Since glaucoma was recognized as an eye disease related to elevated intraocular pressure (IOP) in the 19th century, treatment was mostly directed towards lowering IOP to a “normal” range (1). In the 1960ies, “guarded” filtering procedures were developed in order to reduce complications from “unprotected” operations such as Elliot’s trephination or Scheie’s procedure. Over 50 years, trabeculectomy has been considered the standard filtering procedure for advanced or progressing glaucoma cases to achieve long lasting low IOPs. Trabeculectomy has the advantage of reducing IOP into the low teens or upper single digits which is considered a safe “target” range suited to halt or reduce visual field (VF) deterioration. However, the complications of trabeculectomy are still considerable, such as hypotony, choroidal effusion or subchoroidal bleeding, and long-term success is challenged by excessive wound healing resulting in obliteration of the new outflow routes.
The different approaches of glaucoma treatment—medication, laser, and surgery—have been reviewed for effectiveness and outcome (2-4). In most of these studies, filtering surgery achieved lower IOPs than medication or laser. Surgical studies usually take IOP as a primary outcome measure, and not function such as VF. For the patient, however, the VF is more relevant as it is related to quality of vision.
Up to nowadays, the current view has been that retinal ganglion cell (RGC) function cannot be restored or improved because RGCs irreversibly die by apoptosis. However, improvement of retinal function has already been suggested by George Spaeth (5) as a measure of successful glaucoma treatment. According to his hypothesis, glaucoma is only adequately treated if improvement occurs.
During the past decades, quantitative VF evaluation often was based on global indices such as mean defect (MD), glaucoma index (GI) (6), Bebié curve (7), or selected integral areas of the VF test (8,9). However, global indices hide scotomata in small limited areas, which are typical for early glaucoma. The pointwise linear regression analysis, initially developed at Moorfields Eye Hospital was an approach to overcome these problems (10,11).
The paper by Caprioli et al. (12) evaluated the individual changes of small test locations over time using a similar test strategy. The hypothesis of improvement of the VF was addressed in patients that had sufficient VF follow-up before and after trabeculectomy to allow statistical significant regression analysis (follow-up ≥4 years and ≥4 VF examinations before trabeculectomy, and follow-up ≥4 years, ≥4 VF examinations after trabeculectomy). They also used statistical methods to remove outliers and tested linear versus nonlinear regression for follow-up. Using clinical data, the paper of Caprioli et al. (12) provides evidence that the VF can improve after successful trabeculectomy.
Here is a short summary of the results:
The advantage of using trabeculectomy cases (and not medically treated cases) is (I) the clear time definition of the therapeutic “intervention”; (II) the amount of IOP lowering; and (III) the independence from the compliance as compared to patients who are medically treated.
Four major results can be learned from this paper:
Of course, the high variation of individual test locations that improve versus those that decay is a challenge. In this study, 70% of locations decayed and 30% improved preoperatively, whereas 56% decayed and 44% improved postoperatively. Although this difference was statistically significant, one must be aware that the reproducibility of measurement in an individual patient may be critical from one to the next VF test in a busy routine clinic. Oblique head position, cyclorotation, attentiveness of the patient at different visits etc. may lead to considerable variability.
Retrograde axoplasmic flow inhibition at the level of the lamina cribrosa has been considered the primary step to deprive RGC somata from surviving signals sent from the lateral geniculate nucleus synapses (i.e., neurotrophins). This deprivation triggers irreversible cell death (apoptosis). On the RGC level, electrical transmission of information (propagation of axon potentials) is very sensitive to reduced oxygen supply, but can restore after long periods of oxygen deprivation (13). Electrical signal transmission of the axon is different from axonal survival regulated by its nutritional support from axoplasmic flow. Several observations support the view that ganglion cells can survive while electrical function is suspended when borderline nutritional support still available to the axon. For example, in pituitary tumor, VF defects are partly or completely reversible when pressure by the tumor is removed (14). The photopic negative response of the electroretinogram, a signal associated with the RGC function, improved in those eyes where the IOP was lowered ≥25% (15). In single cell recordings of retinal axons, electrical activity of the axon (action potentials) can resume even after longer periods of IOP induced ischemia (13,16,17). In such a scenario the RGC function may resume after lowering of IOP, in particular when lowered to a low-normal range by surgery.
The paper of Caprioli et al. (12) is a milestone telling us that (I) improvement of compromised VF areas can be achieved and (II) low normal IOPs as obtained by trabeculectomy are a prerequisite to reach this goal.
