Fig. 2. SST analogs such as octreotide act via paracrine and autocrine effects on retinal endothelial cells and RPE cells. SST or an SST analog binds to the SSTR on endothelial cells and inhibits endothelial cell proliferation stimulated by growth factors like VEGF and IGF-1. RPE cells play a crucial role in the regulation of outer retinal homeostasis. Systemic IGF-1 can stimulate IGF-1 receptors present on RPE cells to express VEGF. VEGF then can stimulate retinal endothelial cells via VEGFR1 or VEGFR2. SST or its analogs can block the activation of IGF-1 receptor in RPE cells resulting in a decrease in VEGF expression, which leads to less VEGF-induced endothelial proliferation. This is one of the mechanisms of SST analog inhibition of angiogenesis.

CLINICAL EVIDENCE FOR SST AS A THERAPEUTIC FOR PDR

In vitro and in vivo studies have confirmed that SST analogs are potent inhibitors of GH and IGF-1. Octreotide was found to reduce elevated levels of GH and IGF-1. Octreotide showed a positive effect on DR in several small, controlled trials and case reports.

In a study of 18 patients with persistent PDR with vitreous hemorrhage after laser treatment, a significantly reduced incidence of vitreous hemorrhages and number of vitrectomies was observed in the group treated with octreotide (41). The dose used was 300 g per day in three divided doses. In the treated group of nine patients, 78% showed an improvement in contrast to the control group. In the octreotide group visual acuity was stable, whereas it significantly decreased in the control group. Neovascularization decreased in 85% of the patients in the treated group and was stable in 15%, and in the control group neovascularization increased in 42% and was unchanged in 58%.

We studied the effect of octreotide in type 1 and 2 diabetics with pre-proliferative and early proliferative DR. In the treated group, significantly fewer patients developed high risk characteristics. In only 1 of the 22 eyes was laser treatment required in contrast to 9 of 24 eyes in the control group. An important limitation in our study was the use of the maximally tolerated dose of the drug. Moreover, patients were simultaneously treated with thyroid hormone. This was justified based on octreotide’s ability to suppress TSH with prolonged use and on the enhancement of SSTR expression by thyroid hormone (42).

Recently, two Phase III, multi-center, double-masked, placebo-controlled study trials that included both type 1 and 2 diabetic patients who were ETDRS stages 47–61 were completed. The patients were treated with the long-acting octreotide (Sandostatin LAR, Novartis), which was injected intramuscularly once a month. The studies were initiated in 1999 and completed in 2006; when combined, these studies represent the largest investigation to date in moderate to severe NPDR to low-risk PDR patients. At completion, study 802 evaluated 585 patients at 61 European sites and study 804 evaluated 313 patients at 36 sites in North America and Brazil. The patients received randomized therapy for on average 4 years, with some patients being treated up to 6 years. For these studies, patients were randomized to receive either Sandostatin LAR at 20 or 30 mg in study 802 and 30 mg or placebo in study 804. Ophthalmologic assessments included visual acuity measurements and semiquantitative, stereoscopic, seven-field, color, 30° ETDRS fundus photography. The Wisconsin Central Reading Center graded the fundus photographs according to ETDRS criteria. The primary outcome was DR progression (octreotide vs. placebo) as defined by the ETDRS retinopathy severity scale for one or two eyes. Key secondary outcomes included change in overall visual acuity, which was defined as time to loss of ≥15 letters on the ETDRS visual acuity scale between baseline and follow-up visits. Octreotide has shown efficacy as a treatment for refractory cystoid macular edema and therefore macular edema was an end point in the study (43). Thus macular edema (changes between baseline and follow-up visits) was a secondary outcome.

Similar mean age, gender ratios, body mass index (BMI), and blood pressure characteristics were observed in both the 802 and 804 studies but there was a greater proportion of Caucasians in the European study. In both studies, similar proportions (roughly 75%) of patients had Type 2 diabetes, similar percentage of patients had ≥10-year duration of diabetes, and similar number of patients used insulin for glycemic control. Approximately 60% of patients had DR of ≤ 5-year duration since detection, and almost 10% had nephropathy. Slightly lower proportion of patients in Europe had hypertension (56%) or neuropathy (21%) compared with those in study 804 (69% and 42%, respectively). Similar distribution of retinopathy severity was seen in study 802 vs. 804 as defined by the ETDRS severity scale: 20–25% of patients were already at low-risk PDR at study entry. Similar distribution of ETDRS-rated visual acuity was observed. Most patients scored within or above 70–84 letters (almost 80% overall) and approximately 20% scored within or below 55–69 letters. The primary endpoint was to determine the efficacy of octreotide in delaying time to progression of retinopathy, and this was by ≥ 3 steps on the ETDRS severity scale or by ≥2 steps on the ETDRS severity scale for individual eyes. The secondary end points were to determine the efficacy of octreotide in delaying time to the development

or progression of edema or to the loss of ≥ 15 letters on the ETDRS visual acuity scale. In study 804, the patients received either octreotide 30 mg intramuscularly every 4 weeks or placebo every 4 weeks. In the 804 study, IGF-1 levels in the serum were significantly reduced, with the percent change from baseline being approximately 20%, which remained consistent for the 160 weeks it was examined.

In the study 804, the time to progression of retinopathy was delayed with a p-value = 0.0430 over the 304 weeks of the study. There was no effect of octreotide on the time to development or progression of macular edema (p = 0.8751). Loss of ≥15 letters on the ETDRS visual acuity scale was delayed in the octreotide-treated patients but did not reach statistical significance (p = 0.1054). In contrast, the result of the study 802 was not as encouraging because the primary end point, being time to progression of retinopathy, was not achieved. Interestingly, however, IGF-1 levels were not suppressed during this study, suggesting that the systemic endocrine effects of octreotide may indeed be important to having this drug achieve the optimal results. In study 802, there was a trend for improvement in visual acuity; however, it did not achieve statistical significance.

From an endocrine perspective, octreotide treatment resulted in a reduction of the blood glucose level in patients treated with insulin and they required lower insulin doses. Typically, insulin doses have to be reduced by 25–50% in patients with octreotide treatment. Close daily monitoring of blood glucose levels is mandatory under octreotide treatment because of the risk of hypoglycemia. The side-effect profile was similar to that observed in other large, long-term studies. Diarrhea and tenesmus are common at the beginning of octreotide treatment but rapidly improve. Nausea and vomiting are less common. Hypothyroidism due to TSH suppression and gallstones are additional side effects.

POTENTIAL REASONS FOR MIXED SUCCESS IN CLINICAL TRIALS

The cumulative results suggest that the clinical therapeutic effect of octreotide in DR may be due to both an endocrine effect and a direct effect on SSTR in ocular target tissues. High doses of octreotide required for clinical efficacy in PDR and ultimately other neovascular ocular diseases are likely because of inadequate penetration of the BRB by this peptide drug after systemic administration. Activation of one or more of the other ocular tissue target SSTRs for which octreotide has much lower potency (such as SSTR3) may be as important as SSTR2 activation. The activation of native SSTR2 receptors on endothelial cells inhibits growth factor-stimulated proliferation by a signaling mechanism that is fundamentally less efficient than in other cell types, resulting in higher concentrations being needed. In these antiproliferative studies, IC50 values are 2–3 orders of magnitude greater than those observed with GH release from pituitary cells, typical target cells of SSTR analogs. Moreover, pharmacokinetic studies have not been conducted to determine the relative distribution of octreotide or other SST peptide analogs in the retina or other ocular tissue compartments after systemic administration.

The expression of SSTRs in diverse ocular cells and in endothelial cell types from various beds may differ, and studies have rarely been performed with rigorous quantitation. To date, these studies have not been performed with human retinal endothelial cells (HRECs). Watson et al. reported that SSTR2 receptors were expressed at higher

458 Ljubimov et al.

levels in proliferating relative to quiescent human umbilical vascular endothelial cells (HUVEC). However, reverse transcriptase polymerase chain reaction (RT-PCR) studies were conducted only using probes for the SSTR2 subtype. Furthermore, the antiangiogenic effect has been attributed to SSTR2 activation based on the activity of octreotide as an SSTR2-selective agonist in the endothelial cells and in vitro vascular tissue model systems (44). Octreotide inhibits proliferation of HRECs (31), bovine choriocapillary endothelial cells (BCECs) (45), and HUVECs (44), and has antiangiogenic activity in the chick chorioallantoic membrane (CAM) (44) and human placental vein (HVPM) models (46). However, octreotide also has affinity for SSTR5 and SSTR3 with selectivity of 1 and 2 orders of magnitude higher, respectively, in cloned receptor binding studies (47). This is in sharp contrast with the nanomolar potency of octreotide both in SSTR2 binding affinity and in SSTR2-mediated functional assay, such as the antisecretory effect (e.g., on GH release) in neuroendocrine cells both in vitro and in vivo. We have shown, using SSTR-selective agonists in HREC, that SSTR 3- and SSTR 2-selective agonists had dramatic antiproliferative effects (47). Furthermore, this is highly relevant, as human eye specimens showed expression of SSTR2 in CNV lesions (48).

In vivo studies proved that SST analogs are good therapies for proliferative conditions of the eye and are tolerated with little toxicity even when administered by intravitreal injection (47). Octreotide reproducibly inhibited neovascularization in vivo in many different systems (44, 49–51).

We showed using the oxygen-induced retinopathy (OIR) mouse model and the laser rupture of Bruch’s membrane CNV model that small non-peptide molecules mimic octreotide’s effects. These selective SSTR3 and SSTR2 agonists are less expensive to produce, have efficiencies comparable to octreotide, and are specific for SSTR2 or 3.

This work presents a rationale for further clinical studies of these drugs. Moreover, trials of DR therapies must pay close attention to both the progression and severity of DR, as well as appropriate targeting of stage(s) for intervention, assessment of relevant outcomes, observation over a sufficiently long time period, and adequate sample size.

FUTURE DIRECTION: SST ANALOGS IN COMBINATION THERAPY

There is a continued need to add new pharmacological treatment modalities in order to improve the management of neovascular diseases, both as novel monotherapies and combination therapies. In recent years, there has been a burst in relevant studies. The most interesting examples of new drugs against ocular neovascularization are anti-VEGF therapeutics, bevacizumab (Avastin), ranibizumab (Lucentis) (both from Genentech), and pegaptanib (Macugen) (OSI-Pfizer), for treatment of the wet form of AMD (52, 53). The success of these drugs for AMD may be related to the major role of VEGF in the development of neovascularization in this disorder. Avastin has been also tried for PDR; the most pronounced effect, however, was a decrease of neovascular leakage, again consistent with the role of VEGF (54, 55).

Currently, the major challenge in DR treatment is to be able to stop the progression to vision-threatening PDR; data to this effect on clinical use of octreotide have been discussed above. When PDR still develops, an effective anti-angiogenic therapy substituting or complementing panretinal photocoagulation is badly needed. It should be noted that PDR development could be dependent on more factors than just VEGF

(56–58). It was also shown that retinal endothelial cells respond much more strongly to growth factor combinations than any single factor (59, 60). In a complex, apparently multifactorial, disease as exemplified by DR and PDR, some potentially successful drug monotherapies may rely on targeting master regulators of the angiogenic process, such as HIF-1α or protein kinase CK2 (29).

Another powerful approach is to develop efficient drug combinations. This principle is the mainstream of drug therapy for cancer and AIDS (61–63), but until very recently it was not considered for PDR. In 2004, we pioneered this approach using a mouse OIR model of retinal neovascularization (57). This model is widely used to test antiangiogenic drugs because diabetic animal models, with very rare exceptions (64), fail to reproduce human PDR with preretinal neovascularization unless genetically manipulated (65). In our experiments, protein kinase CK2 inhibitors, emodin or tetrabromobenzotriazole (TBB), were administered alone or in combination with octreotide. Each compound was able to significantly reduce preretinal neovascularization in mouse pups with systemic administration (57, 66). Combination therapy produced an additive effect. Moreover, using only 1 mg kg−1 per day octreotide in combination with a CK2 inhibitor, it was possible to achieve the same degree of inhibition of neovascularization as with 5 mg kg−1 per day octreotide alone (Fig. 3). Since emodin is a component of some laxatives and is known to be essentially nontoxic, (57) these experiments pave the way to clinical trials using its combination with octreotide for inhibiting DR progression.

Fig. 3. Combination therapy with octreotide. Counts of preretinal nuclei as a measure of neovascularization in various groups of mice are shown. Intraperitoneal treatment with 30 mg kg−1 per day emodin reduced retinal neovascularization by about 57%, and with 30 mg kg−1 per day TBB, by 46%. Treatment with 5 mg kg−1 per day octreotide yielded about 67% reduction, and with 1 mg kg−1 per day octreotide, about 50% reduction. Emodin combined with 1 mg kg−1 per day octreotide reduced neovascularization by 69%, and TBB combined with 1mg kg−1 per day octreotide, by 61%. Ten sections per eye from each mouse were counted. Five mice were used per each group in three independent experiments. Vehicle represents emodin solvent since octreotide solvent was just PBS. Bars = mean

± SD. *p < 0.001 vs. vehicle; **p < 0.05 vs. single drug or vehicle. (Reprinted from (66) with permission from the American Society for Investigative Pathology.)