To my knowledge, this is the first prospective study of corneal endothelial cell status after phacoemulsification cataract surgery up to 7 years.

The ECL was significantly greater in group 3 at 7 years although the total ECC at 7 years did not differ between the groups. This can somehow appear contradictory, but since there seem to be a tendency for a converging state among the groups, the ECL may as well not differ in a future time point. The statistics, with a wider range in standard deviation in the ECL calculations, may also have contributed to the results.

Observing the ECC at different time points, one can hypothesize that the initial surgical trauma initiates a long-term ECL that is predestined regardless of the early ECL. The greater cell loss in group 3 up to 3 months is compensated by the lack of further loss up to 7 years, whereas the not significant cell loss in group 1 up to 3 months progressed significantly up to 7 years.

The same pattern was seen in the comparison between clear and not clear corneas where the initial cell loss in the not clear group was compensated by no loss between 3 months and 7 years. On the other hand, a greater loss was found in the clear corneas between 3 months and 7 years with no difference in ECC at this time point.

It is stated that corneal endothelial cells cannot regenerate. However, studies have questioned the existing paradigm concerning corneal endothelial wound healing. Evidence has emerged that healthy peripheral cells in the cornea can compensate for the ECL and restore function. Maybe this is a part of the explanation regarding the results [15].

The time point 7 years is a comparably long interval from 3 months. The time for the follow-up was merely chosen since the opportunity and financing came up at that moment. It is likely that the results could be obtained earlier since the cells likely stabilize at an earlier time point. However, an even longer follow-up could be of interest to establish whether there is a complete steady state. The medium age for cataract surgery in Sweden is 74.6 years and the medium life span is 82.55 years, so it can be encouraging that even if there is an initial large cell loss, it can level out and the cornea can stabilize with a lower risk for decompensation[16.]

A current retrospective study found that a 10-year ECL was 20.62 ± 13.63%[17.] They also showed that the overall cell loss was correlated to the degree of early corneal swelling. In our study, the total cell loss was 23.46 ± 13.94%, but unlike the 10-year study we included brunescent cataracts. In our previous study, nucleus colour was the only pre- or intraoperative variable that independently correlated with the central endothelial cell loss. All four lenses with nucleus colour grade 5 were in group 3, and all these eyes had unclear corneas at the first postoperative day. As discussed above, it can somewhat be reassuring that a hard lens leading to an initial substantial corneal oedema and ECL, in the long term, does not seem to have a lesser ECC. The crucial point in these cases is the preoperative ECC and if it is too low, the patients cornea may decompensate permanently after the cataract operation[18,19,20.] On the other hand, in patients with guttata and a significant postoperative corneal oedema, one can benefit from waiting and observe since the cell loss might not progress further. This is also what is recommended in some literature[20.]

Corneal endothelial cell capacity is not only dependent on the number of endothelial cells but also how the existing cells function. There is no feasible way to check the function of the remaining cells, but the corneal thickness is one way to indirectly evaluate the function. The cell function may partly also be reflected by the changes in HSF, DE and CV. After the surgical trauma, the cells become more variable in shape, but in time they “come to rest” with a more normal function.[6, 21] According to the literature, the morphology parameters will probably stabilize much earlier and did not differ between the groups at 7 years. An equivalent function in the groups can be assumed since there were no differences between the groups in VA and corneal thickness at the 7-year follow-up.

The Orbscan II slit-scan method to measure the corneal thickness was chosen because it was used in the primary study of the three groups, so that the same method was applied. One benefit of this method compared to i.e. ultrasound is that it can be more consistent when comparing different areas of the cornea. The slit-scan method can be less accurate than i.e. ultrasound, Scheimpflug, or OCT imaging, but in this case this method had to be used in order to achieve a correct comparison.

Regarding the sample size, the groups were small but were taken from the first study with three groups of ten patients each. The results should therefore be interpreted with some caution. Nevertheless, to my knowledge, this is the longest prospective follow-up of detailed corneal endothelial cell changes after phacoemulsification cataract surgery.