Profile of angle closure in a tertiary care center in north India

Purpose: To study the demographic and clinical profile of the types of primary angle closure patients presenting at a tertiary care center in North India. Materials and Methods: Clinic records of patients diagnosed as primary angle closure were reviewed. International Society of Geographical and Epidemiological Ophthalmology (ISGEO) classification scheme was used to categorize patients. Demographic and clinical data including prior management was collected and analyzed. Main Outcome measures were age, sex, symptomatology, best corrected visual acuity (BCVA), intraocular pressure (IOP), gonioscopy, optic disc assessment and visual field defects. Logistic regression model and receiver operating curve (ROC) were calculated for predictors of type of glaucoma. Results: Eight hundred and fourteen patients (1603 eyes; males: 380, females: 434) were diagnosed to have various subtypes of angle closure. Mean (±SD) age at presentation was significantly higher for males (57.57 ± 11.62 years) as compared to females (53.64 ± 10.67 years) (P < 0001). Primary angle closure glaucoma (PACG) was most frequently diagnosed subtype (49.38%) followed by Primary angle closure (PAC) (39.68%) and Primary angle closure suspect (PACS) (10.93%) respectively. The three subtypes differed significantly among their mean IOP (on ANOVA, F = 14.04; P < 0001 using Greenhouse-Geisser correction). Univariate analysis was done to find significant predictors for the outcome of PACG. Logistic regression model and ROC containing the significant predictors yielded a very high AUC of 0.93 with strong discriminatory ability for PACG. Conclusion: In our hospital-based study, the significant predictors for the outcome of PACG included male gender, diminution of vision, the presence of pain and worsening grades of BCVA. Nearly half of PACG presented with advanced disease. In spite of one-third of the patients being diagnosed as angle closure prior to referral, only 8.34% had iridotomy (laser or surgical) done.

drainage angle i.e., the posterior (usually pigmented) trabecular meshwork is seen for less than 90° of angle circumference and features indicating that trabecular obstruction by peripheral iris has occurred, such as peripheral anterior synechiae, elevated IOP, iris whorling, "glaucomflecken" lens opacities or excessive pigment deposition on the trabecular surface, with no optic nerve head changes.
Primary angle closure glaucoma (PACG) was labeled if disc and field changes were present with PAC (appositional or synechial) as defined above i.e., a vertical cup to disc ratio (VCDR) of 0.7 or greater or asymmetry between the right and left VCDRs of 0.2 or more, and a visual field defect consistent with glaucoma. [2] If the media opacities obscured optic disc assessment, then an IOP greater than 26 mm Hg and visual acuity worse than 20/400, or evidence of previous glaucoma filtering surgery was considered. The VCDR and IOP criteria described above were based on the 97.5 th and 99.5 th percentiles for "hypernormals" in surveys described by Foster et al. [2] Minimal criteria for labeling a glaucomatous visual field defect were as follows: Glaucoma hemifield test (GHT) outside normal limits, pattern standard deviation (PSD) with P values ,5%, or a cluster of three or more points in the pattern deviation plot in a single hemifield with P values ,5%, one of which must have a P value ,1%. Any one of the preceding criteria, if found again on repeat testing on two tests within one month, was considered sufficient evidence of a glaucomatous visual field defect. [3,4] Visual fields were done for patients with BCVA of 20/200 or better. Advanced field defects were defined as mean deviation greater than -12 dB and on Pattern deviation plot, points below 5% between 37 to 55 with points below 1% ranging from 19 to 36. The diagnosis was confirmed by either of the two senior consultants in all cases. We present our data using both the visual acuity and visual field criteria [World Health Organization (WHO) criteria]. [5] Eyes with advanced field damage and/or VCDR of more than 0.80 and/or BCVA less than 20/400 due to glaucomatous disc damage were classified as advanced glaucoma. [6] Statistical analysis was performed using SPSS statistical software (10.0 version). Data were descriptively analyzed for estimates of quantitative variables using mean and 95% confidence intervals and qualitative variables as proportions. Chi-square test of independence was used to evaluate associations between qualitative variables and Glaucoma Classification groups. Univariate Odds Ratios (with their 95% CI) were calculated for various dichotomous characteristics. ANOVA was used to compare intraocular pressures measured in the eyes affected by three different types of glaucoma (814 patients providing data for 1603 eyes).
Logistic regression analysis was carried out on predictors for PACG found significant on Univariate analysis. All tests were two-tailed and P-values ,0.05 were taken as significant.

Results
Of the 4703 patients seen during the specified span, 1505 (32%) presented with primary glaucoma (both structural and functional damage) and ocular hypertension (OHT). POAG was seen in 595 (39.5%) patients, OHT in 96 (6.4%) while 814 (54.1%) patients had PAC subtypes. Since PACS and PAC do not have structural abnormality, the net proportion of PACG was 40.3%.
Amongst the angle closure subtypes, PACG was the most common seen in 402 patients (n 5 814; 49.4%) followed by PAC in 323 patients (39.7%) and PACS in 93 patients (11%). In patients with different subtypes of angle closure in both eyes, the eye with the higher degree of angle closure was used for categorization. PACG had an overall peak presentation in the seventh decade (34.8%) while PAC was more common in the sixth decade (34.1%). The overall mean age at presentation for males was 57.70 years (95% CI; 56.53-58.86); significantly higher when compared to mean age of 53.56 years (95% CI; 52.55-54.57) in females (P , 0.001). The mean age of presentation also varied significantly among patients stratified by types of glaucoma [ Table 1]. However, on general linear model analysis using age as dependent variable and gender as well as type of glaucoma as fixed covariates, there was no significant group interaction between gender versus type of glaucoma (F 5 1.55; P 5 0.21) as far as mean age of patients was concerned [ Fig. 1].
In addition, one female patient had central retinal artery occlusion (CRAO) and five had central retinal vein occlusion (CRVO). All had a history suggestive of acute angle closure attack in the past in the affected eye followed by a sudden diminution of vision. Mean IOP at presentation was 18.71 6 8.43 mmHg in eyes that had received prior medical therapy [228 patients (35.7%)] and 18.47 6 7.11 mmHg in remaining eyes. Surgical iridectomy or filtering surgery was seen in 33 patients (4.1%) while laser iridotomy (LPI) was noted in 35 patients (4.3%).
Three hundred and fifty six eyes (54.1%) amongst PACG presented with advanced glaucomatous field damage and advanced glaucomatous optic neuropathy (VCDR . 0.8) in 317 eyes (48.2%). Out of 814 angle closure patients, 60 (7.4%) were lost to follow-up and hence visual field analysis could not be performed; baseline fields were unreliable for 79 (9.7%) patients.
On univariate analysis, the significant predictors (explan atory variables) for the outcome of PACG included male gender, diminution of vision, presence of pain and worsening grades of BCVA (above 20/50) [    Age in years** 1.05 1.01-1.08 0.001 # Odds ratios were obtained by means of binary logistic regression analysis using 'PACS' as the reference category. OR, odds ratio; CI, confidence interval; *Reference category for this predictor was "PAC" as there was no patient in PACS group; **Since age is a quantitative predictor, only logistic regression odds ratio are mentioned.
(predicting presence of PACG and using PACS as the reference category) with backward Logistic Regression method, the above mentioned variables continued to be highly significant predictors of PACG in the final model. Furthermore, receiver operating curve (ROC) of model containing these significant predictors yielded a very high area under curve (AUC) of 0.93 with strong discriminatory ability for PACG [ Fig. 3].

Discussion
The epidemiology of ACG has received marked attention in recent times. Difficulty in estimation of the prevalence of ACG stems from the fact that angle closure is an ongoing process and hence the stage at which patient presents to an ophthalmologist is important. Generally patients characterized by an acute episode or striking symptoms get diagnosed, whereas the more chronic forms tend to go unnoticed. [6] PACS, PAC and PACG are now distinct entities that reflect the severity of the anatomical and functional disorder, according to the ISGEO classification scheme. [2] Four population-based studies have been conducted in South India [7][8][9][10] and one in West Bengal [11] as regards prevalence of glaucoma. The results of these studies provide data about the magnitude of the problem. [10] The aim of this study was to ascertain the prevalence of various subtypes of ACG that present to a tertiary care referral centre in North India. Our study being a retrospective analysis has limitations associated with case record review with added issues of selection bias and referral bias, institute being a tertiary care center.
Hospital-based data from North India reports almost equal prevalence of both primary open angle glalucoma and PACG. [12] Sihota et al. studied the clinical profile of the subtypes of primary ACG: Acute, subacute, and chronic, in a prospective study of 500 patients. ACG constituted 45.9% of all primary adult glaucoma seen. In our study, 814 patients (1603eyes) were found to have PAC. ACG constituted 40.3% of all primary adult glaucoma. We also found a marginal female predominance (53.3%) for angle closure, like Sihota et al. (51.4%). [12] Chennai Glaucoma Study (CGS) also reported PAC and PACG to be more common in women. [10] Sihota et al. reported ACG occurred maximally in the sixth decade. [12] In our study, 269 (33.04%) patients were less than 50 years old, of which 54 were less than 40 years of age, which is in agreement with the study by Sihota et al.
We found PACG to be the most common type seen in 402 patients (49.38%) followed by PAC in 323 patients (39.68%). Amongst other studies, only CGS and West Bengal Glaucoma Study used ISGEO classification. In CGS, the prevalence of PACG, PAC and PACS was 0.9%, 0.71% and 6.3%, respectively. In natural course the likelihood of PACS should be more than PAC and similarly, PAC should be higher than PACG. On the contrary, in a clinic-based study there is always a selection bias, patients come with complaints and most of PACS would be asymptomatic and hence would not seek consultation. So in a clinic based study, a higher PACG prevalence is expected.
In the hospital-based study by Sihota et al. 24.8% had acute ACG [12] while at our center, only 31 patients (4.7%) presented with an acute attack. Only four of these patients had evidence of disc damage, rest presented as an acute rise of pressure in PAC patients. We have not taken into account those patients who reported to us after receiving primary anti glaucoma treatment from the referring physician. Hence, this does not project the exact incidence of acute attacks and may account for the discrepancy between the two studies. In addition, there may be differences in the studied populations.
Diminution of vision and pain (periorbital pain or a non-specific headache) were more common in PACG than PAC or PACS. Most patients with PACG presented late in the course of the disease with probable repeated attacks of intermittent angle closure, which would account for pain as the chief presenting complaint. Sihota et al. also documented ocular pain to be most common in the acute and subacute groups, 62.1% and 45.5% respectively. [12] Prior to referral by a general physician or an ophthalmologist, 228 patients (35.96%) had received medical therapy, surgical iridectomy or filtering surgery was done in 33 patients (4.1%) and LPI was noted in 35 patients (4.3%), in our study. This implies that glaucoma was detected in 36% patients (who received anti glaucoma medication); however, LPI or surgical iridotomy was done in only 8.3% of the referred patients. Pilocarpine was prescribed either alone or in combination to 102 patients (44.7%, n 5 228) and was discontinued later without doing LPI. Low rate of LPI, either due to nonavailability of the Nd YAG laser or inability to recognize the condition in PAC and/or PACG patients is a matter of concern.
A large proportion of persons presenting to tertiary care centers in North India with angle closure disease have PAC or PACG. Most of the PACG patients present in the advanced stage of the disease. It is important for an ophthalmologist to identify signs of angle closure as, in India, asymptomatic chronic ACG mimicking POAG is common. LPI should be considered for all PAC or PACG patients. A study conducted at Vellore, based on the natural history of PACS reported that as many as 22% may progress to PAC but none progressed to PACG over a span of five years. Accordingly, LPI may not be warranted for PACS per se. [13] Hence, a comprehensive ophthalmic examination holds the key for timely diagnosis of angle closure disease.
To conclude, in a tertiary care center, PACG is the most common presentation amongst various angle closure subtypes with diminution of vision as the most common presenting complaint. Blindness due PACG is common (20.5%, unilateral). Inspite of one-third of patients being diagnosed as angle closure prior to referral, only 8.34% had iridotomy (laser or surgical) done. Hence, there is a need to emphasize the role of laser iridotomy in the management of angle closure glaucoma.