Failure to Detect an Effect of Pneumatic Displacement in the Management of Submacular Haemorrhage Secondary to Age‐related Macular Degeneration: A Retrospective Case Series

Editor, Submacular haemorrhage (SMH) is an infrequent complication of choroidal neovascularization (CNV) secondary to age-related macular degeneration (AMD) with poor visual prognosis. There is no proven treatment for SMH, and one treatment option, submacular surgical removal, does not increase the chance of stable or improved visual acuity (Submacular Surgery Trials (SST) Research Group 2004. Heriot (1996) reported that pneumatic displacement of SMH with intravitreal injection of tissue plasminogen activator (tPA) was effective in reducing SMH and improving visual acuity. However, the long-term effects seemed to be limited by the underlying CNV and, possibly, by toxicity. Anti-VEGF therapy has also shown promise for treating SMH. Here we summarize our experience with pneumatic displacement without tPA followed by anti-VEGF therapy in the treatment of SMH secondary to AMD and evaluate the efficacy of this therapeutic option. We retrospectively reviewed the medical records of 15 consecutive patients with SMH. The inclusion criteria was acute SMH for <4 weeks, blood under the fovea, reading vision in the involved eye immediately before the haemorrhage and a follow-up period of at least 8 months. Treatment consisted of either combination therapy with pneumatic displacement using perfluoropropane (C3F8) gas followed by intravitreous injection of ranibizumab (Group 1) or monotherapy with intravitreous injection of ranibizumab (Group 2). The groups were based on the surgeons’ treatment preferences. Patients’ clinical characteristics are summarized in Table 1. A complete ophthalmologic examination, including fluorescein angiograms, fundus photographs and optic coherence tomography, was performed prior to and during the treatment course. This allowed us to compare data between the groups to determine the effect of pneumatic displacement on visual function. The mean duration of SMH before treatment was 6.6 days in Group 1 and 13.3 days in Group 2. Group 1 patients underwent injection of pure C3F8 (0.3 ml) into the vitreous cavity through a 30-gauge needle introduced through the pars plana. A prone position was maintained for 48 hr during hospitalization. Ranibizumab (0.5 mg) was injected through the pars plana in both groups, and follow-up was in accordance with the CNV treatment regimen. The mean patient age was 84.9 years in Group 1 (n = 8) and 81.2 years in Group 2 (n = 7). The average SMH size was 4 disc areas in Group 1 and 2.4 disc areas in Group 2. Complete pneumatic displacement of the SMH was seen within 1 week in 3 of 8 cases (37.5%) in Group 1, with partial displacement in 3 eyes (37.5%) (Table 1). The mean number of ranibizumab retreatments was 5.6 in Group 1 and 7 in Group 2. Postoperative complications developed in two eyes in Group 1 (patients 1 and 4). One of these patients presented with a breakthrough vitreous haemorrhage 6 weeks after surgery. Retinal detachment developed after 4 months of follow-up, and a pars plana vitrectomy was performed. The other patient developed a breakthrough vitreous haemorrhage 2 weeks after surgery, which was treated with pars plana vitrectomy. During the 8.5–10 months of follow-up, two eyes (patients 6 and 8) in Group 1 and one eye in Group 2 (patient 7) developed SMH with breakthrough bleeding into the vitreous cavity. These patients declined further treatment. Preoperatively, the mean best-corrected visual acuity (BCVA) was 1.36 -logMAR (SD 0.37) in Group 1 and 1.08 -logMAR (SD 0.44) in Group 2. After a mean follow-up of 13.8 months (range, 8–29 months), the mean BCVA was 1.33 -logMAR (SD 0.50) in Group 1 and 0.87 -logMAR (SD 0.47) in Group 2 (Table 1). The mean follow-up was 11.65 months in Group 1 and 14.92 months in Group 2; however, only 6 of 15 patients had 12 months of follow-up. BCVA improved by 0.03 -logMar units (1.5 letters on the ETDRS chart) in Group 1 versus 0.20 -logMar units (10 letters on the ETDRS chart) in Group 2. Figure 1 shows the course of pre-operative and 12-months postoperative visual acuity of each patient. At the 12-month visit, there was a wider spread in the visual acuity values in Group 1, reflecting a more unpredictable final visual acuity. Overall, visual acuity improved in 3 of 8 (37.5%) patients in Group 1 versus 6 of 7 (85%) patients in Group 2. There was no significant difference between the groups in BCVA at baseline (p = 0.23) or in visual gain at follow-up (p = 0.09) by the Mann–Whitney test. Individual course of pre- and postoperative visual acuity. Untreated patients with SMH secondary to AMD deteriorate two lines at the 12-month follow-up, on average (Submacular Research Group 2004). Pneumatic displacement combined with anti-VEGF drug injection improved visual acuity in 33.3% of patients (Höhn et al. 2010), similar to our results of 37.5% improvement at the 12-month follow-up. Ranibizumab monotherapy increased this percentage to 57% of 21 eyes, and there were fewer complications (McKibbin et al. 2010). We saw a similar trend in our study of 7 eyes. Intravitreal injection of tPA, expansile gas and ranibizumab increased the percentage further to 64% of 81 eyes at the 12-month follow-up examination, and complete displacement of SMH was achieved in 85% (Chen et al. 2007). In our study, only 35.5% of SMH was successfully displaced, which may explain, in part, the lack of an additional effect of the gas and may suggest a role for gas as an adjunct to tPA and ranibizumab treatment. Thus, we found no evidence that pneumatic displacement of SMH had any additive effect. Notably, this conclusion is based on the premise that the study groups were comparable. The primary assignment of patients to treatment groups was unmethodical randomization, based on surgeons’ treatment preferences, and the treatment regimen was nonstandardized. Nonetheless, the only systematic difference between the groups was a longer treatment delay and a smaller average SMH size in Group 2. The preoperative BCVA was similar in both groups and thus independent of the SMH size because of the subfoveal location, and no difference in visual gain between groups was found postoperatively. We therefore believe that the effect of any possible selection bias was minimal. The small sample size resulted in limited power to detect differences in outcome, but there was no trend, reflecting the absence of any substantial clinical effect of gas displacement. Although the uncontrolled and retrospective character of this study and the limited number of included patients limit the interpretation of the results, SMH is an infrequent complication and it is difficult to achieve strong statistical power. Considering these limitations, this study supports the idea that gas injections may not be worthwhile in eyes with SMH, as similar visual outcome may be achieved with a lower complication rate with ranibizumab monotherapy. The results do not support initiation of a larger trial of gas as an adjunct to ranibizumab treatment of SMH. However, additional large prospective randomized clinical trials are needed to clarify the precise role of ranibizumab and gas with adjunctive intravitreous t-PA, which in the right combination may enhance SMH displacement, visual recovery and reduce retreatment need.