The study included 338 eyes with iFTMHs and revealed a spontaneous closure rate of 9.2% at a median of 44 days after the diagnosis. All eyes with spontaneous improvement had small iFTMHs.

Previous retrospective studies using OCT imaging for iFTMH diagnosis reported spontaneous closure in 2,7–3,5% of eyes, while the mean time to spontaneous closure were 2, 2.5, and 5 months [17,18,19]. The time to closure was similar in our large cohort, however, we found a higher rate of spontaneous closure (9.2%). This discrepancy may be explained by the different study populations. In our study, eyes with spontaneous closure had an average hole diameter of 98.9 µm, which was smaller compared to 201 µm, 195 µm and 162 µm found in previous studies. This might suggest that patients with smaller iFTMH are underrepresented in the other studies, leading to a lower spontaneous closure rate.

One of the prospective studies investigating the effect of ocriplasmin in eyes with vitreomacular traction or iFTMH found in the control group a spontaneous closure rate of 15.4% (4/26 eyes with iFTMH) at month three without re-opening in the 24-month follow-up period [8, 23, 24].

In our study population, spontaneous closure was found only in eyes with small iFTMHs, whereas in the mentioned retrospective studies, few eyes had diameters greater than 250 µm [17,18,19]. Smaller iFTMHs have been reported to have a higher likelihood of spontaneous closure. Our study found a significant correlation between hole diameter and spontaneous closure.

There are also reports regarding the spontaneous re-opening between three months and three years after achieving complete iFTMH closure with or without surgical intervention [10, 25,26,27]. In our study population, we found re-opening of spontaneously closed iFTMH in five patients with a median of 4 months after closure. Considering the relatively short time span until reopening, the continuation of the regular follow-up examinations is important. In our clinic, patients were advised to have follow-up examinations at 2-week intervals for 6 weeks after the initial diagnosis. Subsequently, ophthalmological examinations were performed alternately with local ophthalmologists at four-week intervals, provided an OCT examination was available. As all patients in our study had full health insurance, it is unlikely that any missed follow-up examinations were due to monetary reasons. Although we cannot be certain, we assume that patients were referred if any morphological or functional worsening was noticed by the local ophthalmologists. All patients with a re-opening after spontaneous closure underwent macular surgery and showed closure of the iFTMH. Due to the small number of observed re-opened iFTMHs, further statistical analysis of OCT features was not feasible. In the aforementioned retrospective studies and the prospective OASIS study, only one re-opening was described which might be due to the lower number of observed spontaneous closed iFTMHs compared to our study [17,18,19, 23, 25, 28].

The exact mechanism of spontaneous closure of iFTMH remains unknown. One explanation is that the release of vitreous traction leads to decreased vertical forces and to a flattening and convergence of the hole edges [29]. However, since the proportion of eyes with and without vitreofoveal traction was comparable in both groups (30% in the no spontaneous closure group, and 45% in the spontaneous closure group, p = 0.103), other mechanisms should also be considered. Furthermore, closure of the iFTMH with persistent vitreofoveal traction is also reported [18]. Consequently, we would consider the release of vitreofoveal traction favorable for spontaneous closure, although this cannot be generalized for all cases with vitreofoveal traction.

Interestingly, in case reports of secondary full-thickness macular holes associated with cystoid macular edema in uveitis, ocular trauma, post-laser inflammation, or Irvine-Gass syndrome, the resolution of the intraretinal pseudocysts is thought to have contributed to spontaneous closure [10, 26, 30,31,32,33,34,35,36].

The formation of a bridge-like structure at the edges of the hole is thought to be a reparative reaction of the Müller cells preceding the spontaneous closure [37,38,39]. In line with these reports, many eyes with spontaneous closure in our study presented with a bridging phenomenon in the OCT (Fig. 3).

Spontaneous closure of an idiopathic full-thickness macular hole with bridging phenomenon in a 57-year-old lady 52 days after diagnosis

So far, limited information is known about the time course of spontaneous closure of the small iFTMHs. Based on the data presented, approximately 25% of small iFTMHs (diameter ≤ 250 µm) and around 50% of iFTMHs with diameters < 150 µm can close spontaneously within two months of diagnosis. Therefore, in patients with small iFTMHs and relatively good vision a „watchful waiting" approach with short follow-up regime may be offered to some patients after careful considerations of all circumstances. However, a recent meta-analysis estimated a minor vision loss of 0.008 logMAR per month of additional waiting for surgery [20]. As this study included iFTMHs of all sizes, and small iFTMHs have a higher surgical success rate, extrapolating the degree of vision loss for small iFTMHs only is difficult [20, 21].

One of the main limitations of this study is its retrospective study design. However, we believe that our analysis is rather underestimating spontaneous closure rates because patients with spontaneous BCVA improvement may not seek ophthalmological examination in the first place or are more likely to miss follow-up appointments. Moreover, the time of the iFTMH closure is likely overestimated, as spontaneous closure was diagnosed on the day of the OCT examination and not at the time of subjective BCVA improvement. However, it is notable that the time of diagnosis based on the OCT examination is not the time at which the iFTMH actually develops. An alternative to our approach is using the onset of symptoms as the time of diagnosis. Nevertheless, we have decided against this, as not all patients are able to state the onset of symptoms precisely enough (e.g. if the healthy partner eye is dominant). Therefore, caution should be applied when generalizing these results.

In summary, this study revealed spontaneous closure of small iFTMHs in a relatively high percentage of patients within two months of diagnosis in our cohort. Considering that the established standard therapy for iFTHMs is a surgical approach, this information could be relevant to clinical practice and future studies. While our data cannot be used to predict the course of an individual patient's disease with high accuracy, future investigations are required to validate the conclusions drawn from this study and to help establish validated treatment recommendations for patients with small iFTMHs in the clinical routine.