Outcomes of vitrectomy for diabetic tractional retinal detachment in Chicago’s county health system

Purpose To examine outcomes of 23-gauge (23G) pars plana vitrectomy (PPV) for complex diabetic tractional retinal detachment (TRD) in Chicago’s Cook County Health and Hospitals System (CCHHS). Materials and methods This is a retrospective noncomparative study of diabetic TRD cases that underwent PPV at CCHHS. Primary retinal reattachment rate, visual function, and postoperative complications were analyzed. Results Sixty nine consecutive cases were included. Primary reattachment and final attachment were achieved in 68/69 eyes (98.6%). Secondary retinal detachment was noted in 1 eye (1.4%). Vitreous hemorrhage requiring repeat PPV developed in 5 eyes (7.2%) and reoperation due to other complications was required in 4/69 eyes (5.8%). Perfluoropropane (C3F8) gas tamponade was used in 91.3% of eyes and silicone oil in 8.7% of eyes. Mean LogMAR visual acuity significantly improved from 1.84 ± 0.61 to 0.93 ± 0.66, (P<0.0001). Vision was stabilized or improved in 66 eyes (95.7%). Visual acuity of 20/200 or better was achieved in 49/69 eyes (71.0%) and 20/50 or better in 16/69 eyes (23.2%). Conclusions Even in patients with severe and advanced diabetic TRD pathology and unique demographics as seen in CCHHS, modern vitrectomy techniques can provide excellent anatomical and visual outcomes.


Introduction
Tractional retinal detachment (TRD) represents one of the most severe complications and a major cause of vision loss in patients with proliferative diabetic retinopathy (PDR) [1]. Hyperglycemia and molecular events related to diabetes lead to retinal vessel endothelial cell damage, increased vascular permeability, bleeding, retinal vessel occlusion, and subsequently retinal ischemia. An ischemic retina secretes locally active cytokines, such as vascular endothelial growth factor and connective tissue growth factor, which lead to neovascularization and connective tissue formation in the proliferative stage of diabetic retinopathy. This fibrovascular proliferation grows into the vitreoretinal interface and may contract, potentially resulting in TRD [2].
Diabetic TRD is repaired with pars plana vitrectomy (PPV), which allows for simultaneous visualization and manipulation of the retina [3]. In 1981, Michels described posterior membrane segmentation as a surgical technique for the management of TRD by relieving anterior and tangential traction caused by fibrovascular membranes, allowing for reattachment of the retina [4]. The surgical repair of diabetic TRD is among the most challenging surgeries for a vitreoretinal specialist due to the friable nature of the ischemic retina and the presence of extensive fibrovascular membranes [5]. Despite the difficulty of repair, recent advancements in vireoretinal surgery, including smaller gauge instruments, wide angle viewing systems, use of tamponade, self-retaining endoillumination, and the development of medications targeting vascular endothelial growth factor, have greatly improved outcomes in complex TRD cases [6][7][8][9].
Cook County Health and Hospitals System (CCHHS) is one of the largest public county hospital systems in the United States and is an important safety net hospital for the city of Chicago. CCHHS serves a primarily underserved population with 43.8% of outpatients using Medicaid as a payor source and 22.4% of outpatients requiring charity care [10]. Furthermore, the patient population of CCHHS is primarily from minority groups, with 52.4% of patients identifying their race as Black and 27.6% of patients identifying their ethnicity as Hispanic or Latino [10].
CCHHS serves patients with severe diabetes and limited access to preventative medical and retinal screening. These patients commonly present with advanced PDR and severe longstanding diabetic TRD with broadly adherent plaques of fibrovascular proliferation. Surgery for diabetic TRD has a high rate of complications postoperatively, especially for eyes with complex pathology. With just a handful of previous studies investigating PPV for diabetic TRD in a county hospital, limited data are available for outcomes of surgical repair in patient cohorts with such demographics and severity of pathology [5,8,11]. The purpose of this study is to examine the visual and anatomical outcomes and complication rates of 23-gauge (23G) PPV for diabetic TRD at CCHHS.

Material and methods
We conducted a retrospective review of the medical records of all consecutive surgical cases of diabetic TRD performed by a single vitreoretinal surgeon (DS) at CCHHS, between November 2013 and March 2016. Eyes that underwent 23G PPV for primary repair of diabetic TRD and had at least 3 months of follow-up were eligible for inclusion; 8 patients with less than 3 months of postoperative follow-up, or patients with other proliferative vitreoretinal disease, such as proliferative sickle cell retinopathy, were excluded from this study. This study was conducted in accordance with the Declaration of Helsinki. The collection and evaluation of all protected health information was performed in a Health Insurance Portability and Accountability Act (HIPAA)-compliant manner. Ethical approval for this study was obtained from Chicago's Cook County Health and Hospitals System Institutional Review Board (IRB). Informed consent was not obtained because it was waived by the IRB that approved the study.
23-gauge vitrectomy was performed using the Alcon Constellation vitrectomy machine (Alcon Laboratories, Inc., Fort Worth, TX, USA) using a wide-angle viewing system and scleral depression during the vitreous base dissection. 25 and 27-gauge vitrectomy was not used as it was not available at the hospital where the surgeries were performed. Fibrovascular membrane dissection was performed using a combination of one or more of the following without bimanual technique: the vitrector handpiece, internal limiting membrane forceps, curved scissors, vertical scissors, and/or blunt delaminating spatula. Triamcinolone staining and Perfluoro-n-octane (PFO) liquid was used as needed. Endolaser photocoagulation was applied around retinal breaks and in panretinal photocoagulation (PRP) fashion using a flexible curved illuminated laser. Fluid-air exchange was followed by injection of long-acting tamponade of perfluoropropane gas (C 3 F 8 ) (14-16% concentration) or silicone oil. The choice of tamponade was determined by the extent of detachment, the presence and location and size of retinotomy or retinectomy, and the amount of residual traction/fibrovascular membranes. C 3 F 8 gas tamponade was favored in the vast majority of cases except in eyes requiring large inferior retinectomies and with significant residual traction/fibrovascular membranes. All sclerotomies were sutured. Scleral buckle (SB) was used in cases with significant peripheral vitreoretinal traction or with significant scar tissue that required inferior retinectomy. In patients with a preoperative visually significant cataract, cataract extraction with intraocular lens replacement (CE/IOL) was performed at the time of vitrectomy by a cataract surgeon. To our knowledge, all lens calculations were made using A-Scan ultrasound biometry and measurements of the fellow eye.
During the postoperative period, patients were instructed to maintain face-down positioning for up to 3 weeks at least 80% of the time. Patients were seen weekly for the first 3 weeks after surgery to monitor the postoperative course and reinforce and ensure compliance with instructions. Primary outcome measures included postoperative best-corrected Snellen visual acuity (BCVA), rate of primary reattachment with a single surgery, and rates of postoperative complications including secondary retinal detachment, vitreous hemorrhage needing repeat PPV, anterior hyaloid fibrovascular proliferation (AHFVP), fibrinoid syndrome, neovascular glaucoma, and epiretinal membrane formation (ERM) needing PPV.
BCVA was converted to the logarithm of the minimum angle of resolution (LogMAR) for quantitative comparison. A LogMAR value of 1.98 was used for vision limited to counting fingers (CF) and 2.28 for hand motions (HM) [12]. Vision limited to light perception (LP) or no light perception (NLP) could not be converted to LogMAR, as these do not represent visual acuities, and were thus excluded from this analysis. Preoperative and postoperative LogMAR values were analyzed using a paired, two-tailed t-test, and a P-value <0.05 was considered statistically significant.
Results 69 consecutive eyes from 61 patients that underwent 23G PPV for primary repair of diabetic TRD were included in this study. Patient characteristics are described in Table 1. The average age of all patients was 47 years (range, 21-65 years) and average follow-up was 11.4 months (range, 3-28 months). 36 patients were male and 25 were female. 36 patients were Hispanic, 21 were African American, 3 were Caucasian, and 1 patient was Asian.
69.6% of eyes were characterized by detachments involving the macula, 44.9% by combined tractional and rhegmatogenous retinal detachments (TRD/RRD), and 85.5% by high complexity detachments or severe fibrovascular proliferation ( Table 2 along with pre-operative findings, complications, and other anatomic/functional outcomes). TRD/RRD cases included those with preexisting breaks only, a combination of iatrogenic and preexisting breaks, or iatrogenic breaks only. No specific associations were apparent between the presence of retinal breaks and patient age or retinal ischemia. Combined SB was performed in 8/69 eyes (11.6%), combined CE/IOL in 6/69 eyes (8.7%), and combined pars plana lensectomy (PPL) in 3/69 eyes (4.3%), which is summarized in Table 3. Long-acting tamponade was used in all cases, with C 3 F 8 in 63/69 eyes (91.3%) and silicone oil in 6/69 (8.7%). There were 56 eyes that were phakic at the time of operation, of which 9 underwent combined cataract extraction or lensectomy. Of the 47 eyes that were phakic immediately after vitrectomy, 17 (36.2%) required postoperative CE/IOL in the follow-up period. General anesthesia was used in 96% of cases, monitored anesthesia care (MAC) anesthesia was used in 4% of cases, and mean surgical time was 181 minutes (SD 65.4). Mean post-operative day 1 intraocular pressure (IOP) was 20.0 mmHg (SD 9.3), post-operative week 1 IOP was 15.7 mmHg (SD 4.1), and post-operative month 1 IOP was 15.0 mmHg (SD 4.2). Patients with elevated IOP responed to topical therapy, which was tapered as IOP improved. No patients required surgical intervention for elevated IOP. Fig 1 highlights three cases from the cohort, which effectively represent the typical complexity encountered in the patients analyzed in this study.
Retinal reattachment with a single surgery was achieved in 68 of 69 eyes (98.6%) ( Table 2). One patient had persistent shallow chronic subretinal fluid in the macula after the primary repair and repeat vitrectomy was performed for draining of persistent submacular fluid. Secondary retinal detachment was observed in 1 eye (1.4%) at 4.5 months postoperatively and was reattached successfully with one additional surgery with combined vitrectomy and lensectomy with scleral buckle and silicone oil tamponade. Final reattachment was achieved in 68 eyes (98.6%). Postoperative vitreous hemorrhage requiring repeat PPV was noted in 5 eyes (7.2%). Although not systematically quanitified, in combined CE/IOL and SB cases inflammation appeared increased. There was one case of neovascular glaucoma that was managed medically and did not require filtering surgery. No eyes developed endophthalmitis, phthisis, corneal decompensation, or secondary angle closure. No patients were noted to have NLP vision and no patients had complications (e.g. IOL positioning) due to prolonged face-down positioning. Overall, the mean preoperative visual acuity improved from LogMAR 1.84 (SD 0.61, 95% confidence interval (CI) 1.69-1.98) to 0.93 (0.66, 0.77-1.09) after vitrectomy, P<0.0001. Significant improvements in postoperative visual acuity were observed in eyes that received C 3 F 8 but not silicone oil, which is summarized in Table 4. Both eyes with TRD and eyes with TRD/  Vitrectomy for diabetic tractional retinal detachment Improved visual acuity was observed in 61/69 eyes (88.4%). Visual acuity was stabilized in 5/69 eyes (7.2%) and was worse in 3/69 eyes (4.3%). Differences between preoperative and postoperative BCVA are summarized in Tables 2 and 6. Regarding the eyes with stable visual acuity, visual outcomes were limited by macular ischemia/photoreceptor loss due to macula off RD despite retinal reattachment. Of the eyes with worse postoperative acuity, one surgery was complicated by severe hemorrhagic choroidals at the end of surgery as a result of severe bucking due to airway irritation. This patient developed anterior hyaloid proliferation with hypotony and cyclitic membrane 3 months postoperatively and underwent repeat vitrectomy. One patient developed fibrinoid syndrome noted on postoperative day 1, which was managed with systemic and topical steroids without significant response. This patient declined further intervention due to serious comorbidities and eventually developed a white cataract and hypotony. In the third patient, vision was limited by macular ischemia with photoreceptor loss and subsequent dense posterior capsular opacification, ERM, and cystoid macular edema formation under silicone oil for which the patient declined further surgical intervention.
It is reasonable to consider that the results of this study could partially be attributed to differences in disease severity or intraoperative technique. However, the patient population in our series represents an underserved community at a large urban county hospital, and is comprised of mostly Hispanic and African-American patients (93%) who frequently present with more advanced PDR [57]. As expected, these demographic and socioeconomic trends are reflected in the high anatomic complexity of these cases. Preoperatively and intraoperatively, many eyes were characterized by combined TRD/RRD (44.9%) and/or macular involvement (69.6%), and over 80% represented severe cases characterized by extensive broad adherent fibrovascular plaques. Thus, the patients in this series likely represent a more complex population than observed in previous studies.
Our overall surgical approach was consistent with many recent studies, but special attention was focused on meticulous removal of fibrovascular membranes and residual hyaloid with triamcinolone staining, which may contribute to both our favorable outcomes and our longer surgical durations [5,[7][8][9]. Full PRP including the anterior retina with scleral depression could contribute to the lower rates of neovascular glaucoma and postoperative vitreous hemorrhage requiring PPV. It is unlikely that the results can be exclusively attributed to minor surgical differences alone since most studies for diabetic TRDs mention thorough removal of membranes and hyaloid and full PRP. Moreover, a distinction in our study was the use of long-acting tamponade in all cases, in particular gas tamponade with C 3 F 8 14-16% in more than 90%% of cases, and prolonged face-down positioning. In most diabetic TRD studies in the literature, the tamponade is variable, including cases without any tamponade, air tamponade, variable concentrations of sulfur hexafloride (SF 6 ) and C 3 F 8 , or silicon oil and information regarding patient positioning is rarely addressed. Currently there are no strict guidelines for the use and the type of tamponade (short versus long-acting and gas versus silicon oil) or face-down positioning after vitrectomy for diabetic TRD repair. In this series, we favored the use of long-acting C 3 F 8 rather than silicone oil, limiting the use of silicone oil mostly to cases involving large inferior retinectomies. In this study, we observed a significant improvement in BCVA in eyes that underwent C 3 F 8 tamponade but not silicone oil. However, the six eyes that received silicone oil had worse preoperative BCVA and were selected to receive silicone oil tamponade due to their severe pathology needing large inferior retinectomies. This study is not randomized so definitive conclusions for the use of silicone oil versus gas cannot be drawn. Since C 3 F 8 was used in more than 90% of the cases, our data indicates that excellent outcomes can be achieved by using long-acting C 3 F 8 in complex TRD cases, even those with severe FVP and combined TRD/RRD.
Strengths of this study include the large population of eyes with complex diabetic TRD operated on by a single surgeon with consistent techniques at a single institution with unique demographics. Limitations of this study include the retrospective design, lack of grading of vitreous hemorrhage, lack of retinal break descriptions, potential for loss to follow-up bias due to the wide range of follow-up times (3-28 months), and the inability to make definite conclusions about tamponade agent and the need for prolonged face down positioning. Although our data suggests that long-acting C 3 F 8 with prolonged face-down positioning may contribute to improved outcomes, the lack of randomization for the type of tamponade and duration of positioning and the absence of a control group do not allow us to make a definite conclusion.
In conclusion, our findings show that 23G vitrectomy can provide improved outcomes in patients with complex TRD pathology as described in our study representing Chicago's urban population in a large public hospital. Despite our efforts to prevent late-stage diabetic retinopathy and TRD, many patients are unfortunately faced with these debilitating conditions. Although medical and surgical advancements have greatly improved treatment modalities, further research is warranted to better elucidate which techniques lead to the best outcomes for our patients.
Supporting information S1 Data. Pre-operative and final visual acuities. All data used to calculate means, standard deviations, confidence intervals, and P values in manuscript. (XLSX)