Surgical technique for removing vitreous cortex remnants using a diamond‐dusted membrane scraper

The purpose of this paper is to present a new surgical technique to remove vitreous cortex remnants.


Introduction
Vitreoschisis is a consequence of anomalous posterior vitreous detachment (PVD), in which the posterior vitreous cortex splits into two layers, and its outermost sheet remains attached to the retina, forming so-called vitreoschisis-induced vitreous cortex remnants (VCR) (Sebag 2008). This process may occur spontaneously or be induced during surgery. Vitreous cortex remnants are associated with macular diseases, such as macular holes or macular puckers (Sonoda et al. 2004;Sebag et al. 2007). Notably, a recent theory by van Overdam proposed that VCR are present in many cases of primary retinal detachment (RD) and play a crucial role in retinal redetachment (van Overdam 2020). In particular, VCR, by becoming a scaffold for cell proliferation, might promote the development of proliferative vitreoretinopathy (PVR) (Sebag 2008). As the vitreous is transparent, VCR may be better visualized with triamcinolone acetonide (TA) staining, even if this procedure is not routinely performed during pars plana vitrectomy. Consequently, the prevalence of VCR is probably underestimated. Reported VCR prevalences range from 67% to 75% in the macular area (Kimura et al. 2004;Chen et al. 2006;Cho et al. 2018), and 29% to 44% in the extramacular area (van Overdam 2020).
Once visualized, the removal of VCR is challenging due to the risk of damaging the underlying retina.  (Shanmugam et al. 2019).
To overcome these issues, we propose an alternative technique to remove VCR using the Tano Diamond-Dusted Membrane Scraper (DDMS) (Synergetics, O'Fallon, MO, USA).

Materials
The technique was applied to patients affected by primary rhegmatogenous retinal detachment (RRD) referred to our clinic between January 2020 and June 2020. The exclusion criteria were prior vitreoretinal surgery; RRD caused by giant retinal tears; undetected retinal breaks after either dynamic scleral depression or scrupulous vitreous base shaving; PVR grade C (Machemer et al. 1991) or worse; and ocular disorders that could interfere with visual recovery (e.g. age-related macular degeneration, diabetic retinopathy, or glaucoma). Patients were included once the surgery started, and the presence of VCR was confirmed by TA staining. Demographic data are reported in Table 1. The study was conducted in adherence to the tenets of the Declaration of Helsinki.
A standard 3-port vitrectomy with chandelier illumination was performed using the Constellation Platform (Alcon, Fort Worth, TX, USA), and a complete vitrectomy with shaving of the vitreous base was performed. Vitreous visualization was enhanced with TA (Kenacort-A, 40, Bristol-Myers Squibb Srl, Anagni, Italy) after removing the suspension's preservatives. The TA concentration was 20 mg/ml. The extent of VCR was highlighted by reinjecting the TA in the vitreous cavity towards the retinal surface. Perfluorocarbon liquid was injected to secure the retina at the posterior pole.
Subsequently, the DDMS was introduced into the vitreous cavity, and VCR were gently removed from the detached retina. The manoeuvre started by sweeping with the DDMS over the retinal surface until a free edge of the vitreous cortex was created in the mid-periphery.
When the vitreous cortex's flap was obtained, it was enlarged first in a centripetal way (towards the posterior pole) then in a centrifugal way (towards the vitreous base) and circumferentially. The vitreous cortex was meticulously removed in all quadrants from the detached retina. It is essential to note that a concentric ring of VCR was left on the retina's surface, central to the point where the free edge was initiated.
Once the detached vitreous cortex was lifted sufficiently, it was gently aspirated by the vitreous cutter (Fig. 1).
Triamcinolone acetonide (TA) was readministered to check for any persistent VCR. Finally, endorser was applied around the retinal breaks and air tamponade was performed. Video S1 displays a case treated with 23-gauge vitrectomy, although this technique was also suitable for 25-gauge vitrectomy, as DDMS is available in different diameters. In addition, Video S2 shows the same case of VCR removal with threedimensional resolution.

Results
Between January 2020 and June 2020, the same surgeon (G.C.) performed this procedure on 34 eyes affected by primary RRD, with a minimum follow-up of 6 months. The number of TA injections depended on the thickness and consistency of the VCR, which varied among different cases. In particular, when the VCR was thick, TA was applied twice; conversely, when the VCR was thin, multiple reinjections were needed, usually 3-4, and the excess dye was removed each time using a backflush.
Intraoperative complications have not been reported in any case. Only one patient developed retinal redetachment because VCR was not wholly removed from the detached retina, becoming pivotal for PVR development, which was identified during the second operation.
The retinal redetachment rate in our cohort of patients was 2.9% (1/34). This result is comparable with those in other cohorts reported in the literature, such as the 4% (4/91) and 0% (0/68) redetachment rates reported by van Overdam (2020). These results are encouraging compared to the redetachment rates (11% and 14%) in another cohort of our patients affected by RRD and treated with vitrectomy performed by the same surgeon (G.C.) without VCR removal. In this study, the mean follow-up was 7.75 AE 1.42 months, and the patients were comparable, with identical exclusion criteria. However, TA was used to enhance vitreous visualization but not to highlight VCR. Therefore, the rate of VCR was not reported. (Loiudice et al. 2021) In addition, our results are below the mean value of retinal redetachment reported in the literature (21%) (Znaor et al. 2019).

Discussion
We agree with the theory proposed by van Overdam that highlighted the role of VCR as a scaffold for fibrocellular proliferation with consequent PVR development (van Overdam 2020). Consequently, we agree with the benefits of removing VCR in the midperiphery to improve surgical success in RRDs. However, we believe that the vitreous wiping technique has some drawbacks. First, manual cutting of the PVA piece is required; thus, dimensional variability can influence the elastic properties and alter the instrument's feedback. Second, Eyetec TM PVA may not be readily available in all hospital facilities. Third, vitreous wiping requires enlargement of the sclerotomy to introduce the PVA piece into the vitreous chamber.
To overcome the abovementioned issues, we decided to remove VCR with a DDMS. This device was introduced in 1997 to create a free edge of the epiretinal membrane, helping the surgeon remove it (Lewis et al. 1997). Later, internal limiting membrane peeling with a DDMS was also proposed; however, the manoeuvre was prone to damage the underlying neurosensory retina; thus, it is not currently recommended (Kuhn et al. 1998;Hirakata et al. 2010;Leung et al. 2016).
Nevertheless, a DDMS can easily follow the contour of the retina, being a silicone tube, and it can easily create a flap in the VCR thanks to the tip coated with diamond dust. Thus, we believe that the DDMS may be appropriate for removing VCR. It also does not require sclerotomy enlargement and provides constant elastic properties in consecutive surgical procedures, not requiring manual fashioning. Conversely, DDMS provides constant and predictable feedback during multiple surgeries. In our opinion, this feature is crucial for surgeons who need to improve their confidence with this challenging manoeuvre.
We acknowledge that this technique has some limitations. First, very gentle wiping with the DDMS is needed to avoid intraoperative retinal tears. However, the learning curve is short due to predictable feedback, and PFCL provided countertraction. Second, investigations (microperimetry, static visual field) of retinal toxicity are warranted to understand the impact of the manoeuvre on retinal function. In particular, postoperative microperimetry results from two groups of patients affected by RRD of similar duration, one group with and one without VCR removal, should be compared. Third, the extent and indications for VCR removal remain to be determined. Is it advisable to perform VCR removal only on the detached retina, or should it also be performed on attached retina? May it be suggested in all cases or only in selected ones, in particular cases with known risk factors for PVR, such as large retinal breaks or giant tears, vitreous haemorrhage associated with retinal tears, multiple previous eye surgeries, previous trauma to the posterior segment, or preexisting signs of localized PVR such as fixed folds?
In summary, the prevalence of VCR is underestimated, and questions have been raised regarding whether the risks of VCR removal outweigh the benefits. However, in our opinion, VCR plays a crucial role in the success of RRD surgery, as well as complete vitrectomy with vitreous base shaving. Notably, TA should be routinely introduced in vitrectomy for RRD to understand the true prevalence of VCR. Following van Overdam, we suggest removing the VCR to improve the retinal detachment success rate (van Overdam 2020). We understand that this manoeuvre is challenging regardless of the instrument used. Nevertheless, we propose a new technique to achieve this goal. A DDMS can easily follow the contour of the retina without requiring sclerotomy enlargement. It also provides predictable feedback, improving the surgeon's learning curve. We intend to investigate the redetachment rate after VCR removal with DDMS in vitrectomy for RRD and to compare it with those obtained from other techniques, such as vitreous wiping.