Serous retinal detachment after panretinal photocoagulation for proliferative diabetic retinopathy: a case report

A 24-year-old Saudi man with type 1 DM and a history of tobacco smoking presented with bilateral severe non-proliferative diabetic retinopathy (DR). His past medical history was unremarkable except for DM. No ophthalmic interventions, such as laser, intravitreal injections, or surgery, had been given prior to presentation. His social and family history: he had two brothers and three sisters. His mother had a history of type 2 DM and one younger sister had type 1 DM, however, none of them were known to have any DR. Environmental history: he was living in a mainland urban area working as a clerk. He was seen on 12 occasions during 18 months. Compliance with follow-up visits and instructions was variable. On presentation, his temperature was 36 °C and his heart rate was 72. His blood pressure (BP) ranged between 110/60 and 150/90 mmHg, with an average BP of 115/80 mmHg over eight measurements during 16 months. Random blood glucose ranged from 9.8 to 19 mmol/l with an average of 14.6 mmol/l (normal range is 5 to 7 mmol/l in our laboratory). His glycated hemoglobin (HbA1C) was 13.6% (>6.5 is diagnostic of diabetes in our laboratory) or 125 mmol/mol (> 47 is diagnostic of diabetes in our laboratory). Urine analyses and detailed neurological examinations are not routinely performed in our hospital (which is an eye specialist hospital). Our patient was on the following medication prior to presentation and throughout follow-up: Mixtard 30 Novolet (biphasic isophane insulin) subcutaneous injections of 48 Units before midday and 38 Units after midday. His unaided visual acuity was 20/200 in his right eye and 20/300 in his left eye. Fluorescein angiography at baseline (performed with a standard, non-wide-field fundus camera) had demonstrated mild ischemic maculopathy in his right eye with no clinically significant macular edema and severe non-proliferative DR (not shown). His left eye could not be imaged at baseline with the standard fluorescein angiography camera due to poorly dilating pupil. Optical coherence tomography demonstrated a mild macular edema in his left eye (not shown). He was offered, but refused, intravitreal injections in his left eye.

Due to the unfavorable risk factor profile, he received several rounds of PRP in both eyes during the following 1 year after presentation; however, he developed PDR and the neovascularizations did not regress, in spite of the PRP (Fig. 1 upper panel). His unaided visual acuity was 20/200 improving to 20/50 with pinhole in his right eye and 8/200 improving to 20/70 with pinhole in his left eye. There was a discussion with our patient whether to proceed with additional PRP or to inject off-label intravitreal bevacizumab. A potential risk with intravitreal bevacizumab is the progression of fibrosis, which may ultimately lead to tractional RD [2, 3]. There were early signs of fibrosis in his left eye on optical coherence tomography (Fig. 1 lower panel).

He was recommended intravitreal bevacizumab injection, however he refused again. Instead, another round of PRP was attempted in both eyes. He presented 2 days later with bilateral extensive serous RDs involving the macula (Fig. 2 upper panel). His unaided visual acuity was 1/200 in his right eye and 5/200 in his left eye, with no improvement with pinhole correction. At this point, he accepted a single off-label injection of 1.25 mg bevacizumab (Avastin; Genentech, South San Francisco, CA, USA) in each eye. Eleven days after the bilateral injection, the serous RD had regressed completely in his left eye, and there was a substantial regression in his right eye (Fig. 2 lower panel), which subsequently regressed completely over the following 2 weeks. During the follow-up, his unaided vision gradually recovered up to 20/100 improving to 20/40 with pinhole in his right eye and 20/300 improving to 20/60 with pinhole in his left eye. At 3 months after the intravitreal bevacizumab injection, there was partial regression of the neovascularizations and no signs of tractional RD (not shown).

Imaging was performed using spectral domain optical coherence tomography (Heidelberg Engineering, Inc., Heidelberg, Germany) and wide-field fundus photography (Optos PLC, Dunfermline, UK).

PRP was performed using Pattern Scan Laser (PASCAL) photocoagulator (OptiMedia, Santa Clara, California, USA) and Volk QuadrAspheric fundus contact lens (Volk Optical Inc. Mentor, OH, USA). The duration was 20 ms for each spot with a spot size of 200 micrometers and spacing of 0.75 using a 2 × 2 or 3 × 3 grid pattern. The power range was 200 to 1000 milliWatt, adjusted as needed until retinal whitening was seen for each burn.

This case report highlights the clinical course seen during the progression of PDR in spite of several sessions of PRP. After one of these sessions, a bilateral serous RD appeared. This complication was successfully managed with intravitreal bevacizumab. To the best of our knowledge, the use of intravitreal bevacizumab has not been described previously for this complication. Serous RD after PRP for PDR is a rare complication nowadays and we are aware of only two publications from this century using high resolution imaging to describe its features. None of these, however, presented wide-angle imaging of the PDR characteristics. Gharbiya et al. [4] used bilateral intravitreal triamcinolone injections, leading to rapid resolution of the serous RD in a 52-year-old woman with type 2 DM. Azar et al. [5] used two bilateral subconjunctival betamethasone injections, leading to resolution of the subretinal fluid (SRF) and recovery of vision in a poorly regulated 34-year-old woman who had a miscarriage in the 20th week of pregnancy.

There are no studies on the natural history of the resolution of SRF after PRP using optical coherence tomography. Doft and Blankenship (1982) described 13 out of 50 eyes that developed serous RD after PRP, with resolution of the serous RD within 14 days in all eyes; however, this was not documented with optical coherence tomography [6].

The case described by Gharbiya et al. [4] most likely had persisting SRF for 3 weeks after PRP, with almost complete resolution within 1 week after the intravitreal triamcinolone acetonide (TA). In the case described by Azar et al. [5], SRF presented 2 days after the PRP, with resolution of SRF within 1 month after subconjunctival betamethasone injection. Thus we argue that the fast resolution of SRF described in this case may have been due to the intravitreal Avastin (bevacizumab) injection. This is supported by the fact that there was a partial regression of the neovascularizations and no signs of tractional RD at 3 months after the intravitreal bevacizumab (not shown). However, an element of spontaneous resolution of SRF cannot be excluded. The limitation of this study is that it is a single case report and there is no control group. However, serous RD is nowadays a very rare complication after PRP, and therefore its optimal management is unlikely to be supported by any clinical trials on DR.

Serous RD after PRP for PDR is a rare complication nowadays. In this case, it seems that excessive photocoagulation exceeded the energy-absorbing capacity of the retinal pigment epithelium, leading to a disruption of the blood–retinal barrier. A single injection of bilateral intravitreal bevacizumab was sufficient to control the serous RD. This effect may have been due to a reduction of vascular leakage resulting from the mechanism of action of this drug. No complications were noted from the injection. To conclude, caution should be exerted when attempting bilateral PRP. This has to be weighed against issues such as the risk of progression of PDR with complications such as vitreous hemorrhage and patient compliance and access to ophthalmic health care.