Clinical, physiologic, and psychologic comparison of treatment by cromolyn or theophylline in childhood asthma

Clinical, physiologic, and psychologic comparison of treatment by cromolyn theophylline in childhood asthma C. Springer, S. Godfrey,

or

M.D., B. Goldenberg, Ph.D., I. Ben Dov, M.D., and M.D., Ph.D., F.R.C.P. Jerusalem, Israel

Treatment of chronic perennial childhood asthma with cromolyn sodium (CS) or long-acting theophylline @AT) was compared in 13 children by a double-blind, cross-over trial in which each drug was administered for 4 wk. LAT was administered twice daily in individual doses to elicit adequate blood levels (mean 14.0 ? 1.6 SEM, micrograms per milliliters). Powdered CS was administered by inhalation in a dose of 20 mg four times daily. The patients were followed by means of a daily symptom and drug consumption diary and twice daily peak flow measurements. A series of psychologic studies were performed in the third week of each month. Exercise and histamine bronchial provocation tests were performed at the end of each month. There were no sign$cant dtrerences between the month in which CS was taken and the month in which LAT was taken in terms of diary scores, peakjow rates, additional drug consumption, or in bronchial reactivity to exercise or histamine at the end of each month. The results of psychologic tests reflecting visual-spatial planning were significantly better for the four children receiving CS with lower IQ scores (87 to 105) but not for the nine children with higher IQ scores (111 to 134). The possibility that the psychologic pet$ormance of children with lower intelligence may be adversely affected by theophylline requires further evaluation. (J ALLERGY

CUNIMMUNOL76.64-9, 1985.)

Both LAT and CS are effective in controlling symptoms of chronic asthma in children’~ * with little evidence of clinically relevant differences. 3 Although CS is known to be virtually without systemic side effects,’ theophylline preparations have various side effects including central nervous system stimulation (within the therapeutic range of serum concentrations). This stimulation can be manifested as disturbances in delicate muscular coordination,4 speech disturbances,5 depression,6 nervousness, restlessness, insomnia, tremors hyperesthesia,’ and even “madness”.* Clearly, side effects such as these could be a serious problem for school age patients who are commonly treated with LAT preparations, and Furukawa et al.,9 in a recent letter, have reported significant improvement in six children with learning and behavior problems who were taking theophylline when treatment was changed to cs. Objective psychologic studies have suggested that From the Departments of Pediatrics and Child Psychiatry, Hadassah University Hospital, Jerusalem, Israel. Received for publication May 3, 1984. Accepted for publication Oct. 20, 1984. Reprint requests: Prof. S. Godfrey, Department of Pediatrics, Hadassah University Hospital, Mount Scopus, POB 24035, Jerusalem, 91240 Israel. 64

Abbreviations used LAT: Long-acting theophylline CS: Cromolyn sodium

children with severe asthma may suffer from some cognitive problems’“-‘2 but have normal reaction times. In one study it was found that the cognitive problems lay particularly in visualizing and remembering spatial configurations.” It is not certain whether asthma results in neuropsychologic difficulties or whether they occur independently. Almost nothing is known of the effects of drugs used to treat asthma on psychologic parameters, although Suess and Chai13 in a preliminary study reported visual retention and spatial configuration difficulties that they attributed to steroid therapy. We therefore decided to compare the clinical, physiologic, and psychologic effects of CS and LAT therapy in a group of children with chronic, perennial asthma with the use of a double-blind prospective protocol. METHODS Thiieen children with chronic, perennial asthma ages 8 to 13 yr (mean 10.5 yr, Table I) attending the Pediatric

VOWME 76 NlJMBFR 1

Cromolyn

Pulmonary Clinic in Hadassah University Hospital were included in the study. All had moderate asthma defined as requiring continuous daily medication for at least the preceding 6 mo but not requiring continuous steroids or any steroids in the previous 3 mo. The previous regular daily medications were LAT. usually Theo-dur (Key Pharmaceutical. Inc., Miami, Fla.) or CS with additional selective pZsympathomimetic (albuterol) if it were needed. The protocol consisted of a double-blind, randomized, cross-over study with I mo each of treatment by CS or LAT. At the end of each month, the trial drugs were stopped for 2 days to enable studies of bronchial reactivity to be performed so that there was in effect a 2-day washout period between the months. Each child received a tablet twice daily at 12 hourly intervals. either placebo or active LAT, and inhaled four doses of drug through a Spinhaler (Fison’s Corp., Bedford. Mass. ), either placebo or active CS. Before entering the trial, the dose of LAT was determined for each child to elicit serum theophylline concentrations of IO to 20 p&ml 4 hr after receiving the drug. Theophylline blood concentrations were measured again before psychologic tests in the third week of each month and at the end of each month 12 to 16 hr and 36 to 40 hr after stopping the trial drugs. Serum theophylline levels were measured by an enzyme immunoassay (Syva, Palo Alto, Calif.). In addition to CS and LAT, the children were allowed free use of albulerol by metered-dose inhaler except for the 4 hr before any lung function or psychologic test.

Clinical evaluation Each child used a standard asthma diary card in which nighttime and daytime symptoms were scored and recorded throughout the trial. In addition, peak expiratory flow rates were measured morning and evening before taking the medications every day by use of a Wright peak flow gauge. The best of three efforts was accepted on each occasion. All medications taken were recorded in the diary. The results of the study were evaluated in terms of diary scores, peak flow rates, supplemental drug consumption, and symptomfree days.

Psychologic

evaluation

Initially, a screening intelligence test (WISC vocabulary) was administered to rule out children of unacceptably low intelligence. In addition, all parents were interviewed to inquire about any serious psychologic or physical problem that might influence performance on cognitive testing. An attempt to obtain information from teachers was abandoned because of lack of adequate cooperation to make the data reliable. Thirteen children were found to be suitable for the study. A baseline battery of psychologic tests was administered in order to familiarize the child with the testing situation, thereby diminishing the effects of anxiety and eliminating the novelty effects on the first administration. The subjects were then tested 3 wk after the beginning of the administration of either CS or LAT. On the day of the test, the child took his prescribed trial medication but no additional drugs. No child was taking nasal decongestants. The tests lasted 2 hr with a short break and were performed

or theophylline

TABLE I. Patient

data

No.

Age (vrl

Sex

I 2 3 4 5 6 I 8 9 10 11 12 13

11.5 10.5 il.5 9.0 13.0 11.0 9.0 X.0 I 1 .o X.5 12.0 12.0 10.0

M M M F F M M M F M M M F

in cniidhoori

___-.Height (cm) 141 141 143 14:! 163 14x 137 131 143 130 143 IS1 132

asthma

65

. . . ~_. . .- .-...--Wright ikgi

i? 17 “/

at a similar time of day for each trial period. Three groups of tests were selected to attempt to measure attention: visualspatial planning, learning, and memory.‘i”’ and these are summarized in Table II.

Bronchial

reactivity

evaluation

At the end of each month, 12 to 16 hr and 36 to 40 hr after stopping the trial drugs, the child undertook routine lung function tests and studies of bronchial reactivity by both exercise and histamine challenge. l3aseIine and postchallenge FEV was measured with the use of a Fleisch pneumotachograph and integrator (478Q4A Pulmonary Calculator System, Hewlett-Packard Co., Cupertino, Calif.). The histamine challenge was performed by use of a simple modification of the method of Cockroft et al. ” Doubling concentrations of histamine acid phosphate (0.03 to 10 mgi ml) were inhaled for 1 min during tidal breathing at 5-minute intervals until there was a fall in FEV of 20% or until the highest concentration was inhaled. The provocation concentration of histamine required to produce a Fall in FEV of 20% (PCJ was calculated from the log dvsc response curve. The exercise test was performed as described previously.” Each exercise test consisted of 6 min of cycling at a fixed load on a cycle ergometer (Lode Instruments, N. V., Holland) in such a way as to achieve approximately 60% of predicted maximal oxygen consumption.“J Conditioned air of constant temperature and humidity was inspired in all tests throughout the study period. Ventilation heart rate and gas exchange were recorded during the tests, and there were no significant differences between the CS and LAT months. The severity of exercise-induced asthma was calculated as the maximal percent decrease in FEV, (AFEV,) after exercise in relation to the baseline, preexercise value.

Results The mean dose of LAT and serum theophylline levels are illustrated in Table III. The mean theophylline

levels

during

the trial

were

similar

to those

66 Springer

TABLE

II. Psychologic Test

Attention CPT

J. ALLERGY CLIN. IMMUNOL. JULY 1985

et al.

Description

Reference

tests

Automated presentation of two visual stimuli; subject tested for reaction time and accuracy of identification Coding (WISC-R) Match and copy symbols with use of guide of symbols associated with numbers Digit span Repetition of digits for(WISC-R) ward and backward Visual spatial planning tests Porteus mazes Series of paper and pencil maze problems increasing in difficulty Reitan TM Subject draws a line from point to point numbered sequentially PLT Learning a sequence of dots on a grid from visual spatial clues Memory and learning tests Rappel Visual and learning task involving various degrees of visual clues Benton visual re- Reproduction of figures tention test after 10 set of exposure Vocabulary Defining a list of words (WISC-R, pretest only) Mazes (WISC-R, Eight maze problems pretest only) similar to Porteus mazes

CPT = continuous performance PLT = position learning test.

III. Mean ( % SEMI dose of LAT and serum theophylline levels for the month on LAT therapy

TABLE

tests

test; TM = trail

17

15

Theophylline dose (mg/kg/day) Serum theophylline level (pgiml) During preliminary stabilization. Before psychologic tests Before first exercise test Before first histamine test Before second exercise test Before second histamine test

IV. Results of clinical studies daily peak flow rate measurements, (*SEMI

TABLE

15

16

17

18

19

20

15

15

making;

during the preliminary stabilization, and the levels fell as expected at the end of the month when the drug was stopped before exercise and histamine challenges. On an individual basis, all patients had theophylline levels before the psychologic tests between 10.6 and 23.0 pg/ml except for three (patients 2, 9, and 13) in whom levels were 7.5, 6.5, and 8.2 pg/ml, respectively. In the month when CS was being taken, no theophylline was detected in any of the blood sam-

Twice daily peak flow rate (% predicted) Symptom-free days (% total) Symptom score (% maximum) Doses of albuterol (per mo)

19.2

(0.8)

14.8 14.0 9.9 9.4 0.2 0.7

(0.8) (1.6) (1.4) (1.9) (0.1) (0.3)

and mean

SCG

LAT

87 (5)

87 (4)

59 (10)

60 (7)

7 (2)

6 (1)

16 (8)

12 (5)

ples that were collected as part of the double-blind nature of the study. The results of the clinical and peak flow rate measurements are illustrated in Table IV. There were no significant differences with the use of the paired t test between the month of CS administration and the month of LAT administration in terms of peak expiratory flow rates (87 ? 5% SEM predicted and 87 + 4% predicted), symptom free days (59 t 10% SEM total days and 60 * 7% total), symptom score (7 + 2% SEM maximum and 6 + 1% maximum), and additional albuterol (16 + 8 SEM per month and 12 f. 5 per month). On an individual basis all four indices were better during CS administration in three children and better during LAT administration in two children, whereas there was no difference in eight children. There was no relationship between individual plasma theophylline levels measured during the trial and the efficacy of LAT as compared with CS. There was no relationship between the results obtained during the study and the medication received by the patients before entering the trial. It was found that three of the psychologic tests were unhelpful because of learning effects (coding, trail making B, and word list), and they were excluded

VOLUME 76 NLMBE8 1

from the analysis. There were some initial technical problems with the reaction time equipment, and once each with the mazes and Rappel tests so that some children missed some of the individual tests. Only those tests performed by the child during both months have been analyzed. The results in terms of whether the child performed better in 1 mo or the other or no difference between months are illustrated in Table V. Bq use of a rank sign test, there were no significant differences between the LAT and CS for any of the attention tests individually or together. Although the individual test of visual-spatial planning also demonstrated no difference, the three tests taken together were significantly better performed during the month of CS administration (p < 0.05). None of the learning and memory tests demonstrated any significant differences. There was no relationship between medication taken before entering the study and the results of the psychometric tests. It was noted that four of the children had relatively low IQ scores (87 to 105) compared with the other nine ( I 11 to 134) based on four components of the WISC-R tests,ls namely, vocabulary, digit span, coding,. and mazes. The results of testing in these two groups are summarized in Table VI. There was a marked advantage for CS in the lower IQ group in terms of visual-spatial planning tests (p < O.Ol), but none of the other differences in this group or in any of the tests in the higher IQ group were significant. The results for the three visual-spatial planning tests for all children in terms of IQ and the drug eliciting the better result (if there were any) is demonstrated in Fig. 1. It can be observed that patients with the lower IQ performed better during CS administration, whereas there was no difference in the performance of patients in the higher IQ range (over 110). There was no correlation between individual plasma theophylline levels measured before the psychologic tests and the performance of the children in these tests. The results of bronchial provocation studies are summarized in Fig. 2. There was no significant difference between baseline FEV before the challenge on the first day or second day after stopping treatment when the months of CS and LAT administration were compared. On the second day after stopping LAT, the baseline FEV dropped significantly compared with the first day (p < 0.01). Although the FEV also fell after stopping CS, this was not significant. There were no significant differences in the postexercise fall in FEV or the provocation dose of histamine between the month CS was administered and the month LAT was administered whether tested on the first or second day after stopping treatment.

Cromolyn

or theophylline

TABLE V. Number which performance LAT, or neither

in cbildhnod

asrhma

67

of psychologic tests in was better with CS,

Attention tests CPT-reaction time UT-errors Digit span-forward Digit span-backward All attention tests Visual spatial planning tests Mazes TMA PLT All visual spatial planning Memory and learning tests Rappel Benton correct responses Benton errors All memory learning tests All tests

LAT

4 2 4 4 14

2 4 8 7 17

8 8 7

4 4 I

t!

23”

9

h

5 5

5 s

5 15

4 I4

52

40

CPT = continuous performance PLT = position learning test. *Significantly better (p 10.05).

test:

MO

difference

CS

fl ‘I

i

i -2 )! 24

TM

= trail

making;

TABLE VI. Results of psychologic tests in subjects with high and low IQs demonstrating number of tests in which performance was better with CS, LAT, or neither CS

Tests

Attention High IQ Low

Y 5

IQ

Visual spatial planning High IQ Low IQ* Memory and learning High IQ Low IQ *Significant

LAT

difference

between

13 4

No cwerenoe

6 ~3

l? II

Y 0

h 0

7

IO

9

8

4

II

CS and LA’T (p <:I) 011

DISCUSSION

This study confirms the results of a previous study’ that demonstrated that CS and a theophylline derivative were equally effective in controlling symptoms of chronic asthma in children. In that study regular short-acting aminophylline was administered every 6 hr. Although the advent of LAT preparations is clearly advantageous in terms of administration and probable

66 Springer

J. ALLERGY CLIN. IMMUNOL. JULY 1995

et al.

SCG

TESTS BETTER ON: CS Mazes

Neither

LAT

Lat

T.M. PLT t4aresT.M. PLT Mazes TM. PLT

l&O 130 g

120 - i.

z 5 ;.E

HO - 0

ii

l 100

- 0

90

-

80

-

l

l

l

FIG. 1. Relationshii, between IQ and spatial planning tests for the two treatment months illustrating whether the results for each test in each subject was better when one or the other drug or neither drug was administered. M, mazes; T, trail making; f, position learning.

patient compliance, it is interesting that the present results with LAT are almost identical to those of Hambleton et aL3 by use of a short-acting theophylline preparation. All the present patients were relatively well controlled on their usual chronic medication before entering and during the study, and hence they had few symptoms as evidenced by the low diary scores and good, twice-daily peak flow rate measurements. Since they had all been followed in the clinic and had been persistently symptomatic before starting treatment, it was not believed to be practical or ethical to include an additional month of plabeco treatment alone. CS was believed to act by suppressing mediator release from mast cells, and it has been demonstrated to diminish both the immediate and late reactions to allergic bronchial challenge.24y *’ It has been demonstrated that allergic bronchial stimulation increases the general level of airway reactivity to nonspecific stimuli,26 and it has been claimed that treatment for 1 mo with CS results in a sustained reduction in nonspecific bronchial reactivity to hyperventilation but not to challenge with methacholine. *’ In the present study we found no differences in bronchial reactivity to exercise or histamine after 1 mo of treatment with either CS or LAT. In the study of Griffin et al.,*’ the hyperventilation and methacholine challenges were performed less than 2 hr after the last dose of CS , whereas

12-16

3640 Time after

3640 12-16 last dose ( hours 1

FIG. 2. Mean 2 SEM of prechallenge FEV,, AFEV, for exercise, and a fall of FEV, of 20% (PC,) for histamine at 12 and 36 hr after last dose of each treatment.

we waited at least 12 to 16 hr before the challenges. It is therefore possible that the reduction in response observed by Griffin et al.27 was due to a persistence of drug action. Woenne et al.” also noted reduced response to histamine in four of 13 subjects shortly after inhaling CS, but Ryo et a1.29 found no effect from CS on either histamine or methacholine response. It is possible that the duration of CS treatment was not long enough to reveal any suppressive effect, and more prolonged, controlled studies are needed to settle this point. The psychologic tests demonstrated that the only difference between performance when CS and LAT were administered was in visual-spatial planning tests, and this difference was attributable entirely to the children with the lower estimated IQ levels (Fig. 1). Difficulties with visual-spatial planning in children with asthma were noted by Dunleavy tid Baade” who, in a later study,” noted that those children who appeared to have minimal brain damage did worse on all performance tests. We were unable to find any relationship between the results of the psychologic evaluation and any of the other variables in the study, including the level of serum theophylline at the time of testing and the order in which the drugs were administered. Since we had no period without placebo

VOLUME 76 NlJMBER i

medication, we cannot comment on the effect of asthma itself on the performance of the children. On a practical level, difficulties in visual-spatial planning may be related to what McWhirte?’ calls problems in laterality of directionality. These children may have difficulty discriminating letters and numbers. Frosti$’ notes that this could cause difficulty with discriminating the sequencing of letters in a word or the order in which numbers are to be divided and multiplied in a long division problem. Such difficulties would have direct effects on the technical aspects of reading, writing, and mathematics, especially for a child of lower intelligence. This would lend support to the preliminary observations of Furukawa et al.’ However, we did not find improvement in attention or memory when theophylline was omitted as they did, but the coding test that they and we used was found in our study to demonstrate a practice effect and was removed from our analysis. Furukawa et a1.9 did not investigate this problem, although they were aware that it could have influenced the results. In conclusion this study has demonstrated no clinical or physiologic differences resulting from treatment of children with asthma with either CS or LAT for 1 mo. There was an advantage for CS treatment in terms of psychologic tests involving visual-spatial planning in the four children with lower IQs, which, if it is confirmed, would imply that LAT therapy could adversely affect school performances in such patients and should be avoided. In view of the implications of these results for the management of asthma, we believe that a study involving a more detailed examination of the influence of intelligence would be worthwhile REFERENCES I. Godfrey S, Balfour-Lynn I, Konig P: The place of cromolyn sodium in the long-term management of childhood asthma, based on a 3 to 5 years follow-up. J Pediatr 87:465, 1975 2. Weinberger M, Bronsky E: Evaluation of oral bronchodilator therapy in asthmatic chikiren. J Pediatr 84~421, 1974 3. Hambleton G, Weinberger M, Taylor J, Cavanaugh M, Ginchansky E, Godfrey S, Tooley M, Bell T, Greenberg S: Comparison of cromoglycate (cromolyn) and theophylline in controlling symptoms of chronic asthma. Lancet 1:381, 1977 4. Goldstein A, Kaizer S, Warren R: Psychotropic effects of caffein in man. II. Alertness, psychomotor coordination, and mood. J Pharmacol Exp Ther 150:146, 1965 5. McCarthy MM: Speech effects of theophylline. Pediatrics 68:749, 1981 (letter) 6. Murphy MB, Dillon A, Fitzgerald MX: Theophylline and depression. Br Med J 15:281, 1980 7. Rail TW: Central nervous system stimulants. In Goodman LS, Gilman A, editors: The pharmacological basis of therapeutics. New York, 1980, Macmillan Publishing Co, Inc, p 592 8. Wasser WC, Bronheim HE, Richardson BK: Theophylline madness. Ann Intern Med 95:191, 1981

Cromolyn

or theophylline

in childho&

asthma

69

9. Furukawa CT, Shapiro GG, DuHamei ‘I’, Weuner- I,. Plcrson WE, Biennan CW: Learning and behal Ioral prohlemb associated with theophylline therapy. Lancer I:62 I. 1984 (,!etter) 10. Dunleavy RG. Baade LE: Neuropsychologicai c,orreIatcs of severe asthma in children 9-14 years old I (‘