Published Online First: 16 August 2006. doi:10.1136/bjo.2006.097998
British Journal of Ophthalmology 2006;90:1481-1485
? 2006 by BMJ Publishing Group Ltd

EXTENDED REPORT

Retention time for corticosteroid-sparing systemic immunosuppressive agents in patients with inflammatory eye disease

K B Baker1, N J Spurrier2, A S Watkins3, J R Smith3,* and J T Rosenbaum3

1 Department of Medicine, Oregon Health and Science University, Portland, Oregon, USA
2 Department of Paediatrics and Child Health, Flinders University of South Australia and Flinders Medical Center, Adelaide, South Australia, Australia
3 Department of Ophthalmology (Casey Eye Institute), Oregon Health and Science University, Portland, Oregon, USA

Correspondence to:
J R Smith
Biomedical Research Building, Mail Code: 467AD, Oregon Health and Science University, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA;smithjus@ohsu.edu

Accepted for publication 3 August 2006
*    ABSTRACT
 
Background: Multiple immunosuppressive drugs have been used to manage inflammatory eye disease when control cannot be achieved by corticosteroid alone. However, although clinical studies support the effectiveness of most of these agents, comparative studies have not been undertaken. Retention time, a measure of the duration of treatment with any given drug, is a crude indicator of drug effectiveness and tolerability that facilitates such a comparison. The retention time was compared for corticosteroid-sparing immunosuppressive agents in patients attending our tertiary referral inflammatory eye disease clinic.

Methods: The clinical records of all patients attending an inflammatory eye disease clinic at the Casey Eye Institute over a 1-year period (2003) were reviewed. From these records, we collected the following clinical data: age; sex; ocular diagnosis; and use of steroid-sparing systemic immunosuppression, including drugs, duration of treatment and, if ceased, reasons for cessation. Cox regression analysis, adjusted for clustering, was used to compare other drugs against methotrexate.

Results: 107 of 302 (35%) patients seen at the inflammatory eye disease clinic in 2003 had a total of 193 current or past prescriptions for systemic steroid-sparing immunosuppressive agents. The treated group, most of whom had uveitis, included 32 men and 75 women, aged 5?86 years. Most commonly prescribed were methotrexate (66 uses, 34%), ciclosporin (37 uses, 19%), azathioprine (26 uses, 13%), mycophenolate mofetil (22 uses, 11%) and cyclophosphamide (15 uses, 8%). Patients were retained significantly less on ciclosporin (p = 0.004), azathioprine (p = 0.04), mycophenolate mofetil (p = 0.04) and cyclophosphamide (p<0.001) compared with methotrexate. Reasons for cessation included adverse events, lack of effectiveness, success or remission, cost and desire for fertility.

Conclusions: In patients with inflammatory eye disease, methotrexate may offer a superior combination of effectiveness and tolerability over other commonly used corticosteroid-sparing immunosuppressive agents. In this study, there was a twofold risk of not being retained on azathioprine, mycophenolate mofetil and ciclosporin and a fourfold risk of not being retained on cyclophosphamide compared with methotrexate.

Multiple corticosteroid-sparing immunosuppressive agents have been used to treat immune-mediated forms of inflammatory eye diseases, including uveitis, scleritis and orbital inflammatory disease, when vision is threatened, and either the patient cannot tolerate oral prednisone or the disease cannot be controlled. Case series and controlled clinical trials support the use of most of these drugs, but comparative studies, particularly between different classes of drugs, are lacking.1,2,3,4,5,6,7,8,9,10,11,12,13,14,15 Although nowadays the concept of evidence-based medicine recommends that a doctor?s decision making should be based on the results of randomised controlled clinical trials,16 such studies are rare in the specialty of inflammatory eye disease.17 Barriers include the low incidence of these disorders, the heterogeneity of each disease and the lack of validated diagnostic criteria for many subtypes, as well as the high cost of several of these drugs. Consequently, other legitimate methods for differentiating between an ever-increasing number of "conventional" and "biologic" immunosuppressive drugs are required.

In several other medical disciplines, "retention time" has been used to compare drug effectiveness and toxicity.18?20 In essence, the retention time for any given drug, also variously referred to as drug "persistency", "survival" or "discontinuation rate", refers to the time a patient continues to take that drug. When considering treatment of chronic diseases, retention time provides a measure of drug effectiveness and tolerability. Although comparing retention times for different drugs used for the same indication seems to be a rather simplistic approach, it has become an accepted method to assess the relative effectiveness of drugs used for chronic diseases. For example, the measurement of retention time has helped to establish the relative value of ocular hypotensive treatments in glaucoma, as reported in supplements to the American Journal of Ophthalmology (supplement to volume 137, 2004) and the American Journal of Managed Care (supplement to volume 8, 2002). In addition, and of particular relevance to inflammatory eye disease, the measure has been used by rheumatologists to distinguish between disease-modifying antirheumatic drugs for rheumatoid arthritis.21,22

In this study, we compared the retention time for corticosteroid-sparing systemic immunosuppressive agents used to manage inflammatory eye disease. As it was our clinical impression that methotrexate was well tolerated and presented an excellent safety profile, we hypothesised that methotrexate would have a longer retention time than other immunosuppressive agents that we commonly prescribed.


*    METHODS
 
All patients who, during 2003, attended a weekly clinic at the Casey Eye Institute (Portland, Oregon, USA) for the management of inflammatory eye diseases were enrolled in a retrospective study of retention time on corticosteroid-sparing systemic immunosuppressive agents. At this clinic, steroid-sparing systemic immunosuppression is prescribed and monitored in accordance with the guidelines for the use of immunosuppressive drugs in patients with ocular inflammatory disorders.23 Medical charts of all patients were examined to collect prespecified demographic and clinical information that included age; sex; ophthalmological diagnosis and, in the case of uveitis, disease subtype; corticosteroid-sparing immunosuppressive treatments including dates on which a drug was started and, if applicable, ceased; and if any drug had been discontinued, reason for cessation. Charts were analysed retrospectively, and start and cessation dates for all instances of corticosteroid-sparing immunosuppressive treatments were noted for each patient. Many patients received more than one drug, and follow-up information extended as long as 5000 days. If patients were still on one or more drugs at the time that we began data collection (30 June 2004), they were recorded as continuing on the drugs for the purposes of the statistical analysis. Patients who were enrolled in clinical trials at any time during their period of attendance at the clinic were not included in this study to avoid the possibility that time of receiving drugs was dictated by a study protocol. The institutional review board gave approval for this study.


*    STATISTICAL ANALYSES
 
Data were entered into Stata statistical software. Six potential patients were removed from subsequent analyses because they had not been re-evaluated clinically after starting the relevant drug, at the time we began data collection. Retention time for the usage of each drug was calculated by subtracting the date the drug was started from either the date the drug was discontinued (for any reason) or the date that the patient was last reviewed by a clinician. Frequencies for relevant clinical and demographic data were calculated. Survival analyses (Kaplan?Meier method and Cox proportional hazards regression model) were undertaken, as these methods allow for variable lengths of follow-up and account for different end points. Kaplan?Meier survival plots were constructed to allow visual representation of retention (survival) time for the five most common drugs. Median retention times were estimated using the Kaplan?Meier method for each of these five drugs. Cox regression analyses were conducted to assess whether retention time was significantly different between the five most commonly prescribed drugs. In these analyses, methotrexate was set as the comparison agent.

All analyses were adjusted for clustering by subject, because many patients had trials of more than one drug during the study period. Three separate models were developed: (a) all ophthalmological diagnoses and all reasons for drug cessation were included; (b) only data from patients with uveitis were included; and (c) all diagnoses were included, but drug usages where cessation occurred because the underlying disease was considered to have remitted were excluded. These usages were not included because ceasing treatment due to lack of effectiveness or side effects is quite different to discontinuing treatment because disease has remitted.


*    RESULTS
 
A total of 302 people attended a weekly clinic (with JTR) at the Casey Eye Institute?s inflammatory eye disease service on one or more occasions during 2003. Of these 302 people, 113 patients had a history of taking one or more non-corticosteroid systemic immunosuppressive drugs and had never been enrolled in a clinical trial of such treatment. As noted above, six potentially eligible patients had not had a subsequent clinical visit at the time of data collection and were thus excluded from further analyses. Thus, data from 107 patients remained for inclusion in this study. Table 1Go summarises the demographic and clinical data collected on these patients.

The number of uses of immunosuppressive drugs totalled 193 in the 107 participants. Table 2Go gives a description of the number of uses of each specific drug, along with the median length of retention of patients prescribed the most commonly used drugs. Figure 1Go shows the Kaplan?Meier survival curves for the five most commonly prescribed drugs (ie, methotrexate, ciclosporin, azathioprine, mycophenolate mofetil and cyclophosphamide). Visual inspection of these curves suggests that although there is little difference in survival (retention) between azathioprine, mycophenolate mofetil, ciclosporin and cyclophosphamide, methotrexate appears to be associated with greater retention than the other drugs. This was subsequently tested with Cox regression analyses.



Figure 1
 		 Figure 1  Kaplan?Meier survival curves showing probability of being retained on drug versus total analysis time at risk, for all ophthalmological diagnoses (n = 101, number of observations = 166).

 
No significant relationship was found between retention probability and sex (p = 0.2) or age (p = 0.1) by Cox regression analysis using patients on all possible drugs. In 98 of the 193 drug treatments, drugs were ceased, giving an overall percentage of discontinuation of 51%. Reasons for cessation of drug were lack of effectiveness (n = 35), adverse events (n = 25), disease remission or treatment success (n = 13), poor compliance (n = 5), cost (n = 2), other reasons (n = 4) and multiple reasons (n = 9); in five instances, reasons for cessation were not specified in the medical records. Table 3Go provides the information on reasons for cessation of drugs.

The first Cox regression analysis included 101 patients treated with one or more of the five most commonly prescribed drugs, regardless of diagnosis or reason for drug cessation. This analysis, presented in table 4Go, showed that patients were retained for a significantly shorter time on ciclosporin (p = 0.004), azathioprine (p = 0.04), mycophenolate mofetil (p = 0.04) and cyclophosphamide (p = 0.001) compared with methotrexate. There was an increase in the risk of not being retained on a drug compared with not being retained on methotrexate of approximately twofold for ciclosporin, azathioprine and mycophenolate mofetil, and approximately fourfold for cyclophosphamide. Table 5Go shows the analysis, repeated for 63 patients diagnosed with uveitis only, which gave similar results. Finally, because it was possible that our results were biased by including patients who discontinued a drug because of disease remission, we repeated the Cox regression analysis excluding the 15 patients who had ceased treatment with the drug for this reason (including two patients classified as ceasing treatment with the drug for multiple reasons), and found that methotrexate continued to show a more favourable retention rate (table 6Go).



*    DISCUSSION
 
We investigated the retention time for methotrexate versus other corticosteroid-sparing systemic immunosuppressive agents conventionally used to treat inflammatory eye disease in a cohort of 107 patients attending one doctor?s clinic over a 1-year period. Our hypothesis that methotrexate had a relatively longer retention time, implying a superior combination of effectiveness and tolerability over other agents, was confirmed by Cox regression analysis adjusted for clustering. The analysis showed that patients were twice as likely to stop other antimetabolites, azathioprine and mycophenolate mofetil, and the T cell modulator, ciclosporin, and four times as likely to stop the alkylating agent, cyclophosphamide, compared with methotrexate. A similar result was obtained when the analysis was repeated including patients with uveitis only. Finally, because it is common practice to attempt to wean patients from immunosuppression if disease has been quiescent for 6?12 months, we repeated the analysis excluding patients who ceased the drug owing to remission of disease. Despite a smaller sample size, all drugs were significantly less likely to be retained compared with methotrexate.

Methotrexate, an antimetabolite that targets rapidly dividing leucocyte precursors by inhibiting DNA replication at the level of dihydrofolate reductase, has a therapeutic history dating back to the 1950s, originally as a treatment for cancer but later also in inflammatory diseases.24 In agreement with our findings, studies from the rheumatological literature indicate a prolonged retention time of methotrexate, in patients with rheumatoid arthritis. Pincus et al21 studied 532 patients attending seven private rheumatological practices in the US and found that methotrexate was continued significantly longer than other second-line drugs that included azathioprine, gold, hydroxychloroquine and penicillamine. Aletaha and Smolen22 investigated the use of disease-modifying antirheumatic drugs in 593 patients from two hospital rheumatology clinics in Vienna. They found that of single-agent treatments, patients were retained on methotrexate longer than all other drugs, including ciclosporin, sulfasalazine, gold, chloroquine and penicillamine, with the exception of azathioprine. As well as reports of effectiveness for inflammatory eye disease,1?3 methotrexate has the reputation of being well tolerated. As discussed by an expert panel on management of inflammatory eye disease,23 the most common side effect is gastrointestinal disturbance; often, this is alleviated by combination with folic acid or giving the drug subcutaneously. Bone-marrow suppression, cirrhosis and interstitial pneumonitis are serious complications that fortunately occur infrequently. Other factors that could contribute to patient continuation on methotrexate include the relatively low cost of this agent in comparison with other non-corticosteroid immunosuppressive agents and once-weekly frequency of administration.

Although patients were retained on methotrexate for a considerable length of time, our data paint a disappointing picture of systemic immunosuppressive treatment for inflammatory eye disease. Overall, 51% of drugs were discontinued, and approximately two thirds of patients stopped for lack of effect or side effects. This result agrees with the comprehensive report of systemic drug toxicity trends by Tamesis et al,25 who found that drugs including methotrexate, azathioprine, ciclopsporin, cyclophosphamide and chlorambucil were continued by just 32?45% of patients with various ocular inflammatory syndromes. Newer agents promise longer retention times. Two recent large studies suggest that mycophenolate mofetil may be very effective and safe for inflammatory eye disease, although follow-up in both studies is limited. Thorneet al5 described a group of 84 patients treated with mycophenolate for inflammatory eye disease, with a median follow-up of 10 months; the drug was discontinued for ineffectiveness and side effects in approximately 10% and 8% of patients, respectively. Siepmann et al6 have followed a cohort of 106 patients with uveitis taking mycophenolate for at least 6 months, with just four treatment failures and no side effects requiring cessation of drug. Our experience with mycophenolate is less positive, but our impressions are based on only 22 patients. New biologic agents that specifically target key molecules in inflammatory signalling pathways are attracting much attention in our field. In particular, blockade of tumour necrosis factor {alpha} with the monoclonal antibody, infliximab, has been heralded as highly effective for patients with recalcitrant uveitis. However, recent work from our group14 indicates that retention time is short, owing to the frequent occurrence of treatment-limiting side effects.

Our data suggest that methotrexate offers a favourable combination of tolerability and effectiveness for the treatment of uveitis and inflammatory eye disease in general. Our study is limited by the small size of the study cohort and retrospective data collection. In addition, we are studying a heterogeneous collection of diseases. Specific subsets of uveitis, such as Beh?t?s disease or serpiginous choroidopathy, might respond preferentially to a given drug, thus extrapolating from an analysis based on all patients with uveitis could fail to identify the utility of a specific drug for a defined subset. Our data are derived from a single clinical practice, which introduces unavoidable bias relating to the drug preference of the prescribing doctor. If we routinely prescribe methotrexate as the preferred corticosteroid-sparing agent, we might selectively place those patients who are best able to tolerate immunosuppression and artificially extend the retention time for this drug. Although we cannot deny our enthusiasm for methotrexate, the underlying philosophy of the clinic is to take an individualised approach when selecting a cortiosteroid-sparing immunosuppressive, considering specifics of the ocular inflammatory disease, medical history and lifestyle factors, including the wish to consume alcohol, which would relatively contraindicate methotrexate. It is obvious from our data that a wide range of immunosuppressive drugs are prescribed in this clinic.

Although our study was not designed to assess quality of life, another way of assessing differences in effectiveness of drugs is to collect of quality of life data, which includes measurements of the effect of disease on physical, social and psychological well-being. Particularly when drugs are extremely expensive, if quality of life is shown to be enhanced, arguments can be made for funding on the basis of appropriate cost-utility analyses. Future studies on the effectiveness of corticosteroid-sparing systemic immunosuppressive agents could be strengthened by including measurement of quality of life.

This work shows the use of the concept of retention time to evaluate and compare different drug treatments for inflammatory eye disease. Given the difficulties that have hindered randomised controlled clinical trials in this field to date, there is much value to such an approach, particularly if prospectively collected data are pooled from multiple clinical centres over a defined time. In this way, the effectiveness and tolerability of both conventional and novel drugs, singly or in combination, could be evaluated.

 
*    FOOTNOTES
 
* These authors contributed equally to this work and share senior authorship of this manuscript. Back

Published Online First 16 August 2006

Funding: This work was supported in part by Research to Prevent Blindness (Career Development Award to JRS and Senior Scholar Award to JTR) and the Rosenfeld Family Trust.

Competing interests: None.


*    REFERENCES
 

  1. Shah SS, Lowder CY, Schmitt MA, et al. Low-dose methotrexate therapy for ocular inflammatory disease. Ophthalmology 1992;99:1419?23.
  2. Samson CM, Waheed N, Baltatzis S, et al. Methotrexate therapy for chronic noninfectious uveitis: analysis of a case series of 160 patients. Ophthalmology 2001;108:1134?9.
  3. Smith JR, Rosenbaum JT. A role for methotrexate in the management of non-infectious orbital inflammatory disease. Br J Ophthalmol 2001;85:1220?4.
  4. Yazici H, Pazarli H, Barnes CG, et al. A controlled trial of azathioprine in Behcet?s syndrome. N Engl J Med 1990;322:281?5.
  5. Thorne JE, Jabs DA, Qazi FA, et al. Mycophenolate mofetil therapy for inflammatory eye disease. Ophthalmology 2005;112:1472?7.
  6. Siepmann K, Huber M, Stubiger N, et al. Mycophenolate mofetil is a highly effective and safe immunosuppressive agent for the treatment of uveitis A retrospective analysis of 106 patients. Graefes Arch Clin Exp Ophthalmol 2005;244:1?7.
  7. Nussenblatt RB, Palestine AG, Chan CC, et al. Randomized, double-masked study of cyclosporine compared to prednisolone in the treatment of endogenous uveitis. Am J Ophthalmol 1991;112:138?46.
  8. Wakefield D, McCluskey P. Cyclosporin therapy for severe scleritis. Br J Ophthalmol 1989;73:743?6.
  9. Goldstein DA, Fontanilla FA, Kaul S, et al. Long-term follow-up of patients treated with short-term high-dose chlorambucil for sight-threatening ocular inflammation. Ophthalmology 2002;109:370?7.
  10. Buckley CE III, Gills JP Jr. Oral cyclophosphamide in the treatment of uveitis. Trans Am Acad Ophthalmol Otolaryngol 1970;74:505?8.
  11. Messmer EM, Foster CS. Destructive corneal and scleral disease associated with rheumatoid arthritis. Medical and surgical management. Cornea 1995;14:408?17.
  12. Kotter I, Zierhut M, Eckstein A, et al. Human recombinant interferon-alpha2a (rhIFN alpha2a) for the treatment of Behcet?s disease with sight-threatening retinal vasculitis. Adv Exp Med Biol 2003;528:521?3.
  13. Murphy CC, Greiner K, Plskova J, et al. Neutralizing tumor necrosis factor activity leads to remission in patients with refractory noninfectious posterior uveitis. Arch Ophthalmol 2004;122:845?51.
  14. Suhler EB, Smith JR, Wertheim MS, et al. A prospective trial of infliximab therapy for refractory uveitis: preliminary safety and efficacy outcomes. Arch Ophthalmol 2005;123:903?12.
  15. Nussenblatt RB, Peterson JS, Foster CS, et al. Initial evaluation of subcutaneous daclizumab treatments for noninfectious uveitis: a multicenter noncomparative interventional case series. Ophthalmology 2005;112:764?70.
  16. Group TE-BMW. Users? guides to the medical literature. Essentials of evidence-based clinical practice. Chicago, IL: AMA Press, 2002.
  17. Okada AA. Noninfectious uveitis: a scarcity of randomized clinical trials. Arch Ophthalmol 2005;123:682?3.
  18. Varadharajan S, Jumadilova Z, Girase P, et al. Economic impact of extended-release tolterodine versus immediate- and extended-release oxybutynin among commercially insured persons with overactive bladder. Am J Manag Care 2005;11:S140?9.
  19. Mauskopf JA, Paramore C, Lee WC, et al. Drug persistency patterns for patients treated with rivastigmine or donepezil in usual care settings. J Manag Care Pharm 2005;11:231?51.
  20. Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med 2005;353:1209?23.
  21. Pincus T, Marcum SB, Callahan LF. Longterm drug therapy for rheumatoid arthritis in seven rheumatology private practices: II. Second line drugs and prednisone. J Rheumatol 1992;19:1885?94.
  22. Aletaha D, Smolen JS. Effectiveness profiles and dose dependent retention of traditional disease modifying antirheumatic drugs for rheumatoid arthritis. An observational study. J Rheumatol 2002;29:1631?8.
  23. Jabs DA, Rosenbaum JT, Foster CS, et al. Guidelines for the use of immunosuppressive drugs in patients with ocular inflammatory disorders: recommendations of an expert panel. Am J Ophthalmol 2000;130:492?513.
  24. Whitehouse MW. Drugs to treat inflammation: a historical introduction. Curr Med Chem 2005;12:2931?42.
  25. Tamesis RR, Rodriguez A, Christen WG, et al. Systemic drug toxicity trends in immunosuppressive therapy of immune and inflammatory ocular disease. Ophthalmology 1996;103:768?75.