CLINICALLY RELEVANT POSTERIOR VITREOUS DETACHMENT STAGING USING CIRCUMPAPILLARY AND MACULAR VOLUME OPTICAL COHERENCE TOMOGRAPHY

Supplemental Digital Content is Available in the Text.

The application of circumpapillary ocular coherence tomography for the assessment of posterior vitreous detachment (PVD) applies a newly defined grading scale that separates PVD into five distinct stages. Methods and staging presented can aid clinicians in clinical examination and surgical planning.

Abstract

Purpose:

This study was designed to investigate retinal nerve fiber layer circumpapillary optical coherence tomography to determine posterior vitreous detachment (PVD) status and to develop a clinically relevant PVD grading scale based on retinal nerve fiber layer circumpapillary optical coherence tomography to determine the incidence of PVD by age and association with vitreomacular traction disorders.

Methods:

Ophthalmic images and medical records of patients with retinal diseases were retrospectively analyzed by three masked graders using retinal nerve fiber layer circumpapillary optical coherence tomography and macular optical coherence tomography. Based on PVD status, eyes were categorized into five newly defined PVD stages.

Results:

Among 2002 eyes, PVD stages were as follows: A) 25 (1.25%); B) 725 (36.21%); C−) 248 (12.39%); C+) 151 (7.54%); D) 851 (42.51%); X) 2 (0.1%). Posterior vitreous detachment was correlated with advanced age (P < 0.0001). Limited separation or partial separation between lamella within the posterior vitreous cortex (Stage B) was noted early (68% of eyes <18 years). Overall, 34% of eyes >70 years did not exhibit complete PVD. Of 75 eyes with tractional vitreoretinal disorders, 64 (85.3%) were Stage C−/C+, identifying Stage C as the high-risk “complication” stage.

Conclusion:

Imaging analyses using retinal nerve fiber layer circumpapillary optical coherence tomography and macular optical coherence tomography scans in conjunction allow rapid assessment of the PVD stage. These techniques can assist clinicians and surgeons in counseling patients and planning surgical approaches. Observations confirmed the progression of PVD through predictable stages and the progression of PVD with age.

Posterior vitreous detachment (PVD) status is important in the pathogenesis and surgical planning of vitreomacular traction (VMT), idiopathic macular hole (MH), peripheral retinal tears, and rhegmatogenous retinal detachment.^1,2 Leopold Weiss described prepapillary ring-shaped vitreous opacities in both eyes of a myopic patient with the direct ophthalmoscope in 1897³ and Alfred Vogt published the first slit-lamp biomicroscopy description of a prepapillary glial ring, indicating PVD in 1930.⁴ A century later, clinicians essentially use the same technique to identify the eponymous Weiss or Vogt ring using slit-lamp and indirect biomicroscopy aided by magnification with a 90D, 78D, or Hruby lens. Because of the transparent nature of the vitreous and concomitant media opacities, this is often challenging and inconsistent. In an attempt to see the separation of the posterior vitreous cortex (PVC) from the internal limiting membrane of the retina or the presence of a prepapillary glial ring, a contact lens examination with a Goldmann lens or a specialized El Bayadi lens can add magnification and better visualization.⁵ B-scan ultrasonography can also be used to identify PVD.⁶

Optical coherence tomography (OCT) has greatly aided the clinical assessment of the vitreoretinal interface.⁷ In 2001, Uchino et al⁸ utilized OCT to determine PVD status and defined five PVD stages based on horizontal and vertical axial time-domain OCT: Stage 0: no discernible vitreoretinal separation; Stage 1: focal parafoveal separation of the PVC in 1 to 3 quadrants; Stage 2: perifoveal PVC separation in four quadrants with persistent vitreous adhesion to the fovea and optic disc; Stage 3: vitreofoveal separation with persistent attachment to the optic nerve and midperipheral retina; Stage 4: complete PVD (with an inability to visualize the vitreous on OCT). In 2018, Tsukahara et al⁹ applied wide-angle OCT imaging and observed that PVD typically initiates in the midperiphery and distinguished lamellar separation (Stage 1a) from the partial separation of the PVC (Stage 1b). In 2019, Kraker et al¹⁰ modified Uchino's Stage 2 PVD as eyes with vitreoretinal adhesion only within a 1,500-µm diameter of the fovea.

The creation of a PVD is typically paramount for vitreoretinal surgical success and can be challenging to induce in some cases. The surgical technique for creating a PVD is to engage the posterior hyaloid (PVC) in the peripapillary preretinal vitreous with vacuum and, by careful anterior and lateral manipulation of the surgical instrument, create vitreopapillary separation. Indeed, Uchino and colleagues noted that a typical PVD culminates with vitreopapillary separation, implying that visualization of the vitreous at the optic nerve head is necessary for the definitive determination of PVD status. The 6-mm² macular optical coherence tomography (M-OCT) volume scan commonly employed in routine retina clinical practice does not incorporate imaging of the optic disc and therefore is insufficient in accurately determining PVD status.

The impetus of this study was based on the observation that the circumpapillary optical coherence tomography (C-OCT) scan using a Heidelberg OCT 2 (Heidelberg, Germany), generally employed for retinal nerve fiber layer (RNFL) analysis to assess glaucoma risk, provides excellent imaging of the posterior vitreoretinal interface. This study investigates the utility of the RNFL C-OCT (a preset 15-second scanning protocol) for viewing the PVC and devising a more clinically and surgically relevant staging of PVD progression to aid in clinical care and vitreoretinal surgical planning.

Methods

Medical records and images were retrospectively reviewed for 2002 eyes of 1,045 patients who received both high-quality RNFL C-OCT and M-OCT scans between July 2019 and March 2021 at Retina Consultants of Texas. Institutional review board approval was obtained. Retinal nerve fiber layer C-OCT and M-OCT images and clinical data were collected and analyzed. Eyes with a prior history of retina surgery or retinal detachment were excluded. Eyes were graded according to the newly developed PVD grading scale based on hyaloid face separation from the peripapillary retina (determined by the RNFL C-OCT scans) and the fovea (from the macular volume scan [M-OCT]).

Macular OCT and RNFL C-OCT scans were obtained. The preset glaucoma RNFL C-OCT pattern is 20° × 20°, one circular B-scan, 768 A-scans with 100 times image averaging, centered on the optic disc. The macular OCT pattern was 20° × 20°, 49 B-scans, 768 A-scans with 9 times image averaging, in a 6-mm² square centered over the fovea. All images were focused on the retina and thus were not specifically optimized to image the vitreous gel or PVC.

Grading of Posterior Vitreous Detachment

Posterior vitreous cortex separation status was independently graded in a masked fashion by three graders including two experienced vitreoretinal surgeons (D.M.B. and C.C.W.). Retinal nerve fiber layer C-OCTs were graded for separation of the PVC from between the peripapillary retina vessels as well as separation from the vessels—noting carefully which vessels had adherent versus separated PVC. Macular OCTs were graded for separation or adhesion from the fovea.

Detailed analyses of where the vitreoretinal separation occurred in relation to peripapillary retinal vessels revealed that, in every case, separation of the PVC occurred over the papillomacular bundle (PMB) between the arcade vessels before separation from the nasal nerve. This PMB separation always occurred preceding or simultaneous with vitreofoveal separation. In no case did the PVC separate from the fovea or the nasal RNFL antecedent to PMB separation. Based on this observation, we graded eyes using a unique PVD classification, which emphasizes the clinically relevant events of vitreopapillary separation and vitreofoveal separation.

Stage A: Attached Stage

Fully attached PVC with no vitreoretinal separation

Stage A is defined by tight adhesions between the PVC and retina with no discernible signs of separation (including lamellar separation). Furthermore, 360° RNFL C-OCT demonstrated no intravessel separation, and M-OCT revealed no evidence of vitreofoveal separation. Scans were differentiated from full PVD by visualization of vitreous anatomy, including the posterior precortical vitreous pocket, Cloquet canal, the bursa of the Martegiani, and the septum interpapillomaculare (Figure 1A).

Fig. 1.

Vitreous stages as graded on RNFL C-OCT and M-OCT. Stage A: Attached stage. A1. RNFL C-OCT scan depicting the right eye of a 6-year-old male patient exhibiting vitreous Stage A, with vitreous firmly attached to the RNFL (arrows). A2. M-OCT scan of the same eye depicted in (A1) exhibiting PVC firmly attached to the macula (arrows), with visible premacular bursa (arrowhead). Insets in A1 and A2 depict the location of the corresponding OCT scans. (Note: all subsequent figures will follow the same scheme, with corresponding insets.) Stage B: Beginning lamellar vitreous separation. stage B, Attached posterior cortical vitreous with early separation of the posterior cortical lamellae (arrowheads) B1. RNFL C-OCT scan depicting the right eye of a 49-year-old female patient exhibiting B2. M-OCT scan of the same eye showing the PVC firmly attached to the macula with early separation of posterior cortical lamellae in the perifoveal area. Stage C−: PMB vitreous separation with persistent adhesion to nasal papillary retinal vessels and persistent vitreofoveal adhesion. C-1. RNFL C-OCT scan depicting the left eye of a 65-year-old male patient exhibiting Stage C−. Note the “stingray” pattern, demonstrating a vitreous attachment nasally with detachment temporally (arrows). C-2. M-OCT scan of the same eye showing separated PVC within the PMB attached PVC at the fovea (arrows). Stage C+: PMB vitreous separation with persistent adhesion to nasal papillary retinal vessels PLUS vitreofoveal release. C+1. RNFL C-OCT scan depicting the left eye of a 65-year-old male showing Stage C+: full separation within the PMB (upward-facing arrows), with persistent vitreous attachment nasally (downward-facing arrows). C+2. Macular OCT scan of the same eye showing complete separation of the PVC from the macula (arrows). (Note: images for C+ depict the same eye as C− scanned later, after the progression of the PVD stage.) Stage D: “Done” (complete PVD). D1. RNFL C-OCT scan depicting the OD of an 82-year-old female patient OD showing Stage D: dark signal anterior to the disc and RNFL retina. D2. Macular OCT scan of the same eye showing the same dark lack of signal anterior to the macula, confirming Stage D: full PVD.

Stage B: Beginning Stage

Distinct linear separation between lamellae within the PVC between the peripapillary retina vessels without true separation over 25 μm of the posterior hyaloid from the internal limiting membrane of the PMB or fovea

Stage B is defined on the RNFL C-OCT by areas of discernible separation or schisis between lamellae within the PVC without any definite separation of the PVC greater than 25 μm from the internal limiting membrane between tethered peripapillary retina vessels including the major arcade vessels (PMB) on either RNFL C-OCT or M-OCT (Figure 1B). Note that the PMB is defined as the temporal circumpapillary retina encompassed by the temporal vascular arcades as they emanate from the optic nerve head (see Figure 2 for definition and depiction of the PMB). Macular OCT reveals no evidence of vitreofoveal separation.

Fig. 2.

Left: Near-infrared scanning laser ophthalmoscopy images depicting the location of the RNFL C-OCT scans shown on the right, with the PMB highlighted in yellow and indicated with yellow arrows. Right: RNFL C-OCT scan of a 65-year-old female patient with the region within the PMB indicated by yellow brackets with arrows, and separated PVC within the PMB indicated with yellow arrowheads.

Stage C: Complication Stage

Stage C−: Papillomacular bundle vitreous separation with persistent adhesion to nasal papillary retinal vessels with persistent vitreofoveal adhesion

Stage C− is defined by separation of the PVC greater than 25 μm over the PMB as observed on RNFL C-OCT, with persistent vitreofoveal adhesion seen on M-OCT. As the anterior–posterior forces of vitreous contraction increase, VMT and MH can be induced before vitreofoveal separation.

Stage C+: Papillomacular bundle vitreous separation with persistent adhesion to nasal papillary retinal vessels PLUS vitreofoveal release

Stage C+ is defined by full vitreofoveal separation as observed on M-OCT, with persistent nasal vitreopapillary adhesion as observed on RNFL C-OCT. Stage C+ may present on M-OCT as either a detached, visible PVC anterior to the macula or a dark lack of visible PVC anterior to the macula. Concurrently, RNFL C-OCT may show either vitreous attachment to the nasal circumpapillary retina or visible hyperreflective vitreous signal anterior to the circumpapillary retina (Figure 1, C− and C+). (Note: Nasal papillary adhesion was confirmed with volume scanning of the optic nerve in all eyes where the PVC was visualized above but separated from the retina for 360° on the RNFL scan.) (Note that of 2000 eyes graded, no scans showed vitreofoveal attachment with total vitreopapillary separation, supporting the progression of PVD through consistent, distinct stages.)

Stage D: “Done”—Complete Posterior Vitreous Detachment

Stage D is defined by complete vitreoretinal separation. At Stage D, both RNFL C-OCT and M-OCT show the dark absence of any signal anterior to the retina or with pinpoint translucence consistent with vitreous cells or hemorrhage, and no visualization of vitreous anatomy, indicating a complete PVD (Figure 1D).

Stage X: Indeterminate Posterior Vitreous Cortex Status

Stage X is defined by an inability to conclusively determine PVC status.

Statistical Analysis

Statistical analysis of the incidence of PVD stage by age was performed using chi-square tests, with P <0.05 considered statistically significant.

Disorders Involving Vitreous Traction

Retinal nerve fiber layer C-OCT and M-OCT scans were additionally assessed for the presence of vitreoretinal traction including VMT and MH and verified and adjudicated with clinical chart data.

Results

Prevalence of Posterior Vitreous Detachment Stage

A total of 25 eyes (1.25%) were graded as having no vitreoretinal separation (Stage A); 725 eyes (36.21%) early separation between lamellae within the PVC (Stage B); 248 eyes (12.39%) significant vitreous separation of the PVC from the PMB but no separation from the fovea (Stage C−); 151 eyes (7.54%) vitreofoveal separation with persistent adherence at the nasal optic disc (Stage C+); 851 eyes (42.51%) full PVD (Stage D); and two eyes (0.10%) was graded inconclusive (Stage X).

Vitreoretinal Stage by Age

Graded eyes were 56% of women and 44% of men. The mean age was 58.8 years (range 6–101 years old). Posterior vitreous detachment status by age is shown graphically in Figure 3 (details in Table, Supplemental Digital Content 1, http://links.lww.com/IAE/C244). The initial signs of vitreous lamellar separation that preceded hyaloid separation between the interpapillary retina vessels (Stage B) were seen in 36/53 (67.9%) eyes of patients less than 18 years old and 80/84 (95.2%) eyes of patients 19 to 29 years old. Total PVD (Stage D) was present in only 1/2 of the eyes of patients 60 to 69 years old, and 1/3 of the eyes of patients 70 to 79 years old did not have a complete PVD. Sixteen percent of octogenarian eyes and 7% of eyes of patients over 90 years old did not have a complete PVD. Interestingly in the 70 to 79-year age group, 110/399 (27.6%) eyes were Stage C− or C+ with vitreopapillary separation of the PMB but not the nasal RNFL. P < 0.001 for PVD progression by age, indicating a clear correlation between PVD progression and age.

Fig. 3.

PVD status of 2000 eyes using M-OCT and RNFL C-OCT (two “indeterminate” eyes not included in graph).

Concomitant Retinal Conditions

Retinal disorders of patients included age-related macular degeneration (391 eyes, 19.5%) diabetic retinopathy (289 eyes, 14.4%), hypertensive retinopathy (296 eyes, 14.7%), epiretinal membrane (335 eyes, 16.7%), other retinal diagnosis (732 eyes, 36.6%), and no diagnosed retinal pathology (287 eyes, 14.3%).

Vitreomacular Traction and Macular Hole

Figure 4 demonstrates the correlation between VMT and MH to the PVD Stage.

Fig. 4.

PVD status of 2000 eyes using M-OCT and RNFL C-OCT, with VMT/ Macular Hole status eyes highlighted in red.

Twenty-six eyes were diagnosed with MH, and 49 eyes exhibited vitreoretinal traction (VMT or other vitreous traction) observed on M-OCT or RNFL C-OCT, for a combined 75 eyes exhibiting disorders of vitreoretinal traction. None of these eyes were graded as Stage A. Four Stage B eyes (5.3%) exhibited VMT (Figure 5). Forty-eight eyes with VMT or MH with VMT were Stage C− eyes (64.0%) and 16 eyes with MH or VMT were C+ (21.3%). The remaining seven eyes (9.3%) exhibited MH with full PVD (Stage D). Cumulatively, 85% of patients with MH or significant VMT were Stage C− or C+.

Fig. 5.

Left: RNFL C-OCT scan depicting the right eye of a 66-year-old male patient exhibiting Stage B vitreous separation including partial lamellar separation. Right: M-OCT scan VMT exhibited on M-OCT (asterisk), an uncommon finding among Stage B eyes.

Discussion

At birth, the vitreous body is a solid matrix of collagen fibrils insufflated by hyaluronic acid. With aging, the hyaluronic acid dissociates from the collagen fibrils resulting in the aggregation of collagen fibers into bundles and pockets of liquefied vitreous.^11,12 Eventually, the vitreous body collapses and the PVC separates from the retina. The process of synchysis senilis (essentially the separation of the collage fibrils from hyaluronic acid) increases linearly with age and is directly correlated with PVD.¹³ The PVC is most firmly attached at the vitreous base, the disc, the macula, and overlying the retinal arcade vasculature.¹⁴ The strong adhesion at the vitreous base and the papillary margin is mediated by chondroitin sulfate glycosaminoglycan.¹⁵ Foos published an autopsy study in 1973¹⁶ demonstrating that the visible Vogt or Weiss ring had components of avulsed peripapillary glial tissue in every case demonstrating the tenacity of this adhesion. It follows that separation from the papillary margin would be the final step in PVD.

The finding that the PVC is most adherent to the nasal peripapillary RNFL C-OCT is consistent with the highest density of retina vessels nasally, which tether the PVC from the anteroposterior traction forces caused by the age-related collapse of the vitreous body. This study demonstrates that the larger arc of the vitreoretinal interface between the major arcades separates before the nasal peripapillary retina. The finding that residual vitreous attachments in partial PVDs are always on the nasal side of the optic nerve was described by Archimedes Busacca in a large series of autopsy eyes in 1964¹⁷ and confirmed by Foos.¹⁶ It is the critical finding that the nasal peripapillary retina is the most adherent and that the PMB separates before foveal or nasal separation that is the main difference between Uchino staging and the current staging. Uchino, using only horizontal and vertical time-domain OCT available in 2001, counted quadrants of perifoveal separation to delineate between stages. We agree that the separation from the fovea and optic nerve is the most clinically relevant but feel that the PMB vitreoretinal separation—which precedes vitreofoveal or nasal vitreopapillary separation—is a critical stage that should be identified.

The overall incidence of total PVD in this series was 42.5%. Posterior vitreous detachment incidence in 60 to 70 years old was 51%, which correlates with the classic slit-lamp identified PVD studies of Pischel¹⁸ (53% total PVD over age 50) and Lindner¹⁹ (65% over the age of 65 years old). In Foos autopsy study¹³ in a presumably normal population, the overall rate of total PVD was 23.2%. While our population had an incidence of 50% total PVD in the age range 60 to 69, Foos reported 27% PVD in this age range, likely related to selection bias of the tertiary referral retina center from which the current patients were derived. Foos incidence of 67% PVD in the eighth decade is the same as our series incidence of 66%. Future work will include determining the incidence of PVD by the decade in a nonretina referral population by looking at a database of RNFL C-OCTs from glaucoma and primary eye care referral practices. We noted that the beginnings of PVD (namely, separation between lamellae of the PVC) are evident at an early age (68% of eyes less than 18 years old were Stage B). In addition, 34% of eyes aged 70 to 79, 16% of eyes aged 80% to 89%, and 7% of eyes aged 90 and above had no total PVD demonstrating the age range of PVD is much broader than previously realized.

Only two eyes of 2002 (0.1%) were deemed to have an indeterminate PVD status by RNFL C-OCT and M-OCT. Overall, 189 eyes (9.45%) had an incorrect assessment in the clinical record (with a positive PVD or no PVD incorrectly). Augmenting clinical biomicroscopy with RNFL C-OCT should allow clinicians to diagnose PVD more accurately than through biomicroscopy alone. Obtaining RNFL C-OCT in addition to standard M-OCT in pharmacologic and surgical clinical trials should easily provide data to determine the influence of PVD status on drug clearance rates and the efficacy of a studied intervention or therapy.

Posterior vitreous cortex separation in the PMB without vitreofoveal release (Stage C−) represents a higher risk stage. In this configuration, anteroposterior vitreous traction concentrates at the fovea, meaning the risk of VMT and idiopathic MH is elevated until the point of vitreofoveal release.²⁰ Analysis of eyes exhibiting tractional disorders within this study confirmed that eyes graded C−/C+ contained the highest incidence of VMT and MH (64 of 75 eyes, 85.3%). A classic teaching is that epiretinal membranes form following total PVD. As described by other authors, we noted numerous cases of clinically significant ERMs in Stage C− and C+ eyes.²¹ Separation of the PH from the underlying retina allows hyalocytes to proliferate. This can result in epiretinal membrane formation regardless of whether the nasal peripapillary retina (and other regions) have separated or not.

A subgroup of Stage C− and C+ eyes exhibited a particularly striking pattern of peripapillary vitreous attachment, especially common among eyes with VMT and MH. This “stingray sign” (vitreoretinal attachment nasally and elevated vitreoretinal separation temporally), visually resembles a stingray flapping its pectoral fins (Figure 6). The “stingray sign” was seen in 279/399 eyes C− or C+ (69.9%) and 60/64 Stage C/C+ eyes with VMT and/or MH (93.8%) (P = 0.0125). Eyes exhibiting the “stingray sign” should be followed closely for the development of a tractional disorder.

Fig. 6.

Left: Circumpapillary scan depicting the left eye of a 65-year-old female patient at Stage C− vitreoretinal separation within the PMB, with persistent macular adhesion. On the RNFL C-OCT, note the “stingray sign,” showing vitreoretinal attachment nasally and detachment temporally. Arrowheads indicate the PVC; double arrows indicate lamellar vitreous attachment. Right: M-OCT scan of the same eye showing separated PVC within the PMB (arrowheads) attached PVC with VMT at the fovea (asterisk), and separation of the vitreous lamellae temporal to the macula (double arrow).

Combining the M-OCT with RNFL C-OCT available on commonly available OCT devices allows rapid determination of PVD status. Patients with an MH or VMT in one eye can be reassured if a total PVD is determined in the fellow eye (or counseled that they are at risk in the fellow eye if Stage B or C−). Among cases of partial PVD (almost all MH and VMT cases), a surgeon can easily determine the anatomic region with the most separation between the PVC and the neurosensory retina to plan PVD initiation using the safest and most efficient surgical approach. While it may seem safer to attempt to create a PVD nasally, the lack of vitreoretinal separation in the peripapillary nasal retina often makes this approach the most likely to result in instrumentation damage to the retina. If the RNFL C-OCT shows that the greatest PVC separation is along the inferior or superior arcade, this presurgical “treasure map” may facilitate a safer, more efficient surgical plan. The authors have found that often the preop RNFL C-OCT shows the highest separation in the PMB and that PVD is often surgically created very efficiently by engaging the hyaloid face in the PMB, where it is often quite elevated from the neurosensory retina.

While the use of RNFL C-OCT and M-OCT to determine PVD status is rapid, accurate, and efficient, other imaging approaches such as the utilization of widefield OCT (OPTOS) and broad swept-source OCT patterns can likely be used to achieve the same conclusions. Further advances in OCT scanning will likely allow greater depth of field, allowing the creation of three-dimensional reconstructions of the posterior cortical vitreous relationship with the optic nerve, fovea, and peripheral retina, providing even more information for clinicians and surgeons to guide discussions with patients and surgical planning.