Intravitreal pegaptanib sodium (Macugen®) for treatment of diabetic macular oedema: a morphologic and functional study

Abstract

AIMS

To study whether morphologic (foveal thickness, FT) variations of clinically significant macular oedema (CMO) in patients suffering from diabetes following intravitreal pegaptanib sodium (IVP) injection were associated with functional [macular sensitivity (MS) and colour discrimination (CD)] changes.

METHODS

A longitudinal, interventional, non-randomized study was performed. FT was assessed by optical coherence tomography (OCT), MS by microperimetry, best-corrected visual acuity (BCVA) by early treatment diabetic retinopathy study charts (ETDRS) and CD by Farnswoth-Munsell test. The treatment protocol consisted of three consecutive injections (0.3 mg/0.05 ml; baseline, week 6 and week 12). Follow-up checks were scheduled at 18, 24, 36 and 48 weeks, after injections.

RESULTS

Thirty eyes of 30 patients with clinically significant CMO were included for analysis. After IVP a significant decrease of FT occurred with a mean reduction from baseline of 56.9% (P= 0.0001). An improvement of functional parameters was recorded in all patients (BCVA from 18.2 ± 8.5 letters to 25.5 ± 8.4 letters, P < 0.005, MS from 8.6 ± 2.16 dB to 10.6 ± 2.61 dB, P < 0.001, colour analysis from 376.1 ± 125.6 TES to 116 ± 34.6 TES, P= 0.0001). A statistically significant correlation between FT and BCVA as well as MS and CD was also found. Neither ocular nor systemic adverse events were reported.

CONCLUSIONS

Intravitreal pegaptanib significantly reduced FT, with a concomitant improvement of MS and CD. This association emphasizes the efficacy of IVP in the treatment of CMO.

WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT

• Diabetic macular oedema can develop at all stages of diabetic retinopathy (DR), and its prevalence increases with the duration of diabetes, affecting approximately 30% of diabetic patients after 25–30 years of disease.

• Elevated concentrations of VEGF have been identified in aqueous humour and vitreous of patients with DR, and the severity of macular oedema is correlated with its vitreous concentrations.

• Intravitreal anti-VEGFs exert a key role in the treatment of diabetic macular oedema.

WHAT THIS STUDY ADDS

• The present study, performed on 30 eyes of 30 patients, shows that intravitreal pegaptanib injections induced a significant reduction in mean central retinal thickness, with a parallel improvement in visual acuity.

• For the first time a correlation between retinal morphologic and functional parameters in diabetic patients with macular oedema has been ascertained.

• Our study demonstrates that the selective inhibition of VEGF-165 isoform by pegaptanib represents an effective treatment for diabetic macular oedema, as assessed throughout a 48 week follow-up.

Introduction

Diabetic retinopathy (DR), one of the most frequent diabetic microvascular complications, affects 30%–50% of all diabetic patients and represents the main cause of legal blindness in 20–74-year-old persons in developed countries [1]. Clinically significant macular oedema (CMO) can develop at all stages of DR and its prevalence increases with the duration of diabetes, affecting approximately 30% of diabetic patients after 25–30 years of disease [2]. Hyperglycaemia appears to play a key role in the aetiology of DR although recently focus has been directed to the molecular basis of the disease and several biochemical factors other than hyperglycaemia have been considered [3]. These mechanisms act by affecting cellular metabolites and then inducing release of cytokines [4]. Among these, vascular endothelial growth factor (VEGF) is the most representative and its role in angiogenesis and microvascular permeability is well known. As a pro-inflammatory mediator, VEGF may contribute to the development of CMO by causing expression of cellular adhesion molecules and chemotaxia of leucocytes as well as increasing liquid flux from the vasculature to the tissue compartment [5]. Moreover, elevated concentrations of VEGF have been identified in aqueous humour and vitreous of patients with DR and the severity of macular oedema is correlated with the concentration of VEGF in vitreous of CMO patients [6].

There are at least nine different VEGF isoforms, but some of them (e.g. VEGF-A165) play a more critical role in the pathogenesis of ocular disease [7]. VEGF-165 is the predominant pathologic isomer in DR, thus selective blocking of VEGF-165 may exert therapeutic effects on the natural course of CMO. Pegaptanib sodium is a ribonucleic acid aptamer that specifically targets the VEGF-165 isoform and is currently approved in a number of countries worldwide for the treatment of neovascular age-related macular degeneration (AMD) [8]. Clinical evidence suggest a direct effect of pegaptanib on retinal neovascularization in patients with diabetes mellitus [9] and more recently, even on CMO [10].

There is evidence that macular oedema is associated with reduced retinal sensitivity [11]. In this prospective study we evaluated the safety and efficacy of intravitreal pegaptanib (IVP) in the treatment of CMO. We have also examined whether a correlation exists between the degree of macular oedema and the results of functional tests.

Methods

This longitudinal, interventional study was conducted between April 2010 and June 2011 at the Department of Ophthalmology of the University of Naples, Italy. Each patient provided informed consent according to the guidelines of the Declaration of Helsinki and the institutional review board approved the study protocol. Eligible subjects were ≥18-years-old of either gender affected by type 2 diabetes. Inclusion criteria were (i) best-corrected standard ETDRS visual acuity (VA) letter score ≥10 (10/160 or better) and ≤45 (10/32 or worse), (ii) retinal thickening due to CMO involving the centre of the macula with corresponding leakage from microaneurysms evidenced on fluorescein angiography, (iii) optical coherence tomography (OCT) foveal thickness (FT) ≥300 µm and (iv) no history of previous treatment for CMO at any time. Exclusion criteria were (i) untreated high risk proliferative DR, (ii) history of panretinal or focal photocoagulation, (iii) vitreoretinal traction within 1 disc diameter of the fovea confirmed either clinically or on OCT, (iv) CMO previously treated with triamcinolone and other VEGF therapies and (v) atrophy/fibrosis or hard exudates involving the centre of the macula that would preclude improvement in BCVA and (vi) pregnancy.

Before study inclusion, each patient underwent a baseline eye check including BCVA measured at 2 m with standard ETDRS charts, biomicroscopy and fundus examination, applanation tonometry, fluorescein angiography and evaluation of FT by OCT (OCT 3 Stratus; Carl Zeiss, Dublin, CA). FT was defined as the distance between the vitreoretinal interface and the retinal pigment epithelium in the centre of the fovea. Macular sensitivity (MS) and colour discrimination (CD) were respectively tested using microperimetry (MP-1, Nidek Co Ltd, Gamagori, Japan) and the Farnsworth-Munsell 100-hue test. The MP test uses light stimuli which are randomly presented during the examination, as in standard static perimetry and results are reported in decibels (dB). The fixation target and stimuli are projected onto the retina by a liquid crystal colour monitor completely controlled by dedicated software. The location and stability of fixation were classified and evaluated separately. Location was classified as predominantly central, poorly central and predominantly eccentric. Fixation stability was classified as stable, relatively unstable and unstable.

The Farnsworth-Munsell 100 hue test was administered only to the affected eye, under room illumination. The test total error scores were calculated using web-based scoring software designed by Torok (http://www.torok.info/colorvision/fm100.htm).

Because several conditions may influence MS and CD, patients with moderate to dense lens opacity, implanted intraocular lenses, corneal opacities, a history of refractive surgery, glaucoma or ocular hypertension, history of intraocular inflammation such as anterior or posterior uveitis, multifocal choroiditis, a history of retinal detachment, a history of ocular trauma and optic neuropathy were also excluded.

The treatment protocol consisted of three consecutive injections (baseline, week 6 and week 12). The need for re-treatment after this loading dose was determined by the investigator based on protocol-specific criteria, including ≥300 µm in central FT and BCVA decrease of >5 letters. Week 24 was the primary end point of the study. The primary study objectives were to demonstrate efficacy of IVP in the reduction of CMO, assessed by OCT, whereas, the secondary outcome was to verify the occurrence of functional improvements secondary to IVP administration, analyzing BCVA, MP and FM modifications throughout the follow-up. Follow-up checks were scheduled at the following intervals: 18, 24, 36 and 48 weeks, with the option of additional intermediate visits depending on response. All parameters (BCVA, biomicroscopy, fundus, tonometry, FT, FA, MS and CD) were recorded at each check. Safety was assessed by the incidence of ocular and non-ocular adverse events and serious adverse events, including those potentially related to VEGF inhibition.

The secondary outcome was to verify the occurrence of functional improvements secondary to IVP administration, analyzing BCVA, MP and FM modifications.

Intravitreal injections (Macugen®, Pfizer Limited, Sandwich, UK; 0.3 mg/0.05 ml) were performed by retinal specialists.

Statistical analyses

Due to the nature of the main outcome measures of this clinical investigation (i.e. post IVB modifications in BCVA, FT and CD), the sample size calculation, accomplished for the amount of study population (30 cases), provided a value constantly higher than 90%, using the PASS 97 statistical program (NCSS, Inc., Kaysville, UT). Data analysis was performed with STATA (STATA/IC 11.0). Since homogeneity was found between mean and median values for continuous variables, we concluded that the values were normally distributed in the sample. anova test with Bonferroni correction was used to compare baseline vs. subsequent check values. To verify the relationship between variables we analyzed their correlation and the linear model regression. FT was chosen as an independent variable, whereas BCVA, MS and CD as dependent variables. A value of P < 0.05 was considered significant.

Results

Baseline characteristics of enrolled patients are summarized in Table 1. The study included 30 eyes with significant CMO (FT > 300 µm). All 30 eyes completed the follow-up.

Table 1.

Demographics and clinical characteristics of patients with diabetic macular oedema at study entry

Characteristics	n
Number of patients	30
Gender
Male	12
Female	18
Mean age±SD (years)	62.6 ± 6.9
Median age (years)	63
Type of diabetes	Type II
Mean BCVA (letters±SD)	18.2 ± 8.5
Foveal thickness (µm) (mean±SD)	551.5 ± 129.8
Macular sensitivity (dB) (mean±SD)	8.6 ± 2.16
Colour discrimination (mean±SD)	376.1 ± 125.6

The primary outcome was to verify whether IVP induced morphological changes assessed by OCT analysis. At the first check (18 weeks) a significant decrease of FT in comparison with baseline from 551.5 ± 129.8 µm to 246.4 ± 45.8 µm was recorded (−55%, P < 0.001). At 24 and 48 weeks FT measured 269.4 ± 66.2 µm and 237.4 µm ± 41.1 respectively. The mean reduction from baseline was 56.9% (P= 0.0001) (Table 2).

Table 2.

Clinical characteristics of 30 eyes before and after IVP

	FT (µm)	BCVA (letters)	MS (dB)	CD (TES)
Baseline	551.5 ± 129.8	18.2 ± 8.5	8.6 ± 2.16	376.1 ± 125.6
18 weeks	246.4 ± 45.8°	21.2 ± 8.7**	9.5 ± 2.09*	233.2 ± 48.0**
24 weeks	269.4 ± 66.2***	21.4 ± 9.1***	9.6 ± 2.61**	229.6 ± 42.8***
36 weeks	243.6 ± 43.8°°	24.9 ± 9.3***	10.6 ± 2.49**	116.5 ± 40.2°°
48 weeks	237.4 ± 41.1°°	25.5 ± 8.4°	10.6 ± 2.61**	116.0 ± 34.6°°

*P < 0.05; **P < 0.01; ***P= 0.005; °P < 0.001; °°P < 0.0001 (anova test with Bonferroni correction). FT, foveal thickness; BCVA, best corrected visual acuity; MS, macular sensitivity; CD, colour discrimination.

Mean BCVA at baseline was 18.2 ± 8.5 letters. At 18 weeks a significant improvement to 21.2 ± 8.7 letters was recorded (P < 0.005). At the last check a further significant increase of BCVA from baseline was documented (25.5 ± 8.4 letters; +29%, P < 0.005).

Mean MS determined with the MP-1 at baseline was 8.6 ± 2.16 dB and 18 weeks later a mean value of 9.5 ± 2.09 dB was recorded (P < 0.005). At the end of follow-up a further improvement to 10.6 ± 2.61 dB was documented (+19%; P < 0.001) (Table 2).

Colour analysis gave the following results, expressed as total error scores (TES): baseline 376.1 ± 125.6, 18 weeks later a recovery to 233.2 ± 48.0 was registered (P= 0.05) and at the 48 week check a value of about 116 was achieved (−30%; P= 0.0001) (Table 2). Lastly, at baseline eight patients presented unstable fixation, 20 relatively unstable fixation and just two patients had stable fixation, whereas location was poorly central for all of them. At the end of follow-up all patients presented stable and predominantly central fixation (Table 3).

Table 3.

Relation between fixation characteristics and BCVA

		Fixation
		Stability			Location
Week	VA	S	RU	U	Pr.C	Po.C	Pr. EC
0	10/160			8		8
	<10/160 and >10/80		14			14
	<10/80 and >10/32	2	6			8
	10/32
18	10/160		1			1
	<10/160 and >10/80	17			17
	<10/80 and >10/32	12			12
	10/32
24	10/160		5			5
	<10/160 and >10/80	11			11
	<10/80 and >10/32	9			9
	10/32	5			5
36	<10/160 and >10/80	3			3
	<10/80 and >10/32	19			19
	10/32	8			8
48	<10/80 and >10/32	21			21
	10/32	9			9

S, stable; RU, relatively unstable; U, unstable; Pr.C, predominantly central; Po.C, poorly central; Pr.EC, predominantly eccentric.

To verify the relationship between variables we analyzed their correlation with linear model regression. Linear regression between FT and BCVA as well as MS and CD was calculated. FT was chosen as independent variable, whereas BCVA, MS and CD represented the dependent variables. As shown in Figure 1, a statistically significant correlation was found between FT and BCVA, CD and MS, in detail, P= 0.01, r= 0.24; linear regression: Y = 34.16 − 0.03X; P= 0.01, r= 0.22; linear regression model: Y = 48.27 + 0.38X and P= 0.049, r= 0.12; linear regression model: Y = 12.071 − 0.0057X, respectively.

Figure 1.

Linear regression between FT µm and A) BCVA, B) MS and C) FM

No patient required additional IVP up to the 24th week. After 36 weeks five patients underwent additional IVP. On average, study participants received 3.3 pegaptanib sodium intraocular injections each. All 30 patients completed the study. During the study, neither severe decrease of vision immediately after the injection nor systemic adverse events were reported. Ocular and non-ocular adverse events are summarized in Table 4.

Table 4.

Incidence of ocular and non-ocular adverse events (AE) over the follow-up period (48 weeks)

Ocular AEs	Number of patients (%)
Eye pain	12 (40.0)
Ocular hyperaemia	14 (46.6)
Conjunctival haemorrhage	4 (13.3)
Retinal haemorrhage	1 (3.3)
Increased intraocular pressare	1 (3.3)
Reduced visual acuity	1 (3.3)
Non-ocular AEs
Hypertension	1 (3.3)
Nausea/Vomitino	1 (3.3)

Discussion

Angiogenesis plays a critical role in the development of diabetic complications, particularly those complications that involve the eye. DR begins as vascular occlusion and ischaemia and may result in macular oedema. VEGF plays a key role in the aetiology of CMO and its blockade is effective in the treatment of this complex pathology. Commercially available drugs allow for a total or specific isoform block of VEGF depending on the agent used: bevacizumab, ranibizumab or pegaptanib. All these compounds inhibit VEGF angiogenic activity binding to VEGF protein, and thus preventing its receptor activation/interaction. Selective blockade has the advantage of less potential risk, since the physiological functions that are VEGF-mediated are preserved. A frequently employed compound is bevacizumab, a full length humanized antibody that binds to all subtypes of VEGF, approved by the US Food and Drug Administration for the treatment of metastatic colorectal cancer [12]. Recent reports have suggested that bevacizumab may be useful for the treatment of choroidal neovascularization (CNV), CMO, DR and macular oedema associated with retinal venous occlusive diseases [12, 13]. Despite the fact that bevacizumab has not been licensed for intraocular use and its systemic safety has not been completely proved, many physicians have been encouraged to use this drug mainly for its low cost. However current safety data for bevacizumab are incomplete and not yet robust [14, 15]. On the contrary, the efficacy and safety of pegaptanib and ranibizumab in the treatment of CMO is being confirmed in clinical trials with these agents [5, 10, 16] and emerging data also confirm that these two compounds have a more safe ocular and systemic adverse event profile [15, 16]. In addition, both pegaptanib and ranibizumab are approved for intravitreal administration in the treatment of AMD, whereas bevacizumab is used off-label in all ocular diseases, including DR.

In the present clinical trial pegaptanib intravitreal administration induced a significant reduction in mean central FT (P= 0.0001), with a parallel BCVA improvement (P < 0.005). At 24 weeks five out of 30 eyes (16.6%) presented with a FT > 300 µm and a further intraocular injection of pegaptanib sodium was needed. Within these five patients three presented a stable BCVA compared with the 18 week check, whereas two had a slight reduction in BCVA. Consequently, a repeated treatment improved the final BCVA and reduced FT. The findings presented are in keeping with those of other recently published trials evaluating the use of intravitreal anti-VEGF therapies as treatment for CMO, i.e. the READ-2 [17] and the DRCRN [18] studies and 12 month reports of the BOLT [19] and RESOLVE [20] studies.

The functional impact of CMO is generally evaluated by recording BCVA. Microperimetry better defines the function of the macula and provides detailed information about the degree and pattern of macular alteration in CMO [21], with a good OCT correlation [22]. Recently, Deák et al. [23], comparing OCT results with retinal sensitivity assessed with microperimetry, have found a correlation between retinal morphologic alterations and retinal functionality. Microperimetry is able to quantify MS and fixation in an exact, fundus-related fashion, thus adding detailed information about the degree and pattern of macular function alteration. In our series a statistically significant correlation between MS and FT (P= 0.049, r= 0.12; linear regression model: Y = 12.071 − 0.0057X) was documented (Figure 1). Following IVP injection, reduction of macular thickness was accompanied by a significant improvement of retinal sensitivity. A stabilization of fixation was also achieved. In fact, at the end of follow-up, all patients presented stable and predominantly central fixation (Table 3).

Diabetics have poorer colour vision than the general non-diabetic population and, among diabetics, patients with clinically detectable retinopathy have poorer colour vision than those with normal appearing fundi. The colour discrimination defect is even more pronounced in patients with CMO [24]. In fact, macular oedema reduces light transmission to photoreceptors. This might affect the blue rather than the red-green mechanism, as a result of the lower density and number of blue cones in the human fovea. Therefore, colour vision can improve as a result of resolved CMO. In fact, in the presence of macular oedema, visual acuity may be reasonable, whereas the colour discrimination is significantly affected, as in our study population. Thus, the use of visual acuity alone could overlook the presence of or underestimate the functional effects of CMO, whereas visual acuity combined with colour vision and microperimetry would provide a more comprehensive evaluation of retinal involvement. In our series a significant correlation between BCVA and FT (P= 0.01, r= 0.24; linear regression: Y = 34.16 − 0.03X), CD and FT (P= 0.01, r= 0.22; linear regression: Y = 48.27 + 0.38X) was recorded confirming a reduced risk of visual loss in all treated eyes (Figure 1).

Although our present findings supply additional rationale for the use of pegaptanib in CMO, the study shows some limitations. First, the sample investigated is small, second, the short time of follow-up does not provide for a long term pegaptanib safety and efficacy evaluation and third, the lack of a control group reduces the significance of these findings, even if the possibility that the changes observed are due to the natural history of the disease is very remote. However we prospectively evaluated the relationship between morphologic and functional modifications secondary to CMO demonstrating that IVP represents an effective treatment for CMO.

Ocular and systemic adverse effects of pegaptanib are potentially minimized compared with those of the other VEGF-A blockers, such as ranibizumab and bevacizumab, and this is particularly important in diabetic patients who may have an underlying increased risk for systemic vascular events such as stroke and myocardial infarction.

Competing Interests

There are no competing interests to declare.