Real-World Experience in the Use of Immunosuppression for the Management of Inflammatory Eye Disease

ABSTRACT

Purpose

Patients with sight-threatening inflammatory eye disease (IED) are maintained on systemic immunosuppression whilst in long-term clinical remission. There are no clear guidelines on the duration of remission before implementing treatment withdrawal. We present a real-world analysis on the use of immunosuppression in IED in long-term remission and consider strategies for withdrawal.

Methods

Adult IED patients on systemic immunosuppression were categorised into four disease groups: Corneal Transplant Survival Strategies (CTSS), Ocular Surface Disease (OSD), Non-infectious Uveitis (NIU) and Scleritis. Patients with Behçet’s disease were excluded. Data on systemic immunosuppressants and biologics used; duration of treatment; reasons for drug discontinuation; disease activity/remission status; duration of clinical remission with an emphasis on patients who had been in remission for a minimum of 24 months were captured.

Results

Out of a total of 303 IED patients, 128 were on systemic immunosuppression with a clinical remission of their ocular disease for ≥24 months. The median duration of remission was 4–5 years with the longest duration of remission 22 years, and some patients on immunosuppression for up to 23 years. Sixty patients stopped at least one immunosuppressive agent without prior discussion with a health-care practitioner.

Conclusion

Progressive conditions, such as cicatrising conjunctivitis may require lifelong immunosuppression, but patients with NIU and Scleritis and those on CTSS, immunosuppression withdrawal should be considered if they remain in remission for 2 years. Any patient stopping a medication should be contacted immediately for counselling. These data will better inform patients, encourage adherence and aide formal guideline development.

KEYWORDS:

• Corneal transplant
• immunosuppression
• pemphigoid
• scleritis
• uveitis

Inflammatory eye disease (IED) comprises a wide range of potentially blinding conditions that include ocular surface disease (OSD),^1,2 uveitis^3–5 and scleritis.⁶ According to the severity of the condition, patients may require systemic immunosuppression to initially suppress the inflammation, and then to establish long-term control of the inflammatory process, i.e. clinical remission. In many IED patients, these drugs are effective in controlling the inflammation,^1,2,6–10 but have potentially severe unwanted side effects⁷ with associated increased probability with longer treatment duration.¹¹ It is important to recognize that early reduction or discontinuation of treatment could result in a relapse of disease.^12–15

Although there are consensus guidelines on the use of systemic immunosuppression in ocular inflammatory disorders including non-infectious uveitis (NIU),^16–19 there are no formal guidelines for treatment reduction and withdrawal.²⁰ In NIU, this decision is made on an individual patient basis. Minimal recommendations exist on when to discontinue therapy and state that any decision to withdraw systemic immunosuppression in patients that have achieved remission involves a classic risk–benefit analysis and should be based on shared decision-making that includes patient preference, tolerance of and risk resulting from the current treatment, duration of disease control, and the specific cause of uveitis. This considers the risk that the inflammation will recur and the benefits resulting from the convenience of having life-style restrictions lifted by not being on systemic immunosuppression, such as shielding during the SARS-Cov-2 pandemic, regular blood monitoring, polypharmacy from adverse event prophylaxis (bone protection, opportunistic infections).¹⁸

It has also been proposed that NIU patients on non-alkylating agent immunosuppressive drugs should be treated for 24 months after discontinuing oral corticosteroids, while maintaining inactive disease; after that, an attempt can be made to taper and discontinue treatment.¹² Similarly, in a cohort of Behçet disease patients of whom 76% had ocular involvement, the authors suggested discontinuation of anti-TNF treatment, using a step-down approach with close monitoring, should be attempted after approximately two years of complete remission.²¹ The Adalimumab in Juvenile Idiopathic Arthritis-associated Uveitis Trial (ADJUST) expects to deliver information on clinical outcomes to guide clinicians in their decision-making regarding discontinuation of adalimumab;²² and the Treatment exit Options For non-infectious Uveitis (TOFU) registry will offer recommendations on systemic immunosuppressive treatment exit strategies.²³ Both these studies will provide in part, evidence for the build of formal guidelines.

For patients with OSD, there are consensus management guidelines for starting systemic immunosuppression, but these are for conditions with an associated systemic disease, such as Sjogren’s syndrome,²⁴ and ocular mucous membrane pemphigoid (OcMMP)^10,16 and there is very little information on when to reduce and discontinue therapy, although 24-months of clinical remission has been proposed in OcMMP.²⁵ In OcMMP, a stepladder approach to treatment has been recommended.^1,2

In the three National Behçet Centres of Excellence in England²⁶ there are consensus guidelines stating that anti-TNF alpha drugs, but not conventional immunosuppressants, be discontinued in patients with occlusive retinal vasculitis who are in clinical remission for 24 months. Importantly, some patients may not necessarily be on treatment for their uveitis, but for other systemic manifestations of the disease. Similar scenarios may occur in patients with NIU, scleritis and OSD who also have an underlying systemic disease who may either still require systemic immunosuppression for their systemic disease even if their ocular disease has gone into remission, or alternatively may require immunosuppression for the ocular disease whilst their systemic disease is in remission and inflammatory markers have normalized. There are American College of Rheumatology guidelines²⁷ and European League Against Rheumatism (EULAR) recommendations²⁸ on the discontinuation/withdrawal of systemic immunosuppression in rheumatoid arthritis once remission has been achieved, but it is unlikely that these could be transferable to IED patients.

In our clinics, patients are appropriately counselled before starting treatment and blood monitoring is put in place. Patients are evaluated at each clinic visit regarding disease activity, but our impression is, due to a lack of formal guidelines, treatment is rarely discontinued. We may be continuing therapy in some patients that could have previously been withdrawn. As we do not have robust randomized controlled trials evidence on treatment duration, we are unable to tell patients when they start treatment how long they are likely to be on these drugs.

In this study, we present a real-world analysis of the use of long-term systemic immunosuppression in adult patients with sight threatening IED. We wished to establish how long these patients are maintained on systemic immunosuppression whilst in long-term clinical remission. Our primary aim was to determine if there is a cohort of patients in “long-term” remission (≥24 months) where the immunosuppressive treatment could have been reduced and discontinued. These data will allow us to better inform our patients, limit potential drug side effects, reduce drug costs and may assist in the development of formal guidelines of when it might be safe to withdraw treatment.

Methods

We undertook a retrospective, longitudinal cohort study that was registered as a clinical effectiveness service evaluation (#1789) at Sandwell and West Birmingham (SWB) NHS Trust, as IRB/Ethics Committee ruled that approval was not required for this study. The described study adhered to the tenets of the Declaration of Helsinki. All adult (≥18 years) IED patients currently, or had previously been, on systemic immunosuppression, i.e. any immunosuppressant and/or biologic drug and were currently attending the Centre for Inflammatory Eye Disease, Birmingham and Midland Eye Centre, a major UK tertiary regional and supra-regional referral Eye Unit, were included. Details of these patients and their immunosuppressive drugs they are taking are held on a password protected Excel (Microsoft) spreadsheet held on a secure National Health Service server. All patients with Behçet’s disease were excluded as they are seen at the multidisciplinary Birmingham Behçet National Centre of Excellence and data on these patients is held separately.

As the Adult IED Service sees patients with various types of ocular surface and intraocular inflammatory conditions, it was appropriate (and for ease of data analysis) to categorize IED patients on systemic immunosuppression into four distinct groups: Ocular Surface Disease (OSD), Corneal Transplant Survival Strategies (CTSS i.e. systemic immunosuppression required for “high-risk” corneal transplants), NIU and Scleritis (non-infectious).

The following information was captured on all patients: basic demographic data (Ethnic Group, England and Wales: Census 2021), diagnosis, indication for starting therapy, the systemic immunosuppressants and biologics used, duration of treatment, reasons for drug discontinuation, if treatment was still required for any underlying systemic disease, and whether the ocular disease was in clinical remission and if so, for how long, paying specific attention to patients who had been in remission for a minimum of 24 months at the time of data collection.

Clinical remission i.e. no active disease was defined as:

• OSD: Absence of signs of activity and/or increasing damage²⁹ as outlined by Mathewson et al. and/or an Ocular Surface Disease Index (OSDI)³⁰ total score of 0–22, i.e. none to mild where applicable.

• CTSS: Absence of the recognized signs of corneal allograft rejection, including corneal edema, keratic precipitates on the graft, corneal neovascularization, subepithelial corneal infiltrates, an epithelial rejection line, corneal stromal infiltrates, an endothelial rejection (Khodadoust) line, anterior chamber (AC) cells.

• NIU: Anterior segment: no new anterior chamber inflammatory activity, i.e. AC cells ≥1 that required an increase or commencement of topical corticosteroid. Posterior segment: no evidence of new vitreous haze ≥1+, and/or retinal vasculitis, and/or active chorioretinal lesions, and/or cystoid macular edema, and/or papillitis – corroborated on imaging where applicable.

• Scleritis: White and pain-free eye(s).

At each clinic visit, patients were seen either by a Consultant experienced in the management of IED, a Fellow who had completed their 7-year UK or equivalent ophthalmology training, an experienced Senior Trainee (Resident), or a Junior Trainee. Any patient where it was felt the disease had flared was discussed with the Consultant and depending on the grade of doctor was also seen by the Consultant. All patients seen by the Junior Trainee were discussed with, and often also seen by, the Consultant. Regular drug monitoring was organized by our Immunosuppression Nurse Practitioner for every patient, either through the hospital or coordinated with their General Practitioner. Repeat prescriptions of drugs was invariably undertaken by the Immunosuppression Nurse Practitioner as the patients’ General Practitioners usually declined.

Data were collected from each patient electronic record on the Clinical Data Archive (SWB NHS Trust), from Winscribe™ electronic patient letters (Nuance Communications), the Medisoft™ ophthalmology electronic medical record (Medisoft Limited), and Unity electronic health record system (Oracle Cerner) to ensure maximum data collection. Some patients were under combined management with a physician, e.g. rheumatology, gastrointestinal medicine, respiratory, dermatology, and hematology. Data collection continued until July 2022.

Study data were collected and managed using REDCap electronic data capture tools hosted at the University of Birmingham, UK.³¹ REDCap (Research Electronic Data Capture, Vanderbilt, TN) is a secure, web-based software platform designed to support data capture for research studies, providing 1) an intuitive interface for validated data capture; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for data integration and interoperability with external sources.

Statistical analyses

Descriptive statistical analysis was undertaken using REDCap and data exported to an Excel spreadsheet. Additional statistical analyses were performed using SPSS v27 (IBM, New York City, NY). Continuous variables were described using median and interquartile range (IQR). A two-sample t test was used to compare age and a chi-squared test with Yates’ correction for gender, ethnic and patient groups (Fisher’s exact test was used for samples <5). For the Index of Multiple Deprivation scores, the Mann–Whitney U-test was used. A statistical significance of <0.05 was used for all tests.

Results

A total of 303 IED patients were identified from the Excel spreadsheet. Of these, 82 patients were found to be no longer on treatment. There were 133/221 patients on systemic immunosuppression currently attending whose ocular disease was deemed to be in clinical remission for ≥24 months. Of these 133 patients, 5 still required ongoing treatment for an underlying associated systemic disease, leaving 128 patients for analysis. Of the remaining 88 patients on treatment, 20 were not in remission of their ocular disease, 9 in remission <6 months, 13 in remission 6–11 months, 37 in remission 12–17 months, and 9 in remission 18–23 months.

Characteristics of these 128 patients is shown in Table 1 with most patients being in the NIU (62/128, 48%) and OSD (48/128, 38%) groups. There were more females than males in both these groups. In each group, the majority of patients were of White ethnic group. The NIU group had the youngest mean age (47 (range 16–75) years), and the OSD group the oldest (70 (range 30–91) years). There was a wide range in duration of clinical remission and time on treatment. The median time in clinical remission for all patients was 58 months. The Scleritis and OSD groups having the longest duration of remission recorded for any given patient at 220 and 269 months (≡ 18 and 22 years), respectively, but the median duration of remission in each group was similar at 50 (range 32–220) and 59 (range 25–269) months, respectively. The median duration of treatment for all patients was 81 months. In the NIU group, there were patients who were still on treatment after 279 months (23 years). Only 15/128 (12%) patients had the dose of an immunosuppressant reduced (10 NIU, three OSD, two Scleritis), regardless of being in clinical remission for ≥24 months.

Table 1. Characteristics of 128 IED patients on systemic immunosuppression who are in clinical remission for a minimum of 24 months.

	All patients	OSD	CTSS	NIU	Scleritis
Number of patients	128	48	9	62	9
Gender
Cismale:Cisfemale:Other*	55:72:1	20:28:0	5:4:0	25:36:1	5:4:0
Mean Age (years)	58	70	63	47	59
Range	16–91	30–91	33–81	16–75	20–80
Ethnic group
Asian/Asian British/Asian Welsh	15	3	1	10	2
Black/Black British/Black Welsh/Caribbean/African	15	1	0	11	3
Mixed/Multiple ethnic groups	2	0	0	1	0
Other ethnic group	0	0	0	0	0
White	96	44	8	40	4
Median duration on treatment (months)	81	72	63	107.5	79
Range	30–279	32–274	30–110	30–279	33–237
IQR	79	63.25	32	93.25	70
Median duration of clinical remission (months)	58	59	58	58	50
Range	24–269	25–269	26–104	24–173	32–220
IQR	44.75	31	8	53.75	42

IED: inflammatory eye disease; OSD: ocular surface disease; CTSS: corneal transplant survival strategies; NIU: non-infectious uveitis; IQR: Interquartile Range.

*: Transgender.

In the NIU group, the anatomical classification of uveitis was 17 anterior, 4 intermediate, 15 posterior and 26 pan. The underlying diagnosis of each of the 4 groups is shown in Table 2. A total of 2/48 (4%) OSD, 18/62 (29%) NIU, and 3/9 (33%) Scleritis patients were still taking prednisolone (≤7 mg). No patient in the CTSS group was on prednisolone.

Table 2. Underlying diagnosis for 128 IED patients on systemic immunosuppression who are in clinical remission for a minimum of 24 months.

OSD (n = 48)	CTSS (n = 9)	NIU (n = 62)	Scleritis (n = 9)
Ocular mucous membrane pemphigoid (36)	Keratoconus (2)	Idiopathic (30)	Idiopathic (5)
	Fuchs’ endothelial dystrophy (1)	Sarcoidosis (7)	Rheumatoid arthritis (2)
Peripheral ulcerative keratitis (5)	Iridocorneal endothelial syndrome (1)	Birdshot chorioretinopathy (7)	Sarcoidosis (2)
Stevens-Johnson syndrome (3)	Microbial keratitis (5): Herpes simplex virus (3), Acanthamoeba (1), Unknown (1)	Juvenile idiopathic arthritis (6)
Atopic keratoconjunctivitis (3)		Miscellaneous (12)
Pemphigus vulgaris (1)

IED: inflammatory eye disease; OSD: ocular surface disease; CTSS: corneal transplant survival strategies; NIU: non-infectious uveitis.

Demographic data, time on treatment, and reasons for discontinuing an immunosuppressant/biologic on the 82 patients who were no longer on systemic immunosuppression are shown in Table 3. The number of patients who had their treatment stopped because they were in disease remission was 15 (18%). There were 3/18 OSD patients, 2/8 CTSS patients, and 10/43 NIU patients. No patients in the Scleritis group had their treatment stopped because they were in disease remission. The mean duration of treatment before it was discontinued for all 15 patients was 62.5 months and for OSD 70 months (range 25–108), CTSS 50.5 months (range 38–63), and NIU 62.6 months (range 18–186). Since treatment was stopped, no patient has required it to be restarted and the mean duration of clinical remission without systemic immunosuppression for all 15 patients is 30 months and for OSD 38 months (range 8–50), CTSS 21 months (range 18–24), and NIU 33.6 months (range 6–56).

Table 3. Data on 82 patients no longer on systemic immunosuppression.

	OSD	CTSS	NIU	Scleritis
Number of patients	18	8	43	13
Gender
Male:Female	11:7	5:3	21:22	3:10
Mean Age (years)	66	62	48	47
Range	34–86	44–82	21–76	18–86
Ethnic group
Asian/Asian British/Asian Welsh	3	2	10	3
Black/Black British/Black Welsh/Caribbean/African	2	0	7	1
Mixed/Multiple ethnic groups	1	0	1	1
Other ethnic group	0	0	2	2
White	12	6	23	6
Median duration on treatment (months)	41.5	53	33	21
Range	3–126	24–97	1–198	1–40
IQR	56.25	26.25	42	27
Median duration of clinical remission (months)	44.5	54	93	43
Range	11–119	27–93	2–195	15–86
IQR	51.25	39.5	60.5	41
Reasons for discontinuing treatment (there may be more than one reason for each patient)	In remission (3), Adverse drug reaction (7), Abnormal blood result (1), Patient stopped treatment (7), Patient deceased (1)	In remission (2), Adverse drug reaction (2), Abnormal blood result (1), Patient stopped treatment (4)	In remission (10), Lack of efficacy (4), Adverse drug reaction (8), Abnormal blood result (1), Patient stopped medication (24)	Lack of efficacy (1), Adverse drug reaction (4), Abnormal blood result (1), Patient stopped medication (8), Pulse therapy only (1)

OSD: ocular surface disease; CTSS: corneal transplant survival strategies; NIU: non-infectious uveitis; IQR: Interquartile Range.

When looking at the total cohort of 303 patients, the immunosuppressants and biologics for each group are shown in Table 4. The commonest immunosuppressant and biologic used was mycophenolate mofetil and adalimumab. Table 5 shows the indications for starting systemic immunosuppression, with the commonest indication being corticosteroid sparing. The reasons for discontinuing an immunosuppressant or biologic are shown in Table 6. No patient developed a malignancy because of their treatment, but one OSD patient was diagnosed with carcinoma of the breast whilst on mycophenolate mofetil that was felt to be unrelated. We have differentiated between an adverse drug reaction and an abnormal blood result as one represents a symptomatic complication and the other a laboratory complication. Although the commonest reason for discontinuing a medication was an adverse drug reaction (n = 83), we were surprised that there were many occurrences of patients stopping a medication without prior consultation with a health-care practitioner. This occurred on 77 occasions, with 14 patients stopping more than one drug with mycophenolate mofetil being the commonest (Table 6), giving a total of 60 patients. We investigated this further to see if there were any specific characteristics that could establish why this was the case and we compared this group with the remaining 243 patients who had not stopped any medication. The main statistically significant findings were that the patients who had stopped a medication were younger, less likely to be of White ethnic group, and a greater proportion with NIU (Table 7).

Table 4. Immunosuppressants and biologics used for each disease group out of the total cohort of 303 IED patients.

Drug	OSD	CTSS	NIU	Scleritis
Immunosuppressant
MMF (n = 209)	70 (33.5%)	26 (12.4%)	101 (48.3%)	12 (5.7%)
MTX (n = 93)	11 (11.8%)	3 (3.2%)	67 (72%)	12 (12.9%)
Azathioprine (n = 78)	19 (24.4%)	2 (2.6%)	45 (57.7)	12 (15.4%)
Cyclophosphamide (n = 27)	17 (5.6%)	–	7 (2.3%)	3 (1%)
Ciclosporin (n = 23)	5 (21.7%)	4 (17.4%)	11 (47.8%)	3 (13%)
Tacrolimus (n = 22)	2 (9.1%)	18 (81.8%)	2 (9.1%)	–
Dapsone (n = 11)	11 (3.6%)	–	–	–
Sirolimus (n = 3)	–	3 (100%)	–	–
Other (n = 5)
None (biologic only) (n = 5)
Biologic
Adalimumab (n = 43)	1 (2.3%)	–	39 (90.7%)	3 (7%)
Infliximab (n = 15)	2 (13.3%)	–	11 (73.3%)	2 (13.3%)
Rituximab (n = 7)	5 (71.4%)	–	–	2 (28.6%)
Other (n = 21)*
None (n = 242)

IED: inflammatory eye disease; OSD: ocular surface disease; CTSS: corneal transplant survival strategies; NIU: non-infectious uveitis.

MMF: mycophenolate mofetil; MTX: methotrexate.

*etanercept (6), ustekinumab (2), certolizumab pegol (5), dupilumab (1), abatacept (1), tocilizumab (4), tofacitinib (1), upadacitinib (1).

Table 5. Indications for starting immunosuppressant/biologic in total cohort of 303 IED patients.

Drug	Severe disease not fully responsive to corticosteroid	Corticosteroid sparing	Side effects of another immunosuppressant	Lack of efficacy of another immunosuppressant
MMF (n = 209)	9 (4.3%)	167 (79.9%)	14 (6.7%)	19 (9.1%)
MTX (n = 93)	4 (4.4%)	72 (79.1%)	6 (6.6%)	9 (9.9%)
Azathioprine (n = 78)	2 (2.6%)	53 (67.9%)	13 (16.7%)	10 (12.8%)
Cyclophosphamide (n = 27)	1 (3.7%)	15 (55.6%)	–	11 (40.7%)
Ciclosporin (n = 23)	2 (8.7%)	15 (65.2%)	2 (8.7%)	4 (17.4%)
Tacrolimus (n = 22)	–	20 (90.9%)	–	2 (9.1%)
Dapsone (n = 11)	1 (9.1%)	10 (90.9%)	–	–
Sirolimus (n = 3)	–	3 (100%)	–	–
Adalimumab (n = 43)	1 (2.3%)	19 (44.2%)	2 (4.7%)	21 (48.8%)
Infliximab (n = 15)	–	7 (46.7%)	–	8 (53.3%)
Rituximab (n = 7)	1 (14.3%)	–	–	6 (85.7%)

IED: inflammatory eye disease; MMF: mycophenolate mofetil; MTX: methotrexate.

Table 6. Reasons for discontinuing an immunosuppressant/biologic in total cohort of IED 303 patients (there may be more than one reason for each patient).

Drug	In remission	Lack of efficacy	Adverse drug reaction	Abnormal blood result	Malignancy	Patient stopped medication
MMF	16	9	22	3	1	29
MTX	10	12	21	5	–	18
Azathioprine	10	6	22	6	–	16
Cyclophosphamide*	10	5	3	–	–	3
Ciclosporin	4	3	6	1	–	5
Tacrolimus	8	1	4	1	–	2
Dapsone	1	1	1	7	–	–
Sirolimus	1	–	–	–	–	–
Adalimumab	1	11	2	–	–	4
Infliximab	2	8	2	–	–	–
Rituximab**	3	2	–	–	–	–
Total	66	58	83	23	1	77†

IED: inflammatory eye disease; MMF: mycophenolate mofetil; MTX: methotrexate.

*Given as a pulse/treatment cycle in 9 patients.

**Given as a pulse/treatment cycle in 4 patients.

†14 patients stopped more than one drug, giving a total of 60 patients.

Table 7. Characteristics of 60 patients who stopped at least one medication at some stage during their treatment compared with 243 patients who did not stop any medication.

	60 patients who stopped treatment	243 patients who did not stop treatment	p-value
Mean Age (years)	50	56	0.02817*
Range	18–87	18–90
Gender
Cismale:Cisfemale:Other**	31:28:1	108:135:0	0.330128
Ethnic group
Asian/Asian British/Asian Welsh	16 (27%)	40 (17%)	0.101378
Black/Black British/Black Wesh/Caribbean/African	8 (13%)	25 (10%)	0.655082
Mixed/Multiple ethnic groups	3 (5%)	3 (1%)	0.0944
Other ethnic group	0	0	–
White	33 (55%)	175 (72%)	0.016895*
OSD	9 (15%)	75 (31%)	0.021596*
CTSS	4 (6%)	24 (10%)	0.6193
NIU	40 (67%)	122 (50%)	0.031976*
Scleritis	7 (12%)	22 (9%)	0.710525

OSD: ocular surface disease; CTSS: corneal transplant survival strategies; NIU: non-infectious uveitis.

*statistically significant; **: Transgender.

Using the full residential postcode for each patient, we also looked at comparing the Index of Multiple Deprivation (IMD) of patients who stopped a medication with those who did not. The IMD are widely used datasets in England to classify the relative deprivation (essentially a measure of poverty) of small areas.³² The IMD ranks every small area in England from 1 (most deprived area) to 32 844 (least deprived area), and combines information from seven domains (income deprivation, employment deprivation, education, skills and training deprivation, health deprivation and disability, crime, barriers to housing and services, living environment). These multiple components of deprivation are weighted with different strengths and compiled into a single score of deprivation. To help with this, deprivation “deciles” are calculated by ranking the 32 844 small areas in England and dividing them into 10 equal groups. These range from the most deprived 10% of small areas nationally (decile 1) to the least deprived 10% of small areas nationally (decile 10). When comparing the 60 patients who stopped a medication with the 243 patients who did not, the only difference was in deciles 5&6 (Table 8, Quintile 3) that just reached statistical significance, but as there were only 5 patients in the group that had stopped a medication it would be difficult to come to any firm conclusions on the importance of this.

Table 8. Index of Multiple Deprivation scores in 60 patients who stopped at least one medication at some stage during their treatment compared with 243 patients who did not stop any medication.

IMD Deciles and IMD Scores (Decile 1 = most deprived, Decile 10 = least deprived)	60 patients who stopped medication	243 patients who did not stop medication†	p-value
Deciles 1&2 (Quintile 1)	n = 27	n = 85	0.06676
Median Score	49.57	44.53
IQR	12.68	11.09
Deciles 3&4 (Quintile 2)	n = 14	n = 51	0.9682
Median Score	25.87	25.95
IQR	5.95	7.017
Deciles 5&6 (Quintile 3)	n = 5	n = 31	0.04408*
Median Score	20.16	16.71
IQR	2.68	2.781
Deciles 7&8 (Quintile 4)	n = 8	n = 42	0.4195
Median Score	12.48	11.94
IQR	1.27	2.79
Deciles 9&10 (Quintile 5)	n = 6	n = 31	0.578
Median Score	7.2	6.07
IQR	2.77	2.78

IMD: Index of Multiple Deprivation; IQR: Interquartile Range.

Deciles were combined for analysis as quintiles as some deciles contained less than 4 observations.

†The IMD of the postcodes of 3 patients were not found; *statistically significant.

We also looked at whether stopping a medication could be linked to the SARS-Cov-2 pandemic as patients were deemed to be at risk of more severe coronavirus infection if on systemic immunosuppression. Interestingly, 17/60 (28%) patients did stop a medication during the pandemic i.e. from March 2020 to July 2022 (the end of data collection).

Discussion

From a securely held spreadsheet of 303 adult IED patients currently or previously on systemic immunosuppression, we achieved our primary aim by identifying 128 patients in four diagnostic groups who were still on potentially toxic and costly therapies despite their condition being in clinical remission for ≥24 months. The median duration of remission was 4–5 years (depending on the condition) with the longest duration of remission being 22 years, with some patients on treatment up to 23 years. We chose to focus on a minimum of 24 months of clinical remission as that had been mentioned in previous publications.^12,18,21,25 To our knowledge, this is the first study to report this important clinical issue, particularly as no consensus guidelines exist for when to reduce and discontinue systemic immunosuppression in IED.

Of the 82 patients who were no longer on systemic immunosuppression, only 15 had it stopped because they were in disease remission. Unfortunately, the small number of patients and the long duration of time they were on treatment (mean 62.5 months) makes it difficult to identify any characteristics that could help predict in which patients it would be appropriate to consider discontinuing therapy. Nevertheless, the decision to stop treatment appeared justified as no patient has needed to restart it after a mean of 30 months.

There is a balance between reducing and discontinuing medication too quickly with the possibility of relapse,^{13–15,33,34} and prolonging treatment, perhaps unnecessarily, and the risks of unwanted side effects. Although the most commonly used immunosuppressive drugs do not seem to increase overall or cancer mortality in patients with ocular inflammation,³⁵ another report suggested a modest increase in absolute risk of malignancy.³⁶ There are numerous papers on adverse events experienced by IED patients on these medications,^37–42 and although the MUST Trial data suggest that immunosuppression for NIU can be administered relatively safely for at least 7 years⁴³ some of our patients were on treatment for at least 20 years.

There is also a potential cost-saving if treatments can be reduced and discontinued in those patients in long-term clinical remission. In England, the commonly prescribed immunosuppressants are inexpensive, e.g. generic mycophenolate mofetil 500 mg at a dose of 1 g bd costs £186 per year, with methotrexate (15 mg per week) and azathioprine (150 mg per day) cheaper at £57 and £34 per year, respectively.⁴⁴ However, the cost of a biologic is much more expensive, e.g. adalimumab is £704.28 for 2 pre-filled syringes, i.e. £8451 per year⁴⁴ (slightly less for a biosimilar) plus the additional homecare charges for managing, dispensing and delivering the medication. Then, there are the costs of blood tests for monitoring purposes and indirect costs of care (patient or caregiver reduction of work productivity because of the morbidity associated with a given disease).

We have described the widely accepted clinical signs we employed for indicating absence of disease activity, but there is heterogeneity regarding the criteria used to judge disease activity in IED (particularly NIU), with no accepted disease outcome measures^45–47 or biomarkers^48,49 that could be used to aid withdrawal of systemic therapy. The ADJUST trial²² and TOFU registry²³ may provide helpful information.

The patient voice⁵⁰ plays a vital role in identifying the outcome measures that matter to the patient through core outcome set development and reporting these outcomes through appropriate patient-reported outcome measures (PROMs). Although PROMs have been identified for OSD,^51,52 CTSS,⁵³ and NIU^54–57 (but none for Scleritis) there are no accepted core outcome sets that could not only form part of clinical trials but also used in clinical practice to guide treatment reduction and discontinuation. Furthermore, validated patient reported outcomes frequently omit important concepts that matter to patients, highlighting the need for additional PRO items for early-phase clinical trials of new health-care technologies for treating IED.^51,52

A surprising finding was that at some stage during their treatment, 60 patients stopped at least one medication. This group was younger and with less patients of White ethnic group and more with NIU than the 243 patients who did not stop a medication. There appeared no major difference in the IMD between the groups. One reason could be the cost of drug prescriptions, as younger patients may be less likely to be eligible for free prescriptions.⁵⁸ The National Health Service charges in England from April 1, 2023, are single charge: £9.65, 3-month PPC: £31.25 and a 12-month PPC: £111.60. PPCs are prescription pre-payment certificates and a PPC covers all your NHS prescriptions, no matter how many items you need.⁵⁹ Also, those of working age or in higher education, e.g. at university may not be able to get time off work or from their studies or have the funds to attend the hospital or General Practitioner for their regular monitoring blood tests. We also wondered if the SARS-Cov-2 pandemic may have contributed to patients stopping their immunosuppression and identified 17/60 (28%) patients who had stopped at least one medication between March 2020 and July 2022 (end of data collection). We believe that the pandemic may have been an important factor for them stopping medication and there are similar findings for immunomodulatory medicines in other chronic conditions,⁶⁰ yet consensus guidelines recommend not stopping systemic immunosuppression in healthy patients with uveitis during the pandemic.⁶¹

The strengths of this study were that it was performed using data generated in real-life clinical practice, there was input from ophthalmologists experienced in the management in IED, and except for patients with Behçet’s disease, data from all IED patients on systemic immunosuppression were captured. There were some limitations as we unable to provide reasons for why clinicians appeared reluctant to consider treatment withdrawal despite knowing patients were in long-term clinical remission. In the clinic, we failed to document the reasons for patients stopping a medication despite them being counselled appropriately prior to the commencement of therapy.

Our recommendations are that all patients deemed to be in clinical remission for a minimum of 18 months should be identified/flagged in the clinic and/or by the Immunosuppression Nurse Practitioner (or equivalent) so that if they are still in remission at 24 months, a plan is put in place to reduce treatment with an aim to “exiting” immunosuppression treatment between 24 and 36 months. This should be discussed with the patient at the next clinic visit, with consideration of any specific ocular and general medical issues. This recommendation is most likely to be relevant for NIU and Scleritis patients and those on corneal transplant support where discontinuation of treatment has been recommended if stable for 2–3 years.⁶² For certain types of progressive OSD patients such as those with immune-driven progressive cicatrizing conjunctivitis, in the absence of licensed antifibrotic agents for this group of diseases, lifelong immunosuppression may be required but a step-down approach to less potent agents should be considered.^1,2 Once these recommendations are in place we will need to reevaluate those patients to see if there has been disease relapse, but this is outside the scope of this paper and a subject for a future study.

In the 128 patients described in this cohort, discussions should take place at their next clinic visit about reducing and possibly discontinuing treatment. We also need to identify any patient who has stopped a medication much earlier than when they are seen at their clinic visit. Any patient not contacting the Immunosuppression Nurse Practitioner or equivalent health-care professional for a repeat prescription or missing an appointment for a monitoring blood test, should be contacted immediately and counselled about the importance of continuing treatment. If the patient is reluctant to restart a specific medication despite reassurance, then there should be discussions about switching to an alternative medication.

Finally, in the absence of robust clinical trials evaluating optimal exit strategies for immunosuppression in the various IED sub-groups, there needs to be formal consensus guidelines on the reduction and discontinuation of systemic immunosuppression for these patients. The consensus group should include not only ophthalmologists but also other physicians, such as rheumatologists, immunosuppression nurse practitioners, pharmacists, and patients.

Meeting presentations

This work was presented as a Poster at the Association for Research in Vision and Ophthalmology Annual Meeting, 2023 and the Royal College of Ophthalmologists Annual Congress, 2023.

Disclosure statement

Matthew Azzopardi: Health Education England (financial support to attend ARVO 2023), Royal College of Ophthalmologists (financial support to attend Annual Congress 2023).

Yu Jeat Chong: no competing interests to declare

Sreekanth Sreekantam: no competing interests to declare

Robert J. Barry: no competing interests to declare

Natraj Poonit: no competing interests to declare

Saaeha Rauz: UKRI Medical Research Council (grant support), National Institute for Health and Care Research (grant support)

Philip I. Murray: Oxford University Press (book royalties), Scope Eyecare (lecture fee)

Additional information

Funding

S.R. is supported by UKRI Medical Research Council Experimental Medicine Award Ref [MR/X019195/1]. The sponsor or funding organization had no role in the design or conduct of this research.