Flurbiprofen : A non-selective cyclooxygenase (COX) inhibitor for treatment of non-infectious, non-necrotising anterior scleritis

Abstract

Objective

To analyse the safety and efficacy of a non-selective cyclo-oxygenase (COX) inhibitor in the management of non-infectious, non-necrotising anterior scleritis.

Methods

Retrospective chart review of 126 patients with non-necrotising anterior scleritis treated with oral flurbiprofen (Froben®(Abbott Healthcare)) with ( group B, n=61) or without topical steroids (group A, n=65) was performed and time to remission was plotted.

Results

The observed incidence rate was 1.07 (95% CI: 0.57–1.99) per 1000 person-years with failure rate of 0.68 (95% CI: 0.22–2.12) per 1000 person-years in group A and 1.41 (95% CI: 0.67–2.96) per 1000 person-years in group B. The failure rate was 3.97(1.89–9.34) per 1000 person-years with hazard ratio of 10.01 ( 95% CI: 2.52–39.65; p<0.001) for patients with associated systemic disease.

Conclusion

To our best knowledge, this is the first and largest case series on the safety and efficacy of a non-selective COX inhibitor in the management of anterior scleritis.

Introduction

Scleritis is a painful inflammation of the sclera and can be potentially sight threatening.¹ It can be non-granulomatous or granulomatous when associated with conditions such as sarcoidosis or tuberculosis.^1,2 There is a variable degree of scleral edema and deep episcleral vascular plexus inflammation, which leads to violaceous hue of the sclera with inflammatory signs such as rubor (redness), tumour (swelling) and dolor (pain).^1,3–5 Pain and redness represent the main reasons for presentation for a consultation.⁶

Non infective scleritis can be divided into anterior or posterior scleritis. Anterior scleritis can be further classified into necrotising or non-necrotising types, and non-necrotising scleritis can be further sub-classified into nodular or diffuse type, based on the morphology of the scleritis.¹ 25–50% of patients with scleritis can be associated with underlying systemic autoimmune diseases such as rheumatoid arthritis, systemic lupus erythematosus, relapsing polychondritis, Wegner’s granulomatosus or polyarteritis nodosa.^7–10 Most commonly necrotising type of scleritis has been reported to have associated systemic disease and is more aggressive in nature.¹⁰ There is very limited data on association of systemic autoimmune disease with non necrotisng diffuse anterior scleritis. Besides being a potential manifestation of an underlying severe, life threatening systemic disease, scleritis can by itself lead to severe ocular morbidity and affect quality of life due to its associated pain and redness.^{2–4,11–14}

Optimal control of the inflammation in the sclera is essential to alleviate pain and prevent further worsening of disease which can eventually lead to complications such as cataract, glaucoma and exudative retinal detachment.^2,11,13,15 Topical steroids can be effective in milder forms of inflammation such as episcleritis but are usually not helpful in severe inflammation. In contrast to posterior or necrotising scleritis, non-necrotising diffuse scleritis may be mild and may not require oral corticosteroids or immunosuppression as the first line of therapy. As a clinical equipoise, oral non-steroidal anti-inflammatory drugs (NSAID) plus/minus adjuvant local therapy can be the initial therapeutic regimen. Failure of the NSAIDs will justify the use of stronger anti-inflammatory drugs such as oral corticosteroids, immunosuppressive drugs, or eventually biologics agents.¹⁵

Amongst the non-steroidal anti-inflammatory drugs, there are several different options and the clinicians’ decision on one agent will likely depend on their own experience. At our institution, we have been using flurbiprofen (Froben®, Abbott Healthcare) a non-selective cyclo-oxygenase (COX) inhibitor, ¹⁶ for over 20 years. This study is on our retrospective cohort of 126 patients with non-infectious diffuse or nodular, non-necrotising anterior scleritis. To the best of our knowledge and as per literature review; this is the first ever reported study on the use of this agent in scleritis.

Methodology

Institutional Board Review approval was obtained for this retrospective cohort study. All research complied with the tenets of the declaration of Helsinki. The review of clinical case records using the electronic database system and medical case notes was performed on the patients with non-infective, non-necrotising scleritis seen at a tertiary referral eye care centre in the UK. For further statistical analysis, patients were divided in two subgroups: group A included the patients who were treated with oral flurbiprofen only and group B included those who were placed on oral flurbiprofen and topical corticosteroid drops. Statistical analysis was also performed according to systemic status involvement.

The following features were recorded for each patient: visual acuity, age, gender, eye involved, associated systemic disease, type of scleritis, details of first line of therapy regime which was oral flurbiprofen (150 mg or 300mg) with or without topical corticosteroids eye drops, time to relapse or disease remission after a course of flurbiprofen, side effects associated with flurbiprofen and duration of therapy, details about subsequent additional therapy, and disease related ocular complications (corneal involvement, development of anterior uveitis, cataract surgery, and progression to posterior scleritis and number of recurrent episodes). Visual acuity at initial and final follow up was recorded for all the patients and compared for two subgroups and those with and without systemic disease involvement. Clinical resolution of scleritis with oral flurbiprofen with or without topical steroids was the primary outcome measure, and the number of days till this outcome with our first line of therapy- oral flurbiprofen was analysed in our study. The alogrithm adopted by us in treatment for patients with non necrotising diffuse anterior scleritis was oral flurbiprofen for mild cases, addition of topical steroids or only flurbiprofen for mild – moderate cases and if not responding to treatment or in severe cases oral corticosteroids was instituted. In patients recalcitrant to oral corticosteroids or with side effects secondary to steroids, systemic immunosuppressive therapy was added.

Statistical Analysis

The study population was divided in Group A (receiving oral flurbiprofen only) and Group B (oral flurbiprofen with topical corticosteroids). STATA 13.0 software was used for descriptive and analytical statistics. Scatter plot was plotted between log of initial and final visual acuity for all the patients, patients with and without systemic disease and for patients in two subgroups. Failure event for statistical purpose was defined as switching to second line of agent; which was either in the form of immunosuppressive agent or oral corticosteroids. Using time to switch to oral corticosteroid therapy after initiation of flurbiprofen therapy as the intervention, Kaplan Meier survival curve was plotted for all the patients, between two sub groups and for systemic disease association. Person-time at risk of failure was calculated from the point at which oral flurbiprofen therapy was initiated. Person-time at risk of failure ended with the point of initiating alternative second line of therapy and the point where persons taking oral flurbiprofen had resolution of disease and exited therapy (for those patients who did not had failure but had clinical resolution). Qualitative variables were expressed as precentages and quantitative variables were expressed as mean values ± standard deviation (SD) if they followed a normal distribution or as median values (range) if they did not. Association between two groups was tested either using χ² test or with Fischer’s exact test where necessary. Differences between mean were tested using independent samples t-test or Mann-Whitney test. Relapse risk was summarized as the incidence per person year (analyzed by patient); a 95% confidence interval (CI95) was generated assuming a Poisson distribution. Kaplan-Meier curves were created to evaluate failure to achieve to remission for scleritis using our proposed therapy in either group A or group B and also for patients with or without systemic disease. Bivariate analysis of time to relapse was performed through the study of Kaplan-Meier curves and analysis with the Breslow test. Multivariate analysis was performed with Cox proportional hazards models, after confirming non-violation of proportional risks assumption through the analysis of Schoenfeld residuals, log-log graph and predicted survival. Cox proportional hazard ratio were performed to establish the level of significance.

Results

The case records of 126 patients seen over a period of five years with non-infective, non-necrotising scleritis were analysed. Based on the treatment received, there were 65 patients in group A (oral flurbiprofen only) and 61 patients in group B (oral flurbiprofen with topical steroids). Demographics of all the patients and for both subgroups are presented in Table 1. Diffuse type of non-necrotising scleritis (Figure 1A, 1B, 1C) was the most common presentation in both groups (80% in group A and 73.77% in group B). The rationale for using a NSAID as the first line was based on the less aggressive nature of this type of scleritis and the smaller risk of associated visual loss. The starting level of flurbiprofen, either 50mg or 100mg three times a day, was based on the severity of the presentation, clinical findings and the level of pain. Patients who received oral corticosteroids or other types of oral NSAIDs as their first line of therapy were not included in this study.

Table 1.

Demographics and clinical characteristics of all the patients, and subgroups based on only systemic flurbiprofen (150mg or 300mg) and systemic flurbiprofen (150mg or 300mg) with topical steroids

	Total ( n= 126) (Oral Flurbiprofen)	Group A (n=65) (Oral Flurbiprofen)	Group B (n=61) (Oral Flurbiprofen + topical steroids)	P value
Age (Mean (SD))	53.17(15.02)	53.00 (16.53)	53.36 (13.35)	0.9649
Gender (M:F)	43:83 (65.87% females)	24:41 (63.08 % females)	19:42 (68.85 % females)	0.531
Right Eye	56 (44.44%)	32 (49.23%)	24 (39.34%)	0.309
Left Eye	58 (46.03%)	29 (44.62%)	29 (47.54%)
Both Eyes	12 (9.52%)	4 (6.15%)	8 (13.11%)
Diffuse Scleritis	97(76.98%)	52 (80%)	45(73.77)	0.406
Nodular Scleritis	29(23.02%)	13 (20%)	16(26.23%)
Systemic involvement	18(14.29%)	10 (18.46%).	8 (9.84%)	0.716

Figure 1.

Figure 1A and 1B: Slit lamp photographs of the non-necrotising diffuse scleritis

There were 10 patients in group A with systemic association. There were 6 (9.23%) patients with rheumatoid arthritis, 1 (1.54%) patient with systemic lupus erythematosus, 4 (6.15%) patients with arthritis, and 1 (1.54%) patient with other connective tissue disease disorder. Group B had eight patients with systemic association. Each patient had one of the following etiology: Takaysu’s arteritis, systemic vasculitis, relapsing polychondritis, ankylosing spondylitis, arthritis, and other connective tissue disorder.

The characteristics of the patients in Groups A and B are shown in Table 1. Ultrasound B-scan of the globe using orbital mode was done in 21 out of 65 patients (32.31%) in group A and 36 out of 61 (59.01%) patients in group B with the intention of excluding posterior scleritis. There was no evidence of posterior scleritis in any of the patients in our study subset.

Oral flurbiprofen was stopped after disease remission in 62(95.38%) patients in group A and 54(88.52%) patients in group B. The average length of time that patients were on oral flurbiprofen only or oral flurbiprofen with slow tapering course of topical steroids until either disease remission or switching to one of the agents in second line of therapy was 74.07±68.40 days(IQR: 14–390). Average number of days patients were on oral flubriprofen only (group A) was 67.47±59.92 days (IQR:14–265) and on oral flurbiprofen with topical steroids (group B) for 81.09±76.29 (IQR:14–390). Three patients from group A (4.61%) and 7 patients from group B (11.47%) required corticosteroid therapy or immunosuppressive therapy due to non-response or recurrent episodes of scleritis ( Fisher’s exact test: p=1.000).

Table 2 presents the visual acuity for patients in both groups. Scatter plot for pre- and post-treatment visual acuity shows a linear correlation (Fig 2A) amongst all 126 patients under this alternate form of therapy and group-wise comparison for both group A and B show equally good linear correlation with 95% confidence interval as plotted on the scatter plot (Fig 2B). One patient in each group developed cataract (1.54% and 1.64% respectively). Two patients in each group (3.08% and 3.28% respectively) had an episode of anterior uveitis during the disease course. Two patients in group B (3.58%) developed some corneal complications such as corneal epithelial abnormalities and dry eyes.

Table 2.

Visual Acuity pre-treatment and post-treatment in group A and group B

Visual Acuity	Group A		Group B
	Pre Treatment	Post Treatment	Pre Treatment	Post Treatment
6/6	26(40%)	39(60%)	35(57.38%)	40(65.57%)
6/9	20(30.77%)	15(23.08%)	12(19.67%)	15(24.59%)
6/12	9(13.85%)	5(7.69%)	5(8.20%)	1(1.64%)
6/18	4(6.15%)	3(4.62%)	4(6.56%)	1 (1.64%)
6/24	3(4.62%)	1(1.54%)	0	1 (1.64%)
6/36	3(4.62%)	0	0	1 (1.64%)
6/60	2(3.08%)	2(3.08%)	3(4.92%)	0
HM	1(1.54%)	0	1(1.64%)	2(3.28%)
CF	0	0	1(1.64%)	0
PL	0	0	0	0

Figure 2.

Figure 2A: Scatter Plot of log values of Initial (Pre-treatment) visual acuity and Final (Post-treatment) visual acuity combined for both groups

Figure 2B: Comparative scatter plot for both groups A and group B and scleritis patients with and without systemic disease

The time to stop the oral flurbiprofen before switching over to another therapy was plotted using Kaplan Meier curve for the entire cohort of 126 patients (Fig. 3A), for both the groups (Fig. 3B) and also for patients with underlying systemic disease (Fig. 3C). The failure event was defined as the initiation of alternate second line agent due to non-response to oral flurbiprofen or combination of oral flurbiprofen with topical steroids. We also computed survival rate per person year for both the groups. The observed incidence (failure) rate for 126 patients was 1.07 (95% CI: 0.57 to 1.99) per 1000 person-years with failure rate of 0.68 (955 CI: 0.22–2.12) per 1000 person-years in the treatment group A for three failures and incidence rate of 1.41 (95% CI: 0.67–2.96) per 1000 person-years with 7 failures in the treatment group B. However, the failure rate was 3.97(1.89–9.34) per 1000 person-years for patients with associated systemic disease as against failure rate of 0.39 (0.12–1.22) per 1000 person-years for patients without underlying systemic disease (p <0.001). The cox-proportional hazard ratio was 1.35 (95% CI: 0.66–2.75; p=0.39) for patients in group A. Patients with underlying systemic disease had a hazard ratio of 10.01 ( 95% CI: 2.52–39.65; p<0.001). On follow up, flurbiprofen was hence not found to be effective in anterior diffuse non necrotising scleritis with associated systemic disease.

Figure 3.

Figure 3A: Overall kaplan Meir Survial graph

Figure 3B: Kaplan-Meier survival estimates for group A and group B

Figure 3C: Kaplan-Meier survival estimates by systemic disease association

The major side effects noted in group A & B were gastrointestinal intolerance found in 7 (10.77%) and 5(8.20%) patients, respectively. One patient each had nausea and headache in group A, and there was one patient each with nausea, headache, and skin rash in group B.

Discussion

Non-infectious, non-necrotising anterior scleritis is a painful inflammatory condition of the sclera, and it often challenges the specialists for optimal control of pain and inflammation. Aim of therapy in patients with scleritis is to control inflammation, preserve vision, alleviate pain, maintain quality of life, and prevent disease and therapy related complications.⁶ There has been a significant change in options for both local and systemic therapy for scleritis. Rational approach in scleritis management will be tailored treatment based on the type, severity of the scleritis, associated systemic diseases, the patients age and gender and the preference for the array of anti-inflammatory agents available to the treating specialist.

Non-necrotizing anterior scleritis can be treated with a variety of local and systemic therapy,^17,18 including topical corticosteroids or NSAIDs, oral NSAIDS¹⁹ or corticosteroids, and sub-conjunctival²⁰ or orbital floor steroid injections.²¹ In some recalcitrant cases or in patients with need for prolonged corticosteroid therapy, immunosuppressive agents such as methotrexate or mycophenolate mofetil may be required.^22,23 Very severe cases may prompt physicians to use biologics such as Anti TNF-α inhibitors, or other biologics such as Rituximab, for optimal control of inflammation.^{2,6,17,18,22–27}

Many different NSAIDs have been used worldwide for initial control of pain and inflammation, and each physician may have a different experience or preference for particular agents. NSAIDs have been shown to be successful in controlling inflammation in 92% of cases as the first line therapy.¹⁷ Unfortunately, no randomized superiority or non-inferiority control trial has been published about the safety and the efficacy of any specific anti-inflammatory agent in patients with non-infective, non-necrotising anterior scleritis. The current literature and preferred practice patterns are based on small case series or anecdotal case reports. Hence, it may be necessary to try different options before finding the appropriate drug that will optimize the outcome for each individual patient. In this report, we have attempted to analyse the safety and the efficacy of flurbiprofen in patients with non-infectious, non-necrotizing diffuse anterior scleritis.

Flurbiprofen, a propionic acid derivative, is a NSAID with anti-inflammatory and anti-pyretic activity.¹⁶ It has been widely used in the treatment of inflammation and pain in patients with rheumatoid arthritis, osteoarthritis and ankylosing spondylitis. In addition, the use of this oral agent has been extended to treat mild to moderate pain in patients with dysmenorrhoea, bursitis, tendonitis and soft tissue trauma. Topical form of this drug is often used preoperatively in cataract surgery to prevent miosis during surgery.

Per its classification, flurbiprofen is one of the non-selective COX inhibitors blocking both COX-1 and COX-2 pathway.¹⁶ Flurbriprofen leads to non-selective, reversible inhibition of cyclo-oxygenase (COX), which is responsible for prostaglandin synthesis pathway. It is claimed to be one of the most potent non-selective COX inhibitors²⁸ and leads to a decrease in availability of prostaglandins at the site of inflammation. Flurbiprofen is rapid and completely absorbed after oral administration reaching the peak plasma concentration in 30 to 240 minutes.¹⁶ Renal elimination is the second most important pathway of metabolism after liver, therefore due caution has to be exercised in patients with altered kidney function. Although relatively safe, the most common side effects reported with prolonged use of this agent are nausea, vomiting, abdominal pain, gastrointestinal discomfort and bleeding, peptic ulcer, abnormal liver function test and prolonged bleeding time. Additionally, it can lead to allergic reaction, with pruritus, rash and tinnitus.¹⁶

Topical prednisolone acetate or dexamethasone has been used as first line therapy in mild cases of anterior scleritis by numerous clinicians and researchers. One study reported a success rate of 47% with isolated topical therapy alone after two months of treatment without additional need for oral NSAIDs or corticosteroids.¹⁹ However, most clinicians agree that topical therapy on its own is not very effective and may not lead to complete disease resolution.^10,29 In our case series, we have used topical corticosteroid along with oral NSAIDs in one of the subgroup of patients and observed good response, however it did not showed improved efficacy or added to the overall effect. Topical steroids may play a role in helping in pain control or intermittent flare up of scleritis, but it does not seem to affect course of the disease independently. Based on failure rate per 1000 person years, there were only few patients requiring further alternative therapy and hence both the treatment approaches using oral flurbiprofen with or without topical steroids were quite successful. However, for the patients with underlying systemic disease this therapy may not work very efficiently. Our Cox-proportional hazard model revealed that, although not statistically significant, patients on combination therapy of oral flurbiprofen and topical steroids might have slightly prolonged time to failure against those on oral flurbiprofen only. Also, based on the cox – poroportional hazard analysis, we have demonstrated that after the initial use of disease modifying agents like flurbiprofen, the reminiscent disease activity can be managed on topical steroids.

In a recently published study by Sainz de la Maza et al, the authors have documented the effective role of oral NSAIDs in 144 out of 392 patients with scleritis. ¹⁷ These patients were diagnosed with idiopathic, non-necrotising, diffuse scleritis with low degree of inflammation with or without ocular complications.¹⁷ When prescribing NSAID, it is important to realize that a full analgesic effect may be obtained within days to a week of the start of the use of the drug, whereas an anti-inflammatory response may not be achieved for up to three weeks.

There are numerous reports on use of subconjunctival steroids in patients with diffuse non-necrotising scleritis. Several authors have demonstrated the safety and efficacy of this form of local therapy.^30–33 The multicenter prospective study about the safety and the efficacy of subconjunctival steroid injection had shown resolution of scleritis in 89.7% of cases. ²⁰ The effect was sustained for two years in 50% of the patients without need of additional oral anti-inflammatory therapy. Although there were no reported cases of scleral perforation or melt, 20% patients were found to have raised intraocular pressure.²⁰ Elevation in intraocular pressure (IOP) and local scleral necrosis are the two major potential side effects prohibiting the use of this agent as first line stand alone therapy in patients with non-necrotising, non-infective scleritis.

Oral corticosteroid is the current first line therapy in patients with severe non-necrotising and necrotising scleritis. It is also used as second line therapy in patients non-responsive to oral NSAIDs.^17,29,34 However, oral corticosteroids have their own myriad of side effects, and long term, oral corticosteroids therapy can be detrimental to the quality of life of patients with scleritis.³⁴ Precaution should be exercised while co-administering oral corticosteroids and oral NSAIDs as it has high potential for gastric ulceration.³⁵

Bauer et al have reported their experience of using selective Cox-2 inhibitor (Celecoxib) in 24 patients with diffuse anterior scleritis.³⁶ Twenty-two patients in their group showed a significant improvement without any major side effects. One of the major advantages with selective COX-2 inhibitor over non-selective ones is the lesser incidence of gastrointestinal side effects. Our study of 126 patients using a non-selective COX inhibitor had similar results with minimal side effects. Both of these agents should be further explored for their safety and efficacy in future prospective multicenter randomized control trials for management of non-infectious, non-necrotizing scleritis.

To the best of our knowledge & literature review, this is the first ever reported and the largest case series (but Sainz de la Maza published on 144 cases)of use of this specific agent (non-selective COX inhibitor) as oral non-steroidal anti-inflammatory agent in the management of non-necrotizing, anterior scleritis. Similar case series though has been reported by Bauer et al but that is using selective COX inhibitor.³⁶ However, from both our study using non-specific COX inhibitor and study by Bauer et al using specific COX inhibitor³⁶ and also from large series of use of oral NSAIDs in scleritis by Sainz de la Maza et al¹⁷ it is quite evident that non-steroidal agents can be quite effectively used in selected cases of diffuse non-necrotizing scleritis hence obviating the need for corticosteroids or immunosuppressives.

The results from this series suggest that oral flurbiprofen may be a safe treatment for selected subgroups of scleritis patients. We have demonstrated prolonged time to failure using this alternate form of therapy while obviating the need for oral corticosteroids in large proportion of patients. Failure to respond should alert to the possibility of an underlying disease since these patients are less likely to respond as demonstrated by our data. Along with the disease severity, individual factors such as the patient’s tolerance and response must be analyzed, and a tailored therapy should be offered to the patients with non-necrotizing anterior scleritis instead of a doctrined pre-meditated approach. Ideally, the agent of choice should be the one that clinicians are familiar with and that will provide a good indication of success or failure. Larger multicenter studies need to be conducted to further analyze the efficacy of both selective and non-selective COX inhibitors in the management of different severity grades of scleritis.