Posteriorly inserted vitreous base: pre-operative characteristics, intra-operative findings, and outcomes after vitrectomy

Abstract

Purpose:

To determine the pre-operative characteristics, intra-operative and post-operative complications, and outcomes of eyes with posteriorly inserted vitreous base (PIVB).

Methods:

In this retrospective, observational, consecutive case series at two academic centers, 37 patients were studied who had PIVB noted during vitrectomy. PIVB was defined as the insertion of the posterior hyaloid membrane posterior to the vortex veins. Fifteen eyes were analyzed in a histopathologic study of donor eyes to determine the average distance of the ora serrata from the vortex veins as this distance is uncertain.

Results:

PIVB was identified during vitrectomy in 31 eyes with rhegmatogenous retinal detachment (RRD, 84%), 4 with macular hole (11%), 1 with vitreous hemorrhage, and 1 with epiretinal membrane. Adjunctive buckle was used in 24%; 54% had 360-degree laser. Average number of tears seen pre-operatively in those with RRD was 3.1. 30% had new breaks identified intra-operatively. 41% had lattice degeneration; new breaks were found in 40% of eyes with lattice. 13% of RRDs developed proliferative vitreoretinopathy. Average distance from the ora serrata to the vortex veins was 7.6mm.

Conclusions:

Any eye undergoing vitrectomy may have PIVB, but those with a high number of retinal breaks and lattice near the equator may be at highest risk. Re-detachment and proliferative vitreoretinopathy still occur despite knowledge of the disorder and adjuvant treatments.

SUMMARY STATEMENT:

The vitreous base is critical in the pathogenesis and repair of retinal detachment. Eyes with posteriorly inserted vitreous base have a high number of retinal breaks and are at risk of additional breaks intra-operatively—recognition of this as an entity is important for all eyes undergoing vitrectomy.

INTRODUCTION

The advent of pars plana vitrectomy over the last 45 years has revolutionized the ophthalmic surgeon’s ability to treat retinal diseases. A combination of improved endoillumination, the ability to more thoroughly remove vitreous and manipulate the retina, and wide-angle surgical viewing systems has enabled observation of the peripheral vitreoretinal interface that was previously limited to histologic assessment. An example of a condition seen during vitrectomy is the posteriorly inserted vitreous base. This condition was first observed in histologic examination of human donors eyes^1–4 and is a concept often taught in retina fellowships due to its increased case complexity and anecdotal reports of a higher rate of failure after retinal detachment repair.

Posterior vitreous detachment (PVD) status and insertion of the vitreous base is relevant for every vitrectomy, and critical to consider in eyes with rhegmatogenous retinal detachment (RRD) as some seem to be at higher risk of experiencing additional breaks intra-operatively. An unusual vitreoretinal interface alerts the clinician to look for breaks more posteriorly than the typical location. A surgeon may try to pull the posterior hyaloid anteriorly in an attempt to bring it to the normal peripheral location but due to irregular and/or posterior attachment, may induce retinal breaks.

The vitreous base encompasses a 2–6 (average 3.2) mm wide circumferential area that straddles the ora serrata and serves as the strongest attachment of the vitreous to the retina (Figure 1).^5,6 As the vitreous collagen fibrils attach firmly in radial orientation to the internal limiting membrane at the vitreous base, retinal breaks typically develop at the posterior edge of the vitreous base.⁷ A posteriorly inserted vitreous base is generally defined as a wider than average vitreous base that straddles the ora serrata, and the posterior edge is more than 2–3 mm posterior to the ora serrata. Thus, retinal breaks typically occur more posteriorly than the usual location in eyes with posteriorly inserted vitreous base. Regardless of the exact location and dimensions, the vitreous base is an area where the posterior hyaloid membrane cannot be elevated during surgery. Posteriorly inserted vitreous base has been described histologically in human donor eyes^1,3,4 but is poorly recognized in the clinical literature. We are not aware of any case series or reviews of eyes with posteriorly inserted vitreous base before death or enucleation. Presently, it is unclear how a posteriorly inserted vitreous base may affect surgical decisions and clinical outcomes.

Figure 1.

Eye diagrams showing normal and posteriorly inserted vitreous base anatomy. Left depicts normal vitreous base insertion 2–3 mm posterior to the ora serrata. Right depicts a vitreous base that is inserted posteriorly, over 6 mm posterior to the ora serrata.

The goals of this study were to describe the posteriorly inserted vitreous base as a recognizable entity, enumerate the intra-operative findings that can result from it, and relate clinical outcomes of eyes with posteriorly inserted vitreous base undergoing vitrectomy. In this study, we defined the clinical entity of a posteriorly inserted vitreous base as one that inserts at or posterior to the equator, and we use the vortex veins as a surrogate for the equator. We also assessed the distance of the ora serrata to the vortex veins in a series of donor eyes. We believe this condition is important to recognize, and this manuscript provides a basis for which future studies can further elucidate its incidence and optimize treatments.

METHODS

This retrospective study consisted of consecutive patients who had a posteriorly inserted vitreous base noted during pars plana vitrectomy at the University of Iowa Hospitals and Clinics from 2010 to 2014 (EHS) and Massachusetts Eye and Ear Infirmary (DE) from 2011 to 2014. It was standard of care for us to dictate the presence of posteriorly inserted vitreous base in the operative report when it was identified. We defined the vitreous base as posteriorly inserted if the posterior hyaloid membrane could not be elevated anterior to the vortex veins, which approximates the equator of the eye and is estimated to lie between 6 to 8mm posterior to the ora serrata (Figure 1). Cases described here either had a pre-existing complete posterior vitreous detachment (PVD) or underwent intraoperative creation of a PVD using aspiration by the vitrectomy handpiece. In all cases, the PVD was created with careful attention to the intraoperative occurrence of new retinal breaks at the usual posterior edge of the vitreous base.^2,3,8 A wide-angle BIOM viewing system (Oculus Surgical, Port St. Lucie, Florida) was utilized for all surgeries. As the surgeons were intimately aware of the potential challenges that posteriorly inserted vitreous base could present, all cases had close shaving of the vitreous anterior to the posterior insertion of the vitreous base with scleral depression. No additional breaks were encountered during vitreous shaving.

As a number of conditions can result in pre-existing altered vitreo-retinal interface, we excluded patients with a history of diabetes, ocular trauma to the affected eye, retinopathy of prematurity, prior treatment for any peripheral vitreoretinal pathology in the affected eye (i.e. laser or cryo-retinopexy, prior vitrectomy or scleral buckle), and less than two and a half months post-operative follow-up. Institutional Review Board approval was obtained from the University of Iowa and Massachusetts Eye and Ear Infirmary for this project, which adhered to the principles of the Declarations of Helsinki.

After identifying cases through a review of operative reports, data were extracted from review of patient charts from the pre-operative visit and at consecutive follow-up visits. Data collected at baseline, pre-op and follow-up visits included past medical history, smoking history, past ocular family history, systemic medications, best corrected Snellen visual acuity (VA), refractive error with particular attention to high myopia (i.e. spherical equivalent ≥ −6 diopters), presence of PVD pre-operatively, presence of lattice degeneration, location of pre-existing breaks, indication for surgical intervention, gauge of surgical instrumentation, new retinal tears noted intra-operatively, intraoperative treatment, gas or oil used, number of operations required, and postoperative complications. Adjunctive scleral buckle consisted of placement of a 42 band with 6 mm horizontal mattress spacing in each quadrant with the anterior suture placed 1–3 mm posterior to the spiral of Tillaux. Also noted, if applicable, were the dates and details of any subsequent vitreoretinal procedures (intravitreal injections or surgeries). Data of the fellow eye was also collected at baseline, pre-op and follow-up visits.

The VAs recorded in our cohort were pre-op, post-op month 3, and at final follow-up. To analyze this information, visual acuity data were converted to logMAR (logarithm of the minimal angle of resolution) values (logMAR = log [1/Snellen visual acuity] to allow for statistical analysis; count fingers was assigned a LogMar of 2.3 and hand motion was assigned a LogMAR of 3.3.⁹ No patients had a pre-op or post-op VA of light perception or no light perception.

Statistical analysis

Statistical analysis of extracted data included paired two tailed t tests comparing logMAR values from the affected pre-operative VA and VA at last follow-up. Extracted data was categorized based upon pre-operative indication (retinal detachment, macular hole, epiretinal membrane, or vitreous hemorrhage). Retinal detachment was further sub-categorized between macula-on and macula-off retinal detachments.

Vortex vein donor eye study

Paraformaldehyde-fixed eyes from fifteen donors, ages 72–92, were selected from a collection of donor eyes at the University of Iowa Institute for Vision Research. Eyes were initially fixed in 4% paraformaldehyde in phosphate buffered saline (PBS) for 2 hours, followed by washing and intermediate storage in PBS. For long term storage, eyes were maintained in 1% paraformaldehyde in PBS. Vortex veins were identified by the displacement of choroidal pigment in a spiral pattern, or by the presence of red pigment (from pooled erythrocytes that collected post-mortem) in the same pattern. The linear distance between the center of each vortex vein and the ora serrata was determined using a Vernier caliper (Electron Microscopy Sciences, Hatfield, PA). When possible, multiple distances were collected in different quadrants of each eye.

RESULTS

The mean age of patients with posteriorly inserted vitreous base (n=37) was 60 years with median age 61 years (range = 38–90). More cases (n=16; 43%) were in the 60–69 year old age group than any other group. Twenty-three were male (62%) and 14 were female (38%). Two were current smokers and 7 were former smokers within the last 20 years. Pre-operatively, 23 (62%) eyes had a Weiss ring, 18 (47%) were pseudophakic, and 7 had high myopia (not all pseudophakic eyes had this information available). The primary indications for vitrectomy were rhegmatogenous retinal detachment (RRD; n=31; 84%), macular hole (MH; n=4; 11%), vitreous hemorrhage (VH; n=1), and isolated epiretinal membrane (ERM; n=1). Of the 31 patients with RRD, 1 also had an ERM, 4 (13%) had proliferative vitreoretinopathy (PVR), and 2 had VH. One patient with isolated VH had concern for retinal tear on ultrasound pre-operatively that was confirmed intra-operatively. Extensive demographic, pre-operative and intra-operative details for all patients are in Table 1.

Table 1.

Demographic and clinical information on all eyes included in living patient study

ID	Age	Gender	Indication	Lens status	Lattice	# of pre-existing retinal tears	quadrant(s) of pre-existing retinal tears	Weiss ring seen pre-op?	# of new tears from complete PVD induction	Quadrant of new tears from PVD induction	Adjunct SB, 360 EL, or both	Reason for re-operation (if needed)
1	68	M	Mac on RRD	Pseudophakic	No	8	All quadrants	yes	2	Inferonasal, Inferotemporal	Both
2	90	F	Mac off RRD	Pseudophakic	No	1	Superotemporal	no	13	All quadrants	Both	Anterior PVR, hypotony and oil removal
3	51	M	Mac off RRD	Pseudophakic	No	1	Inferonasal	no	0		SB (+focal laser only)
4	61	F	Macular Hole	Pseudophakic	No	0	None	no	0		360 EL
5	67	F	Macular Hole	Pseudophakic	Yes	2	Inferotemporal, Inferonasal	yes	-		Focal laser
6	63	F	Mac off RRD with PVR B	Phakic	No	3	Superotemporal, Superonasal, Inferotemporal	yes	-		360 EL
7	65	M	Mac on RRD	Pseudophakic	No	4	All quadrants	yes	-		360 EL
8	67	M	Mac on RRD	Pseudophakic	Yes	3	Superotemporal, Inferotemporal	yes	-		360 EL
9	59	M	Mac on RRD	Pseudophakic	Yes	12	All quadrants	yes	4	All quadrants	360 EL
10	67	M	Mac off RRD with PVR C and macular hole	Pseudophakic	No	3	All quadrants	no	4	All quadrants	Both	Epiretinal membrane and oil removal
11	57	F	Mac on RRD	Pseudophakic	No	1	Superotemporal	yes	2	Inferotemporal, Superonasal	360 EL
12	70	M	Mac off RRD	Phakic	No	2	Superotemporal	yes	-		Focal laser	Re-detachment (2 weeks after initial surgery)
13	42	F	Mac off RRD	Phakic	Yes	6	Superotemporal	no	3	Inferotemporal	360 EL
14	71	M	Mac on RRD	Pseudophakic	No	5	Superonasal, Inferonasal	yes	-		360 EL
15	64	M	Epiretinal membrane	Pseudophakic	No	1	Inferotemporal	yes	1	Inferotemporal	360 EL
16	50	F	Mac on RRD	Pseudophakic	Yes	1	Superonasal	yes	-		360 EL
17	61	F	Mac on RRD	Pseudophakic	Yes	3	Superotemporal, Superonasal, Inferonasal	yes	1	Inferonasal	Focal laser
18	53	F	Mac on RRD	Phakic	Yes	2	Inferotemporal, Inferonasal	no	0		Both
19	38	M	Mac on RRD	Phakic	No	2	Superonasal	no	0		Focal laser
20	64	M	Mac on RRD	Phakic	No	3	Superotemporal, Superonasal	no	0		Focal laser	Severe epiretinal membrane
21	64	M	Macular Hole	Phakic	Yes	0	None	yes	-		Focal laser
22	57	M	Mac off RRD	Phakic	Yes	3	Superotemporal	yes	3	Superotemporal	360 EL
23	65	M	Mac on RRD	Phakic	No	1	Superonasal	yes	-		360 EL
24	65	M	Mac on RRD	Pseudophakic	Yes	4	Superonasal, Inferonasal	yes	-		SB (+focal laser only)	Re-detachment (secondary to PVR)
25	54	F	Mac off RRD	Phakic	Yes	1	Superonasal	no	2	Superotemporal	SB (+focal laser only)
26	55	F	Mac on RRD	Pseudophakic	Yes	1	Inferotemporal, Inferonasal	no	0		Focal laser
27	48	M	Mac off RRD	Phakic	No	2	Superotemporal, Superonasal, Inferotemporal	no	0		Focal laser
28	57	F	Mac on RRD	Phakic	No	5	Superotemporal, Inferotemporal, Inferonasal	no	0		360 EL
29	44	M	Mac off RRD	Phakic	Yes	1	Inferonasal	yes	4	All quadrants	Both	Epiretinal membrane
30	56	M	Mac off RRD	Phakic	Yes	4	Superotemporal, Superonasal, Inferotemporal	yes	-		360 EL
31	74	F	Mac off RRD	Pseudophakic	No	1	Superotemporal	no	0		360 EL
32	55	M	Mac off RRD	Phakic	Yes	2	Superotemporal, Inferonasal	no	0		360 EL	Re-detachment (secondary to PVR)
33	56	M	Mac on RRD	Phakic	No	6	Superotemporal, Superonasal	yes	-		360 EL
34	49	F	Mac off RRD	Phakic	No	2	Superotemporal, Superonasal	yes	-		Focal laser
35	66	M	Mac off RRD	Pseudophakic	No	4	All quadrants	yes	-		360 EL	Re-detachment (secondary to PVR)
36	63	F	Vitreous Hemorrhage	Phakic	No	3	Superotemporal	yes	-		Focal laser
37	66	M	Macular Hole	Phakic	No	0	None	yes	-		Focal laser

M = Male; F = Female; Mac = Macula; RRD = rhegmatogenous retinal detachment; ERM = epiretinal membrane; EL = endolaser; SB = scleral buckle; PVD = posterior vitreous detachment

*Note that for tears occuring at junction between quadrants, both quadrants were included (ex: 9 o’clock tear right eye = superotemporal and inferotemporal quadrants)

Endolaser was used in all 37 patients, with 360-degree laser demarcation performed in 20 eyes (54%). Adjunctive scleral buckle to support the vitreous base was utilized in 9 patients (24%). Tamponade employed was C₃F₈ gas (n=29; 78%), SF₆ gas (n=5; 14%), or silicone oil (n=3; 8%). The average number of tears seen on pre-op examination in eyes with RRD was 3.1. Eleven eyes (30%) had one or more new retinal breaks encountered after PVD induction (i.e. not present pre-operatively); these localized to the inferotemporal quadrant in 8 eyes (22%), inferonasal quadrant in 6 (16%), superonasal quadrant in 4 (11%), and superotemporal quadrant in 6 (16%). Six of the 11 eyes (55%) that had new breaks found intra-operatively had breaks in more than one quadrant. Four of the 11 eyes had new breaks encountered in all quadrants. For eyes with RRD that were found to have new breaks identified intraoperatively (10 of 31 RRD cases), the average number of new breaks was 3.8 (i.e. higher than the 3.1 seen in all eyes pre-op with RRD). Interestingly we found that 7 eyes with a Weiss ring seen on pre-op exam developed more tears intra-op despite utmost care.

Lattice degeneration was identified in 15 eyes (41%) and new retinal breaks were found intra-operatively in 6 of these eyes (40%) during PVD induction (four of which had a Weiss ring seen on pre-operative examination). Of the eyes without lattice (n=22; 60%), new retinal breaks occurred in four (18%) during PVD induction (2 of which had a Weiss ring seen on pre-operative examination). There was a trend (p=0.07) toward higher rate of new retinal breaks occurring in eyes with lattice compared to those without lattice.

Visual acuity results classified by pre-operative indication are demonstrated in Table 2. In summary, eyes with macula-off RRD had significant improvement in VA at last observation (p=0.001) compared to pre-op. Eyes with macula-on RRD had no change in last observation VA (p=0.28). While VA was close to showing a trend to improvement after surgery for the MH group (pre-op VA 20/159, post-op VA 20/35), the small n was the likely reason significant improvement was not found.

Table 2.

Surgical indications and visual acuity results for eyes with posteriorly inserted vitreous base

Surgical indication	# of eyes	Median pre-op VA	Median post-op VA at last follow up	Median last follow up (mo)	Mean last follow up (mo)	P-value for last follow up VA
Mac on RRD	16	20/25	20/25	6	5.5	0.28
Mac off RRD	15	20/500	20/50	6	5.8	0.001
Macular hole	4	20/159	20/35	6	7.3	0.183
Vitreous hemorrhage	1	CF	20/25	3	3	N/A
Epiretinal membrane	1	20/40	20/20	16	16	N/A

VA=visual acuity; CF=counts fingers; RRD=rhegmatogenous retinal detachment; mo=month

Eight eyes, all with history of RRD, needed re-operation for complications for PVR-related re-detachment (4), ERM (3), or retinal re-detachment without PVR (1). Detailed data on eyes that needed re-operation for these complications are shown in Table 3. Several eyes had risk factors for post-op PVR seen pre-operatively, including two with vitreous hemorrhage and four with primary PVR. Of the four eyes that had recurrent RRD from PVR, one started macula-on but ended up with VA of 20/200 (pre-op 20/25); another that started macula-off ended with VA of 20/80 (pre-op counting fingers); one had a pre-op VA of 20/25 but ended up 20/200; and the fourth had hypotony and anterior PVR with oil needing to be removed but was hospitalized due to unrelated illness and failed to follow-up after that. All patients with PVR-related RRD had adjunctive scleral buckle placed or 360-degree endolaser at the time of primary PPV. The incidence of PVR-related recurrent RRD in this cohort was 13% (4 of 31 eyes).

Table 3.

Details on eyes with posteriorly inserted vitreous base needing re-operation for complication(s)

Study ID	Reason for re-operation	# vitrectomies needed for retinal reattachment	Time from 1st PPV to re-op (days)	Pre-op VA	Final VA	Follow-up time from first to final PPV (days)	Membrane peel at initial op?	Details for subsequent surgeries in affected eye
2	Anterior PVR, hypotony, oil	lost to follow-up		20/250	CF		No	hospitalized due to illness then lost to follow-up
10	Oil and severe ERM	lost to follow-up		HM	20/600		Yes	lost to follow-up
12	Recurrent mac-off RRD	1	21	CF	20/30	21	No	PPV + SB + CE/IOL + endolaser + C3F8
20	Severe ERM (to counts fingers VA)	1	177	20/30	20/20	177	No	PPV + membrane peel + CE/IOL
24	Tractional mac-off PVR-associated re-detachment	1	48	20/25	20/200	48	Yes	PPV + membrane peel + SB + C3F8 + endolaser
29	Visually significant ERM	1	110	20/50	20/20	110	Yes	PPV + membrane peel + SB + C3F8 + endolaser
32	Tractional mac-off PVR-associated re-detachment	2	134	CF	20/80	477	Yes	PPV + membrane peel + SB + C3F8 + endolaser (134 days from primary procedure); PPV + membrane peel (477 days from primary procedure)
35	Macular starfold and PVR-associated re-detachment	1	39	20/25	20/40	39	Yes	PPV + membrane peel + C3F8 + endolaser

Pre-op = before first vitrectomy; RRD = rhegmatogenous retinal detachment; VA = visual acuity; PVR = proliferative vitreoretinopathy; CF = counting fingers visual acuity; HM = hand motions; PPV = pars plana vitrectomy; PPV = pars plana vitrectomy; SB = scleral buckle; ERM = epiretinal membrane; CE/IOL = cataract extraction and intraocular lens; C3F8 = perfluoro-octane gas; SF6 = sulfur hexafluoride gas

Vortex vein donor eye study

The average distance for all measurements was 7.6 mm (standard deviation 1.32 mm). The distance between the vortex veins and ora serrata in the superotemporal quadrant was significantly higher than in the superonasal quadrant (p<0.01). The inferotemporal distance showed a trend toward being higher than the inferonasal quadrant, but this difference did not reach statistical significance (p=0.09). Details for the donor eyes and measurements are in Supplemental Table 1 (available online).

DISCUSSION

While this is the first detailed clinical description and study of a series of eyes with posteriorly inserted vitreous base, this is not a new entity. Posteriorly inserted vitreous base has been described in donor eyes since at least 1954¹. The concept of posteriorly inserted vitreous base has been passed on to generations of ophthalmologists (especially pathologists and retina specialists) for decades, but it is unclear why this condition has not been clinically described in peer-reviewed literature earlier. The best way to observe this condition is when the vitreous is physically manipulated, which is why the first reports were in donor eyes during gross dissections. While one could surmise in the pre-vitrectomy era based on the number and location of breaks in a retinal detachment that there may be an anomalous vitreous base insertion,¹⁰ it would be difficult to know for certain how far posteriorly the vitreous base inserted without handling it. Pars plana vitrectomy, with its endoillumination and microscopic detail, allows keen observation and manipulation of the posterior hyaloid. While the advent of vitrectomy has afforded vitreoretinal surgeons unparalleled ability to repair some detachments that could not be fixed in the age of scleral buckles, encountering an eye with a posteriorly inserted vitreous base can be challenging, as many experienced surgeons have suspected that these eyes might have a relatively lower rate of single surgery success.

An eye with a posteriorly inserted vitreous base is important to recognize, as there could be a higher risk for re-detachment. While some eyes in our series had pre-existing ocular risk factors for the development of recurrent RD due to PVR such as pre-existing PVR and VH, and many of these eyes underwent adjunctive scleral buckle to support the base and/or 360-degree endolaser just posterior to the insertion of the vitreous base, the occurrence of re-detachment in some eyes (16% of those presenting with RRD) indicates that these eyes may be at higher risk for recurrent detachment. Though this is the largest series of patients with posteriorly inserted vitreous base to date, it was not powered to determine the best way to reduce re-detachment. Our clinical experience suggested that a scleral buckle supporting the vitreous base with at least a 42 band, giving 4 mm wide indentation, was worth considering in eyes with posteriorly inserted vitreous base. However, our histopathologic study shows that the vortex veins, which we used to approximate the equator intra-operatively, is on average 7.6 mm posterior to the ora serrata (8.5 mm in the superotemporal quadrant), thus an adjunctive scleral buckle that is wider than 4 mm (and/or placed more posteriorly to support the posterior edge of the vitreous base) may be useful to prevent re-detachment. The buckles placed in our patients may not have fully supported the posterior edge of the vitreous base and therefore may have had limited effectiveness. Research using wider and/or more posteriorly placed scleral buckles (keeping in mind that complications can develop with too posterior placement of scleral buckle over vortex veins) is necessary to determine whether this can lower the incidence of re-detachments for eyes with posteriorly inserted vitreous base.

While we felt that proper endoillumination of the vitreous and close shaving with scleral depression obviated the need for use of adjunctive triamcinolone to highlight the vitreous, ‘chromo-vitrectomy’ could be studied in the future to reduce post-operative complications.^11,12 In addition, use of scleral buckle without vitrectomy could also be considered in eyes with a higher chance of having a posteriorly inserted vitreous base based on the presence of posteriorly located retinal breaks. The relatively high number of breaks seen pre-operatively (and more resulting intra-operatively) in our series suggests that those presenting with RRD with at least 3 retinal breaks and/or lattice near the equator may be at highest risk of having posteriorly inserted vitreous base.

It is not surprising that the majority of patients in our series with posteriorly inserted vitreous base had a pre-operative indication of RRD, but it is noteworthy that patients with other indications such as macular hole, epiretinal membrane and vitreous hemorrhage also had posteriorly inserted vitreous base. This suggests that posteriorly inserted vitreous base is intrinsic to certain eyes themselves, instead of being associated only with RRD. It is unclear if there is a genetic association for posteriorly inserted vitreous base but a specific inheritance pattern upon questioning of patients was unrevealing.

Creation of a PVD can cause new retinal breaks, but it was interesting that new breaks were observed intra-operatively even in eyes that had a Weiss ring noted pre-operatively. Thus, while Weiss ring on pre-operative examination indicates detachment of the hyaloid face from the optic nerve head, its presence may not always be an accurate indicator of the hyaloid being separated from the retina to its most anterior extent. This could explain why some eyes had additional breaks seen after ‘complete’ PVD induction with the cutter despite presence of Weiss ring. This supports the notion that eyes with posteriorly inserted vitreous base have a different vitreoretinal interface than eyes with normal insertions that might be elucidated better at the cellular/molecular level. It also emphasizes the need for heightened awareness of the vitreoretinal interface during PVD induction, and the need to meticulously examine the vitreous base for new breaks prior to fluid-air exchange or before cannula removal in cases not undergoing fluid-air exchange.

This study has the typical shortcomings of a retrospective series, such as a low number of patients, no comparative cohort, lack of standardized treatment, and subjective nature that is based on technique and skill with PVD creation. However we believe this is an important condition for vitreoretinal surgeons to recognize. Future studies could look at the use of triamcinolone staining of vitreous, utility of adjunctive treatments like a wider scleral buckle, and a careful determination of incidence of posteriorly inserted vitreous base in eyes with RRD and other conditions.

In this paper we detailed the characteristics of patients with posteriorly inserted vitreous base found independently by two different surgeons at different institutions. While we only noted the eyes that had posteriorly inserted vitreous base and did not count eyes undergoing vitrectomy without the condition, we estimate that the rate is less than 5–10% of cases of RRD. Since it is useful to be knowledgeable of the concept of the posteriorly inserted vitreous base and its risk factors, mitigating the risk of re-detachment with adjuvants like scleral buckle may bring single surgery success to published rates of non-complex RRDs.^13–15 As eyes with posteriorly inserted vitreous base needing vitrectomy seem to have different anatomy and post-operative results than those without, we hope that continued education of trainees and further research into this condition will enable retina specialists to better treat patients in the future.

Supplementary Material

Supplemental Table 1

Supplemental Digital Content: Supplemental table 1.pdf