Conjunctival Allergen Challenge A Clinical
Approach
to
Studying Allergic Conjunctivitis
Mark B. Abelson, MD; Wiley A. Chambers, MD; Lisa M. Smith
\s=b\ To evaluate antiallergic agents, we conducted five allergen challenge studies of increasing refinement. The final study design that evolved included two baseline visits, when skin test-positive subjects were administered a bilateral ocular allergen challenge. At the first visit, the threshold dose of reactivity was determined by increasing allergen doses at 10-minute intervals. At the second baseline visit, 3 days later, the responsive subjects were challenged with the final, highest dose used on visit 1 to assure that the allergic reaction was reproducible and not a cumulative effect of multiple allergen doses. The responsive subjects then returned 3 days later for the drug efficacy evaluation. After a slit-lamp examination, subjects were pretreated with the test drug in one eye and the placebo in the fellow eye in a randomized, double-masked fashion. After 10 minutes, subjects were challenged bilaterally with the allergen dose identified on the previous visits. Postchallenge evaluations of hyperemia, itching, chemosis, eyelid swelling, and tearing were performed at 3, 10, and 20 minutes. Subjects were rechallenged 4 hours after drug administration to assess duration of action.
were again performed at the same intervals as after the initial challenge. A total of 396 subjects were given a baseline allergen challenge; 83.6% responded with a moderate (2+) ocular allergic reaction. Of the 266 given a second baseline challenge, 87.2% responded positively again, suggesting that ocular challenge was highly correlated with skin reactivity and reproducible with a second challenge. No statistically significant difference in redness and itching was found when both eyes were challenged with the same dose of allergen. There was a significant decrease in the allergic reaction of a placebo-pretreated eye compared with its baseline response, and the response to rechallenge was less than the initial one. Of the 950 challenges administered, only two mild systemic reactions occurred, demonstrating the safety of this incremental challenge method. We conclude that this method of allergen challenge is safe, reproducible, and symmetric, but a refractory period may exist when the eye is challenged a third and fourth time in an 8-day period. We now separate the challenge visits by 7 days.
1\/Tany ocular allergic diseases can be
neal vascularization, recurrent cor¬ neal infiltrates, and melting tenden¬ cies associated with hair loss and ec¬ zema in atopic dermatoconjunctivitis, eyelid edema, lichenification in con¬ tact- and drug-related dermatocon¬ junctivitis, and, finally, the classic signs and symptoms of conjunctival hyperemia, edema, and itching associ¬ ated with acute allergic conjunctivitis. The variability of the signs and symptoms of acute allergic conjunc¬ tivitis, diurnally and from day to day, is clearly related to the amount of con¬ junctival allergen contact. This vari¬ able and intermittent stimulus and its resulting response do not provide a suitable baseline for evaluating the
classified by the appearance of certain unvarying clinical character¬ istics, ie, cobblestones and/or limbal infiltrates in vernal conjunctivitis, corAccepted for publication August 22,1989. From the Department of Immunology, Eye Research Institute, and Department of Ophthalmology, Harvard Medical School, Boston, Mass (Dr Abelson and Ms Smith), and the Division of Anti\x=req-\ Infective Drug Products, Center for Drug Evaluation and Research, Food and Drug Administration, Rockville, Md (Dr Chambers).
This article contains the views of the authors and not necessarily those of the Food and Drug Administration. Reprint requests to 20 Staniford St, Boston, MA 02114 (Dr Abelson).
Slit-lamp examinations
(Arch Ophthalmol. 1990;108:84-88)
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
efficacy of potential
agents. More difficulties
antiallergic are encoun¬
tered if
one uses a drug vehicle as a control, since this may attenuate the allergic response in several ways: (1) it
may act
as a
tear substitute that
can
minimize
antigen penetration to the conjunctiva; (2) it may act as a diluent to the antigen; or (3) it may enhance the conjunctival transit time of the antigen. In many of our clinical envi¬ ronmental studies, these mechanical properties of placebo controls have been shown to elicit a "drug effect" in approximately 70% of the treated sub¬ jects. To precisely evaluate and compare anti-inflammatory agents, mast cell stabilizers, antihistamines, vasocon¬ strictors, phosphodiesterase inhibi¬ tors, and anti-IgE compounds, a con¬ trolled and reproducible situation is required. Our previous models of acute allergic conjunctivitis, histamine,1 48/ SO,2 or arachidonic acid,3 administered topically to the eye, have been helpful in studying allergic ocular disease but they do not have the clinical validity of antigen-induced mast cell degranulation. We have, therefore, modified the conjunctival provocation test of Moller et al4 for the evaluation of antiallergic compounds in a study design that min¬ imizes the potential variables of the allergen challenge model.5 We report data obtained from this standardized technique and various study designs that have proved to be safe and repro¬ ducible in 396 subjects. SUBJECTS, MATERIALS, The
subject selection
AND METHODS
was an
important
step in performing these antigen challenge
studies. Only normal, healthy volunteers who were not using any type of topical agent or systemic medication that might have interfered with results were allowed to enter the studies. Inclusion criteria were as follows: (1) able and willing to give consent; (2) between the
ages of 18 and 65 years; (3) either sex and of any race; (4) willing and able to make required study visits; (5) able to follow in¬
Table 1.—Scoring System to Measure the Signs and of Allergic Conjunctivitis
structions; (6) history of symptoms of a clinically active allergic conjunctivitis; (7) positive diagnostic test (skin or radioallergosorbent tests) for allergic disease; if the subject had no historical documentation of this, a skin prick test was performed; (8)
Redness, =
=
=
=
0 None Mild 1+ =
=
Subjects were asked to refrain from the of systemic medications that could affect test measurements, particularly antihistamines, decongestants, and nonste¬ roidal anti-inflammatory agents. If sub¬ jects were taking medications that would not affect test measurements, the dosage
3+
=
(detectable with slit lamp, conjunctiva separated from sclera) Moderate (visually evident, raised conjunctiva, especially at the limbal area) Severe (ballooning of conjunctiva)
None Mild tearing (eyes feel slightly watery) Moderate tearing (blows nose occasionally) Severe tearing (tears rolling down cheeks) Itching (to be graded by subject) None 0 =
or
Concomitant Medications
=
0 1+ 2+ 3+
bacterial or viral ocular in¬ fection; (2) presence of dry-eye syndrome, blepharitis, follicular conjunctivitis, iritis,
rheumatoid arthritis that can be associated with dry-eye syndrome; (6) use of any top¬ ical ophthalmic solutions during or at least 2 weeks prior to the study, including tear substitutes; subjects who had taken oph¬ thalmic medications that require longer than a 2-week washout; and (7) exhibiting signs and symptoms of allergic conjunctivi¬ tis on taking the baseline examinations prior to entry into the study on days 1, 4, or 8.
2+
Tearing
a
preauricular lymphadenopathy; (3) wearing of contact lenses during the course of the study; (4) taking any systemic med¬ ication that might have interfered with the study; (5) presence of any significant illness that could interfere with the study, partic¬ ularly any autoimmune disease such as
eyelid swelling
None 0 Mild 1+ Moderate 2+ Severe 3+
willing to avoid disallowed medications during the study period; and (9) a success¬ ful challenge, inducing at least moderate itching (2+) and redness. Exclusion criteria were as follows: (1) presence of
*
Symptoms
=
=
=
=
1+ 2+
3+
=
Mild
(intermittent tickling sensation)
Moderate (continual awareness but without the desire to rub) Severe (continual awareness with the desire to rub the eyes)
Incapacitating Itching (subject insists on rubbing eyes) Photographs depicting these gradations of redness are used as 4+
*
=
=
=
Table 2.—Summary of the Five Allergen
Challenge Study No.
Method Bilateral Bilateral Unilateral Unilateral Bilateral
Baseline Visits 1 1 2
2 2
Subjects Screened 33 63 147 69 109
a
guide.
Challenge Studies
Nonresponders to Allergen
Others Discontinued
Completed
4
21
45
2
100
18
16
75
use
and administration of these medications kept constant throughout the-dura¬ tion of the study. All concomitant medica¬ tion was thoroughly documented. were
Materials
Three types of antigens were used in these studies: Kentucky bluegrass (Poa pratensis), short ragweed (Ambrosia artemisiaefolia), and allergenic extract of cat hair and dander (Felis domesticus). Serial dilutions were made from the stock antigen solutions (bluegrass and ragweed: 100000 antigen units [AU]/mL; cat extract: 50 000 AU/mL) for incremental use in the allergen challenge. Six dilutions of bluegrass and ragweed were used, doubling in concentra¬ tion with each step, from 19 to 625 AU/25pL dose. Four dilutions of cat extract were used: 125,250,625, and 1250 AU/25-mL dose. Methods: Final
Study Design
Initial Visit.—Demographic data, medical
and medication history, and informed con¬ sent were obtained from each subject. If prior written documentation of allergy, identified by positive skin or radioallergosorbent tests, was unavailable, a skin test for the suspected allergen was then per¬ formed. The standardized disposable prick test needle (Morrow Brown, HMB Medical
Concepts, High Field House, Derby, En¬ gland) was saturated with stock solution of allergen (ragweed or grass: 100 000 AU/mL; cat: 50000 AU/mL), and the swabbed fore¬ arm of the subject was pierced. A positive
reaction was any definitive wheal and flare present after 10 minutes. The size of the re¬ action was then drawn and the form kept on file as written documentation. If the results of the skin test were positive for any one of the three allergens, an oph¬ thalmologic examination was given to ex¬ clude subjects currently exhibiting clini¬ cally significant signs and symptoms of al¬
lergic conjunctivitis (any itching or conjunctival redness >1+) (Table 1). Based on the subject's allergic sensitiv¬ ity, one of the three allergens was used for the conjunctival allergenic challenge. If the subject did not respond to the first allergen tested, he or she was challenged with an¬ other of the allergens to which he or she had elicited a positive skin reaction. A bilateral ocular allergen challenge was performed in which 25 pL of the diluent, and then 25 pL of increasing doses of allergen, were administered bilaterally with an Eppendorf digital pipette (Brinkman Instru¬ ments, Ine, Westbury, NY) and a sterile tip. A positive reaction was defined as at least a redness and itching score of 2+ in both eyes within 10 minutes of the last adminis¬
tered dose. Visit 2.—Those subjects who responded on visit 1 and who had met all other inclu¬ sion and exclusion criteria were entered
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
into visit 2, day 4. At this visit, a baseline slit-lamp examination was performed to exclude any subjects with signs or symp¬ toms of allergic conjunctivitis. Subjects with nonallergic eyes were then adminis¬ tered the final dose of allergen identified as the threshold dose for 2+ responsiveness on visit 1. If this allergen dose again elicited a 2+ itching and redness response in both eyes, the subject was entered into visit 3,
8. Visit 3.—Those subjects who had re¬ sponded at visits 1 and 2 with 2+ itching and redness to the same dose of allergen and who had met all other entry inclusion
day
and exclusion criteria were entered into the
study on day 8. Subjects were excluded who had baseline signs or symptoms of allergic conjunctivitis. Ten minutes after drug treatment in one eye and placebo in the contralateral eye, bilateral allergen chal¬ lenge was performed. Postchallenge evalu¬ ations occurred at 3, 10, and 20 minutes. Four hours after drug administration, sub¬ jects were rechallenged and reexamined to assess the drug's duration of action. To maximize the sensitivity and accuracy of the allergen challenge model, new infor¬ mation gained from each allergen challenge study was incorporated into subsequent protocol designs. While the final study de¬ sign that has evolved is outlined in detail above, the first four studies must be men¬ tioned briefly to understand the evolution¬ ary process. In the first two
studies, a bilateral ocular
Fig 1.—The conjunctival, episcleral, and ciliary hyperemia elicited by ocular allergen challenge to a sensitized subject.
Fig 2.—Chemosis elicited by ocular allergen challenge subject.
to a sensitized
allergen challenge was given at the baseline visit and 1 week later at the drug evaluation visit. No rechallenge was performed after drug administration. To further refine the study population, a second baseline challenge visit was added to the next two studies. In addition, we mod¬ ified the allergen challenge procedure at these baseline visits, performing it unilat¬ erally instead of bilaterally, in a method similar to that of Moller et al." These changes provided two advantages: (1) any subjects who had reached the moderate al¬ lergic response (2+) on visit 1 only with the cumulative allergen dose received with re¬ petitive challenges would be eliminated and (2) only subjects who had a reproducible reaction, even at the lowest dose, would be included in the test drug evaluation. Bilat¬ eral challenge was performed at the drug evaluation visit, and a rechallenge was ad¬ ministered either 2 or 4 hours after drug instillation to evaluate the drug's duration
of action. Since unilateral baseline challenge did not require a moderate response (2+) in both eyes, and we were concerned about the symmetry of the response, we reverted to bilateral baseline challenge in the final study. The second baseline challenge visit was retained. Table 2 gives a summary of the five studies, with differences in protocol design outlined, and it shows the number of sub¬ jects screened and entered. The sample size varies depending on the type of study de¬ signs used. When the evaluation of these data subsets required statistical analyses, the paired two-tailed t test and SignedRank Sum Test were used. RESULTS
Safety A total of 396 skin test-positive sub¬ jects who had a history of allergic con¬ junctivitis were entered and chal¬ lenged on the initial baseline visits of the five studies. For the completion of these studies, approximately 950 ocu¬ lar challenges were performed. Only 2 cases of systemic reactions occurred.
72-Hour Rechallenge
First Challenge
Fig 3.—The number of skin test-positive subjects who had a history of allergic conjunctivitis and who responded positively to ocular allergen challenge, and the number of ocular-positive responders who reacted positively a second time from the total number of subjects challenged in our
laboratory.
In one asthmatic patient, mild bron¬ chial constriction, without wheezing, occurred 15 minutes after ocular aller¬ gen challenge and was immediately treated with intramuscular epineph¬ rine hydrochloride and diphenhydramine hydrochloride. Respiratory allergic signs disappeared within 24 hours. One case of urticaria occurred approximately 8 hours after challenge and lasted for approximately 24 hours. One case of exaggerated local periorbital swelling was also reported. This occurred approximately 12 hours after challenge and lasted for approx¬ imately 3 hours. Similarity to Clinical Disease The clinical relevance of this model was supported by the appearance of all
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
signs and symptoms of allergic conjunctivitis: itching, tearing, slit¬ lamp evidence of conjunctival vasodilation, chemosis, and eyelid swelling (Figs 1 and 2). the
Relationship to Positive Dermatologie Reactivity A positive skin test was predictive of a positive ocular challenge in 331 (83.6% ) of 396 subjects who had a his¬
tory of allergic conjunctivitis.
Symmetry of Allergic Reaction Evidence for a bilaterally similar ocular response was necessary to use the contralateral eye as the control. Therefore, at the baseline visits of the final study (completed number of sub¬ jects, 75) both eyes were simulta-
neously challenged and data from each eye were recorded individually. The allergic response was found to be sim¬ ilar in both eyes, with no statistically significant differences noted for itch¬ ing (P .32) or for conjunctival (P m .16), episcleral (P .60), or cili¬ ary hyperemia (P .85). =
=
=
Reproducibility and Refractoriness
Response On day 1, of the 331 subjects who showed a positive ocular response to challenge, 266 were given a second baseline challenge before the drug test day. Of these 266 subjects, the positive ocular allergic response was reproduc¬ ible in 232 (87.2% ) of the subjects (Fig of
3).
In study 3, in which 55 subjects were initially challenged, it was possible to determine the change in the redness and itching scores from day 1 to day 4 to day 8. The actual allergic scores (redness and itching) obtained on day 1 were reproducible on day 4, with no
statistically significant found
differences
(itching: .727; redness: .537). After placebo pretreatment on day 8, the redness and itching scores were significantly decreased compared with results obtained on day 4 (itching: .0001; redness: .0001). Data were also analyzed from pla¬ cebo eyes challenged twice in one day. Thirty-four subjects who received pla¬ cebo in one eye and drug in the con¬ tralateral eye (the remainder of the subjects of this study were adminis¬ tered active drug in both eyes) were rechallenged at 2 hours in study 3. Redness (P .021) and itching (P .006) were significantly decreased in the placebo eyes at rechallenge com¬ pared with the initial scores. Fortyseven subjects received placebo in one eye and drug in the contralateral eye and were rechallenged after 4 hours in study 4. A significant decrease in itch¬ ing (P .039), but not redness (P .163), was noted in the placebo eyes at rechallenge. =
=
=
=
=
=
=
=
COMMENT
We have been performing ocular al¬ lergen challenge studies for the past year for the evaluation of antiallergic agents. With the completion of each study, more information was gained about the ocular response to challenge, causing us to reevaluate and modify our original protocol design. At first, in studies 1 and 2, only one baseline chal¬ lenge was required for entry into the drug test evaluation. However, the question arose as to whether the reac¬ tion elicited on the drug day, when only the final highest dose of allergen was
administered, was similar to that on the screening day, when multiple doses were given until a positive reaction was elicited. Thus, an interim baseline challenge was incorporated, when the subjects were challenged only with the final threshold dose identified at the initial visit. If subjects did not react to this dose alone, they were excluded from the study. Studies 1 and 2 also used bilateral allergen challenge at the baseline visit for determination of the threshold dose. To minimize the cumulative ef¬ fect of repetitive challenges on the baseline visit, we switched to unilat¬ eral challenge in studies 3 and 4. This method, similar to that of Moller et al,4 involved alternating eyes with each increasing dose. This minimized the effect of multiple challenges to the same eye and better defined the threshold dose. We then became concerned whether the population of subjects entered into the drug day responded to the allergen in a bilaterally symmetric fashion—a crucial point since the contralateral eye was used as a control. Since the in¬ terim challenge day already excluded any subjects who had responded to a cumulative dose at the initial visit, we decided to revert to bilateral challenge with an interim visit for study 5. This design ensured that (1) the subjects used were bilaterally symmetrical responders, since a moderate reaction (2+) was required in both eyes, and (2) all subjects who were not true responders were eliminated, ie, those who ac¬ tually required the total cumulative dose of allergen instilled on day 1 to achieve a moderate reaction. This study design optimizes the rel¬ evance and predictability of this model for the evaluation of potential an¬ tiallergic agents. It has been shown to be safe in 396 subjects, with systemic reactions reported in only 2 patients. Only 1 case of respiratory distress was reported in approximately 950 chal¬ lenges performed. While this does rep¬ a rare occurrence, patient safety dictates that anaphylactic ther¬
resent
apy must be available in the clinic.
The similarity of the effects of ocu¬ lar challenge to clinical allergic dis¬ ease was apparent. To the examining physician, the pattern of vasodilation, the appearance of the chemosis, eyelid swelling, and tearing, and the subjec¬ tive report of itching was similar to that seen in a patient presenting with clinical ocular allergy. The bilateral symmetry in allergic response observed in these subjects supports the use of internal controls (ie, right eye compared with left eye in
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
the same person) for the assessment of new compounds. The disadvantages of evaluating drugs in the clinical set¬ ting—the innate variability of allergic disease, the subjective nature of symp¬ toms like itching, and the positive ef¬ fects of the placebo—are overcome when internal controls are used in the allergen challenge model. Internal controls have provided a level of sensi¬ tivity from which statistically rele¬ vant data on the efficacy of new com¬ pounds can be derived. We have found that the response to allergen challenge was consistent and reproducible within certain time frames. Of subjects given two baseline challenges, 87% reacted positively a second time, showing that the re¬
sponse to allergen challenge was re¬ producible. Further analysis showed that no statistically significant differ¬ ences
existed in the actual redness and
itching scores of all subjects chal¬ lenged on days 1 and 4. Refractory periods may have been identified in subjects who were chal¬ lenged three times, 3 days apart, and also in subjects who were challenged twice in the same day. However, in these cases, the third challenge was not in naive eyes,
as were the first two; it was in placebo-pretreated eyes, with active drug administered to the fellow eyes. Thus, the decreased response could have been due to the moisturiz¬ ing placebo enhancing the tear film's ability to act as a barrier to antigen attachment. Another possibility is that a slight contralateral effect of the drug was present. This did not inter¬ fere with drug evaluation since the difference between the eyes was usu¬ ally dramatic. Washing away of the allergen was not a factor, since the placebo was never given after, but al¬ ways before, the allergen challenge. Nevertheless, it is possible that this decreased response is a real phenome¬ non relating to an inability of the mast cell to degranulate with the same in¬ tensity after either too short a period between challenges or with too many challenges within a longer time frame. This was a dose-dependent phenome¬ non. In pilot studies, an allergic reac¬ tion of equal magnitude has been elic¬ ited at each of these time points when the dose of the allergen was increased. This suggests that complete mast cell depletion was not a factor. The recog¬ nition of a transient refractory period may help to explain some of the intersubject variability seen in clinical ocu¬ lar allergic disease. The possibility that previous con¬ junctival mast cell experience can al¬ ter its ability to function with future
allergenic stimulation suggests that bilateral challenge should be used at the two baseline visits to maintain an equal allergenic exposure for the eval¬ uation of the drug on day 8. However, in the two studies in which we used unilateral baseline challenges, the agents evaluated were considered ef¬ fective because the differences be¬ tween the eyes were dramatic. Thus, it is possible that while unilateral base¬ line challenges have not been shown to be a factor in the evaluation of the strongly active agents, bilateral chal¬ lenge may maintain the more symmet¬ rical conjunctival mast cell experience needed for the evaluation of margin¬ ally active agents. The finding that a positive skin test was highly predictive of ocular allergic sensitivity (84%) suggests that this simple procedure could be used by ophthalmologists as a diagnostic tool
for ocular allergy. In European coun¬ tries, ocular antigen challenge has been used clinically for years as a method for positively diagnosing the specific allergen responsible for a pa¬ tient's allergic conjunctivitis and also for testing a patient's sensitivity to antigen during immunotherapy.6·7 The reticence in the United States to use ocular challenge for diagnostic pur¬ poses and the inconclusive results ob¬ tained from conjunctival scrapings for eosinophils in mild ocular allergy8 ren¬ der the skin test a potential diagnostic tool when ocular allergy is suspected but not obvious.
signs and symptoms—itching, hyperemia, tearing, chemosis, and eyelid swelling—are elicited by allergen challenge. A highly symmetric, dosedependent reaction is elicited bilater¬ ally, the characteristics of which are independent of the type of allergen used. A refractory period may have
CONCLUSION
This investigation was supported by a grant from Harry and Evelyn Axelrod, Andover, Mass. The Kentucky bluegrass and short ragweed
Using this standardized method and precise grading system, ocular aller¬ gen challenge is safe and reproducible for the evaluation of new antiallergic ophthalmic agents. All of the primary
been identified, and future work in this area may lead to a greater understand¬ ing of the kinetics of mast cell respon¬ siveness. Our latest allergen challenge study design now separates the chal¬ lenges by 7 days, instead of 3, in an ef¬ fort to minimize the potential effect of a
refractory period.
were
provided by Hollister-Stier, Spokane, Wash, provided by ALK Amer-
and the cat extract was ica, Inc, Milford, Conn.
References 1. Abelson MB, Allansmith MR, Friedlander MH. Effects of topically applied ocular decongestants and antihistamines. Am J Ophthalmol.
1980;90:254-257. 2. Udell IJ, Abelson MB. Animal and human
ocular surface response to a topical nonimmune mast cell degranulating agent (compound 48/80). Am J Ophthalmol. 1981;91:226-230. 3. Abelson MB, Butrus SI, Kliman GH, Larson DL, Corey EJ, Smith LM. Topical arachidonic
a model for screening anti-inflammatory agents. J Ocular Pharmacol. 1987;3:63-75. 4. Moller C, Bjorksten B, Nilsson G, Dreborg S.
acid:
The precision of the conjunctival provocation test. Allergy. 1984;39:37-41. 5. Abelson MB, Smith LM. The conjunctival provocation test: a new method for the evaluation of therapeutic agents. Invest Ophthalmol Vis Sci. 1988;29 (suppl):45. Abstract. 6. Moller C, Dreborg S, Lanner A, Bjorksten B.
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
Oral immunotherapy of children with rhinoconjunctivitis due to birch pollen allergy. Allergy. 1986;41:271-279. 7. Moller C, Dreborg S. Cross-reactivity between deciduous trees during immunotherapy, I. in vivo results. Clin Allergy. 1986;16:135-143. 8. Abelson MB, Madiwale NA, Weston JH. Conjunctival eosinophils in allergic ocular disease. Arch Ophthalmol. 1983;101:555-556.
Approach
to
Studying Allergic Conjunctivitis
Mark B. Abelson, MD; Wiley A. Chambers, MD; Lisa M. Smith
\s=b\ To evaluate antiallergic agents, we conducted five allergen challenge studies of increasing refinement. The final study design that evolved included two baseline visits, when skin test-positive subjects were administered a bilateral ocular allergen challenge. At the first visit, the threshold dose of reactivity was determined by increasing allergen doses at 10-minute intervals. At the second baseline visit, 3 days later, the responsive subjects were challenged with the final, highest dose used on visit 1 to assure that the allergic reaction was reproducible and not a cumulative effect of multiple allergen doses. The responsive subjects then returned 3 days later for the drug efficacy evaluation. After a slit-lamp examination, subjects were pretreated with the test drug in one eye and the placebo in the fellow eye in a randomized, double-masked fashion. After 10 minutes, subjects were challenged bilaterally with the allergen dose identified on the previous visits. Postchallenge evaluations of hyperemia, itching, chemosis, eyelid swelling, and tearing were performed at 3, 10, and 20 minutes. Subjects were rechallenged 4 hours after drug administration to assess duration of action.
were again performed at the same intervals as after the initial challenge. A total of 396 subjects were given a baseline allergen challenge; 83.6% responded with a moderate (2+) ocular allergic reaction. Of the 266 given a second baseline challenge, 87.2% responded positively again, suggesting that ocular challenge was highly correlated with skin reactivity and reproducible with a second challenge. No statistically significant difference in redness and itching was found when both eyes were challenged with the same dose of allergen. There was a significant decrease in the allergic reaction of a placebo-pretreated eye compared with its baseline response, and the response to rechallenge was less than the initial one. Of the 950 challenges administered, only two mild systemic reactions occurred, demonstrating the safety of this incremental challenge method. We conclude that this method of allergen challenge is safe, reproducible, and symmetric, but a refractory period may exist when the eye is challenged a third and fourth time in an 8-day period. We now separate the challenge visits by 7 days.
1\/Tany ocular allergic diseases can be
neal vascularization, recurrent cor¬ neal infiltrates, and melting tenden¬ cies associated with hair loss and ec¬ zema in atopic dermatoconjunctivitis, eyelid edema, lichenification in con¬ tact- and drug-related dermatocon¬ junctivitis, and, finally, the classic signs and symptoms of conjunctival hyperemia, edema, and itching associ¬ ated with acute allergic conjunctivitis. The variability of the signs and symptoms of acute allergic conjunc¬ tivitis, diurnally and from day to day, is clearly related to the amount of con¬ junctival allergen contact. This vari¬ able and intermittent stimulus and its resulting response do not provide a suitable baseline for evaluating the
classified by the appearance of certain unvarying clinical character¬ istics, ie, cobblestones and/or limbal infiltrates in vernal conjunctivitis, corAccepted for publication August 22,1989. From the Department of Immunology, Eye Research Institute, and Department of Ophthalmology, Harvard Medical School, Boston, Mass (Dr Abelson and Ms Smith), and the Division of Anti\x=req-\ Infective Drug Products, Center for Drug Evaluation and Research, Food and Drug Administration, Rockville, Md (Dr Chambers).
This article contains the views of the authors and not necessarily those of the Food and Drug Administration. Reprint requests to 20 Staniford St, Boston, MA 02114 (Dr Abelson).
Slit-lamp examinations
(Arch Ophthalmol. 1990;108:84-88)
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
efficacy of potential
agents. More difficulties
antiallergic are encoun¬
tered if
one uses a drug vehicle as a control, since this may attenuate the allergic response in several ways: (1) it
may act
as a
tear substitute that
can
minimize
antigen penetration to the conjunctiva; (2) it may act as a diluent to the antigen; or (3) it may enhance the conjunctival transit time of the antigen. In many of our clinical envi¬ ronmental studies, these mechanical properties of placebo controls have been shown to elicit a "drug effect" in approximately 70% of the treated sub¬ jects. To precisely evaluate and compare anti-inflammatory agents, mast cell stabilizers, antihistamines, vasocon¬ strictors, phosphodiesterase inhibi¬ tors, and anti-IgE compounds, a con¬ trolled and reproducible situation is required. Our previous models of acute allergic conjunctivitis, histamine,1 48/ SO,2 or arachidonic acid,3 administered topically to the eye, have been helpful in studying allergic ocular disease but they do not have the clinical validity of antigen-induced mast cell degranulation. We have, therefore, modified the conjunctival provocation test of Moller et al4 for the evaluation of antiallergic compounds in a study design that min¬ imizes the potential variables of the allergen challenge model.5 We report data obtained from this standardized technique and various study designs that have proved to be safe and repro¬ ducible in 396 subjects. SUBJECTS, MATERIALS, The
subject selection
AND METHODS
was an
important
step in performing these antigen challenge
studies. Only normal, healthy volunteers who were not using any type of topical agent or systemic medication that might have interfered with results were allowed to enter the studies. Inclusion criteria were as follows: (1) able and willing to give consent; (2) between the
ages of 18 and 65 years; (3) either sex and of any race; (4) willing and able to make required study visits; (5) able to follow in¬
Table 1.—Scoring System to Measure the Signs and of Allergic Conjunctivitis
structions; (6) history of symptoms of a clinically active allergic conjunctivitis; (7) positive diagnostic test (skin or radioallergosorbent tests) for allergic disease; if the subject had no historical documentation of this, a skin prick test was performed; (8)
Redness, =
=
=
=
0 None Mild 1+ =
=
Subjects were asked to refrain from the of systemic medications that could affect test measurements, particularly antihistamines, decongestants, and nonste¬ roidal anti-inflammatory agents. If sub¬ jects were taking medications that would not affect test measurements, the dosage
3+
=
(detectable with slit lamp, conjunctiva separated from sclera) Moderate (visually evident, raised conjunctiva, especially at the limbal area) Severe (ballooning of conjunctiva)
None Mild tearing (eyes feel slightly watery) Moderate tearing (blows nose occasionally) Severe tearing (tears rolling down cheeks) Itching (to be graded by subject) None 0 =
or
Concomitant Medications
=
0 1+ 2+ 3+
bacterial or viral ocular in¬ fection; (2) presence of dry-eye syndrome, blepharitis, follicular conjunctivitis, iritis,
rheumatoid arthritis that can be associated with dry-eye syndrome; (6) use of any top¬ ical ophthalmic solutions during or at least 2 weeks prior to the study, including tear substitutes; subjects who had taken oph¬ thalmic medications that require longer than a 2-week washout; and (7) exhibiting signs and symptoms of allergic conjunctivi¬ tis on taking the baseline examinations prior to entry into the study on days 1, 4, or 8.
2+
Tearing
a
preauricular lymphadenopathy; (3) wearing of contact lenses during the course of the study; (4) taking any systemic med¬ ication that might have interfered with the study; (5) presence of any significant illness that could interfere with the study, partic¬ ularly any autoimmune disease such as
eyelid swelling
None 0 Mild 1+ Moderate 2+ Severe 3+
willing to avoid disallowed medications during the study period; and (9) a success¬ ful challenge, inducing at least moderate itching (2+) and redness. Exclusion criteria were as follows: (1) presence of
*
Symptoms
=
=
=
=
1+ 2+
3+
=
Mild
(intermittent tickling sensation)
Moderate (continual awareness but without the desire to rub) Severe (continual awareness with the desire to rub the eyes)
Incapacitating Itching (subject insists on rubbing eyes) Photographs depicting these gradations of redness are used as 4+
*
=
=
=
Table 2.—Summary of the Five Allergen
Challenge Study No.
Method Bilateral Bilateral Unilateral Unilateral Bilateral
Baseline Visits 1 1 2
2 2
Subjects Screened 33 63 147 69 109
a
guide.
Challenge Studies
Nonresponders to Allergen
Others Discontinued
Completed
4
21
45
2
100
18
16
75
use
and administration of these medications kept constant throughout the-dura¬ tion of the study. All concomitant medica¬ tion was thoroughly documented. were
Materials
Three types of antigens were used in these studies: Kentucky bluegrass (Poa pratensis), short ragweed (Ambrosia artemisiaefolia), and allergenic extract of cat hair and dander (Felis domesticus). Serial dilutions were made from the stock antigen solutions (bluegrass and ragweed: 100000 antigen units [AU]/mL; cat extract: 50 000 AU/mL) for incremental use in the allergen challenge. Six dilutions of bluegrass and ragweed were used, doubling in concentra¬ tion with each step, from 19 to 625 AU/25pL dose. Four dilutions of cat extract were used: 125,250,625, and 1250 AU/25-mL dose. Methods: Final
Study Design
Initial Visit.—Demographic data, medical
and medication history, and informed con¬ sent were obtained from each subject. If prior written documentation of allergy, identified by positive skin or radioallergosorbent tests, was unavailable, a skin test for the suspected allergen was then per¬ formed. The standardized disposable prick test needle (Morrow Brown, HMB Medical
Concepts, High Field House, Derby, En¬ gland) was saturated with stock solution of allergen (ragweed or grass: 100 000 AU/mL; cat: 50000 AU/mL), and the swabbed fore¬ arm of the subject was pierced. A positive
reaction was any definitive wheal and flare present after 10 minutes. The size of the re¬ action was then drawn and the form kept on file as written documentation. If the results of the skin test were positive for any one of the three allergens, an oph¬ thalmologic examination was given to ex¬ clude subjects currently exhibiting clini¬ cally significant signs and symptoms of al¬
lergic conjunctivitis (any itching or conjunctival redness >1+) (Table 1). Based on the subject's allergic sensitiv¬ ity, one of the three allergens was used for the conjunctival allergenic challenge. If the subject did not respond to the first allergen tested, he or she was challenged with an¬ other of the allergens to which he or she had elicited a positive skin reaction. A bilateral ocular allergen challenge was performed in which 25 pL of the diluent, and then 25 pL of increasing doses of allergen, were administered bilaterally with an Eppendorf digital pipette (Brinkman Instru¬ ments, Ine, Westbury, NY) and a sterile tip. A positive reaction was defined as at least a redness and itching score of 2+ in both eyes within 10 minutes of the last adminis¬
tered dose. Visit 2.—Those subjects who responded on visit 1 and who had met all other inclu¬ sion and exclusion criteria were entered
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
into visit 2, day 4. At this visit, a baseline slit-lamp examination was performed to exclude any subjects with signs or symp¬ toms of allergic conjunctivitis. Subjects with nonallergic eyes were then adminis¬ tered the final dose of allergen identified as the threshold dose for 2+ responsiveness on visit 1. If this allergen dose again elicited a 2+ itching and redness response in both eyes, the subject was entered into visit 3,
8. Visit 3.—Those subjects who had re¬ sponded at visits 1 and 2 with 2+ itching and redness to the same dose of allergen and who had met all other entry inclusion
day
and exclusion criteria were entered into the
study on day 8. Subjects were excluded who had baseline signs or symptoms of allergic conjunctivitis. Ten minutes after drug treatment in one eye and placebo in the contralateral eye, bilateral allergen chal¬ lenge was performed. Postchallenge evalu¬ ations occurred at 3, 10, and 20 minutes. Four hours after drug administration, sub¬ jects were rechallenged and reexamined to assess the drug's duration of action. To maximize the sensitivity and accuracy of the allergen challenge model, new infor¬ mation gained from each allergen challenge study was incorporated into subsequent protocol designs. While the final study de¬ sign that has evolved is outlined in detail above, the first four studies must be men¬ tioned briefly to understand the evolution¬ ary process. In the first two
studies, a bilateral ocular
Fig 1.—The conjunctival, episcleral, and ciliary hyperemia elicited by ocular allergen challenge to a sensitized subject.
Fig 2.—Chemosis elicited by ocular allergen challenge subject.
to a sensitized
allergen challenge was given at the baseline visit and 1 week later at the drug evaluation visit. No rechallenge was performed after drug administration. To further refine the study population, a second baseline challenge visit was added to the next two studies. In addition, we mod¬ ified the allergen challenge procedure at these baseline visits, performing it unilat¬ erally instead of bilaterally, in a method similar to that of Moller et al." These changes provided two advantages: (1) any subjects who had reached the moderate al¬ lergic response (2+) on visit 1 only with the cumulative allergen dose received with re¬ petitive challenges would be eliminated and (2) only subjects who had a reproducible reaction, even at the lowest dose, would be included in the test drug evaluation. Bilat¬ eral challenge was performed at the drug evaluation visit, and a rechallenge was ad¬ ministered either 2 or 4 hours after drug instillation to evaluate the drug's duration
of action. Since unilateral baseline challenge did not require a moderate response (2+) in both eyes, and we were concerned about the symmetry of the response, we reverted to bilateral baseline challenge in the final study. The second baseline challenge visit was retained. Table 2 gives a summary of the five studies, with differences in protocol design outlined, and it shows the number of sub¬ jects screened and entered. The sample size varies depending on the type of study de¬ signs used. When the evaluation of these data subsets required statistical analyses, the paired two-tailed t test and SignedRank Sum Test were used. RESULTS
Safety A total of 396 skin test-positive sub¬ jects who had a history of allergic con¬ junctivitis were entered and chal¬ lenged on the initial baseline visits of the five studies. For the completion of these studies, approximately 950 ocu¬ lar challenges were performed. Only 2 cases of systemic reactions occurred.
72-Hour Rechallenge
First Challenge
Fig 3.—The number of skin test-positive subjects who had a history of allergic conjunctivitis and who responded positively to ocular allergen challenge, and the number of ocular-positive responders who reacted positively a second time from the total number of subjects challenged in our
laboratory.
In one asthmatic patient, mild bron¬ chial constriction, without wheezing, occurred 15 minutes after ocular aller¬ gen challenge and was immediately treated with intramuscular epineph¬ rine hydrochloride and diphenhydramine hydrochloride. Respiratory allergic signs disappeared within 24 hours. One case of urticaria occurred approximately 8 hours after challenge and lasted for approximately 24 hours. One case of exaggerated local periorbital swelling was also reported. This occurred approximately 12 hours after challenge and lasted for approx¬ imately 3 hours. Similarity to Clinical Disease The clinical relevance of this model was supported by the appearance of all
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
signs and symptoms of allergic conjunctivitis: itching, tearing, slit¬ lamp evidence of conjunctival vasodilation, chemosis, and eyelid swelling (Figs 1 and 2). the
Relationship to Positive Dermatologie Reactivity A positive skin test was predictive of a positive ocular challenge in 331 (83.6% ) of 396 subjects who had a his¬
tory of allergic conjunctivitis.
Symmetry of Allergic Reaction Evidence for a bilaterally similar ocular response was necessary to use the contralateral eye as the control. Therefore, at the baseline visits of the final study (completed number of sub¬ jects, 75) both eyes were simulta-
neously challenged and data from each eye were recorded individually. The allergic response was found to be sim¬ ilar in both eyes, with no statistically significant differences noted for itch¬ ing (P .32) or for conjunctival (P m .16), episcleral (P .60), or cili¬ ary hyperemia (P .85). =
=
=
Reproducibility and Refractoriness
Response On day 1, of the 331 subjects who showed a positive ocular response to challenge, 266 were given a second baseline challenge before the drug test day. Of these 266 subjects, the positive ocular allergic response was reproduc¬ ible in 232 (87.2% ) of the subjects (Fig of
3).
In study 3, in which 55 subjects were initially challenged, it was possible to determine the change in the redness and itching scores from day 1 to day 4 to day 8. The actual allergic scores (redness and itching) obtained on day 1 were reproducible on day 4, with no
statistically significant found
differences
(itching: .727; redness: .537). After placebo pretreatment on day 8, the redness and itching scores were significantly decreased compared with results obtained on day 4 (itching: .0001; redness: .0001). Data were also analyzed from pla¬ cebo eyes challenged twice in one day. Thirty-four subjects who received pla¬ cebo in one eye and drug in the con¬ tralateral eye (the remainder of the subjects of this study were adminis¬ tered active drug in both eyes) were rechallenged at 2 hours in study 3. Redness (P .021) and itching (P .006) were significantly decreased in the placebo eyes at rechallenge com¬ pared with the initial scores. Fortyseven subjects received placebo in one eye and drug in the contralateral eye and were rechallenged after 4 hours in study 4. A significant decrease in itch¬ ing (P .039), but not redness (P .163), was noted in the placebo eyes at rechallenge. =
=
=
=
=
=
=
=
COMMENT
We have been performing ocular al¬ lergen challenge studies for the past year for the evaluation of antiallergic agents. With the completion of each study, more information was gained about the ocular response to challenge, causing us to reevaluate and modify our original protocol design. At first, in studies 1 and 2, only one baseline chal¬ lenge was required for entry into the drug test evaluation. However, the question arose as to whether the reac¬ tion elicited on the drug day, when only the final highest dose of allergen was
administered, was similar to that on the screening day, when multiple doses were given until a positive reaction was elicited. Thus, an interim baseline challenge was incorporated, when the subjects were challenged only with the final threshold dose identified at the initial visit. If subjects did not react to this dose alone, they were excluded from the study. Studies 1 and 2 also used bilateral allergen challenge at the baseline visit for determination of the threshold dose. To minimize the cumulative ef¬ fect of repetitive challenges on the baseline visit, we switched to unilat¬ eral challenge in studies 3 and 4. This method, similar to that of Moller et al,4 involved alternating eyes with each increasing dose. This minimized the effect of multiple challenges to the same eye and better defined the threshold dose. We then became concerned whether the population of subjects entered into the drug day responded to the allergen in a bilaterally symmetric fashion—a crucial point since the contralateral eye was used as a control. Since the in¬ terim challenge day already excluded any subjects who had responded to a cumulative dose at the initial visit, we decided to revert to bilateral challenge with an interim visit for study 5. This design ensured that (1) the subjects used were bilaterally symmetrical responders, since a moderate reaction (2+) was required in both eyes, and (2) all subjects who were not true responders were eliminated, ie, those who ac¬ tually required the total cumulative dose of allergen instilled on day 1 to achieve a moderate reaction. This study design optimizes the rel¬ evance and predictability of this model for the evaluation of potential an¬ tiallergic agents. It has been shown to be safe in 396 subjects, with systemic reactions reported in only 2 patients. Only 1 case of respiratory distress was reported in approximately 950 chal¬ lenges performed. While this does rep¬ a rare occurrence, patient safety dictates that anaphylactic ther¬
resent
apy must be available in the clinic.
The similarity of the effects of ocu¬ lar challenge to clinical allergic dis¬ ease was apparent. To the examining physician, the pattern of vasodilation, the appearance of the chemosis, eyelid swelling, and tearing, and the subjec¬ tive report of itching was similar to that seen in a patient presenting with clinical ocular allergy. The bilateral symmetry in allergic response observed in these subjects supports the use of internal controls (ie, right eye compared with left eye in
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
the same person) for the assessment of new compounds. The disadvantages of evaluating drugs in the clinical set¬ ting—the innate variability of allergic disease, the subjective nature of symp¬ toms like itching, and the positive ef¬ fects of the placebo—are overcome when internal controls are used in the allergen challenge model. Internal controls have provided a level of sensi¬ tivity from which statistically rele¬ vant data on the efficacy of new com¬ pounds can be derived. We have found that the response to allergen challenge was consistent and reproducible within certain time frames. Of subjects given two baseline challenges, 87% reacted positively a second time, showing that the re¬
sponse to allergen challenge was re¬ producible. Further analysis showed that no statistically significant differ¬ ences
existed in the actual redness and
itching scores of all subjects chal¬ lenged on days 1 and 4. Refractory periods may have been identified in subjects who were chal¬ lenged three times, 3 days apart, and also in subjects who were challenged twice in the same day. However, in these cases, the third challenge was not in naive eyes,
as were the first two; it was in placebo-pretreated eyes, with active drug administered to the fellow eyes. Thus, the decreased response could have been due to the moisturiz¬ ing placebo enhancing the tear film's ability to act as a barrier to antigen attachment. Another possibility is that a slight contralateral effect of the drug was present. This did not inter¬ fere with drug evaluation since the difference between the eyes was usu¬ ally dramatic. Washing away of the allergen was not a factor, since the placebo was never given after, but al¬ ways before, the allergen challenge. Nevertheless, it is possible that this decreased response is a real phenome¬ non relating to an inability of the mast cell to degranulate with the same in¬ tensity after either too short a period between challenges or with too many challenges within a longer time frame. This was a dose-dependent phenome¬ non. In pilot studies, an allergic reac¬ tion of equal magnitude has been elic¬ ited at each of these time points when the dose of the allergen was increased. This suggests that complete mast cell depletion was not a factor. The recog¬ nition of a transient refractory period may help to explain some of the intersubject variability seen in clinical ocu¬ lar allergic disease. The possibility that previous con¬ junctival mast cell experience can al¬ ter its ability to function with future
allergenic stimulation suggests that bilateral challenge should be used at the two baseline visits to maintain an equal allergenic exposure for the eval¬ uation of the drug on day 8. However, in the two studies in which we used unilateral baseline challenges, the agents evaluated were considered ef¬ fective because the differences be¬ tween the eyes were dramatic. Thus, it is possible that while unilateral base¬ line challenges have not been shown to be a factor in the evaluation of the strongly active agents, bilateral chal¬ lenge may maintain the more symmet¬ rical conjunctival mast cell experience needed for the evaluation of margin¬ ally active agents. The finding that a positive skin test was highly predictive of ocular allergic sensitivity (84%) suggests that this simple procedure could be used by ophthalmologists as a diagnostic tool
for ocular allergy. In European coun¬ tries, ocular antigen challenge has been used clinically for years as a method for positively diagnosing the specific allergen responsible for a pa¬ tient's allergic conjunctivitis and also for testing a patient's sensitivity to antigen during immunotherapy.6·7 The reticence in the United States to use ocular challenge for diagnostic pur¬ poses and the inconclusive results ob¬ tained from conjunctival scrapings for eosinophils in mild ocular allergy8 ren¬ der the skin test a potential diagnostic tool when ocular allergy is suspected but not obvious.
signs and symptoms—itching, hyperemia, tearing, chemosis, and eyelid swelling—are elicited by allergen challenge. A highly symmetric, dosedependent reaction is elicited bilater¬ ally, the characteristics of which are independent of the type of allergen used. A refractory period may have
CONCLUSION
This investigation was supported by a grant from Harry and Evelyn Axelrod, Andover, Mass. The Kentucky bluegrass and short ragweed
Using this standardized method and precise grading system, ocular aller¬ gen challenge is safe and reproducible for the evaluation of new antiallergic ophthalmic agents. All of the primary
been identified, and future work in this area may lead to a greater understand¬ ing of the kinetics of mast cell respon¬ siveness. Our latest allergen challenge study design now separates the chal¬ lenges by 7 days, instead of 3, in an ef¬ fort to minimize the potential effect of a
refractory period.
were
provided by Hollister-Stier, Spokane, Wash, provided by ALK Amer-
and the cat extract was ica, Inc, Milford, Conn.
References 1. Abelson MB, Allansmith MR, Friedlander MH. Effects of topically applied ocular decongestants and antihistamines. Am J Ophthalmol.
1980;90:254-257. 2. Udell IJ, Abelson MB. Animal and human
ocular surface response to a topical nonimmune mast cell degranulating agent (compound 48/80). Am J Ophthalmol. 1981;91:226-230. 3. Abelson MB, Butrus SI, Kliman GH, Larson DL, Corey EJ, Smith LM. Topical arachidonic
a model for screening anti-inflammatory agents. J Ocular Pharmacol. 1987;3:63-75. 4. Moller C, Bjorksten B, Nilsson G, Dreborg S.
acid:
The precision of the conjunctival provocation test. Allergy. 1984;39:37-41. 5. Abelson MB, Smith LM. The conjunctival provocation test: a new method for the evaluation of therapeutic agents. Invest Ophthalmol Vis Sci. 1988;29 (suppl):45. Abstract. 6. Moller C, Dreborg S, Lanner A, Bjorksten B.
Downloaded From: http://archopht.jamanetwork.com/ by a University of Pittsburgh User on 10/01/2013
Oral immunotherapy of children with rhinoconjunctivitis due to birch pollen allergy. Allergy. 1986;41:271-279. 7. Moller C, Dreborg S. Cross-reactivity between deciduous trees during immunotherapy, I. in vivo results. Clin Allergy. 1986;16:135-143. 8. Abelson MB, Madiwale NA, Weston JH. Conjunctival eosinophils in allergic ocular disease. Arch Ophthalmol. 1983;101:555-556.
