The present study tried to elucidate the risk factors for visual field deterioration in a cohort of new-diagnosed exfoliation glaucoma patients in Sweden. The risk factors studied were recorded both at diagnosis and during the three-year follow-up period.
Previous studies have extensively described risk factors for developing glaucoma [8–11]. In addition, risk factors for progression in glaucoma damage have also been described before . Interestingly, exfoliation seems to be a risk factor for fast progression in previous studies compared with other glaucoma types. However, evidence about if the described risk factors apply for new-diagnosed exfoliation glaucoma patients is lacking. In several previously published studies, exfoliation glaucoma patients were excluded due to the low prevalence of exfoliation glaucoma in other populations and because the disease develops differently. Exfoliation glaucoma patients showed a more aggressive and fast development of visual field damage than primary-open glaucoma patients . At our Department, around 60% of all new-diagnosed glaucoma patients suffer from exfoliation glaucoma, making results from this study very interesting from the clinical point of view.
Age showed to be a common risk factor for progression in the three models studied. The average age at inclusion was 71.41 (±7.52) years. This result accords with previous studies in which the average age at glaucoma diagnosis was around 70 years old . Only three patients were excluded from the study due to their disability to perform visual fields. All of them were about 70 years old, meaning that no patient was indirectly excluded due to age. Considering the β coefficients at the multivariate linear regression analysis, the MD values increased by 0.092 dB/year. The MD increase can be due to age (cataract) or glaucoma progression. However, age was also positively associated with visual field progress when considering the rate of progression (ROP). ROP is based on the visual field index (VFI) which corrects for cataracts and evaluates more accurate glaucoma damage. The β coefficient at the multivariate linear regression analysis showed an ROP deterioration of 0.053%/year. This increased ROP value means that a patient diagnosed with glaucoma at 80 years old had a faster visual field deterioration than a patient diagnosed at 70 years old. The average ROP was 2.60%/year for the whole cohort. In the hypothetical case that a patient aged 70 progressed at 2.60%/year, 80 would progress at 2.65%/year. These results must be interpreted with caution. The results apply to the range studied, let's say 47-88 years.
Furthermore, the model was linear; this assumption can be suitable for the years studied but probably not for older ages. Other models, like exponential, probably suit better with reality in higher ages. Further studies are needed to clarify the best models for different age groups. The study showed increased progress with increased age. The increased progress with age could be that the number of ganglion cells diminished with age, so the visual fields deteriorate faster as age increases.
Several previous evidence described the IOP as the most critical risk factor for glaucoma development. The present study showed an association between IOP at diagnosis, IOP reduction in absolute and relative values, and visual fields' deterioration in new-diagnosed exfoliation glaucoma patients. The average IOP was high at diagnosing, 33.14 mmHg (± 6.66). Exfoliation glaucoma is a high IOP glaucoma. Exfoliation material obstructs the trabecular meshwork producing high IOP levels. The average ROP in the cohort was 2.60%/year during the three years' follow-up. Based on the results from the multivariate analysis for ROP (β coefficient), every mmHg IOP in the baseline represented an increased ROP of 0.05. Theoretically, this means that a patient with a baseline IOP of 25 mmHg and an ROP of 2.60%/year would progress at 2.65%year if the IOP were 35mmHg or at 2.70%/year if the IOP was of 45 mmHg. As pointed out above, this extrapolation must be taken with caution. It probably applies for the IOP levels registered in this study: 22-55 mmHg and probably not for other IOP levels.
Furthermore, the model assumed that the relationship between IOP values and ROP is linear. This is not known; probably each mmHg has a different impact on ROP if the IOP is around 30mmHg than if it is around 50mmHg. Further studies are needed to clarify this issue.
The IOP reduction is another interesting issue to discuss. The present study included both an absolute and a relative value for IOP reduction after diagnosis. Patients included in this study were followed and treated according to our guidelines for treatment and follow-up of glaucoma patients. All patients got what is called “target IOP” after diagnosis. This “target IOP” is an uncertain IOP value that usually is around 18-20 mmHg, and it's established by the ophthalmologist who made the diagnosis. According to our guidelines, three years after follow-up with at least 5-6 reliable visual fields, this "target IOP" must be re-evaluated. In the present study, the IOP three years after follow-up was 17.88 mmHg (±2.99). Even if this value can be considered acceptable, 2/3 of patients showed progression in the GPA analysis. Therefore, it seems that the "target IOP" should be lower than 18mmHg, probably around 16mmHg (or even lower) to diminish the visual field's deterioration. A similar approach can be made related to the relative IOP reduction. A common assumption is that a 20% IOP reduction is enough to stop the progression of the disease . In the present study, the average IOP reduction was 47.28% (±14.28). However, 2/3 of glaucoma patients showed progression according to the GPA analysis. The aggressive nature of these new-diagnosed exfoliation glaucoma patients probably needs a higher relative IOP reduction to stop the progress of the disease.
The absolute IOP reduction showed to be significantly associated in the MD and the VFI but not in the GPA model. Meanwhile, the relative IOP reduction was significantly associated with the VFI and the GPA but not the MD model. This means that both ways to control IOP reduction are needed. However, in our clinical practice, we usually forget the importance of taking care of both the absolute and the relative IOP reduction. We usually focus more on the absolute values, and sometimes we feel that it is difficult to find the IOP value at the diagnosis and calculate the relative IOP reduction. The present study showed the importance of checking both values to establish a more accurate level of IOP that can stop the visual field's deterioration.
Three variables showed a significant correlation with visual field progression only in the VFI model. The mentioned variables were family history, smoking, and SLT. Family history for glaucoma was measured in the present study as present or absent. Only closest relatives were considered; mother and/or father and/or siblings. The information was self-reported, reporting bias might be admitted. Family history for glaucoma was reported by n= 39 (42.85%) meanwhile n=52 (57.15%) reported no family history (Chi-square, p=0.17). Most of the patients were born in Sweden, and their parents were also born in Sweden. Only two patients had Finnish parents. Genetic mechanisms behind exfoliation glaucoma have been described in previous studies . However, no previous evidence regarding family history and a poorer prognostic for glaucoma has been found.
Smoking has been described among risk factors for glaucoma development . In the present study, smoking was measured as present or absent based on a question: "Did you smoke more than 50 cigarettes in your life?” Also, this question was self-reported, addressing the possibility of reporting bias. Interestingly, a quite high number of patients reported that they have smoked (n=41; 45%). Smoking is not so common among the new generations in Sweden. Furthermore, the study did not ask if the patients performed “snuff” (chewing moist powder tobacco), common in Sweden. The increased visual field deterioration among smokers can be explained by decreased blood circulation in small vessels due to nicotinamide effects.
The SLT was also recorded as present or absent to facilitate the analysis. Interestingly, the association was negative; let's say that patients treated with SLT showed a lower progression according to the VFI model. The decision to treat or not to treat with SLT was based on the clinicians that met the patients. SLT treatment has shown promising results in exfoliation glaucoma patients . However, the number of patients treated with SLT in the present study was relatively low (n=16). Based on the current study results, it's possible to recommend SLT often to exfoliation glaucoma patients.
The number of medicines was also associated with visual field progression in the present study. In the VFI model, the association was negative; let's say an increased number of medicines rendered a lower progression. At three years of follow-up, the patients were treated on 2.66 (±0.89) drugs. Based on the β coefficient of the multivariate linear regression, each medicine reduced the ROP by 0.32%/year. The higher number of drugs induced a higher reduction in the IOP, thus diminishing progression. It's also possible that medications can reduce progression by other mechanisms apart from IOP reduction (neuroprotection).
The GPA (dichotomous) model also shows the relationship between the number of medicines and progression was also demonstrated in the GPA (dichotomous) model. The logistic regression analysis showed a significant association (p=0.002) between the number of medications and the detection of progress/no progress of the visual fields in the GPA. A posthoc analysis showed that the average amount of medicines in the no progression group was n=2.1 ± 0.8; meanwhile, in the progression group was n=3 ± 0.7. The estimated Odds Ratio (OR) coming from the regression analysis was 4.79. This means an increased risk of being treated with more medicines if you are suffering from progressing glaucoma than if your glaucoma is not progressing. A posthoc analysis was performed to clarify the issue. The patients were divided into groups 1: 1 or 2 medicines and group 2: 3 or 4 (only one patient was treated with five medications and was excluded). Then an OR was measured between the two groups showing a value of 8.47. Therefore, individuals with progressive glaucoma are eight times more likely to be treated with 3 or 4 medicines than individuals that suffer from non-progressive glaucoma. This is quite logical during the follow-up period if the ophthalmologist suspected progression wanted to decrease IOP even more. Also, it must be considered that high IOP at the diagnosis will increase progression itself and require an increased number of medications.
Cataract surgery was performed in n=14 patients during the three years follow-up period. At the baseline, n=23 patients were already cataract operated. Interestingly, cataract surgery showed association with progression only in the MD-based model. Cataract surgery improved the values in MD in the visual fields. The improvement of visual fields due to cataract surgery was not observed in the two other models (VFI and GPA). Nowadays, it's a general agreement not to use MD to evaluate glaucoma progression since cataract surgery can alter the parameter. MD was included in the present study because it has been included in several studies before.  The other reason is that the most common glaucoma classification (Hodapp’s classification) is MD values . The included patients in this study mainly belonged to the early and moderate glaucoma groups.
The study has certain limitations. As pointed out above, most patients belonged to early (MD=0-5.9 dB) and moderate (MD=6-12 dB) glaucoma groups. Very few patients with advanced glaucoma were included. Patients were excluded if they showed an MD ≥18 dB and/or VFI ≤40%. This criterium was chosen to avoid "floor effects” in which further loss of visual field defects can no longer be detected [19, 20]. This can be a limitation; unfortunately, advanced glaucoma patients cannot be followed with visual fields. Progression assessment would not be accurate when the visual fields are so damaged.
Another limitation of the study is that no morphological assessment of the optic nerve was done. The only parameter included was the cup/disc ratio (C/D), but this is a quite subjective manner of optic nerve assessment. In addition, the use of optic nerve OCT was not so common when the study began. Nowadays, OCT technology is part of the tools for glaucoma follow-up at our Department. However, it seems that OCT also has "floor effects" for advanced glaucoma patients . Still, there is no consensus about if morphology or function is the best to evaluate glaucoma progression .
In conclusion, the present study showed that age was the most critical risk factor for progression in new-diagnosed exfoliation glaucoma patients in Sweden. The other factors to be considered are IOP at diagnosis and IOP reduction both in absolute and relative number. Factors with lower impact in progression were positive family history, smoking, and SLT. Therefore, SLT treatment should be recommended in this patient group to slow down progression. Exfoliation glaucoma is an aggressive form of glaucoma that must be monitored often to avoid advanced visual field damage and blindness.