How would this aphorism stand up to the study of therapeutics and their development? Just as well. While the history of medicine is full of perilous trial-and-error discoveries of therapies, in the 20th century and beyond our approach has been more selective, more precise, and more able to predict a response at onset of treatment as well as monitor the duration of its action. To do this, drug development methods need to distill the disease to its essence in a discrete time frame where clinically and statistically significant effects can be compared to an active or negative control.
The modern approach to drug discovery rests on the foundation of randomized, double-masked clinical trials in predefined disease populations, yet the nature of certain diseases sometimes makes defining a drug effect extremely difficult. In these cases, a clinical model that mimics the disease may provide a superior platform for investigating the activity of drugs in a clinical setting. Those diseases that are best served by a model include conditions with: 1) a strong subjective component with regard to symptoms; 2) inherent temporal, locational and behavioral variability, both between and within subjects; and 3) significant placebo efficacy. In ophthalmology, two such diseases are ocular allergy and dry eye.
To describe the utility of modeling, let’s first drill down on the characteristics of each of these diseases. With the exquisitely sensitive cornea to reckon with, ocular surface diseases are all defined by ocular pain or discomfort. Itching is the pathognomonic symptom of allergic conjunctivitis, while other allergy symptoms, including tearing and swelling, have subjective components that are halfway between a sign and symptom.1 In contrast, dry-eye disease has a plethora of symptoms related to us by the patient: dryness; scratchiness; grittiness; burning; stinging; itching; ocular fatigue; and, rarely, photophobia. How can a clinician accurately measure the daily incidence and severity of these subjective symptoms? The subject diary has been the conventional method of collecting subjective data outside of the clinic, but it is fraught with inconsistencies and, when it is completed, is often done so cursorily in the parking lot just before the doctor’s visit. The elegance of a disease model in this setting lies in the fostering of symptoms’ appearance directly in the presence of the clinician, who no longer has to jog the patient’s memory to retrieve an accurate recapitulation of the disease process. This is a reverse case of “now you have it, now you don’t,” in which consistent, reproducible, real-time subject grading provides invaluable data on how a drug can alter disease symptoms.
Using a clinical model it’s possible to identify a potential study group with disease severity that’s not so mild or so severe that a drug effect cannot be readily observed. In addition, subjects with a similar, moderate severity of disease are more likely to display a clinically significant magnitude of response to an effective therapy. In this way models allow us to reduce the noise inherent in a variable disease and provide a true measure of a treatment’s clinical value.2-4
Temporal variability is also a common feature of many diseases. Is the disease worse or better in the morning or evening? Does it subside at night? In the case of allergy, two possibilities are common. Allergy sufferers who are sensitive only to airborne pollens will be less symptomatic in the evenings while in bed (with eyes closed) away from exposure. This is also true for dry-eye sufferers with their lids closed, allowing the ocular surface time to reset tear-film imbalances and avoid the high cost of tear evaporation. In contrast, a subject with mite allergies might find bedtime to be his worst nightmare when it comes to allergic symptoms.
The dry-eye patient generally feels good in the morning, after the ocular surface has been protected throughout the night, and progressively worsens with time awake.5 This observation highlights the importance of noting and controlling for the time of office appointments when doing a clinical trial, as a dry-eye subject will be much more naturally symptomatic at an evening appointment. A second form of temporal variability involves the seasons: Is the disease characterized by seasonal fluctuations? Pollen allergies are an obvious example of this, but the worsening of dry-eye symptoms in the low humidity of indoor winter environments is another seasonal accent to this complex disease.
The third type of disease variability is locational. Geographic differences in ocular surface disease incidence or severity are typically due to differences in climate: Arid regions are worse for dry eye while temperate climates and long growing seasons can exacerbate pollen sensitivities. There’s also growing evidence that urban pollution contributes to both chronic dry eye and allergic inflammation.6-8 On the opposite end of the spectrum, the absence of dust mites at higher elevations can provide a needed respite to both allergy and asthma sufferers.
Locational variability is far more difficult to control in a clinical setting, however. A person in today’s world typically moves between four or more distinct environmental settings: the home; public or private transportation; the workplace; and various outdoor or public spaces between the others. Each of these environments has its own level of pollutants, pollen and other potential allergens, and its own characteristic relative humidity, temperature, ventilation and lighting. All of these factors can affect the signs and symptoms of dry eye or ocular allergy. Similarly, the drastic changes in environment and location that may occur on weekends can also greatly alter a person’s disease, be it freedom from reading, writing and staring at a computer eight hours a day for a dry-eye sufferer, or outdoor activities like hiking or cycling for the pollen allergy sufferer. This behavioral variation adds an additional layer to the complexity of tracking ocular surface disease process and treatment.
It’s important to remember that all these types of variability are uncontrolled and uncontrollable not only across subjects (inter-subject variability), but also within subjects (intra-subject variability), creating a situation of waxing and waning signs and symptoms without consistency, a highly variable signal that can make the identification of a drug effect difficult or impossible.
The Placebo Effect
The difficulty in conducting reliable clinical trials is impeded further when signs and symptoms respond to placebo treatment. In some conditions, the placebo in a topical drug trial may be similar to the current treatments used by trial subjects. For dry-eye treatment trials, the majority of subjects are already using tear substitutes ad lib, and while they may only provide transient symptomatic relief they are, for many patients, the best available therapy. Nevertheless, a negative control is needed to compare efficacy across double-masked, randomized, placebo-controlled trials. This is usually provided by the drug vehicle, whose comfort, particularly in the case of dry eye, is maximized in terms of osmolarity and wettability. It is highly predictable, then, that this vehicle will provide significant benefit to dry-eye sufferers used to treating their disease with tear substitutes.
Placebo effects also impact ocular allergy trials, although their effect is a bit more complex. Tear substitutes or vehicle placebo can provide multiple benefits to the allergy sufferer and thus confound clinical trial findings. They wash away environmental allergens, minimizing their contact with surface antibodies and mast cells. They also dilute the in-place mediators released by previous exposure, including histamine, prostaglandins, leukotrienes, chemokines and cytokines. Finally, the wettability effects of topical placebos provide comfort and relief from ocular surface damage and inflammation, even in the absence of the well-known comorbidity of allergy and dry eye,6 leading to decreased symptomology.9 With these obstacles to overcome, it’s a wonder that any drug has been approved based on environmental studies.
Of course, use of models in ocular surface drug development is a subject we’ve been involved in for quite some time.10 Now that we’ve described some of the general factors involved in the decision to use a model-based protocol to test drug efficacy, let’s take a closer look at the specific aspects of two models that we have designed and refined to respond to each of these impediments to well-founded and reproducible clinical science.
Conjunctival Allergen Challenge
While the instillation of allergen to the eye has been performed for decades to study allergic disease, the fine control that has evolved in every aspect of this protocol has led to the approval of 19 anti-allergic drugs since its acceptance as a validated method for drug development in 1990.10
In the CAC, pre-selected subjects with a history of ocular allergy and a positive skin test are administered baseline challenges on two separate visits with the allergen to which they are sensitized, establishing a reproducible, and consistent, bilateral response of moderate severity to a pre-determined dose. The step-wise increases in allergen also bring all subjects to approximately the same reaction, avoiding the large fluctuations caused by differences in sensitivity. Subjects can then be administered drug in the prescribed dosage regimen in a double-masked, randomized, placebo-controlled fashion. While this might appear to be prevention rather than treatment of an allergic response, the Food and Drug Administration accepts this method since it reflects the nature of allergic disease, which is actually episodic bursts of mediator release in response to exposure, and whose treatment can be seen as a temporary breach in these episodes. After drug treatment, the subject is again challenged and signs and symptoms are graded with scales developed in the past 25 years that are tailored to the nuances of itching, redness and swelling specific to allergy. The itching scale in particular allows the subject to grade his or her symptoms in front of the clinician with confidence that data are being collected in real time.9-11
This challenge method identifies effects at onset, and also the duration of effect, which is almost impossible to ascertain under natural conditions. Thus, the CAC model creates a discrete allergic reaction in all subjects in-office, under clinically observable conditions. In this way, variability originating from inherent differences in allergic sensitivity, time of day, season, location and behavior are all minimized. With these variables taken care of, the beneficial effects of placebo are reduced and a true drug effect can be accurately defined.11
The CAE Model
The ocular surface is exquisitely in tune with its environmental conditions, and manipulation of factors such as temperature, wind or relative humidity can provide a means to induce tear-film instability and a desiccating stress on the ocular surface. In the controlled adverse environment, subject responses to an adverse environment challenge while performing a visual task are used as a baseline, and aid in identifying a defined population of subjects with comparable disease signs and symptoms. Like the CAC, the CAE can also be used to assess test-agent efficacy by measuring changes in the dry-eye status of subjects from baseline to post-challenge through slit lamp evaluations, validated scoring of dry-eye redness by the investigator and by computer,12 staining with fluorescein and lissamine, tear-film breakup times13 and Schirmer’s testing. The subject’s symptomology is graded individually and with a variety of validated grading systems, again in a clinical setting and in real time, preventing the vagaries of diary completion and with no hindrance from subjective memory. Finally, precise methods have been developed to assess blink behavior,14,15 tear-film dynamics (OPI)16 and visual function during tasking (IVAD),17 all of which can be combined with CAE challenge in a before-and-after protocol. Often, CAE challenge-based studies are conducted in concert with environmental collection of data, since treatment with active dry-eye molecules typically requires treatment durations of one month or more. Like the CAC model, the CAE model minimizes the inter- and intra-subject variability of inherent disease, environment and behavior. This dampening of background noise again aids in minimizing the beneficial effects of placebo when evaluating a drug.2-4,9-11
Overall, models allow us to focus on the potential efficacy of test compounds in shorter, more defined time frames. This allows for a streamlining of the development process with benefits to both the developers and consumers alike. Even when there is a desire for large-scale trials using more traditional protocols, disease models can provide vital proof-of-concept findings to speed the best therapeutics to the patients.
Though it’s true “life is short and art long,” is it possible that clinical models have let us sidestep the ancient wisdom, the idea of advancements in medicine evolving in incremental steps over time? Or have we simply seized the fleeting opportunity to test clinical science with the tools available today, based on a foundation of knowledge built over centuries? Medicine and clinical science have unquestionably benefited from disease models, which provide the opportunity for the experiment and judgment of Hippocrates with the velocity and precision that today’s fast-paced society requires.
Dr. Abelson is a clinical professor of ophthalmology at Harvard Medical School. Ms. Smith is a medical writer at Ora Inc.
1. Leonardi A, Bogacka E, Fauquert JL, et al. Ocular allergy: Recognizing and diagnosing hypersensitivity disorders of the ocular surface. Eur J Allergy Clin Immunol 2012;67:1327-1337.
2. Ousler GW III, Gomes PJ, Welch D, Abelson MB. Methodologies for the study of ocular surface disease. Ocul Surf 2005;3:3:143.
3. Abelson MB, Knight E. Dry eye therapy: Evaluation of current directions and clinical trials. Adv Exp Med Biol 1994;350:431-6.
4. Abelson MB, Ousler GW III, Nally LA, Emory TB. Dry eye syndromes: Diagnosis, clinical trials and pharmaceutical treatment-‘improving clinical trials’. Adv Exp Med Biol 2002;506(B):1079-86.
5. Walker PM, Lane KJ, Ousler GW III, Abelson MB. Diurnal variation of visual function and the signs and symptoms of dry eye. Cornea 2010;29:6:607-12.
6. Gomes PJ, Ousler GW III, Welch DL, Smith LM, Coderre J, Abelson MB. Exacerbation of signs and symptoms of allergic conjunctivitis by a controlled adverse environment challenge in subjects with a history of dry eye and ocular allergy. Clin Ophthalmol 2013;7:157-65.
7. Barnes CS, Alexis NE, Bernstein JA, et al. Climate change and our environment: The effect on respiratory and allergic diseases. J Allergy Clin Immunol Pract 2012;1:2:137-141.
8. Warm K, Lindberg A, Lundback B, Ronmark E. Increase in sensitization to common airborne allergens among adults: Two population-based studies 15 years apart. Allergy Asthma Clin Imunol 2013;9:1:20.
9. Abelson MB, Loeffler O. Conjunctival allergen challenge: Models in the investigation of ocular allergy. Curr Allergy Asthma Rep 2003;3:4:363-8.
10. Abelson MB, Chamber WA, Smith LM. Conjunctival allergen challenge. A clinical approach to studying allergic conjunctivitis. Arch Ophthalmol 1990;108:1:84-8.
11. Abelson MB. Comparison of the conjunctival allergen challenge model with the environmental model of allergic conjunctivitis. Acta Ophthalmol Scand Suppl 1999;228:38-42.
12. Rodriguez JD, Johnston PR, Ousler GW 3rd, Smith LM, Abelson B. Automated grading system for the evaluation of ocular redness associated with dry eye. Clin Ophthalmol 2013;7:1197-1204.
13. Abelson MB, Ousler GW 3rd, Nally LA, Welch D, Krenzer K. Alternative reference values for tear film break up time in normal and dry eye populations. Adv Exp Med Biol 2002;506(B):1121-5.
14. Johnston PR, Rodriguez J, Lane KJ, Ousler G, Abelson MB. The interblink interval in normal and dry eye subjects. Clin Ophthalmol 2013;7:253-9.
15. Rodriguez JD, Ousler GW III, Johnston PR, Lane K, Abelson MB. Investigation of extended blinks and interblink intervals in subjects with and without dry eye. Clin Ophthalmol 2013;7:337.
16. Ousler GW III, Hagberg KW, Schindelar M, Welch D, Abelson MB. The Ocular Protection Index. Cornea 2008;27:5:509-13.
17. Torkildsen G. The effects of lubricant eye drops on visual function as measured by the Interblink Interval Visual Acuity Decay test. Clin Ophthalmol 2009;3:501-506.