Key points are not available for this paper at this time.
Purpose: To determine the prevalence of macular abnormalities detected by optical coherence tomography (OCT) in patients with retinitis pigmentosa (RP), and to report the results of pars plana vitrectomy (PPV) in five patients with RP. Methods: OCT images of the macula of 622 eyes of 323 patients with RP were evaluated. All patients had a complete clinical examination, and PPV was performed on two RP patients with a macular hole (MH), two patients with vitreomacular traction (VMT) syndrome, and one patient with cystoid macular oedema (CME). Results: A macular lesion was detected by OCT in 46 eyes (7.4%) of 37 patients (11.5%). CME was detected in 34 eyes (5.5%) of 26 patients, an epiretinal membrane in four eyes of four patients, VMT in five eyes of four patients, and full-thickness MH in three eyes of three patients. PPV was performed on five of these eyes. The MH was closed with vision improvement after a single surgery in one eye, and was closed after repeated PPVs with a decrease of vision in another eye. In both patients with VMT, the visual acuity decreased after the surgery. The macular morphology and visual acuity did not improve postoperatively on the one patient with CME who underwent PPV. Conclusion: When OCT is used, macular abnormalities are present in 7.4% of patients with RP, which is lower than the prevalence reported earlier. Although vitrectomy can improve the macular morphology in some patients with RP, improvement of visual function may be limited most likely because of the long-standing retinal dysfunction. Ten to twenty per cent of patients with retinitis pigmentosa (RP) have been reported to have cystoid macular oedema (CME) (Fishman et al. 1977a; Fishman et al. 1977b; Hirakawa et al. 1999). This is important because CME can significantly reduce the visual acuity in patients with RP whose vision is usually well preserved until the late stages of the disease. More recently, it was reported that the prevalence of CME detected by optical coherence tomography (OCT) in patients with RP was 38% (Hajali et al. 2008). This higher prevalence was explained to be as a result of the higher sensitivity of OCT than either fluorescein angiography or funduscopic examinations to detect CME (Apushkin et al. 2004). Other macular diseases, such as macular holes (MHs), epiretinal membranes (ERMs), and the vitreomacular traction (VMT) syndrome, are less frequently detected in patients with RP. Giusti et al. (2002) examined 236 patients with RP ophthalmoscopically and found a MH in 25 patients. However to the best of our knowledge, the exact prevalence of these diseases in patients with RP determined by OCT has not been reported. There have been several attempts to treat the CME in patients with RP. Carbonic anhydrase inhibitors (CAIs) have been shown to benefit some of these patients, but the effects of CAIs are limited because they are either ineffective or cause unacceptable side-effects (Cox et al. 1988; Fishman et al. 1989, 1993; Chung et al. 2006; Apushkin et al. 2007). Vitrectomy with inner limiting membrane (ILM) dissection was reportedly to be effective in both reducing the CME and improving the visual acuity in patients with RP (Garcia-Arumi et al. 2003). We report our findings of OCT examinations of the macula in 322 patients with RP to determine the prevalence of macular pathology in patients with RP. We also report the anatomical and functional outcomes of vitreoretinal surgery for macular diseases in five of these patients with RP with macula abnormalities. The medical records of 622 eyes of 323 patients (162 men and 161 women) with RP who were examined at the Eye Clinic of the Chiba University Hospital from November 2003 to October 2007 were reviewed. The diagnosis of RP was made from the clinical history, funduscopic appearance, visual fields, fluorescein angiography, and the results of full-field electroretinograms (ERGs) recorded with the International Society of Clinical Electrophysiology of Vision (ISCEV) parameters. Among the 323 patients, 245 patients were typical RP, 43 were the sine-pigment type of RP, and 30 patients were the central type of RP. Patients with systemic syndromes associated with RP-like fundus appearance, and those with cataracts or vitreous opacities which may affect the OCT examinations were excluded. None of the patient had diabetes mellitus, ocular inflammatory diseases, or primary retinal vascular diseases. Eyes that had undergone cataract surgery were also excluded. OCT was performed with the OCT 3000 model (Zeiss Humphrey, Sun Leandro, CA, USA) with horizontal and vertical scans in each patient. The scan length was usually 5 mm. The study protocol adhered to the tenets of the Declaration of Helsinki. All patients had typical RP and their clinical characteristics are summarized in Table 1. Vitreoretinal surgery was performed on two patients with stage 4 MH, two patients with VMT, and one patient with CME. Surgery was performed on these five patients because they had noticed a recent reduction in visual acuity presumably because of the macular pathology. Patients were fully informed of all relevant aspects of the surgery, and all patients provided a written informed consent. The treatment was mainly conventional pars plana vitrectomy with a 20-gauge system combined with phacoemulsification and aspiration with intraocular lens implantation because all patients had cortical cataracts. The ILM was removed using indocyanine green or triamcinolone acetonide, and sulphur hexafluoride (SF6) gas tamponade was used in the two MH cases. A posterior vitreous detachment was created and the ILM was not dissected in the other three cases. The OCT images showed abnormal macular changes in 46 eyes (7.4%) of 37 patients (11.5%; Table 2). CME, defined by cysts visible on OCT not accompanied by vitreous traction, was the most frequently observed macular alteration in 34 eyes (5.5%) of 26 patients (Fig. 1A). The mean visual acuity in logMAR units was 0.30 in the RP patients with CME and 0.33 in those without CME and this difference was not statistically significant (p = 0.82, unpaired t test). An epiretinal membrane (ERM) was observed in four eyes (0.6%) of four patients (Fig. 1B), VMT in five eyes (0.8%) of four patients (Fig. 1C), and a full-thickness MH in three eyes (0.5%) of three patients (Fig. 1D). Horizontal optical coherence tomographic images of patients with retinitis pigmentosa associated with macular complications. (A) Cystoid macular oedema. (B) Epiretinal membrane. (C) Vitreomacular traction syndrome without posterior vitreous detachment. (D) Full-thickness macular hole. The surgical outcomes of five patients who underwent vitreoretinal surgery are summarized in Table 1. In the MH cases, patient 2 obtained a MH closure and vision improvement by a single operation, while patient 1 required repeated PPVs for MH closure, and the visual acuity was reduced postoperatively probably because of retinal thinning and atrophy of the retinal pigment epithelium at the macula. In patient 2, the multifocal ERGs recorded from the central 10 degrees of the macular area were subnormal before surgery, and non-recordable at 3 months postoperatively. The third patient with MH chose not to have surgery. In both of the VMT cases, the visual acuity decreased after surgery, and the cystoid changes remained in patient 3 and foveal thinning appeared in patient 4. Patient 5 with a CME had been treated with systemic CAI for several months but did not have any improvement, and then the patient chose to have a surgery. The macular morphology and visual acuity did not improve at 1 year postoperatively. Before the introduction of OCT to the clinic, macular oedema was diagnosed to be present by fluorescein angiography or slit-lamp examination with a contact lens. Its prevalence among patients with RP was reported to be 10–20% (Fishman et al. 1977a; Fishman et al. 1977b). Hirakawa et al. (1999) reported the OCT findings of the macula in patients with RP for the first time and they found CME in 6 (13%) of 46 patients independent of the fluorescein angiograms. Sandberg et al. (2005) examined 162 patients with typical RP with OCT and found that the visual acuity was related to retinal thickness by a second-order polynomial. However, they did not report the prevalence of CME. With OCT, the prevalence of CME in patients with RP was reported to be 47 (38%) of 124 patients (Hajali et al. 2008). In our study, we examined the OCT images in a larger cohort of patients with RP and found CME in 26 (8%) of 323 patients. We excluded patients with VMT, which generally showed intraretinal cystoid spaces or subretinal fluid, from those with CME, because it is apparent that pathophysiological mechanism of VMT is a mechanical traction by the posterior vitreous membrane and may be different from that of CME. Even if we had included patients with VMT to those with CME, the prevalence of CME would be increased to 9.2% which is still significantly lower than the reported values. The lower rate of sensitivity in our study could be because we used a time domain OCT which has a lower resolution than that of the Fourier domain OCT. Although the pathogenesis of CME in RP has not been determined, it has been postulated that CME is because of a dysfunction of the RPE leading to a failure of the pumping mechanism and leakage of fluid through the RPE (Cox et al. 1988; Fishman et al. 1989). Carbonic anhydrase inhibitors (CAIs), such as acetazolamide or methazolamide, affect the RPE pumping mechanism and reduce the CME. In addition to attempts to treat CME with CAIs, it has been reported that vitrectomy and ILM removal effectively reduced the macular oedema accompanied with the improvement of visual acuity in patients with RP (Garcia-Arumi et al. 2003). Authors have suggested that the vitrectomy can remove any tractional forces on the retina and also remove the inflammatory factors in the vitreous. These changes should improve the biochemical and fluid exchange at the vitreoretinal interface. However, it is obvious that vitrectomy does not affect the RPE function directly, and potential retinal toxicity of illumination during surgery and indocyanine green for ILM removal should be considered. Because similar successful results of vitrectomy for CME in patients with RP have not been reported, and our case (patient 5) was not successful, we believe that careful consideration is necessary before vitrectomy is performed for CME in patients with RP. We found other macular complications related to abnormalities of the vitreoretinal interface, such as MH, ERM, and VMT, in 12 eyes (1.9%) of 622 eyes. To the best of our knowledge, the prevalence of macular complications among RP patients has not been reported. It was reported that RP patients have a degenerative vitreous, including collapse of vitreous gel, and posterior vitreous detachment, and such changes may cause macular complications related to the vitreoretinal interface (Pruett 1975; Fishman et al. 1977a; Fishman et al. 1977b). Recently, Jin et al. (2008) reported that the visual acuity was significantly improved after MH surgery in three patients with RP and not improved in one RP patient with a MH and retinal detachment. In our cases, the visual acuity was improved to 0.7 from 0.4 in one patient with MH (patient 2) after a single operation; however, the visual acuity was decreased in another patient (patient 1) who required repeated surgeries. The phototoxicity of the illumination of the retina and the indocyanine green may have deteriorated macular function which was evident by the postoperative reduction in the macular responses measured by multifocal electroretinography. Surgical results were also not satisfactory in another two patients with VMT, probably because the dystrophic retinas did not tolerate the surgical trauma, and vitreous tractions are not the sole pathophysiological mechanisms for such macular complications. In conclusion, OCT is a powerful tool to detect macular complications in patients with RP especially in eyes with reduced vision. Because the surgical outcomes were not always favourable, vitreous surgery should be carefully considered for such macular lesions. The authors would like to thank Dr. Duco Hamasaki of Miami University for editing the manuscript. This study was supported by a grant from the Ministry of Health, Labor and Welfare of Japan.
Hagiwara et al. (Mon,) studied this question.