreceptor dysfunction, can be attributed to the disruption of the ellipsoid layer. The observation that the ellipsoid layer later recovers suggests that this layer is not destroyed or absent and is likely just edematous, causing an attenuation of the signal on OCT.
Ocriplasmin has a much higher incidence of significant ERG depression than previously reported. Despite our small series, we found this to occur in 30% of our study participants. We also found that the ERG changes and the ellipsoid layer changes on OCT seem to be related. Our study suggests that when ocriplasmin successfully closes the macular hole, it is then also more likely to cause marked depression of the ERG along with symptoms such as dyschromatopsias, photopsias, and darkening of vision. The depression of the cone-mediated ERG seems to return within a few months, whereas the rod-mediated ERG appears to remain depressed much longer. Additional studies with long-term follow-up are needed to evaluate the full and possibly permanent effects of ocriplasmin.