Analysis Of Prognostic Factors Of Idiopathic Epiretinal Membrane Surgery

Objective To evaluate risk factors about preoperative fundus fluorescein angiography(FFA)results, intraoperative situations and pathologic characteristics of surgical specimens related to the postoperative visual function in idiopathic epiretinal membrane(IERM).Methods Part I Leakage on preoperative FFA as a significant predictor in IERM surgery The prospective study focused on the patients with IERM who were referred to the Shanghai Tenth People’s Hospital Affiliated to Tongji University School of Medicine for operation between January 2014 and March 2015. According to well‐established study criterion, a total of 44 patients(44 eyes) included and classified into three groups based on the vascular fluorescein leakage around FFA 10 min preoperatively: Group A:macular no fluorescein leakage; Group B: macular fluorescein leakage, but not involving 200μm of diameter range of macular center; Group C: macular fluorescein leakage, involving 200μm of diameter range of macular center. All patients received a surgical procedure involved a standard 23‐gauge pars plana vitrectomy and membrane peeling. During this process, retinal superficial hemorrhage was well documented.Best‐corrected visual acuity(BCVA) and OCT were performed before and 1, 3, 6 and 12 months after surgery. Multifocal electroretinography(mf ERG) and fundus fluorescein angiography(FFA) were conducted at 6 months postoperatively.Part II Intraoperative risk factors associated with visual function outcomes of Pars Plana Vitrectomy in Idiopathic Epiretinal Membrane According to well‐established study criterion, a retrospective observational study was carried out on 37 eyes of 37 patients with epiretinal membrane peeling surgery for IERM between January 2014 and January 2015. Intraoperative situations during membrane peeling were documented, involving complexity of operation, superficial hemorrhage and the state of indocyanine green(ICG)staining. Best‐corrected visual acuity(BCVA) and OCT were performed before and 1, 3, 6 and 12 months after surgery.Multifocal electroretinography(mf ERG) and fundus fluorescein angiography(FFA) were conducted at 6 months postoperatively. The patients were divided into two groups based on postoperative BCVA((BCVA≥0.5 and BCVA < 0.5) at 6 months after surgery.The BCVA was converted to logarithm of the minimum angle of resolution(log MAR)equivalents for statistical analysis. The relationship between intraoperative factors and postoperative visual acuity was analyzed by multiple logistic regression analysis.Part III The correlation between pathologic analyses of surgical specimens with visualrecovery prognosis after IERM Peeled membranes from patients after IERM surgeries within January 2014 to January2015 were collected. The average age of patients was 66.31±7.19. The surgical procedures were video‐recorded. The morphological analyses were performed with optical microscopy. Confocal laser scanning microscope(CLSM) was used to count the number of cells and detect the immunofluorescent expressions of collagen type I(Col‐I), collagen type VI(Col‐VI), glial fibrillary acidic protein(GFAP), α‐Smooth muscleaetin(α‐SMA) and fibronectin(FN). Best‐corrected visual acuity(BCVA) were performed before and 1, 3, 6 and 12 months after surgery. Patients were classified into two groups based on postoperative BCVA at 12 months after surgery: Group A: BCVA≥0.5; Group B:BCVA<0.5.Results Part I Leakage on preoperative FFA as a significant predictor in IERM surgery The mean preoperative BCVA(log MAR) of 44 patients was 0.75±0.19(range from 0.52 to 1.30). Based on inclusion criteria, 12 patients(27.27%) classified in Group A; 16 patients in Group B(36.36%); 16 patients in Group C(36.36%). During peeling operation, Group A showed 10 patients with no hemorrhage, 2 patients with superficial foveal hemorrhage. Group B demonstrated 8 patients with no hemorrhage,8 patients with foveal superficial hemorrhage. Group C illustrated 4 patient with no hemorrhage, 12 patients with foveal superficial hemorrhage(χ2=7.619,γ=2,P=0.022).At 12 months postoperatively, BCVA(log MAR) were: Group A, 0.06±0.06; Group B,0.14±0.10; Group C, 0.67±0.16(F=59.471,P=0.00). Data for CFT‐1mm(um) were: Group A, 401.00±55.92; Group B, 411.63±38.76; Group C, 402.63±61.42. Among 44 patients,20 individuals were under 400μm of the values of CFT‐1mm, with visual acuity(log MAR)12‐month postoperatively were 0.29±0.33, which exhibit no statistical difference from CFT‐1mm above 400μm. Also, 10 patients experienced foveal morphology recovery after surgery, with visual acuity(log MAR) 12‐month postoperatively were 0.19±0.29,which exhibit no statistical difference from unrecovered patients. The mf ERG revealed that the average latency of N 1 wave at ring 1at 6‐month after surgeries were: Group A,14.97±2.24ms; Group B, 14.93±1.55ms; Group C, 18.39±1.96ms(F=8.275, P=0.003).Part II Intraoperative risk factors associated with visual function outcomes of Pars Plana Vitrectomy in Idiopathic Epiretinal Membrane All of 37 patients completed follow‐ups for an average duration of 14.41±2.33 months.Among 37 patients, 28 patients(75.7%) were in desirable visual acuity group and 9patients(24.3%)were in undesirable visual acuity group. Statistical analysis revealed that superficial hemorrhage during membrane peeling was associated with poor visual acuity after surgery(OR,7.221;95%CI,1.775‐29.372;P=0.006**). Complex peeling was positively increased with superficial hemorrhage(γ=0.336,P=0.042*) and ICG staining(γ=0.593,P=0.000**) significantly. The mf ERG revealed that the average latency of N 1 wave at ring 1 in eyes with superficial hemorrhage(16.88±1.27 ms, N=10) was longer than that in eyes without superficial hemorrhage(12.80±4.21 ms, N=18)at 6 months postoperatively(t= ‐2.187, P=0.042). FFA showed 8 in 10 eyes with superficial hemorrhage appeared leakage on the macular fovea.Part III The correlation between pathologic analyses of surgical specimens with visualrecovery prognosis after IERM35 specimens were collected intraoperatively and patients were underwent follow‐ups of 1‐, 2‐, 6‐, 12‐month. 23 patients in Group A(BCAV≥0.5) and 12 patients in Group B(BCVA< 0.5). Under observation of low‐magnification optic microscopy, 5 samples in Group A showed dense membrane(5/23, 21.74%); 10 samples in Group B showed thick membrane(10/12,83.33%). The remaining samples indicated transparent‐thin membrane(OR, 18; 95%CI, 2.937‐110.307; p =0.002). Under observation of high‐magnification CLSM, 3 samples in Group A showed cell density above 50cells/field(3/23,13.04%) whereas 8 samples in Group B showed cell density above 50cells/field(8/12,66.67%)(OR, 5.927; 95%CI, 1.739‐20.202; p =0.004). In Group A, 3samples exhibited Col‐I positive staining(3/23, 13.04 %), 12 samples exhibited Col‐VI positive staining(12/23,52.17%),22 samples exhibited GFAP positive staining(22/23,95.65%), 3 samples exhibited α‐SMA positive staining(13/23.56.52%),18 samples exhibited FN positive staining(18/23,78.26%). In Group B, 5 samples exhibited Col‐I positive staining(5/12, 41.67 %)(OR,2.031; 95%CI,0.808‐5.106, p =0.132), 1 sample exhibited Col‐VI positive staining(1/12,8.33%)(OR,0.143; 95%CI,0.040‐0.508, p =0.003),10 samples exhibited GFAP positive staining(10/12, 83.33%)(P=0.268), 10 samples exhibited α‐SMA positive staining(P=0.116), 9 samples exhibited FN positive staining(9/12,75.0%)(P=0.981).Conclusions1. The fluorescein leakage of preoperative FFA from IERM patient is related with intraoperative superficial hemorrhage, which can predict intraoperative risks of IERM surgery.2. The fluorescein leakage of preoperative FFA from IERM patient is related with postoperative visual function recovery, which can predict postoperative prognosis of IERM surgery.3. The fluorescein leakage of preoperative FFA from IERM patient are inconsistent with retinal edema from OCT imaging, the values of CFT‐1mm from postoperative OCT are unrelated with visual recovery postoperatively.4. Superficial hemorrhage in IERM peeling is a risk factor for the poor postoperative visual function5. Complex peeling contributes to superficial hemorrhage as well as the positive staining of ICG.6. The dense‐thick membrane is related to undesirable recovery of visual acuity.7. The density of cells is also related to undesirable recovery of visual acuity.8. The positive expression of Col‐ VI is the protective factor for desirable recovery of visual acuity.