Article Text
Abstract
Aims To assess the effectiveness of treatment to the inner sclerostomy sites at the time of vitrectomy for proliferative diabetic retinopathy (PDR) in reducing the incidence of late recurrent postoperative vitreous cavity haemorrhage (POVCH).
Method Retrospective study of a consecutive series of 82 eyes undergoing vitrectomy for PDR by a single surgeon treated with either cryotherapy or argon laser directly to the inner sclerostomy site at the completion of surgery (treatment group). These were compared with a previous consecutive series of 82 eyes operated on by the same surgeon who did not have inner sclerostomy site treatment (control group). The occurrence of any POVCH was recorded within the first 6 months of surgery.
Results The composition of the two groups was similar in terms of age, indication for surgery and a variety of other preoperative factors. There were 64 patients in the control group and 65 in the treatment group. There was a significant reduction in the incidence of late recurrent POVCH in the treatment group. Late recurrent POVCH occurred in 12 (15%) eyes in the control group compared with five (6%) in the treatment group (p=0.03). The number of eyes requiring revision surgery within the first 6 months for late recurrent POVCH was four (5%) in the control group and two (2.5%) in the treatment group (p=0.31).
Conclusion This study suggests that inner sclerostomy site treatment is effective in reducing the occurrence of recurrent late POVCH in patients undergoing vitrectomy for PDR. A randomised controlled study is needed to clarify this.
- clinical trial
- cryotherapy
- entry site treatment
- laser
- neovascularisation
- postoperative vitreous cavity haemorrhage
- proliferative diabetic retinopathy
- retina
- treatment surgery
- vitrectomy
- wound healing
Statistics from Altmetric.com
- clinical trial
- cryotherapy
- entry site treatment
- laser
- neovascularisation
- postoperative vitreous cavity haemorrhage
- proliferative diabetic retinopathy
- retina
- treatment surgery
- vitrectomy
- wound healing
Pars plana vitrectomy is an established and successful treatment for the complications of proliferative diabetic retinopathy (PDR).1 2 It is required within 1 year in up to 10% of patients presenting with PDR.3
Unfortunately, postoperative vitreous cavity haemorrhage (POVCH) is a significant complication occurring in approximately 20–40% of cases, although the reported range is large, namely 5–80% of cases, dependent on case definition and follow-up.4–9
Patients with POVCH have a delay in their visual recovery and, in some cases, this results in worse vision than was present preoperatively. Non-clearing POVCH necessitates revision surgery in approximately a third to a half of those cases experiencing POVCH and approximately 10% of all patients undergoing surgery.10 11
POVCH can be either persistent, being present from the first postoperative day, or recurrent, occurring later during follow-up, commonly at 2–4 months postoperatively. Although persistent haemorrhage can also be contributed to by ongoing recurrent haemorrhage, its cause is often discrete.
Persistent haemorrhage can result from operative and postoperative oozing of new vessels, remnants or dissected tissue, or directly from the sclerostomies used to perform surgery. Leaching of red blood cells can also occur from retained old haemorrhage in residual anterior vitreous.
Early recurrent haemorrhage, which can contribute to persistent haemorrhage, can occur from clot dissolution and lysis from dissected tissue. Late recurrent haemorrhage can result from traction on residual new vessels or indeed postoperative new vessel growth. Recent studies have shown, however, that a common cause of recurrent haemorrhage is new vessel growth at the inner sclerostomy sites associated with fibrous traction.12–18 This has been termed ‘entry site neovascularisation’ (ESNV).19 It is thought to be an aberrant wound healing response related to the presence of retinal and pars plana ischaemia.15 21 It is a separate entity from the rarer occurrence of anterior hyaloidal neovascularisation, a more widespread growth of new vessels along the residual anterior hyaloid face.15 The presence of ESNV is difficult to observe clinically because of the very anterior location but can be confirmed at the time of revision surgery with deep scleral indentation or endoscopic techniques. It can also be localised with anterior segment high-resolution ultrasonography of the inner sclerostomy sites.13 14 17 18
Yeh et al,21 in a retrospective case series of 23 patients, recently reported a reduction in the rate of recurrent POVCH from 11.5% down to 4.3% by the application of cryotherapy to the sclerostomy sites at the completion of surgery. Since the beginning of 2005 we have been following a similar protocol of sclerostomy site treatment. We report the effect of this treatment on the rate of POVCH in a cohort of patients undergoing vitrectomy for the complications of PDR and compared the results with a previous cohort, who did not have sclerostomy site treatment, by the same surgeon.
Methods
A consecutive series of patients who had undergone primary pars plana vitrectomy for the complications of PDR by one surgeon (DHWS) at Sunderland Eye Infirmary from March 2005 to October 2008 was prospectively identified.
All applicable institutional and governmental regulations were followed during this research. The study was classified as a service evaluation by the local ethics committee and as such did not require formal ethical approval.
Patients were excluded from the analysis if follow-up was for less than 6 months. Patients were also excluded if silicone oil was used.
The operative technique was with three-port pars plana vitrectomy using 20, 23 and 25 g instruments (Alcon Accurus system, Alcon, Fort Worth, Texas, USA). Sclerostomies were made as follows: 20 g—tunnelled incisions were used as previously described;18 23 and 25 g—acutely oblique entry into the sclera was made with the trochar using the Accurus one step entry system and continued until the sclerostomy cannula abutted the sclera. The trochar was then moved to 30° to the globe and final entry made.
After core vitrectomy, delamination and removal of all posterior hyaloid face and fibrovascular membrane was carried out. Careful attention was taken to detect and remove areas of vitreoschisis,20 and staining for residual vitreous gel with diluted triamcinolone was performed routinely in all cases.
Haemostasis was obtained using raised infusion pressure and endodiathermy as needed. The extent of fibrovascular proliferation and vitreoretinal adherence present and requiring peeling/delamination was graded according to disc areas and also classified as by Yorston et al.4 Peripheral vitreous was removed as much as possible, especially around the inner sclerostomy wounds with deep scleral indentation in an attempt to create a pseudo aphakic condition, as described by McLeod,1 and to reduce the occurrence of vitreous incarceration in the sclerostomies that could act as a ‘scaffold’ for new vessel growth.
Indirect or endolaser peripheral retinal laser photocoagulation was carried out to the peripheral retina up to the ora serrata in all eyes. The density of burns was approximately one burn width apart other than for one clock hour behind the inner sclerostomy wounds, where near confluent treatment was applied from the ora posteriorly for approximately one disc diameter.
In all eyes treatment to the sclerostomies was carried out as follows:
In the first year of the series, after partial fluid/air exchange (enough so that the sclerostomy sites were covered by air), indirect argon laser was applied directly to the inner sclerostomy sites using deep sceral indentation over an area of approximately 2 mm in diameter centred on the sclerostomy and extending posteriorly to the ora serrata. If this was not possible because of limited view then cryotherapy was applied as below.
In the subsequent years of the series cryotherapy was applied as the first choice. Again after partial fluid/air exchange external cryotherapy was applied under direct non-contact wide field viewing with endoillumination via the opposite sclerostomy. The endpoint was a freeze extending over approximately 2 mm in diameter (the time taken to achieve this ranged from 6 to 22 s with a mean of 9 s). Two applications were used—one over the sclerostomy itself and one immediately behind it on the intervening pars plana and ora serrata. The time taken to achieve the endpoint was recorded. This time was then used to guide treatment performed without visualisation to the other sclerostomies.
At the end of surgery, and after sclerostomy guard removal in narrow gauge surgery cases, sclerostomy closure was checked by inflating the eye to a pressure of 30 mm Hg using a 30 g needle. If any sclerostomy leak was present an 8/0 absorbable suture was used to secure the wound. Phacovitrectomy was carried out in cases with visually significant cataract obscuring the operative view.
Patients were followed up at 1 day, 2 weeks, 1, 2, 4 and 6 months postoperatively and then at two 6-monthly intervals as needed according to the clinical circumstances. The clarity of the vitreous cavity at each time point was assessed. If a gas bubble was present the clarity of the retinal view below the gas was assessed. Patients were also asked to report disturbances to their vision in between visits and were reassessed at these times.
Ultrasound of the 10 MHz posterior segment and 20 MHz anterior segment was carried out by a single experienced ultrasonographer (RO) if a patient experienced either an observed POVCH or symptoms suggestive of POVCH. Scans were performed in the supine position using the Innovative Imaging Incorporated (I3) system (Innovative Imaging Systems, Inc., Sacramento, USA) as previously described.18
A variety of preoperative and intraoperative characteristics were recorded for the patients. The amount of preoperative pan retinal photocoagulation administered was graded as full, good, moderate, light or none. The presence of active neovascularisation was documented. Eyes with grade 2 vitreous haemorrhage (defined as per the Diabetic Retinopathy Vitrectomy Study (DRVS))22 at 1 month postoperatively that had persisted since the first postoperative day were defined as having persistent POVCH. Eyes that developed new POVCH after being clear at the first postoperative day were classified as having recurrent POVCH and the severity and time point at which this occurred was recorded.
The cases undergoing sclerostomy site treatment were compared with a previous consecutive cohort of patients that had undergone vitrectomy for PDR by the same surgeon from January 2002 to March 2005. The surgical technique in the previously described cohort was identical except that no sclerostomy site treatment was carried out. In particular, the technique of peripheral retinal laser was identical.
Statistical analysis
The two groups were compared using non-paired Student's t tests for normal continuous variables and Fisher's exact test for categorical data. For non-normal continuous observations the Mann–Whitney test for two independent samples was employed.
Results
Eighty-two eyes were included in each cohort. There were 64 patients in the control group and 65 in the treatment group. The groups were well matched in terms of basic demographic variables (Table 1). There were significantly more patients in the treatment group who had narrow gauge surgery. There were also significantly more patients in the treatment group on treatment with cholesterol-lowering agents as well as aspirin. Gas tamponade was used less frequently in the treatment group than the control group.
The number of eyes experiencing POVCH in each group are shown in table 2 (control group) and table 3 (treatment group). There was a significant reduction in the incidence of late recurrent POVCH, with this occurring in 12 (15%) eyes in the control group compared with five (6%) in the treatment group (p=0.03). The numbers of eyes requiring revision vitrectomy with vitreous cavity washout (VCWO) within the first 6 months in the control group was four (5%) and two (2.5%) in the treatment group (p=0.31). There was no significant difference between the groups in the incidence of persistent POVCH (one eye in each group), and there was no significant difference in the total number of eyes requiring VCWO between the groups (three eyes in the study group and five in the control group; p=0.36).
There were no cases of severe uveitis or endophthalmitis and no cases of postoperative retinal detachment in either group.
Discussion
POVCH after diabetic vitrectomy is a significant problem. Comparing results between different publications is difficult because of varying case mixes and importantly because of varying definitions of what constitutes a POVCH. Furthermore, the follow-up period is important. POVCH can occur years after the original surgery although most recurrent POVCH occurs within the first 6 months.17 18 21 This study was principally interested in recording the incidence of late recurrent POVCH, which we defined as any observed POVCH occurring after 3 weeks postoperatively. However, we have also presented all POVCH as accurately as possible in terms of degree, duration, need for revision surgery and timing.
In this study, in patients undergoing diabetic vitrectomy in which cryotherapy or argon laser was applied directly to the inner sclerostomy sites, the incidence of late recurrent POVCH was reduced from 15% to 6% compared with a previous cohort of patients without sclerostomy site treatment. Both groups of patients had extensive peripheral retinal ablation particularly posterior to the sclerostomy sites applied using an indirect argon laser or endolaser. The results concur with those of Yeh et al,21 who also reported a reduction in the rate of POVCH from 11.5% to 4.3% using a similar protocol of sclerostomy site treatment. Yeh et al21 also included a third group in their study who did not receive either sclerostomy site cryotherapy or peripheral retinal cryotherapy, in whom the incidence of recurrent POVCH was higher at 32%.21
Yeh et al21 hypothesised that cryotherapy to the sclerostomy sites impeded tissue proliferation and fibrovascular ‘ingrowth’ at the inner sclerostomy site by inhibiting cellular migration through the sclerostomy wounds and causing focal atrophy of the ciliary epithelia. This combined with peripheral retinal ablation, to areas of previously untreated inaccessible areas of peripheral retinal ischaemia, would reduce angiogenic factors in the eye, reducing the occurrence of ESNV and subsequent late recurrent POVCH.7 21 23
The strengths of this study compared with the study by Yeh et al21 were that this was a consecutive series of 82 patients by one surgeon compared with selected patients by two surgeons of 23 eyes by Yeh et al21 Furthermore, the study by Yeh et al21 included three (13%) patients who had silicone oil inserted, which would mask any recurrent POVCH, and 11 (48%) patients who had a scleral buckle applied, which could also plausibly confound the results. None of our patients had silicone oil or a scleral buckle applied. The technique of sclerostomy site treatment also differed between our study and theirs. In the first year of our study patients had argon laser applied to the inner sclerostomy site as first choice and had cryotherapy applied if laser was not possible. When applying cryotherapy we precisely defined and timed the endpoint reaction by direct wide field viewing of the first inner sclerostomy site. We then used this time to treat the other sclerostomies without direct viewing of the inner sclerostomy site. We specifically treated the sclerostomy site itself and the area between the sclerostomy site and the peripheral retina so that it was confluent with the area of peripheral laser ablation. Yeh et al21 used a protocol of two applications of cryotherapy to each sclerostomy site for 6 s for all patients without observation of the endpoint or exact anatomical location of the freeze.
The sclerostomy site treatment did not reduce the incidence of late recurrent POVCH to zero but it approximately halved the rate. There are two possible explanations for this. First, the sclerostomy site treatment may not have been 100% effective at preventing ESNV, or second, not all late recurrent POVCH is secondary to ESNV. None of the patients in the treatment group with late recurrent POVCH had ESNV identified either at the time of revision surgery or on ultrasound. In the control group, however, which has been partly reported elsewhere,18 approximately 50% of the patients with POVCH had ESNV identified either on ultrasound or at the time of revision surgery, and others have reported that ESNV is found in 62%17 and 88%21 of eyes experiencing late recurrent haemorrhage. It is unlikely that the sclerostomy site treatment is truly 100% effective at preventing ESNV; however, we think it is also unlikely that ESNV accounts for all late recurrent POVCH, and that the other cases of POVCH are due to other causes that are not prevented by sclerostomy site treatment.
This study and that of Yeh et al21 strongly suggests that inner sclerostomy site treatment reduces the occurrence of recurrent POVCH after vitrectomy for PDR, and its effect is additive to that of peripheral retinal ablation with either cryotherapy or laser alone. There are, however, a number of potential weaknesses in this study. The study was not randomised and we compared our treatment group with a previous sequential control group of patients operated upon by the same surgeon without scerostomy site treatment. The surgical technique was identical between the two groups except for the use of narrow gauge surgery. Narrow gauge vitrectomy was used in 51% of the patients in the treatment group and only 2% in the previous control group. It is not known if the use of narrow gauge sutureless vitrectomy is associated with a change in the incidence of POVCH. The aetiology of ESNV is not fully understood, although it is considered to be a variant of sclerostomy site healing.15 Peripheral retinal ischaemia occurs in severe PDR and is likely to be important in the pathogenesis, but it is possible that sclerostomy size is related to the incidence of the condition. The choice of which vitrectomy gauge to use in an individual surgery was based principally on the availability of equipment during the study period—20 g was used exclusively initially and then 23 and 25 g exclusively in the latter half of the study. The use of guarded sclerostomies with the 25 and 23 gauge systems used will have reduced the instrument trauma of the sclerostomy sites themselves. However, in previous studies we18 and others13 have found no correlation of ESNV with the degree of instrument use through the sclerostomy. Hypotony has frequently been reported by authors after narrow gauge surgery, which would potentially increase the rate of immediate POVCH. At the end of all surgeries we carefully assessed sclerostomy closure by inflating the eye to a pressure of 30 mm Hg and suturing sclerostomies if needed. Perhaps as a result of this there were only two patients in the study with an intraocular pressure of less than 8 mm Hg on the first postoperative day and there was a low incidence of persistent POVCH in this study. The incidence of vitreous incarceration in narrow gauge surgery compared with 20 g surgery is not known but is known to occur.24 Care was taken in all patients to remove peripheral vitreous as much as possible, as described in the Methods section. Shaikh et al25 reported an increased incidence of new vitreous haemorrhage in patients undergoing 25 gauge surgery compared with 20 gauge, and fibrous ingrowth has been observed in pig models of sclerostomy site healing in narrow gauge surgery as in 20 g surgery.26 Nagpal et al,27 Oshima et al28 and Lo et al29 have described POVCH occurring after 23 g and 25 g surgery at a frequency that is broadly equivalent to 20 g surgery. It is possible, therefore, that narrow gauge surgery could have influenced the occurrence of ESNV and POVCH, but the direction of this effect is unknown.
The two groups also differed in various aspects of medical treatment the patients were receiving. It is possible that these treatments could have confounded the results. Indeed, Soto-Pedre et al,30 in a retrospective study, reported that the use of antihypertensive agents in the 3 months before surgery was associated with a reduced incidence of POVCH.
The effect of statins and aspirin on the rate of POVCH is not known. The two groups were well matched, however, for a variety of measurable factors of systemic control, including random glucose and haemoglobin A1c on admission.
We changed our protocol from treating sclerostomy sites with indirectly delivered argon laser when possible to all patients receiving cryotherapy to their sclerostomy sites to allow a more consistent and repeatable treatment endpoint. The change in sclerostomy site treatment modality, however, did not appear to alter the efficacy of the treatment effect. Cryotherapy can be applied more consistently and easily to the inner sclerostomy site and is our preferred modality. Sclerostomy site treatment was applied under air in all cases for three reasons. First, it was found that the cryotherapy endpoint was more quickly and repeatably reached under air than under fluid infusion. Second, the peripheral view of the inner sclerostomy site was improved under air, and finally the self-sealing nature of the sclerostomy could be more easily assessed and confirmed under air compared with fluid. The use of air to allow inner sclerostomy site treatment could plausibly have confounded our results, and one recent paper found a reduction in persistent POVCH with air tamponade.31 However, two previous studies, one using SF632 and one using air tamponade33 postoperatively did not find any protective effect of these short-term tamponade agents on any form of POVCH, in particular with no effect on late recurrent haemorrhage. We only used a 30–40% air fill, when tamponade for retinal breaks was not needed, and no specific posturing instructions were given.
The study we have presented is retrospective and non-randomised and the results must be interpreted with caution, but it suggests that sclerostomy treatment in patients undergoing vitrectomy for the complications of PDR reduces the incidence of late recurrent POVCH by approximately 50%. A randomised controlled study of sclerostomy site treatment is needed to answer definitively the question of whether it reduces the incidence of late recurrent POVCH after diabetic vitrectomy, but entry site treatment with cryotherapy is quick and easy with low morbidity and expense.
References
Footnotes
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.