Introduction

Currently, rhegmatogenous retinal detachment (RRD) continues to be an important cause of visual loss. The fundamental principles involved in reattachment of a retina include identification of all retinal breaks, and relief of the vitreous traction. Traditionally, scleral buckling (SB) was viewed as the gold standard treatment for uncomplicated RRD. Pars plana vitrectomy (PPV) was traditionally reserved for treatment of eyes with complications, such as those showing giant retinal tears or exhibiting significant proliferative vitreoretinopathy (PVR). In the 1980s, the indications for PPV in RRD patients were broadened to include less complicated instances, and the term “primary vitrectomy” was introduced by Klöti.¹

As the necessary instrumentation for, and safety of, PPV continue to improve with developments in microscope technology, intensified endoillumination, and wide-angle viewing systems, the indications for vitrectomy in RRD have been further expanded to include most patients with RRD. Indeed, PPV is more useful than SB in eyes requiring simultaneous cataract extraction or those of pseudophakic status.^2–8

Compared to SB, PPV offers several advantages. The view of the retinal periphery is enhanced, identification of retinal breaks is rendered easier, achievement of complete intraoperative retinal attachment is possible, the risks of hemorrhage or retinal incarceration inherent to the external drainage procedure applied during SB is eliminated, and the technique is less likely to cause a refractive change. In addition, the recent introduction of small-gauge vitrectomy has shifted the paradigm of standard vitreous surgery to microincision vitrectomy surgery that is less invasive, affords fast recovery, and is sutureless (the acronym for such surgery is MIVS). As a result of these advances, vitreoretinal surgeons now have more procedural choices when treating RRD patients. Further, recently trained vitreoretinal surgeons may be more familiar with application of PPV (compared to SB) in challenging situations.

In this chapter, the use of primary vitrectomy for treatment of RRD, and the advantages and disadvantages of employing different gauges of vitrectomy system will be explained and evaluated.

Pathogenesis of Rhegmatogenous Retinal Detachment

The vitreous is firmly attached to the vitreous base, an area 3–6 mm in diameter that straddles the ora serrata surrounding the retina. The posterior border of the vitreous base is located both more posteriorly and temporally in older individuals. Therefore, retinal tears may occur at a higher frequency at the temporal periphery in such patients, following posterior vitreous detachment (PVD).

Three factors predispose to development of RRD: (1) the existence of a liquefied vitreous gel; (2) tractional forces that precipitate a retinal break; and (3) the presence of a retinal break through which fluid may access the subretinal space.

Liquefaction of the vitreous occurs naturally upon aging, developing more rapidly in eyes with significant myopia, surgical or non-surgical trauma, and/or intraocular inflammation. Apart from liquefaction, alterations in the extracellular matrix of the vitreous facilitate detachment of the posterior vitreous from the underlying retina. PVD usually presents as an acute event, and is more prevalent in patients older than 50 years, with frequencies as high as 53%.⁹ PVD often precipitates RRD because the tractional forces necessary to generate retinal breaks are produced upon development of PVD, although not all retinal breaks progress to detachment. The reported incidence of retinal tears in patients with acute symptomatic PVD varies from 8–46%.¹⁰ In a study in a general population,¹¹ 18% of eyes with retinal breaks developed retinal detachment. Risk factors for progression included fresh, symptomatic, horseshoe-shaped tears; breaks suggestive of the presence of subclinical retinal detachment (RD); and pseudophakia/aphakia.

Categories of Rhegmatogenous Retinal Detachment

No internationally accepted RRD classification has been established. The various reports suggest that the type of break, the status of the vitreous, and the need for cataract surgery are all important when defining the categories of RRD.

Retinal breaks are traditionally classified as round holes, tears, or resulting from retinal dialysis. Accordingly, RRD can be categorized as: (1) round hole retinal detachment; (2) retinal detachment occurring secondary to retinal tears; and (3) retinal detachment attributable to retinal dialysis. This categorization, and the status of the vitreous, are important when exploring management options, in particular whether SB or vitrectomy should be employed.

Retinal holes are full-thickness retinal defects that usually occur as a result of localized atrophic intraretinal abnormalities or lattice degeneration, and are not associated with vitreoretinal traction. Typical RD patients with round holes show one or more areas of limited retinal detachment, and may thus be optimally treated with either laser demarcation or segmental SB, rather than vitrectomy.

Retinal tears are usually produced by PVD and subsequent vitreoretinal traction. As persistent vitreous traction at an edge usually causes detachment to progress, the vast majority of RD patients in this category require surgical treatment. Among the various methods available, release of traction via PPV is preferred by many vitreoretinal surgeons.

Retinal dialysis is most often associated with blunt ocular trauma, and rarely occurs spontaneously. Dialyses are most common in the inferotemporal quadrant. The vitreous remains firmly attached to the entire peripheral retina, and vitreoretinal traction caused by gravity results in slow retinal detachment. Vitrectomy in an eye with an attached vitreous gel can be technically difficult, and may introduce unnecessary surgical complexity. Therefore, vitrectomy is less desirable than are other treatment options.

The number of aphakic/pseudophakic patients with RRD has increased significantly over the past decades. After cataract surgery, vitreous liquefaction accelerates, causing premature PVD and subsequent RD. Aphakic/pseudophakic eyes tend to have small (and sometimes multiple) anterior breaks. The status of the posterior capsule also influences the speed of vitreous liquefaction. One study found that Nd:YAG laser posterior capsulotomy increased the risk of RD after cataract extraction up to 4.9-fold.¹² Because of the characteristics of the breaks and the status of the vitreous in aphakic/pseudophakic eyes, vitrectomy is preferred by many surgeons and has a good success rate.¹³

Patient selection for primary vitrectomy

SB and/or pneumatic retinopexy serve as the first treatment option(s) in patients with localized detachment in one quadrant together with single neighboring breaks. Young age and anteriorly located small holes in phakic patients encourage the use of SB. The success rates of final reattachment are 90–95%.¹⁴

PPV is indicated for patients with wide and bullous RD, and for older patients with a liquefied vitreous. The presence of RD with marked traction with different anterior posterior depth of breaks, the presence of breaks in multiple quadrants, or the absence of an apparent retinal break in a pseudophakic patient, are all good candidates for the use of PPV. RD of preoperative PVR grade C, giant tear-induced RD, and macular hole RD, are all commonly treated using PPV.

Recent developments in vitrectomy instruments, including small-gauge systems, wide-angle viewing systems, and endoilluminators, as well as adjuvants, including triamcinolone acetonide suspension, perfluorocarbon liquids, and intraocular tamponades, led the choice of surgical technique for the treatment of RRD with medium-complexity shift more and more towards PPV.^15–17 During the last 5 years, several surgeons have reported that primary vitrectomy is the method of choice in 40–80% of all patients with RRD.^17,18

Several trials comparing PPV to SB in patients undergoing RRD surgery have been reported.^19–23 In general, PPV was favored for treatment of pseudophakic eyes with unseen breaks, eyes featuring ocular hypotony, or eyes showing prolonged macular detachment. One large-scale prospective multicenter randomized trial, scleral buckling versus primary vitrectomy in rhegmatogenous retinal detachment study (the SPR Study), also supported the superiority of PPV in the pseudophakic/aphakic eyes, in terms of a primary success rate and a rate of retina-affecting secondary procedures. This study, however, showed no differences in phakic eyes.¹⁹

The results to date suggest that SB or PPV is a good treatment option for primary phakic RRD, whereas PPV may be preferable to SB when pseudophakic RRD is to be treated. It is important to consider individual patient factors, surgeon preference, and availability of equipment, to make an optimal treatment decision.

Principles of vitrectomy

The principles of vitrectomy to treat RRD are release of tractional forces that precipitated the retinal break, and the closure and reattachment of breaks to the underlying retinal pigment epithelium. The surgical procedure requires: (1) removal of the vitreous gel and preretinal tractional membrane; (2) intraoperative flattening of the detached retina; (3) application of retinopexy; and (4) placement of a tamponade in the vitreous cavity.

Abnormal vitreoretinal traction (either perpendicular or tangential) increases vitreous mobility caused by PVD, and atypical posterior extension of the anterior vitreous base predisposes to formation of retinal tears. Therefore, removal of the vitreous gel and any abnormal preretinal structure releases the tractional force causing retinal breaks and detachment.

After release of abnormal vitreoretinal traction, the detached retina must be reattached. To stabilize and flatten the detached retina, a heavy liquid is initially applied; this is subsequently replaced by sterile air. If the retina is mobile and becomes flattened in air, a nonexpansive gas–air mixture is used to achieve a postoperative gas tamponade. Although silicone oil is not routinely used in instances of uncomplicated RRD, use of silicone oil should be considered if eyes have multiple inferior breaks, if PVR is present, or if the eye is an only seeing eye. Retinopexy has been used to create permanent retinal–RPE adherence. Both forms of retinopexy (cryopexy, and the laser photocoagulation) cause tissue destruction and cellular proliferation and should be used as little as possible. Continuous laser retinopexy is preferable to cryopexy, because cryopexy probably causes more tissue damage and a greater proliferative reaction than laser retinopexy. However, if peripheral breaks in phakic eyes are to be treated, cryopexy is preferred, because the endolaser probe can touch the crystalline lens. The risk of postoperative PVR may be minimized if retinal pigment epithelial cells dispersed after cryopexy are thoroughly aspirated via a heavy liquid–air exchange.

Surgical techniques

Primary vitrectomy is commonly performed using a wide-angle viewing system attached to an operating microscope. The surgical steps of primary vitrectomy using sutureless MIVS are described below (Figs 102.1–102.4).

Fig. 102.1 Detached retina with posterior vitreous detachment is shown. Core vitrectomy is performed.

Fig. 102.2 A bubble of perfluorocarbon liquid (PFCL) has been injected to displace posterior subretinal fluid. While holding down the detached posterior retina, peripheral vitreous base is safely shaved and the flap of the retinal break is cut to release the vitreoretinal traction.

Fig. 102.3 Endolaser retinopexy is applied around the retinal break under the PFCL bubble.

Fig. 102.4 Air–PFCL exchange is performed and air is replaced with SF₆, C₂F₆ gas.

Core vitrectomy

First, the central vitreous is removed (Fig. 102.1). When PVD is absent, or even when PVD has occurred, triamcinolone acetonide can be injected to afford better visualization of the vitreous gel, to facilitate PVD, and to completely remove the residual vitreous. The posterior vitreous membrane can be removed using a diamond-dusted scraper; this prevents secondary macular pucker (Fig. 102.5).

Fig. 102.5 Triamcinolone acetonide was injected for a better visualization of the residual vitreous gel. The posterior vitreous membrane can be removed using a diamond-dusted scraper.

Sutureless microincision vitrectomy surgery

Transconjunctival sutureless MIVS using 25-gauge (25G) instrumentation provides several advantages over conventional 20G surgery, including a shorter surgical time, less surgically induced inflammation, and a reduced risk of postoperative corneal astigmatism. These factors ultimately lead to improved patient comfort and faster visual recovery. The 2009 Preferences and Trends (PAT) surveys conducted by the American Society of Retinal Specialists reported that nearly 80% of respondents commonly employ small-gauge MIVS systems.²⁴

The original MIVS procedure, introduced in 2002,²⁵ had some limitations. The excessive flexibility of the first 25G instruments rendered it difficult for surgeons to adequately dissect the peripheral vitreous, to shave the vitreous base, and to rotate the eye. These limitations were particularly troublesome when operating on eyes with RRD. In addition, postoperative leakage of gas or silicone oil from the sclerotomies in the subconjunctival space was a further concern.

The introduction of 23G instruments solved several of these problems. The greater rigidity of such instruments, better illumination, and improved fluidics have facilitated manipulation of the peripheral retina, which is important when RRD is to be treated. MIVS technology continues to develop. Newly introduced 25G instruments, the 25G+ system, are substantially more rigid than their earlier counterparts. Further, an increase in instrument inner diameter has improved both infusion flow rates and illumination. In addition, a new trocar entry system and continued development of instrumentation have enhanced the performance of both 23- and 25G systems. Figures 102.6 to 102.9 show a surgical view of transscleral sutureless vitrectomy using the 25G+ system and a wide-angle viewing system.

Fig. 102.6 Core vitrectomy has been performed and a small bubble of PFCL was injected to hold down the posterior retina.

Fig. 102.7 After peripheral vitreous base dissection, more PFCL has been injected to the level of peripheral tear. Subretinal fluid was first displaced anteriorly by PFCL, and then aspirated through the peripheral retinal break.

Fig. 102.8 Endolaser retinopexy was applied around the retinal breaks under the PFCL bubble.

Fig. 102.9 Air–PFCL exchange was performed while draining subretinal fluid through the peripheral retinal break.

The most recent vitrectomy system features a pneumatic dual-drive cutter with an ultrahigh cut rate of 5000 cpm that effectively reduces vitreoretinal traction.^26,27 In this instrument, the port is also 50% closer to the tip than is the case with conventional cutters, and the inner lumen diameter is enlarged. Therefore, surgeons may work remarkably close to a detached retina, with minimal movement of mobile tissue. Other features include IOP compensation via direct control of infusion pressure, and direct control of the duty cycle (this controls the proportion of time during which the port is open, in terms of the total time of the cutting cycle). These features afford excellent control of flow rate with a constant cut rate, thus increasing precision without sacrificing speed.

A new scleral entry system has been introduced to create a smaller and tighter wound. This system uses an MVR-style blade that features a trailing edge, so that the cannula can be inserted via a linear incision. Such incisions are associated with a need for less force during trocar insertion, and retain shape well, resulting in less postoperative leakage than is the case when chevron incisions are made.^26,27

Another recent advance in the field is the development of wide-angle viewing systems. These systems provide a better field of view compared to that afforded by contact lenses, improving surgeon recognition of peripheral fundus pathology and allowing lesions to be treated safely and efficiently. A wide view of the fundus has become increasingly important during treatment of eyes with RRD employing 23–25G MIVS, because it is difficult to rotate the globe and indent the peripheral sclera during vitrectomy. Additionally, wide-angle viewing systems afford a relatively good view of the fundus in eyes with small pupils or mild corneal opacity. Further, surgeons do not need to change the contact lens during fluid–air exchange at the end of vitrectomy. Several types of wide-angle systems are available, either contact-lens or noncontact instruments. The noncontact systems are currently preferred.

However, transconjunctival MIVS may not always be the ideal in certain circumstances when treating RRD; peripheral vitreous dissection with scleral indentation may be difficult in the absence of conjunctival peritomy, especially in patients with a shortened fornix and a narrow palpebral fissure. In some patients, for whom MIVS is not optimal for completing surgery, one port can be enlarged to 20G and a procedure such as silicone oil injection can be performed. Such hybrid surgery is also ideal in more complicated circumstances, for example when silicone oil is to be removed or when a dislocated intraocular lens is to be repositioned.

Several groups have reported good surgical outcomes when RRD was treated with transconjunctival MIVS, either the 25- or 23G systems (Table 102.1).^{7,8,15,28–32} Although a few reports found that such MIVS afforded a lower single-operation success rate (SOSR) than did 20G vitrectomy^7,29 most authors attained success rates (91.7–97.4%) similar to that of 20G surgery.^8,28,30,31 The major reported advantages of MIVS include no or mild pain on the first postoperative day in the majority of patients,²⁹ a shorter operation time,⁸ and faster postoperative visual recovery.³³ Although some instances of transient hypotony have been reported following MIVS, no patient experienced significant postoperative leakage or endophthalmitis.^{7,8,15,28–33}

Table 102.1 Surgical outcomes of microincision vitrectomy surgery for treatment of RRD

Surgical outcomes

After the occurrence of any type of RD, approximately 40% of patients will not achieve reading ability, 10–40% will need more than one surgical procedure, and approximately 5% will suffer permanent anatomical and functional failure. The reported primary success rates for RRD repair by PPV range from 64–94%.^{3–5,17,19,22}

In the time since the first report of the use of PPV without concomitant SB to treat RD appeared in 1985,³⁴ numerous case series have been reported (Table 102.2).^{6,19,22,34–36} Although primary vitrectomy had historically been indicated principally for treatment of pseudophakic RD with superior retinal breaks, several studies found that the technique was also useful in phakic RD patients and in those with inferior breaks.^22,35 In general, the SOSR and the improvement in visual acuity appear to be comparable to those achieved using SB, in a wide variety of patients. The poorest outcomes were reported in series containing patients with chronic detachment and evidence of PVR. Exploiting the recent advances in surgical technology including MIVS, several recent case series reported that the SOSR improved to over 90% and that the final average visual acuity was 20/40 or better.

Several retrospective series and prospective clinical trials comparing SB, PPV, and/or combined SB/PPV, found no statistically significant difference in SOSR among the various procedures (Table 102.3).^{4,19,20,22,23,37–40} However, a few studies reported that PPV was superior in terms of either anatomic or visual outcome in comparison with SB alone.^19,20,22,40 In addition, several reports comparing PPV and combined PPV/SB also found that the use of a combined procedure did not significantly influence anatomical or visual outcomes.^17,38,39

A meta-analysis of 29 studies on patients with pseudophakic RD found that both PPV and combined PPV/SB were associated with a higher SOSR and better visual acuity outcomes than was SB alone⁴¹; another review of nine reports comparing PPV to SB found no statistically significant difference with respect to SOSR or visual outcomes.⁴²

Between 1998 and 2003, a prospective, multicenter randomized trial of SB versus PPV to treat RRD was conducted in Europe, and the data were reported in 2007.¹⁹ This is the first prospective large-scale randomized multicenter clinical trial with good numbers of centers, surgeons, and patients. Patients were classified on the basis of lens status (pseudophakia/aphakia or phakia), and the group results were separately analyzed. In the pseudophakic/aphakic group, patients treated with PPV showed a significantly greater SOSR and a lower rate of reoperation than the SB group. In the phakic trial, no difference was seen in terms of anatomical success or postoperative PVR between the two groups, although a greater rate of cataract progression was evident in the PPV group.

Although the data are not entirely consistent, most surgeons agree that PPV without buckling effectively treats pseudophakic RD and results in better long-term visual and anatomical outcomes than does conventional SB.⁶ The rationale for use of primary PPV alone to treat pseudophakic RD patients is that visualization of the peripheral retina is good, the causative retinal breaks can be properly identified and treated, buckling-related complications are absent, and the risk of PPV-induced cataracts is absent. When it is considered that the need for re-operation is a (negative) indicator of the quality and efficacy of surgical techniques, the SPR study showed that the risk of such surgery was significantly reduced when PPV was used, compared to SB, in pseudophakic eyes, but increased after use of PPV compared with SB in phakic eyes. The results also showed that the prognostic factors differed between the pseudophakic and phakic subgroups.⁴³

Prognostic factors

Identification of patients at high risk of failure at the time of initial presentation would allow surgeons to individually tailor patient management. Reported risk factors for surgical failure after PPV in the repair of RRD include duration of symptoms, older age, the extent of RD, macular detachment, involvement of inferior quadrants, absence of detectable retinal breaks, high myopia, hypotony, and PVR-related risk factors such as pseudophakia/aphakia, uveitis, vitreous hemorrhage, and preoperative PVR.^{6,17,22,41–45}

The extent of RD can be considered to serve as a surrogate marker of RD duration, although the location of retinal breaks may affect the speed of detachment after onset. Nevertheless, such patients may have a higher level of retinal glial upregulation and RPE cell dispersion, and may also develop PVR, which is an important risk factor in terms of surgical failure.⁴⁶

A number of studies have found that previous lens extraction was a risk factor for development of PVR and was associated with worse outcomes.^47,48 The association between pseudophakia/aphakia and surgical failure may be related to the observations that (1) pseudophakic/aphakic patients often have small, sometimes multiple anterior breaks which may be missed at the time or surgery; or (2) such patients tend to develop new retinal breaks after vitrectomy, caused by residual vitreous base traction. Many features of pseudophakic eyes, including anterior or posterior capsular fibrosis, the presence of cortical remnants, poor pupillary dilatation, vitreous opacity, and optic aberrations secondary to IOL placement per se, may seriously compromise the surgical view, resulting in poor outcomes.

Macular detachment and subsequent structural change detected by OCT appear to cumulatively reduce final visual acuity.⁴⁵ An experimental study showed that rapid retinal reattachment reversed the retinal degeneration associated with detachment.⁴⁹ As complete intraoperative attachment of the macula is achieved in most patients during vitrectomy, visual recovery after surgery is faster. Delayed or incomplete visual recovery after SB for macula-off RD may be associated with macular subretinal fluid persistence, which can also be assessed using OCT.^50,51

Inferior breaks have also been reported to be associated with surgical failure following PPV for RRD.^3,22,35 The functional study of the SPR group showed that inferior detachment with breaks below the 4 and 8 o’clock positions was a significant risk factor in pseudophakic/aphakic subtrial.⁴⁴ However, other studies have failed to show any advantage of combined PPV with SB, compared with PPV alone, in treatment of inferior retinal breaks.⁴⁵

Complications

The most frequent intraoperative complications are relatively high rates of iatrogenic retinal breakage (0.78–24%),^{6,20,22,23,37,38,40} and crystalline lens damage (0.03–9%).^21,37,52 Several less serious complications include retinal incarceration at retinotomy sites (0.6–2.9%), corneal abrasion (0.6%), scleral rupture (0.2%), and choroidal effusion (0.5%), as reported in one national audit.⁵³

Several complications may be encountered after primary vitrectomy in RRD patients. Transient or persistent intraocular pressure increases have been reported in 15–24% of patients.^{6,36–38,40,53} Phakic patients are at significant risk of developing nuclear cataracts; this is especially true of patients older than 50 years (21–86%).^{19,37,40,52–54} Other anterior segment problems, occurring less frequently, include hemorrhage, subluxation or capture of intraocular lenses, posterior capsular opacification, and synechia.^{3,6,19,31,36–38,40,52,53}

Major postoperative complications include retinal redetachment, with or without PVR, usually by new or missed breaks, or reopening of former breaks. As discussed above in terms of surgical outcomes, recent advances in surgical technology, including the development of MIVS and wide-angle viewing systems, have reduced the frequency of anatomic failures. Apart from recurrent detachment, several posterior segment complications may occur; these include cystic macular edema, macular pucker, and macular hole.^{3,6,37,53–56} Although very rare, endophthalmitis has been reported,⁵³ as has sympathetic ophthalmia. The incubation period for the latter condition after vitrectomy varies from 4 weeks to 2 years.^57–59

Recently, as MIVS has become more popular, the risks of postoperative hypotony and endophthalmitis have become of concern. However, recent case series have found that endophthalmitis did not occur more frequently than after conventional vitrectomy, and postoperative hypotony following MIVS was transient and resolved spontaneously in most patients.^{7,8,22,23,28,30–32}

Perspectives

Delineation of the optimal surgical approach(es) for the treatment of RRD requires further work. Future studies should evaluate surgical interventions from multiple perspectives, not only clinical but also socioeconomic. In addition to final anatomic and functional outcomes, the rate of recovery and the need for secondary procedures should be considered. As the paradigm of vitrectomy has shifted from conventional 20G PPV to sutureless MIVS using either 25G or 23G, it is now imperative to assess patient outcomes including postoperative pain and discomfort, time away from work, and quality-of-life. Relevant societal outcomes include procedural cost and cost-effectiveness; these issues cannot be ignored as technology advances rapidly.