Somatostatin analogues as a treatment option for cystoid maculopathy in retinitis pigmentosa

Introduction

Retinitis pigmentosa (RP), the most common form of inherited retinal dystrophies, initially presents with night blindness and a gradual decline in peripheral vision and progresses to central vision loss.1 Cystoid maculopathy (CM), characterised by cyst formation in the intracellular spaces of the retina primarily in the outer-plexiform layer, is a common complication in RP, occurring in ~50%.2 3 Cystoid spaces can exhibit either leaking or non-leaking characteristics.4 Potential mechanisms described for CM in RP include disruption of the blood–retina barrier which causes leakage from the perifoveal retinal capillaries, as well as retinal pigment epithelium (RPE) pump mechanism failure and/or müller cell oedema and dysfunction. Disruption of retinal architecture due to underlying RP might explain mechanisms for non-leaking cystoid spaces.4 5 CM leads to immediate or gradual visual impairment due to retinal thickening and fluid collection, which distort the photoreceptor architecture ultimately accelerating photoreceptor atrophy.5 6 CM treatment is important for preserving photoreceptors, thereby optimising visual acuity, especially for RP patients depending on central vision.1

Several treatments are available for reducing CM in RP patients. The first choice of treatment is carbonic anhydrase inhibitors (CAI) orally or topically which reduces CM in ~42% and increases visual acuity in ~40% of the patients.7 8 Intolerable side effects such as potassium deficiency, tingling of limbs (67%–86%), fatigue (17%–43%) and gastrointestinal symptoms (17%–43%) are often a reason to stop this treatment. Topical application of CAI can reduce CM by 30%–81% and improve visual acuity by 19%–37%.7 8 Topical or intravitreal steroids can reduce CM in almost all patients with improvement of visual acuity in ~20%. Important side effects of steroids are high intraocular pressure (>10%) and cataract (>10%), especially when used chronically, which make this a less attractive treatment option.7 8 Vitreoretinal surgery has been shown to significantly reduce foveal thickness (FT) and improve best-corrected visual acuity (BCVA) in a subset of patients, but it is invasive and can lead to complications such as retinal detachment or macula oedema.7 The above-mentioned treatments are unfortunately not sufficiently effective for all patients or may cause intolerable side effects.

Since ~20 years, somatostatin analogues (SA) have been relatively newcomers in the treatment of refractory CM. Natural somatostatins are 14 and 28 cyclic neuroendocrine peptides, therapeutic analogues are smaller peptides with a longer half-life.9 In the retina, somatostatin is produced by the healthy RPE and somatostatin receptors are expressed at the apical site of the RPE and the neuroretina. The effects of somatostatin are neuroprotection, antiangiogenesis and protection against oxidative stress.9 The first proofs of successful treatment of patients stems from case series for dominant CM, or CM complicating uveitis, and diabetic retinopathy.10–13 To the best of our knowledge, the literature on SA treatment and RP has not yet been reported.14 As the pathophysiology of CM by RP patients differs from uveitis and diabetic retinopathy, we aimed to obtain insights into the effectiveness of SA treatment in RP patients with CM.

Methods

In this retrospective case series, 28 RP patients (Erasmus MC, University Medical Center n=12, the Rotterdam Eye Hospital n=16) with refractory CM with unsatisfactory response to CAI were included. Clinical data were collected from medical charts and registered in the Dutch RD5000 national database for inherited retinal dystrophies, as this study was part of the RD5000 study.15 All patients provided informed consent for participating in the RD5000 study and the use of their clinical data. Patients were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

CM was defined as the presence of cysts within the intracellular spaces of the retina, predominantly located in the outer plexiform layer, confirmed on OCT. Refractory CM with unsatisfactory response to CAI was defined as no changes in CM or insufficient changes in CM on OCT after CAI initiation as assessed by the treating ophthalmologist.

The prescribed SAs were octreotide long-acting release (LAR) (octreotide, sandostatine) and lanreotide (somatuline). For octreotide, LAR the dose was either 20 mg/month or 30 mg/month and administered intramuscularly. For lanreotide, the dose was either 90 mg/month or 120 mg/month and administered deep subcutaneously.

Risks and potential benefits were thoroughly discussed with patients before SA treatment initiation. Prescription of SA, pretreatment screening and close monitoring of patients during treatment were conducted by the internist. All patients underwent laboratory testing including hematogram, glucose and liver biochemistry (routine blood tests). The optimal dose of the SA was determined based on the extent of CM, the patient’s medical history, screening results and the risk of side effects and could be adjusted throughout treatment. The doses referred to in this study represent the ultimate prescription at which the patient is stabilised for the management of CM. SA treatment was discontinued before 12 months if there was no reduction in CM or if intolerant side effects occurred.

Patients underwent genetic testing via whole exome sequencing. Clinical data (medication data, BCVA, slit lamp examination and ophthalmoscopy) and OCT data were collected from medical charts at baseline, and 3, 6 and 12 months after initiating SA treatment. FT and foveal volume (FV) were extracted from OCT scans using the built-in algorithm of the devices (Spectralis OCT, Heidelberg Engineering, Heidelberg, Germany; HS-100 OCT, Canon, Tokyo, Japan; Stratus OCT, Carl Zeiss Meditec, Dublin, California, USA and RTVue OCT, Optovue, Fremont, California, USA), as indicators of CM extent. Additionally, in three patients, fluorescein angiography (FA) was performed to examine leakage.

Statistical analysis

Primary outcome measurements were reduction in FT (μm) and FV (mm³) as measure of decrease in CM and increase in BCVA (logarithm of the minimum angle of resolution (logMAR)). Secondary outcome measures were the occurrence of side effects and the prognostic value of genotype, dose of SA and simultaneous use of CAI.

Linear mixed models were applied to assess differences in FV, FT and BCVA over time. Three models were constructed, with each of these variable serving as the dependent variable and follow-up time as a fixed factor. Independent variables included SA dosage, genotype and CAI usage. The models were adjusted for age and gender and accounted for the correlations between the eyes of individual patients if both eyes were included.

All statistical analyses and figures 1–3 creation were performed by using R statistical package V.4.6.1 for Windows (http://www.r-project.org). A p<0.05 was considered statistically significant. Figure 4 was created using PowerPoint V.2016 for Windows.

Request permission

Figure 1

Reasons for start of or switch to SA treatment. 21 patients (75%) had no effect of CAI treatment, 4 patients (14.3%) had both intolerance and no effect of CAI treatment, and another 2 patients (7,1%) were intolerant for CAI treatment. One patient (3.6%) started with SA treatment as first treatment for CM, as the patient was considered too young (<18 years) for systemic CAI treatment. CAI, carbonic anhydrase inhibitors; CM, cystoid maculopathy; SA, somatostatin analogue.

Request permission

Figure 2

Distribution of SA treatment regimens. LAR, long-acting release; SA, somatostatin analogue.

Request permission

Figure 3

Clinical changes in mean FT (A), mean FV (B) and mean BCVA (C) after start with SA treatment (n=28). At baseline, the mean FT was 409±136 µm, the mean FV was 0.31±0.10 mm³ and the mean BCVA was 0.46±0.35 logMAR. At 3 months the mean FT was 334±100 µm, the mean FV was 0.26±0.08 mm³ and the mean BCVA was 0.45±0.36 logMAR. At 6 months, the mean FT was 328±110 µm, the mean FV was 0.25±0.08 and the mean BCVA was 0.44±0.40 logMAR. At 12 months, the mean FT was 334±119 µm, the mean FV 0.25±0.04 mm³ and the mean BCVA was 0.45±0.38 logMAR. Error bars represent SD. BCVA, best-corrected visual acuity; FT, foveal thickness; FV, foveal volume; logMAR, logarithm of the minimum angle of resolution; SA, somatostatin analogue.

Request permission

Figure 4

Patients (A, B and C) show successful treatment of SA treatment from baseline to 12 months follow-up. Patient (D) shows no effect of SA treatment up to 12 months follow-up. NA, not available; SA, somatostatin analogue.

Results

In total, 28 RP patients (52 eyes, males n=11 (39%)) with CM were included. The median age at treatment initiation was 39 years (IQR 30–53). Median follow-up was 12 months (range 6–12). The genetic cause of RP was known in 16 cases (57.1%). Causative mutations in the USH2A gene (n=4, 14.3%) were most prevalent. All demographic data are presented in table 1.

27 patients had previously been treated with oral CAI for CM; 9 of these patients had also used topical CAI. One patient previously received only topical CAI at a young age (<18 years). Reasons for starting or switching to SA are shown in figure 1. Figure 2 shows the distribution of the SA treatment regimens. Most patients had been prescribed Octreotide LAR 20 mg/month (28.6%) or 30 mg/month (50%). 9 patients continued using CAI after initiation with SA in a lower dose. Another 3 patients started additional CAI, due to a small rest of CM 3 months post-SA treatment initiation. SA treatment was terminated in 4 patients within 12 months: 3 due to no effect and one due to intolerable gastrointestinal side effects, including nausea.

FT and FV were reduced in 25 (89%) patients, with 31 (60%) of the eyes showing a decrease of >20%. At baseline, the mean FT was 409±136 µm and mean FV was 0.31±0.10 mm³. At 12 months, the mean FT decreased to 334±119 µm; and the mean FV to 0.25±0.04 mm³ (figure 3A,B). Linear mixed model analysis showed a significant decrease of log FT and log FV between baseline and 3, 6 and 12 months after starting SA treatment (p<0.001, p<0.001, p<0.001). The different dosages of octreotide LAR and lanreotide, the simultaneous use of CAI and genotype did not significantly affect log FT and log FV after 12 months. The largest decrease of FT and FV was observed after 3 months. Mean BCVA was 0.46±0.35 logMAR at baseline, and 0.45±0.39 logMAR after 12 months, with no significant improvement (figure 3C). Figure 4 illustrates the course of CM on OCT images of 4 patients during 1-year follow-up.

In 3 patients, an FA was performed to obtain information about the underlying mechanism of the CM. All three patients showed no signs of retinal vasculitis, ischaemic vascular regions or leakage of the optic disc. In the late phase of the FA, one out of the three patients showed leakage distributed radially in the Henle’s layer forming the classic petalloid leakage pattern, indicating clinically significant macular oedema.

Discussion

This study evaluated SA treatment for CM in 28 RP patients (52 eyes). In 60% of the eyes, SA treatment effectively decreased the FT and FV (>20%). Notably, SA treatment was not the primary treatment choice for all patients, highlighting its effectiveness in addressing persistent CM. The largest decrease of FT and FV occurred within the first 3 months, implying that these months should be critical for the evaluation of treatment response. After 3 months, the effect was more subtle but a decrease in FT could still be observed after 12 months. Despite substantial structural improvement with notable reductions in FV and FT, functional enhancement was not significant, likely due to the chronicity of CM in most patients. Long-standing or frequently recurring CM has limited potential for visual improvement due to either macular ischaemia, structural retinal damage change or a combination of both.16 Other factors like the development of cataracts or cone function loss associated with RP could also influence BCVA.6 However, the lack of visual improvement should not encourage clinicians to defer from prompt therapy, as this increases the chance of stability of functional vision in long term.

SAs have been safely used for treatments of neuroendocrine tumours and acromegaly for years.17–19 A dilemma is that SA can cause side effects, such as asymptomatic cholelithiasis in 5%–55%.20 The most commonly reported, but mostly transient side effects include gastrointestinal disturbances such as diarrhoea, abdominal pain, nausea and vomiting and hyperglycaemia.21 While cholecystectomy need is rare (<1%), it is crucial to discuss this risk pretreatment.20 Patients should be screened by an internist for predisposing symptoms such as obesity, diabetes, high familial risk or pre-existing gallstones and regular monitoring during SA treatment is advised.

CM is a common complication in RP patients (15%–55%), yet its exact cause remains unclear.2 3 22 The aetiology of CM can be due to a generalised retinal vasculitis, and/or purely leakage caused by local inflammation due to photoreceptor cell apoptosis, and/or impaired RPE cell function. Our observation, including FA in three patients, showed no signs of generalised vasculitis, but merely clinically significant central macula oedema in the late phase. Another study using OCT-angiography to visualise vascular changes in RP and CM, did not observe vascular disruption either.23 We assume, therefore, that CM in RP is a local leakage process without vascular involvement, and therefore, differs from cystoid macular oedema seen in inflammatory eye diseases. The emergence of CM in RP patients might also be linked to the increased oxygen levels in the retinal vasculature compared with controls.24 Due to the degeneration of rod-photoreceptors, oxygen uptake decreases, leading to higher levels of oxygen in the tissue. These higher oxygen levels can generate intracellular oxygen-free radicals, potentially inducing CM and damaging surviving cells, primarily cones, causing cone death.25 26 Furthermore, the metabolic function is more affected in RP patients with CM, compared with RP without CM. The role of this mechanism needs further investigation.24 25 This information is important for clinicians considering therapy for CM in RP as different therapy strategies may be necessary.27

Somatostatin is a cyclic neuroendocrine peptide that acts as an inhibitor of hormonal release, which is used for the treatment of acromegaly and other active hormone producing neuroendocrine tumours.17 Various modes of action of somatostatin in the eye are described. First, somatostatin has a positive effect on apical to basal direction-oriented fluid transport and may cause a rebalance of fluid ion transport at the RPE level. Second, there may be a direct anti-inflammatory effect on the RPE. The somatostatin receptors are upregulated on activated blood vessels and suppression may help restore the inner blood–retinal barrier. Third, somatostatin has a function in the immune modulation and angiogenesis.9 17 Another target of somatostatins that can be speculated is the upregulation of NRF2. NRF2, a transcription factor that is controlling the oxidative stress response, has been demonstrated to protect cone photoreceptors and is slowing down vision decline in mouse models of retinal degeneration. NRF2 activates multiple defences against oxidative stress. Moreover, overexpression of NRF2 in the RPE maintains RPE structure and survival in retinal degeneration mice.28 These properties substantiate the somatostatin receptor as a potential therapeutic target in the treatment of CM. In the last years, the use of generic SA increased, which resulted in an decrease of costs and made this drug more accessible.29

In the clinical management of RP, prediction of who responds to therapy is valuable for decision-making. When focusing on causal genes, USH2A and RHO, were most prevalent in our study. These genes were also highly associated with CM.30 Our analysis found no difference in FV/FT reduction among genotypes. However, the groups were too small for firm conclusions. Second, the dosage of SA did not influence the amount of decrease in FT/FV over time. This may indicate that a higher dose does not always ensure a better effect, and that the effective dose can vary per patient and is also contingent on individual-specific factors. A randomised study from 2010 already showed efficacy of octreotide in patients with postsurgical CME, with similar dosage as used in this study.31 Additionally, the concurrent use of CAI did not significantly affect FT/FV reduction. Our clinical observations suggest that, for certain patients, oral CAI alone lacks efficacy in treating CM. Initiating SA while continuing oral CAI may produce effects attributed to SA or their synergistic combination.

Apart from the small sample, the limiting factors of this observational, retrospective case series include the use of different dosages of SA, the combination of therapies and the heterogeneity of genotypes. Future prospective studies, preferably randomised clinical trials, will be needed to apply SA in a very standardised fashion to draw more profound conclusions.

In summary, this study explored the effect of SA in RP patients with CM, reporting a significant decrease in FT and FV without serious side effects within 12 months. Given the progressive loss of peripheral visual field in RP patients and their reliance on central vision, maintaining an intact macular architecture is imperative. Therefore, an early switch to SA treatment may prevent permanent photoreceptor damage if this is not achieved.