We report the retrospective data from 45 eyes that received an ICL V4c implantation for correction of myopia in a major German university clinic over the time span of 7 years. We reviewed all patient charts and included those with a follow-up of 5 years including subjective refraction, uncorrected and corrected visual acuity, endothelial cell count, intraocular pressure, and assessment of possible complications. With a mean preoperative SE of  − 10.13D ± 3.39, we treated highly myopic patients and were able to correct them to emmetropia with a SE of 0.05D ± 0.44 at 1-month postoperatively with a slight trend of myopization in the long-term follow-up. This could be due to changes/growth of the natural lens with consecutive myopic shift of the refraction. Alternatively, a possible mechanism would be axial growth. But since no measurements were made, we cannot answer this.

Comparing our postoperative spherical equivalent to the current literature, Wan et al. [21] reported similar outcomes to ours in 2020 for 137 eyes 6 months postoperatively with a mean SE of  − 0.03D ± 0.07. Kamiya [22] reported a SE of 0.01D in eyes with myopia ≤ − 6D and 0.02D in highly myopic eyes > − 6.0D 12 months after the implantation of the ICL in more than 350 eyes. Additionally, the predictability outcomes were excellent with 93% and 94% of eyes being within ± 0.5D of target refraction. Kojima compared the V4c to the slightly bigger V5 ICL and had a postoperative SE of 0.05D ± 0.07 after 6 months for 23 patients [23]. These results do exceed ours, with 71% of eyes being within ± 0.5D and 94% being within ± 1.0D, but our results were taken 5 years postoperatively. Shimizu et al. [24] and Alfonso et al. [25] reported similar results with 95–98% of eyes being within 0.5D of target refraction. Five years results similar to ours are reported by Chen et al. [26] with 79% of eyes being within 0.5D of target. Some authors report more myopic outcomes in the long-term follow-up [27, 28]. Those more myopic outcomes could be due to a progressive myopization in young patients or changes of the lens nucleus that could be an early stage of cataract formation. This was the interpretation of the myopic outcome of  − 0.9D ± 0.95 by Yan et al. [29] after a 2 years follow-up. Other trials describe myopization as early as 1 year after ICL implantation [30]. With a myopic shift of  − 0.5D in our patients during our follow-up of 5 years, this is a trend that we were able to reproduce. When comparing long-term results, repeatability needs to be kept in mind in terms of efficacy and stability which can be seen in publications from Liu et al. [14, 31] and Alfonso et al. with a follow-up of up to 25 months and up to 5 years with only 69% and 67% of eyes being within ± 0.5D. Data of a 10-year follow-up by Choi et al. [32] report a myopic outcome with a mean SE of  − 0.69D and a UDVA that decreased from 0.06 logMAR to 0.13 logMAR accordingly at the end of the follow-up while CDVA remained stable.

Postop CDVA was  − 0.02logMAR ± 0.09 and stayed at this level during the 5-year follow-up. The uncorrected visual acuity was a little bit reduced due to the discussed myopization at 5 years postoperative. Comparing those results to Shimizu et al. in their trial at 6 months [24] and after 5 years [10], they report a better VA compared to our patients with a UDVA of  − 0.2 logMAR and  − 0.17 logMAR, respectively, and CDVA of  − 0.25 and  − 0.24 logMAR. But the patients in this trial did have a better CDVA preoperatively compared to our patients with  − 0.17 logMAR compared to 0.05 logMAR. Similar results are reported by Kojima et al. in 2018 [23]. Results comparable to ours are reported by Wan [21], Fernandez-Vega-Cueto [27], and Lisa [33].

Since ICLs are implanted in young patients, the long-term follow-up is of utmost importance due to axial length growth in highly myopic patients and possible changes in the refractive power of the lens. This is shown by the 5 years data of Alfonso et al. that showed a worse UDVA compared to ours (0.13 logMAR) but a similar CDVA (0.02logMAR) in 147 eyes [31]. Similar results are reported by Cao et al. [34] This was reproduced by our data as well with a UDVA of 0.15 logMAR after 5 years. But still some papers report a lower overall VA compared to our like Rizk et al. [35]

With 4 eyes (9%) losing 2 lines of CDVA but 44% gaining at least one line, we reached a very good safety index (SI 1.16) and sufficient efficacy index (EI 0.78) at 5 years postoperatively. This matches similar retrospective trials like Kamiya et al. [36], who report an EI of 1.18 and SI of 0.89 at 8 years postoperatively, with a loss of 1 line CDVA in 8% of eyes or Chen et al. [26] with SI of 1.03–1.32 and EI of 0.83–0.83 for different stages of myopia. These results are comparable to other trials of Chen et al. [28] or Martinez-Plaza et al. [37] Better SI and EI are reported by others with a SI of up to 1.67 [29] and EI of up to 1.5 [38] As already mentioned, the UDVA is influenced by the postoperative refraction that can change during the follow-up which could explain the EI of Fernandez-Vega-Cueto [27] or Alfonso [31] at 3 and 5 years postoperatively with 0.9 and 0.87 EI. However, when discussing the visual acuity of the follow-up, the retrospective nature of this trial must be considered. While patients in prospective trials will often be pushed to their visual limits and VA testing usually relies on forced choice testing, this could be a limitation of our trial and a possibly explanation of the number of eyes losing CDVA in our patients. But comparable results with CDVA loss of 1 or more lines in 17% of eyes were reported in other trials as well [38]. Packer et al. report even higher rates during their 11 years follow-up with 36% losing 1 line CDVA after 5 and 50% after 11 years, which could be due to lens opacification or corneal changes [39].

Due to the anterior chamber depth and possible changes in flow of the aqueous humor, the ECC needs to be monitored. In our study, the overall endothelial cell loss from preoperative to 5 years was 291 (9.66%), which would be 1.9% per year and therefore little above the range of a physiological cell loss in healthy eyes [40]. However, this could possibly be due to an initial cell loss caused by the procedure itself. Since we do not report short-term data, this cannot be verified by us. But initial cell loss due to the surgical trauma was seen in other trials, with a loss of 7.1% [41] after 1 year or a loss of 8.5% [25] after 6 months. After this initial loss of cells, most patients return to the physiological cell loss as described in the 5-year follow-up of Shimizu et al. [10] with a loss of 5.4% compared to an initial cell loss of 2.8% [24] for the first 6 months. Cell loss comparable to this was also published by Kohnen et al. in a 10-year follow-up after anterior chamber pIOL implantation [42]. This is an interesting finding since other papers report lower decreases of the ECC in ICL eyes with a cell loss of below 1% [15, 22]. But other extremes exist as well with Ganesh et al. [43] reporting a ECC loss of 9% after 1 year in 30 eyes or 22% after 5 years. With the mentioned cell loss of 1.9%/year, our results do compare to the literature as described and show that the ECC loss is above the range of the physiological loss of cells. Therefore, monitoring of the cell count is still highly important.

In our patients, a mean pIOL size of 13.02 ± 0.34 was implanted (mean WTW: 12.0 mm ± 0.40, mean ACD: 3.21 mm ± 0.31). The mean vault in our patients was 425 µm ± 204 with a range from 100 to 940 µm. This is similar to the current literature that reports vaults of 389 µm (90–700 µm) [14] or 405 µm (100–980 µm) [33]. However, the postoperative vault is also depending on pupil size, and this could possibly influence our data [44]. Still, all measures were taken in the same room at the same, low mesopic light conditions. New formulas developed to improve ICL calculation seem to reach better results, especially when using data of swept source OCT of the anterior segment [45]. Calculating the vault depends on different ocular parameters like, e.g., corneal diameter, anterior chamber depth, or axial length. Varying formulas are known and show varying results. Formulas depending on OCT seem to be the most promising [46]. With only 2 eyes having a hypo and 3 eyes having a hyper vault, the rate of eyes not being within the wished range is low in our trial. However, due to the retrospective nature of our trial, the timepoint of the vault being measured is rather inconsistent, which makes it hard to compare it to the current literature. If the vault is too small, the residual refraction could be myopic and vice versa. Additionally, it could cause cataract formation located at the anterior capsule of the lens. The rate of postoperative complications was very low in our patients. Two of the 241 eyes (0.8%) that had the pIOL implanted had cataract formation at a mean follow-up time of 21 months postoperatively which compares to or outperforms most studies that describe a rate of cataract formation below 5%, [35, 46] like a recent publication by Gonzalez-Lopez et al. [47] that found anterior cataract in one of 24 eyes (4.17%) with low vault. However, none of the eyes that finished the 5-year follow-up had a clinically significant cataract. Elevated tension was only seen at one week postoperative but could be treated by eye drops and did not increase compared to preoperative tension during the follow-up. This also is comparative to most trials [13, 15, 28, 31]. Chronic iritis and/or pigment dispersion was not seen in our patients but is described in other studies [35].

The main limitation of our trial is the retrospective nature of the study leading to an unstandardized postoperative follow-up time. However, thanks to strict postoperative standards, we believe the procedures and measurements that we reviewed are still comparable to current practices. Additionally, we only included procedures without complications.