By targeting patients with and without ocular symptoms, we aimed to identify subclinical ocular changes and clinical features of ocular irAEs. The prevalence of ocular irAE anti-PD-(L)1 treatment is reportedly < 1%1, most frequently manifesting as uveitis (1%) and dry eye (1–24%)2. As most of the reported PD-1 inhibitor-related ocular complications occur within a few weeks to months after infusion2, we performed a year follow-up.
None of the patients had visual impairments. One patient reported ocular pain after the ICI treatment. Patient #1 (Table 4) had dryness in both eyes before ICI treatment with a tear break-up time (BUT) of 4 seconds in both eyes with slight superficial punctuate keratopathy. The symptoms were worse for 2 weeks after every ICI treatment. The patient visited our clinic one week after the second ICI treatment, complaining of ocular pain in both eyes. BUT was also 4 second in both eyes with slight keratopathy at that time. We cannot conclude that the cause of ocular pain was deterioration of dry eye. However, the ICI treatment might have been influential in causing ocular pain. Including this patient as having dry eye as an ocular irAE, our result is consistent with the statistics (n = 1, 3.7%)2.
No patient showed inflammatory change in both anterior or posterior segment. Exudative changes or retinal pigment epithelium rippling was not observed. We expected that choroidal thickness might increase in some patients after ICI treatment, as seen in VKH. However, the CCT did not show a significant change in our study.
Six participants experienced systemic irAEs. One of the patients (#1) complained of ocular pain. The other patients did not have ocular symptoms. Both CCT and aqueous flare did not show any significant differences before and after treatment in these six cases (Fig. 2).
In normal eyes, PD-L1 is detected in corneal epithelial cells, corneal endothelial cells, iris/ciliary body cells, and retinal pigment epithelial cells8. PD-L1 expressed on human ocular cells is thought to play a role in controlling ocular inflammation by inhibiting the production of proinflammatory and Th2 cytokines by activated T cells8. There is a possibility that ICI might collapse the immune control system at the blood-ocular barrier and increase the aqueous flare. This was not observed in this study (Table 3 and Fig. 1). However, several factors are related to aqueous flare measurements. Application of mydriatic agent9, pupil size10, and diurnal variation11 of aqueous flares might have affected the test results. Aqueous flare intensity reportedly decreased significantly 1 h after the application of a mydriatic agent9. In the present study, we measured the flare intensity 30 min after the application of tropicamide 0.5% and phenylephrine 0.5%, which may have decreased the flare intensity. We did not dilate the pupil from the second visit, as the patients were undergoing chemotherapy and were not willing to undergo dilation. We understood the discomfort of having dilated pupils. Considering that the baseline was measured with a dilated pupil that decreases aqueous flare and that there were no significant differences in aqueous flare between the baseline and after ICI treatment, we can suggest that aqueous flare did not increase after ICI treatment.
Most of the reported PD-1 inhibitor-related ocular complications occurred within a few weeks to months after infusion2, while there was no change in the ophthalmological findings in our observation.
The present study has several limitations. One is the small sample size. Inclusion of various types of cancer and ICIs is another. In addition, some participants died, which shortened the follow-up period. The timing of the ophthalmologic test after ICI treatment was one month after the treatment. However, this timing could have missed ocular changes that occurred just after the treatment.
We tested the participants without mydriasis after ICI treatment, considering the patients’ conditions. However, matching the condition was needed to evaluate the differences in aqueous flare before and after ICI treatment.
Here, we described a one-year follow-up of the patients after ICI treatment. As long as we know, this is the first study to prospectively observe ophthalmologically. There was no patient who had ophthalmological change, which suggest that ophthalmological irAE is dose independent. Thus, we might be able to conclude from this study that we cannot predict whether the patient have risks of presenting irAE by following up ocular findings. However, this is still a study with small sample, so we are continuing this prospective study to validate the results of the present study using a larger sample size.