Bull. Org. mond. Sat 1969, 40, 425-454 Bull. Wid Hithi Org.f

Codeine and its Alternates for Pain and Cough Relief* 3. The Antitussive Action of Codeine- Mechanism, Methodology and Evaluation

NATHAN B. EDDY, M.D.,1 HANS FRIEBEL, Dr. med.,2 KLAUS-JORGEN HAHN, Dr. med.8 & HANS HALBACH, Dr. med. Dr.-Ing.4

This report-the third of a series on codeine and its alternates for pain and cough relief-presents a detailed review of the physiology and pathophysiology of cough, the methods for the experimental and clinical measurement of the antitussive action of drugs, possible mechanisms of action of antitussive agents, and includes a compilation of experi- mental results and clinical experience with codeine as an antitussive.

CONTENTS

INTRODUCTION ...... 425 PHARMACOLOGICAL COMPONENTS OF THE ANTITUSSIVE EFFECT ...... 434 MECHANISM AND SIGNIFICANCE OF COUGH . . 426 ANTITUSSIVE ACTION OF CODEINE ...... 441 METHODS FOR THE STUDY OF ANTITUSSIVE RESUMt ...... 451

ACTION ...... 428 REFERENCES ...... 452

INTRODUCTION respect to the symptomatic management of cough, with only minor variations, and this situation has Codeine was first used as a therapeutic antitussive been influenced little by the introduction of sub- agent soon after its separation as an alkaloidal entity stances claimed to be superior in effectiveness and (Martin, 1834) and some of the earliest accounts of safety (Domenjoz, 1952; Winter & Flataker, 1954; dosage and of the circumstances of its use are sum- Pellmont & Biichtold, 1954).5 marized by Krueger, Eddy & Sumwalt (1943). The number of compounds alleged to possess anti- Codeine has indeed dominated the picture with tussive action comparable with, or superior to, that of codeine has increased greatly in recent years (see * This review of the analgesic and antitussive effects of codeine and its alternates is being published in the Part 4) but the popularity of codeine continues and Bulletin of the World Health Organization in five instal- its annual consumption remains at a high level, paral- ments. The first, devoted to an assessment of codeine as a pain reliever, was published in Bull. WId Hlth Org., leling the high frequency of need for relief of cough 1968, 38, 673-741. The second instalment, on alternates as a symptom, especially chronic cough. Bickerman to codeine as an analgesic, was published in Bull. WId Hlth Org., 1969, 40, 1-53. The third presented here is (1960) estimated, not so long ago, that 13% of the devoted to the evaluation of codeine as an antitussive population of the United States ofAmerica was more agent. The fourth instalment will deal with potential than 65 years old and that there were not less than alternates for cough relief. The final instalment will con- sist of a discussion and summary of the preceding reviews. 1 300 000 chronic coughers. Watt (1958) estimated 1 Consultant, National Institutes of Health, Bethesda, that in the United Kingdom of Great Britain and Md., USA. Northern Ireland 10% of all prescriptions were for 'Professor of Pharmacology and Toxicology, Univer- sity of Heidelberg, Germany. Present address: Chief, Drug cough relief. The need for antitussive preparations Safety, World Health Organization, Geneva. 6 Heroin was introduced 50 years earlier as a superior ' Department of Medicine, University of Heidelberg, antitussive but obviously proved itself to be less safe and, Germany. in spite of persistent claims as to its specificity, it has been 'Director, Division of Pharmacology and Toxicology, banned from therapeutics in most countries. No further World Health Organization, Geneva, Switzerland. reference to it will be made.

2312 -425 426 N. B. EDDY AND OTHERS

is clear and codeine is our standard of reference. biguus. The authors assumed connexions between However, two facts have given impetus to the search these neurones and the receptors characterized by for an adequate substitute-substances without anal- Widdicombe (1954). Under normal breathing condi- gesic action can exhibit antitussive potency and such tions these neurones are silent but become activated substances may be entirely devoid of morphine-like by noxious stimulation and may play an important dependence-producing properties (Isbell & Fraser, role in the central processing of cough-evoking stim- 1953). uli (Engelhorn & Weller, 1965). In recent years much attention has been directed The existence of other connexions of the cough to the mechanism of cough and its relief and to centre must be assumed and has been partially de- adequate methodology for the measurement of the monstrated. Section in the forward part of the pons, latter. This work will be discussed first and will be for example, weakens the cough response in the cat. followed by details of the antitussive action of Spread of excitation of the cough centre to the vomit- codeine and later (in Part 4) by a discussion of the ing centre is well known, connexions with the respi- potential alternates. ratory centre must exist, and considerable voluntary control, even suppression of cough is possible (Kuhn & Friebel, 1960a, 1960b; Engelhorn & Weller, 1961; MECHANISM AND SIGNIFICANCE OF COUGH Friebel & Hahn, 1966; Hahn & Friebel, 1966). Reviews of this topic have been prepared by Bucher Efferent impulses from the cough centre, co-ordi- (1958), Bickerman (1960), Friebel (1963) and nating movements of the glottis, the diaphragm and Widdicombe (1964). Cough is a protective reflex, the abdominal and intercostal muscles in the act of the purpose of which is to remove irritating materials coughing, pass over the recurrent, the phrenic, the or sensations from the respiratory tract, and as such intercostal and other nerves. its pathways include sensitive receptors, afferent nerve Cough is understood as an explosive expulsion of fibres, central connexions with a " relay station in air resulting in the " generation of a high-linear-velo- the arc for cough reflex" (May & Widdicombe, city air stream with high kinetic energy available for 1954), and efferent nerves. Experiments of Larsell the acceleration and displacement of an object in the & Burget (1924) and Widdicombe (1954) demonstra- airway" (Ross, Gramiak & Rahn, 1955). These ted two types of receptors: one in the trachea, sensi- authors assume the explosive expulsion of air to be tive to mechanical stimuli, the other throughout the the product of substantial elevation of intrapleural tracheobronchial tree, responding chiefly to irritating pressure and sudden opening of the glottis. Equali- chemical stimuli. The afferent nerve fibres, which zation of pressure in the trachea with the outside air conduct the action potentials arising in these recep- creates a pressure differential along the bronchial tors, pass along the superior laryngeal nerve and the tree and the bronchi narrow passively. This allocates vagus, entering the medulla with the vagus root. a key position in the expulsion of air to the move- Section of the vagus below the point at which it is ments of the glottis. These movements, at least the joined by the superior laryngeal nerve does not sub- initial closure of the glottis, are centrally controlled stantially affect the cough threshold of anaesthetized independently of intrathoracic pressure and are car- cats (Sell, Lindner & Jahn, 1958), indicating probably ried out even if no air-flow occurs. The other factor that the stretch receptors in the lung (Bucher & Jacot, to bring about the air expulsion is the force of the 1951) take little part in the cough reflex. Cough- muscular contractions. Even if the movements of exciting impulses from the pharynx are carried in the glottis are inhibited, or the expired air goes out the glossopharyngeal nerve. through a tracheal fistula, remarkable blasts of air The existence of a specific cough centre located can be produced (Floersheim, 1959). in the dorsolateral region of the medulla close to The act of coughing is often, but not necessarily, the respiratory centre was suggested by Borison initiated by an inspiration (Staehelin, 1914; Trende- (1948) and by Chakravarty et al. (1956). This sug- lenburg, 1950; Widdicombe, 1954, 1961; Kuhn & gestion was confirmed by Engelhorn & Weller (1961, Friebel, 1960a,- 1960b; Friebel & Hahn, 1966; Hahn 1965), who localized a central mechanism for the & Friebel, 1966). This is not consistent with a very adaptation of the activity of the respiratory system to close correlation between the central portion of the acute changes in its environment in the lower pons cough reflex arc and the respiratory centre. Engel- and medulla. Its nervous components, E, neurones, horn & Weller (1965) suggested that the EQ neurones were localized in the caudal end of the nucleus am- belonged to an independent sensitive mechanism for CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 427 the protection of the lung against noxious stimula- recorded. Engelhorn & Weller (1965) observed in tion. Friebel & Kuhn (1962) in this connexion show- anaesthetized cats an inhibition of the activity of ed that the cough and respiratory centres reacted dif- expiratory neurones (Ea) on account of the activity ferently to drugs. They demonstrated that 2-amino- of cough-specific E, neurones. indane stimulated respiration in conscious and The newer information on the organization and narcotized guinea-pigs but depressed the cough reflex function of the cough reflex mechanism, therefore, mechanism. In fact the cough reflex mechanism con- supports the hypothesis of its independence, in the tains peculiarities in its function and organization main, of the respiratory mechanism, in respect to which seem to exclude a close association with the res- cough-provoking stimulation, summation of afferent piratory mechanism. Its functions can be looked upon tussive impulses and transformation of these im- as summing the cough-inducing stimuli, co-ordina- pulses for co-ordination of the respiratory muscu- ting the respiratory musculature to initiate the cough lature to produce the explosive cough response. reaction and interrupting the respiratory rhythm in The functions of the cough reflex as a protective favour of the cough reaction (Kuhn & Friebel, device are: 1960b). (1) warning against noxious irritation and damage Summation of afferent impulses towards the pro- in the airways from inhaled chemicals or foreign duction of cough was suggested by Stefko & Bensofn bodies; and (1953) and Bucher (1958). Experiments of Friebel & Hahn (1966) in conscious guinea-pigs gave evidence (2) clearing the airways by removal of the offend- of a remarkable ability of the cough centre to store ing irritant. the afferent impulses. The animals reacted with an Cough is a symptom in a wide variety of diseases. explosive expulsion of air, on the average 2.8 seconds Inflammation of the respiratory tract can be accom- after an appropriate electrical stimulation of the panied by increased irritability of the sensory recep- tracheal mucosa was initiated. A response could tors and intrapulmonary tumours can evoke cough still occur with weaker stimulation up to a latency of by pressure on or involvement of sensitive receptors 16 seconds. If the stimulation was too weak to evoke and nerve fibres. The cough response via the auricu- a cough reaction in 16 seconds, it did not produce lar branch of the vagus on account of inflammation, cough at all. The authors concluded that the sum- etc., of the external auditory meatus is well known. mation of cough-inducing impulses served two pur- Cough may be an allergic response, a consequence poses. It intensified the protective function of the of circulatory disturbances or it may be psychogenic. cough reflex and prevented unlimited activity of the Cough arises most commonly from a stimulus within cough reflex mechanism. Variability in the latency the respiratory tract, from infections or mechanical was the only specific modification of the intensity and irritation within the lumen (Bickerman, 1960b) or characteristics of the single cough response produced from inhalation of polluted air or tobacco smoke by variation in the intensity, duration and frequency (Phillips, Phillips & Thompson, 1956; Herzog, 1967). of stimulation. Obviously, the central part of the From the clinical standpoint, cough is probably the cough reflex arc was able to collect variable afferent most important mechanism of bronchial catharsis impulses and transform these into efferent reactions (Bickerman, 1960b). If cough is present for a consid- of relative uniformity (Friebel & Hahn, 1966). erable time, adaptation may occur and the patient The efferent impulses of the cough reflex activated may become almost unaware of its occurrence; or, the respiratory muscles which normally carry out the the apprehensive person may exaggerate its intensity. movements of breathing, so that the normal respira- When the cough is incapable of, or ineffective in, tory rhythm was interrupted or postponed (Kuhn & removing the provoking stimulus, tussive insuffi- Friebel, 1960b). The moment of interruption varied. ciency results. In many suppurative diseases of the In 85 of 100 cough reactions recorded from electri- lung, cough is unable to remove the thick, tenacious, cally stimulated conscious guinea-pigs, the explosive mucopurulent secretions which form constantly in blast began at the height of inspiration, in 12 it began response to the infection. The dry, hacking, irritative during expiration and in three at the moment of cough of some acute infections of the respiratory complete expiration. An interruption of inspirations tract, such as laryngitis, tracheobronchitis and per- has never been observed (Hahn & Friebel, 1966). tussis, may by itself become a major factor in retard- Rarely, an acceleration of the normal expiration, ing recovery by causing loss of sleep, inadequate food leading to a typical cough reaction, has been intake, nausea and emesis. In obstructive disease of 428 N. B. EDDY AND OTHERS the tracheobronchial tree, such as tumours, foreign visual or acoustical observation or instrumental bodies and chronic asthma, the cough reflex may recording of cough reactions. Winter & Flataker become self-perpetuating and the resultant severe (1954) used ammonia vapour and sulfuric acid aero- paroxysms of unproductive coughing may cause sol; Friebel, Reichle & von Graevenitz (1955) intro- serious debilitation of the chronically ill patient duced the inhalation of sulfur dioxide mixtures; (Bickerman, 1960b; Banyai & Johannides, 1956). G6sswald (1958) used citric acid; Silvestrini & Maffii Other consequences ofsevere protracted coughing are (1959) employed acrolein vapour; and Kroepfli (1950) pulmonary emphysema, fractures of ribs, cough syn- blew soap powder against the tracheal mucosa. Each cope, urinary incontinence, post-operative disruption of these agents elicits cough by stimulation of chemo- of abdominal incisions and spontaneous pneumo- receptors and each has its advantages and limitations. thorax. The necessity of cough-depressing therapy Receptors sensitive to cough-provoking irritants for unproductive cough is acknowledged (Friebel, are spread throughout the trachea and bronchi 1963). (Larsell & Burget, 1924; Widdicombe, 1954). Thus the inhalation of gases, vapours and aerosols can METHODS FOR THE STUDY OF ANTITUSSIVE ACTION influence many of them, whereas the introduction of solid material such as soap powder will reach only Measurement of antitussive potency in animal experi- a few in the trachea. But vapours, etc., inhaled via ments the nose and pharynx will also stimulate receptors There are many reports of attempts to assess anti- therein, perhaps eliciting sneezing as well as coughing tussive potency in animals; 70 reports in respect of and making discrimination difficult if the response is codeine, arranged according to animal and method, recorded instrumentally, as it was by Winter & Fla- have been summarized in Table 27, yielding a wide taker (1954). Rosiere, Winder & Wax (1956) and range of effective doses, from 0.3 to 100 mg per kg. LaBarre & Plisnier (1959) therefore directed the The variation is related mainly to the method, but air-flow through an opening in the trachea and had this has varied in detail, even when it was said to be to use anaesthetized animals or others with a chronic the same, in respect of the stimulus used to evoke tracheal fistula. cough, and analysis of the data has varied with the Irritants like citric acid, acetic acid, ether, acrolein, investigator. Generally, however, one can assume ammonia, sulfuric acid or sulfur dioxide have differ- that the method and analysis were the same when ent cough-eliciting potency and the concentration for comparison was made in a particular case between cough threshold varies widely from less than 1.0% a new agent (or agents) and codeine. It is worth (sulfur dioxide) to 20.0% (citric acid) (Friebel, while, we believe, to discuss the methods, at least Reichle & von Graevenitz, 1955; G6sswald, 1958). in principle, the conditions of their application, and The duration of their action is also different and their significance, reliability and interpretation as depression of cough evoked by inhalation of sulfur means of measurement. dioxide requires a larger dose of codeine than cough induced by ammonia if the guinea-pig is the test ani- Intrapleural injection. Systematic exploration of mal (Friebel, Reichle & von Graevenitz, 1954; Winter cough-depressing activity began with Ernst (1938), & Flataker, 1954). On the other hand, dogs show who reported cough thresholds in cats with artificial less difference according to the chemical irritant used. chronic pleuritis. The inflammation was produced It would be worth while to know to what extent the by intrapleural injection of iodine (Lugol's solution), effective dose of codeine for a particular species after which percussion of the chest wall produced a depends upon the physiological characteristics of the cough reaction. Though Ernst (1939) claimed the species or upon the stimulus and the details of the results were reproducible, the method has been used method. only infrequently with no effort towards its refine- Not only does the irritant potency of the chemical ment. stimulant vary but also its effectiveness varies accord- Irritating chemicals. Eichler & Smiatek (1940) ing to the concentration and absolute amount at the were the first to use irritating chemicals to stimulate site of stimulation. Amdur (1957), for example, coughing. Their basic procedure has been variously showed that the effect of sulfuric acid aerosols on modified. Usually the animals have been exposed to bronchial receptors more nearly approached that of the mixture of irritating gas, vapour or aerosol with sulfur dioxide gas the smaller the droplets of the air, in a chamber suitable for one animal and for aerosol. Also the smaller the droplets the more tran- CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 429 sient the effect and the faster the recovery. This is vagal loop and through these, stimuli were applied important for the stability of the cough threshold, in the unanaesthetized animal (dog). The method of since the usual procedure is to determine the thresh- electrical stimulation which is most often employed old before and at intervals after drug administration. was devised by Domenjoz (1952). The superior A change in normal sensitivity of the receptor with laryngeal nerve was isolated in the anaesthetized cat time would influence the accuracy of the judgement and stimulated briefly; the number and duration of of the drug's effect. Therefore, limited stimulation stimuli required adjustment in the particular animal with adequate intervals for recovery have to be taken to elicit cough reliably. The technique of electrical into account, as they have been by some authors. stimulation with modifications has been employed Friebel, Reichle & von Graevenitz (1955) recorded in anaesthetized guinea-pigs (Krause, 1958), in in 3.8% of guinea-pigs a decrease in sensitivity unanaesthetized guinea-pigs (Kuhn & Friebel, 1960a, between two applications ofsulfur dioxide 30 minutes 1960b) and in unanaesthetized dogs (Stefko & apart. If soap powder is blown into the trachea a Benson; 1953). The latter authors implanted elec- considerable adaptation of the receptors has to be trodes in the tracheal wall, eliciting cough by stimula- taken into account (Kroepfli, 1950) because the solid tion of adjacent nerve fibres and receptors, and irritant is scarcely removed from the site of applica- Chakravarty et al. (1956) provoked cough by means tion. of electrodes placed stereotactically in the medulla One advantage of a chemical irritant as the cough of the decerebrate cat. stimulus is that it permits the use of many animals Various forms of electrical current have been used, in a single experiment, facilitating statistical analysis with changes in the shape of the single impulse (spike, of the results. This no doubt explains the frequency square or undulatory), and with variations in voltage, with which such stimuli have been employed. current, frequency and duration. There have been explorations, too, of interrelationships between volt- Mechanical stimulation. Kase (1952) was the first age and frequency ofstimulation for producing cough to describe a method for the production of cough by most effectively (Schroeder, 1951; Stefko & Benson, mechanical stimulation of the tracheal mucosa suit- 1953) and most able for recently of the interrelationship the measurement of antitussive action. The between intensity, duration and frequency of thresh- most sensitive area for such stitnulation is at the old stimuli by Friebel & Hahn (1966). Variation in bifurcation ofthe trachea. It may be reached through these parameters within described limits had no the fenestrated wall of the trachea by various devices, appreciable effect on the intensity or characteristics bristles, catheters, etc., but these must be applied of the cough response, but the latent period between cautiously to avoid damaging the mucosa and the beginning of stimulation and the occurrence of causing a change in sensitivity and adaptation to the the cough could be decreased (Friebel & Hahn, 1966). stimulation (Larsell &I Burget, 1924; Widdicombe, 1954). The strength of the stimulus cannot be con- Electrical stimulation to produce cough has no trolled as accurately as with chemical or electrical physiological equivalent; it also requires operative stimulation and the animal must be prepared opera- preparation of the animal and testing immediately tively. If mechanical stimulation is attempted imme- after the operation can be influenced by the anaes- diately after operation, sensitivity may be impaired thetic. Nevertheless the method has been used by by the anaesthetic and later pulmonary infection may many and has the advantage of exact measurement ensue. The chief advantage of mechanical stimula- of stimulus strength and quality. tion is its simplicity and some of its drawbacks are Avoidance of anaesthesia in antitussive testing was avoided by the technique of Tedeschi et al. (1959). one of the conditions stressed by Schroeder (1951). These authors implanted an iron slug within the tra- A cough-depressing action of some of the anaesthetic chea of a dog which could be made to vibrate against agents used in animals has been observed or referred the tracheal mucosa by bringing a strong electromag- to by Schroeder (1951), Domenjoz (1952), Tedeschi net against the outside of the throat, thus eliciting a et al. (1959), Friebel & Kuhn (1962), Hahn & Friebel cough response. After some time, chronic inflamma- (1966) and others. One might expect the effective tion with epithelial desquamation and loss ofsensitiv- dose of the antitussive, codeine for example, to be ity resulted. smaller in experiments performed under anaesthesia, but the data of Table 27 do not give such an indica- Electrical stimulation. This method was introduced tion. by Schroeder (1951). Electrodes were attached to a Some of the methods for antitussive experiments 430 N. B. EDDY AND OTHERS require the preparation of a tracheal fistula and In most experiments with rats and guinea-pigs, this causes the air passages to be exposed to dry drugs have been given subcutaneously; in cats intra- laboratory air instead of air highly humidified by venous injection has been usual, whilst in dogs, all passage through the nose and throat of the animal. routes have been used. Intravenous injection has the Under this unnatural physiological condition there is advantage of requiring small doses and usually has evaporation and thickening of the bronchial secre- a transitory effect, permitting comparison ofrepeated tion, impaired activity of the ciliated epithelium and doses; if several injections are given the possibility poorer transport of secretions (Perry & Boyd, 1941; of summation of subliminal stimuli or of tachyphy- Dalhamn, 1960). Even ifthe inspired air is humidified laxis cannot be disregarded. The mode ofadministra- to 30%, ciliary movement stops within 3-5 minutes tion takes on added importance when the results of opening the trachea. The impairment of mucus are considered in relation to man. In antitussive transport favours the formation of an immovable therapy in man, the oral route is usual, whilst in layer of secretion and a change in sensitivity to cough- experiments in animals it is the exception. Therefore, provoking stimuli (Gordonoff, 1938). If the tracheal if antitussive effect is shown in the latter, one also fistula is maintained chronically, as proposed by Kase needs to know something of absorption and meta- (1952, 1954, 1955) there is liability to respiratory bolism after oral administration, factors which infections (Rosiere, Winder & Wax, 1956), again have been largely disregarded in predicting clinical influencing the threshold for cough stimulation. effectiveness. Similar considerations must be borne in mind in con- nexion with other operative procedures, implantation Measurement ofantitussive potency in experimentally of electrodes, exposure of the laryngeal nerve, etc. induced cough in man The catalogue of factors which may influence cough Cough may be induced experimentally in man by threshold before and after drug administration is inhalation of an aerosol containing acetylcholine, long, varying with the working procedure, and they ammonia or citric acid, or by intravenous injection play a significant role in the variations in results of lobeline or paraldehyde. shown in Table 27. Hdglund & Michaelsson (1950) published the first There are also qualitative differences in the animal report on experimental cough provoked in healthy species used in the experiments. Guinea-pigs, cats volunteers by inhalation of an irritating chemical and dogs are to be preferred. Mice are too small for compound. The subjects respired through tubes fit- discrimination between sneezing and coughing. Rats ted with valves to prevent expiration into the system. respond readily to sulfur dioxide inhalation by both Ammonia vapour (2%) in the desired quantity was sneezing and coughing and, when tested repeatedly added to the inspired air by means of a syringe. The over a three-month period, showed decreasing sensi- momentary closure of the glottis in response was tivity. Guinea-pigs, on the other hand, maintained a recorded by a pneumograph around the subject's constant sensitivity throughout 48 weeks (Friebel & chest. The method of Hoglund & Michaelsson, more Kuhn, 1964). In general, the smaller species require or less modified, has been used by Trendelenburg higher doses per kg of codeine than do the larger (1950), Blix (personal communication, 1952), Hahn species to produce the same effect; cough elicited & Willbrand (1952) and Gravenstein, Devloo & in guinea-pigs by ammonia is an exception to this Beecher (1954), with all except the latter showing rule. On account of their size, cats and dogs are elevation of the cough threshold by codeine. Tren- amenable to all methods for antitussive testing and delenburg and some others relied on subjective obser- to all routes of drug administration, but unanaesthe- vation of apparent closure of the glottis, believing tized cats and dogs quickly learn to control partially that this was always in accord with mechanical their breathing, restricting the inhalation of irritating recording. Closure of the glottis was taken as the chemicals, so that they have to be checked carefully threshold response whether or not an explosive act to determine whether under undrugged conditions the of coughing followed and, according to Trendelen- cough response remains the same. A further draw- burg, the amount of ammonia causing this response back to the use of dogs is their ability to be trained was the indication of threshold. When this had been to the experimental procedure, and to develop a determined test substances were administered and voluntary cough response; the better the relation measurements of threshold continued every five between handler and animal the more ready the res- minuttes until the threshold quantity of ammonia ponse (Schlez, 1965). had returned to the original amount required. CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 431

Gravenstein, Devloo & Beecher (1954) could not observed a decrease in frequency of the citric-acid- be certain why they failed to confirm the results of induced cough of 22.4% at the second and fourth the others. They called attention to two points, hours after 5 mg of codeine phosphate orally, a firstly that Hoglund & Michaelsson and the others decrease of 22.4-32.5 % at the first second and third had not used a placebo as control or used double- hours after 15 mg, and of 22.4-43.1 % at each hourly blind conditions, and secondly they found that the observation after 30 mg of codeine. Changes in ammonia-mixture container lost ammonia readily. expiratory airflow differed occasionally from differ- A lower concentration ofammonia would be less irri- ences in cough frequency, and the latter yielded the tating and if it were used in later trials after an anti- better dose-response curves. tussive agent it would make the latter appear effective. The method of Bickerman and his associates has This would not allow the threshold to return to its been employed by Rasch (1957), Archibald, Slipp & original value. Shane (1959), Morris & Shane (1960), Groszman, There are also other points at issue in the method. Birker & Casimir (1961), Calesnick, Christensen & While closure of the glottis (sudden pause in breath- Munch (1961), Tannenbaum (1965) and Silson (1965), ing) was chosen as the indication of response, it who demonstrated a reduction in coughing after drug remains to be proven that this can be accepted as an administration, and by Gravenstein, Devloo & adequate indication for a cough reaction and whe- Beecher (1954) and Sevelius, Lester & Colmore ther that reflex is influenced by the same drugs and (1965) who did not. Gravenstein, Devloo & Beecher by the same doses as the cough reflex. Trendelenburg said that the Bickerman technique did not distin- (1950), as well as Hahn & Wilbrand (1952), ob- guish between a noisy, explosive cough and one of served wide variations in threshold values depending little intensity, that the quantity of aerosol reaching upon weather conditions and psychological factors. the lungs could not be accurately controlled or deter- Trendelenburg, however, did not think initial thres- mined, and that double-blind conditions were diffi- hold values influenced drug effects. Gravenstein, cult to maintain because of recognition by the sub- Devloo & Beecher (1954) found the threshold more jects of other effects (sedative action, etc.). Sevelius, constant and reproducible, but observed no signifi- Lester & Colmore nebulized different strengths of cant change after medication. citric acid solution, 5.0-25 %, to determine the con- A method closely related in principle to that of centration which consistently produced maximum ammonia inhalation was introduced by Bickerman cough response, a different threshold from that used & Barach (1954) and by Bickerman et al. (1956). by Bickerman. In spite of emphasis on selection and Cough was induced by'a 25 % solution of citric acid training of normal volunteers, in all of their six deeply inhaled from a nebulizer through a tight-fitting subjects the cough response to the same dose of face mask and a constant oxygen flow of five litres irritant changed from hour to hour by as much as per minute. They recorded the number of coughs 50%. The subjects also definitely became adapted elicited and the expiratory air flow. The subject to the citric acid irritation and coughed progressively took three inhalations, resting two or three minutes less frequently on repeated exposures. Without between each. In drug potency tests the response was proper placebo control data a " false positive " drug recorded at zero time and hourly to four hours after effect could be ascribed to what was only adaptation. medication. Bickerman et al. (1957) later spoke of These authors were unable to distinguish between " trained " subjects, who worked with the method 15 and 30 mg of codeine and placebo. This statement through many months and showed a reasonable concerning adaptation to the stimulus is at variance degree of adaptation (" learning " factor) to the pro- with Bickerman's experience, but Sevelius, Lester & cedure. This group worked under double-blind con- Colmore required stronger citric acid concentrations ditions and was challenged with concentrations of and it is not clear whether or not residues of citric citric acid from 1.25-25.0 % to determine a threshold acid on the tongue and in the throat were removed stimulus, arbitrarily set as that concentration of citric by rinsing as they were by Bickerman. Sevelius acid which evoked a total of not less than four coughs and his associates maintained that Bickerman's when inhaled five times. Since this threshold level procedure was considerably more complicated varied from subject to subject and sometimes in the than theirs since he examined 153 " normal " same subject from day to day, it was redetermined subjects to obtain 21 presenting relatively constant for each subject at the beginning of each test run. reactivity to the cough-provoking aerosol. Such Under these conditions Bickerman et al. (1957) selection not only reduces variation in the sen- 432 N. B. EDDY AND OTHERS sitivity of the cough reflex mechanism related to tive. Haslreiter's experience was similar, a posi- smoking, chronic respiratory infections, etc., but will tive cough response to acetylcholine in most also reduce the influence of another factor, voluntary patients with chronic bronchitis, etc., and a nega- or subjective control over the response, which may tive response in all healthy persons. Both Prime play a role in all experimentation on humans. The (1961) and Sevelius, Lester & Colmore (1965) found cough response can be voluntarily suppressed and that cigarette smokers, admitting to only a little initiated. habitual coughing, responded positively to acetyl- Results obtained with the Bickerman method have choline. not always correlated well with clinical effectiveness Haslreiter (1959) and Prime (1961) used acetylcho- of the agents tested, in the treatment of pathological line to attempt evaluation of cough depressants. The cough. Bickerman & Itkin (1960) failed to demon- subject breathed his individual concentration (Hasl- strate a cough-depressing effect with normethadone, reiter, 1959) or 1.0% (Prime, 1961) acetylcholine though such an effect has been well established in aerosol until coughing began or at most for 11/2 min- animal experiments and clinical experience. On the utes and the number of coughs ensuing during the other hand, Woolf & Rosenberg (1964), using a tape- next 20 minutes was counted. The subject then took recorder system as an objective measure of coughing, an inert tablet or one containing the cough depres- found laevopropoxyphene inactive clinically, while sant and 20 minutes later the test was repeated. The Bickerman et al. (1957) and Calesnick, Christensen active and inactive tablets were administered in ran- & Munch (1961) found it effective against artificially dom order to each subject. Prime compared the produced coughing. cough rates before and after medication with positive Recently Calesnick & Christensen (1967) have used results; Haslreiter determined the number of patients the citric-acid-aerosol method with a different end- who did not coug4 after the test drug. Their results point, the latent period from beginning of inhalation did not correspond. Sevelius, Lester & Colmore under pressure through a tight face mask to cough (1965) could not confirm Prime's results either. In response and regarded a lengthening of this period nine subjects they were unable to distinguish between as indicative of cough suppression. They used only 30 mg of codeine phosphate and placebo, adminis- males because females were too sensitive to reflex tered double-blind in random order with 60 or closure of the glottis which would prevent inhalation 120 minutes allowed for absorption. Prime admitted ofa sufficient amount ofthe aerosol (25 % citric acid). to them that his later experience was in accord with They also excluded persons who did not give a good theirs. It should also be mentioned that acetylcholine cough response before drugging or who failed to antagonists (hexamethonium, forexample), not other- show an antitussive effect with an exploratory dose wise antitussive, abolish the cough response to acetyl- of 40 mg of codeine orally. Graded doses were given choline (Tiffeneau, 1955). with citric acid challenge hourly throughout four As indicated earlier, cough can be provoked in man hours. A good dose-response curve was obtained by intravenous injection of paraldehyde or lobeline. with 5.0-40.0 mg of codeine. Gravenstein, Devloo & Beecher (1954) found that Prime (1961), looking for a controllable agent for the usual dose required to produce a satisfactory artificial cough production, developed an observation cough response varied from 0.25 to 0.5 ml in different of Tiffeneau (1955). The inhalation of a nebulized subjects (Konar (1959) used 1.0 ml). The paraldehyde solution of acetylcholine regularly induced more or was injected into a vein of the forearm through a less coughing in certain persons. Prime also tried continuous 5.0 % dextrose solution drip. The result ether and sulfur dioxide; both were effective, but the was a feeling of constriction of the chest and an taste and smell were objectionable and the latter as explosive cough which relieved the feeling. The used was so powerful that none wanted to repeat the coughing bout lasted 30-60 seconds and might be experience. In non-smokers, and persons without followed by an afterburst of coughing. When the respiratory infection under 30 years of age, Prime effective dose had been determined and repeated at found acetylcholine even at 1.0% completely ineffec- 30 minutes, the antitussive was administered and tive, confirming the observations of Tiffenau (1955) the paraldehyde repeated at half-hour intervals for and Halsreiter (1959). However, in patients with 2 or 21/2 hours. Gravenstein, Devloo & Beecher chronic bronchitis, emphysema, asthma, fibrous investigated the effect of drugs on the response to tuberculosis, pulmonary sclerosis, or acute respira- paraldehyde, and showed a difference between heroin, tory inflammation acetylcholine was 80-90% effec- 2.5 mg subcutaneously, and placebo, where the sub- CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 433 jects were used as their own control, but not with Evaluation ofantitussive action against cough ofpatho- codeine, 10 mg orally or 30 mg intravenously, or logical origin morphine, 5 mg subcutaneously, when the subjects The crucial test for an antitussive is the study of were not their own controls. Konar (1959) also failed occurring to obtain a positive effect on the response to paral- its effect in patients with spontaneously dehyde with codeine or morphine. Gravenstein and cough of various origin. Most of the reports in this his associates deemed it unwise to continue with the connexion are based on clinical experience rather paraldehyde test because of pain at the site of injec- than controlled clinical trial. Reliable, unbiased tion and phlebitis, fear on the part of the subject of evaluation and comparison should come from the the feeling of chest constriction, and variability in latter, but the former is supplementary and essential. the response. Covering a larger number of patients of more hetero- geneous character, the general clinical observation Lobeline administered intravenously excites contributes to the knowledge of side-effects, unusual cough, presumably because of a direct effect on the adverse reactions and useful indications of applica- cough centre (Chabrier, Giudicelli & Thuillier, 1950). tion and contraindication. The effective dose was 3.0-10.0 mg, and this dose caused a cough reaction of 10-20 seconds duration Here, as elsewhere, the clinical trial must fulfil some followed by about 60 seconds of polypnoea. The basic requirements of clinical pharmacology, double- authors found that 80-100 mg of pholcodine blind procedure to avoid error by suggestion, inclu- decreased the response to lobeline. They said, sion of placebo or standard drug of established effi- however, that lobeline had to be used cautiously cacy for comparison, and randomization of drug because of possible toxic reactions. Hillis & Kelly administration to facilitate statistical analysis. This (1951) were unsuccessful in attempts to suppress the is not the place to discuss the design of specific clini- cough response to 5.0 mg of a-lobeline, intravenously cal trials but certain features of the procedure, as in normal subjcts, with codeine, morphine or heroin. applied to the evaluation of antitussives, must be The cough response was blocked by hexamethonium considered for better judgement of the significance iodine. of the data to be presented in respect to specific Bickerman & Barach (1954) said that artificial agents. In the main, two approaches to evaluation cough production should meet the following condi- ofefficacy, etc., have been developed-analysis of the of the sub- subjective judgement of effect by patient and/or tions: uniform and consistent response of changes in ject to the same threshold stimulus; the eliciting agent observer, and objective measurement its appli- number, severity, etc., of bouts of coughing during should be non-toxic and relatively simple in or cation so that observations may be made on many specific periods of time with and without drug the results should be reproducible. placebo-and variations in each approach have subjects; and to investi- None of the methods described meets these condi- appeared almost from investigation tions completely. The reviewers believe the citric- gation. acid-aerosol technique to be the most promising, As much as 31 years ago, Davenport (1938) admin- best developed, and least disturbing to the subject, istered codeine and closely related substances double- but further efforts to develop methods for antitussive blind to coughing tuberculous patients and concluded evaluation should be encouraged. Artificially pro- that the optimal oral dose of codeine was only 10 mg, duced cough, in volunteer subjects, is not the same and Himmelsbach et al. (1940) using placebo for as cough of pathological origin at least in strength, comparison, demonstrated worsening of the cough timing and persistence of cough induction, and one when the change was to the placebo from morphine cannot always expect to obtain the same drug res- or codeine. It has been in the past 15 years, however, ponse in both situations. Nevertheless, the experi- since the discovery of the antitussive property of ments with artificial cough are worth while for deter- dextromethorphan, that particular attention has been mination of carry-over of effectiveness in this and directed to quantification in evaluative efforts. other respects from laboratory animals to man, but Assessment of efficacy is of necessity a comparison should be accompanied by studies of absorption, between groups, as comparable as possible in size distribution and disposition for further information and composition, receiving different agents or in the on species parallelism. Then one has the background individual receiving two or more agents. He may be for interpretation of laboratory results and for clin- asked to compare his condition at administration ical application. with that at fixed time intervals thereafter, or his 434 N. B. EDDY AND OTHERS condition during one period of medication, a day or reduction in bouts of coughing. Woolf & Rosenberg a week or longer, with that during another equal (1964) also compared their patients' judgement of period. He may be asked simply to judge improve- change in coughing with the results recorded auto- ment, no change or worsening, to express his prefer- matically. An actual reduction in the number of ence for the agent used at one time over another, or coughs by 50 % was recognized in only 36 of 90 cases. to judge the severity of the symptoms (number or The authors concluded that the majority of the severity of coughs, etc.) or of the change in the symp- patients did not appreciate the extent to which their toms, whether slight, moderate, great or none. In cough varied from day to day and were frequently either of the latter judgements, numerical scales may unaware of the effect of antitussive preparations. be employed where the highest number means the Gravenstein, Devloo & Beecher (1954) reached the least effect, if the symptom is assessed directly, or the opposite conclusion, that their patients felt better and greatest effect ifchange is beingjudged. Alternatively thought they coughed less. results have also been expressed as a percentage of Various methods of statistical analysis of results, the group of patients reporting, according to the whether subjective or objective, have been used. It criterion employed. All of these methods have been may be of interest to mention Armitage's sequential used in the evaluation of antitussives and are analysis, since its applicability was first demonstrated probably best illustrated with specific results in the in an antitussive study (Snell & Armitage, 1957). It evaluation of codeine (see below). has the advantage of making evaluation possible Davenport (1938) mentioned the unsatisfactory use while the trial proceeds, avoiding undue continuation of kymographic recording of chest movements for or premature termination. No doubt conflicting determination ofthe frequency ofcoughing. Graven- results in clinical trials are influenced by using too stein, Devloo & Beecher (1954) introduced a sound few patients. Improper dosage selection can also recorder, connected to a microphone, from which bias the results by destroying the double-blind feature numbers of bouts or numbers of individual coughs through the appearance of recognizable effects unre- could be counted. The sound-induced electrical im- lated to the one under examination. pulses were registered kymographically. This auto- Finally, it is of interest to review the role of the matic record correlated well with that of a person placebo and the degree ofeffect which it has produced monitoring the cough by direct count. With minor in antitussive testing. Table 26 summarizes some variations the method has been employed by Krug placebo results, and is restricted to those reports (1964), Kleibel (1964), Kleibel, Steinhoff & Wild which could be interpreted in terms of the percentage (1965), Sevelius & Colmore (1967), and others. of the group of patients who improved. From these Woolf & Rosenberg (1964) used a tape-recorder to data an average analgesic effect of placebo of 22.8 % register the actual sound of the cough, and this had may be calculated-somewhat less than the average the advantage for recording over prolonged periods figure (35.2%) calculated by Beecher (1955) as the that the on-off switch was automatically operated by average percentage analgesic effect of placebo. One the sound. Others have overcome the prolonged other point was brought out by Renovanz & Liebrich observation difficulty by restricting the record to a (1965). On the basis of the patients' judgement of specified period, uniform from day to day and based improvement in frequency and intensity of coughing, upon survey of the coughing pattern throughout the placebo was less effective when compared with 24 hours. codeine, 29.0 % and 32.3 % reductions in number and The subjective and objective approaches to evalu- intensity of cough, respectively, than when compared ation of antitussive action have been compared with with no medication, 40.4% and 60.9 % reductions in different results. Gravenstein, Devloo & Beecher number and intensity. (1954) could not detect changes in cough counts to back up judgement of an antitussive effect of codeine compared with placebo. Nicolis & Pasquariello PHARMACOLOGICAL COMPONENTS OF THE ANTITUSSIVE (1962a, 1962b) recorded a significant reduction in the EFFECT number of coughs by codeine versus placebo when the patients' judgement of effect, based on a four- Specific cough-depressant actions point scale, showed no difference. In another study, however, Pasquariello & Nicolis (1962) showed good The mechanisms of antitussive action have been correlation between patients' drug preference and discussed extensively by Bucher (1958), Bickerman CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 435

TABLE 26 ANTITUSSIVE EFFECT OF PLACEBO MEDICATION

Patients IPercentage of patients Reference No. Type improved

134 Adults and children; cough of various origin 5.2 Carter & Maley (1957) 44 Cough due to common cold 7.0 Parish (1959) 60 Chronic cough of various types 14.0 Haslreiter (1959) 65 Chronic cough 39.2 Di Gregorio (1960) 43 Chronic cough 35.6 Di Gregorio (1960) 25 Cough due to acute inflammation 28.5 Di Gregorio (1960) 50 Children; acute bronchitis 22.2 Abruzzi (1962) 50 Adults; non-productive cough 26.7 Abruzzi (1962) 18 Chronic cough of various types 23.0 Mulinos, Nair & Epstein (1962) 26 Tuberculous 6.0 Mulinos, Nair & Epstein (1962) 30 Chronic cough 39.9 Woolf & Rosenberg (1962) 46 Young patients; viral infections 7.5 Amler & Rothman (1963) 35 Cough of various origin 29.0 Schofield (1963) 127 Acute cough of various origin 40.4 Renovanz & Liebrich (1965)

(1960), Friebel (1960, 1963), Salem & Aviado (1964) presumably on receptors in the pharyngeal area. The and others. According to the structural organization inhalation of aerosols containing 1 % cocaine or 2 % of the reflex arc, specifically-acting antitussives might procaine acting on receptors in the pharyngeal, tra- exert this effect on the mechanoreceptors or on the cheal and bronchial regions, has been recommended chemoreceptors along the afferent fibres in the by Martini (1957) for the treatment of asthmatics medullary cough centre and, theoretically at least, troubled by irritative cough. In animal experiments, upon the responding structures. suppression of cough, provoked by mechanical stim- ulation of the trachea, by local application of a pro- Reflexogenic receptor area. Such a statement as caine solution has been demonstrated (Widdicombe, the following suggests that specific antitussive action 1954). Thus reduction of cough impulses at their site at the receptor site is known and the validity of such of origin by locally applied drugs is possible, perhaps action proven-" This new antitussive may well act usual, but none of the agents mentioned possesses a both centrally by depressing the cough centre and specific antitussive effect. " Local anaesthetics are peripherally by reducing the tussal impulses from only moderately selective" (Bucher, 1958). local irritation " (Chen, Biller & Montgomery, 1960) Attempts to obtain inhibition of receptors sensitive -but the drugs of choice for such peripheral inhibi- to cough-provoking stimuli by oral or parenteral tion would be local anaesthetics having special affin- administration of drugs have been made by Bein & ity for receptors responsive to tactile and chemical Bucher (1957), Ther & Lindner (1958), Sell, Lindner irritation. Direct application of local anaesthetics is & Jahn (1958) and others. The first attempts were the routine provision against undesired cough reac- made with local anaesthetics. Kase (1955) tested the tions during bronchoscopy (Frey, Hugin & Mayrho- cough-depressing effect of cocaine and procaine in fer, 1955). Tetracaine or lidocaine in one form or unanaesthetized dogs repeatedly stimulated by bris- another are inhaled or sprayed over the laryngeal and tles pushed against the tracheal bifurcation. The tracheal surfaces. Ethoform (p-aminobenzoic acid drugs were injected into the femoral vein or into the ethyl ester), as an ingredient of antitussive lozenges, cistema cerebello medullaris. The results were as is used against irritative pharyngeal cough, acting follows: 436 N. B. EDDY AND OTHERS

Drug Dose (mg/kg) Cough depression two sets of receptors, but a dose of 2.0 mg per kg IV Intra- No. No. of was sometimes lethal and was too close to the other- cistern positive experiments wise effective dose to suggest any sort of cough- Cocaine specific endoanaesthetic action. With tetracaine the hydrochloride 3.0 0 2 dose causing inhibition of the cough centre was 7.0 3 3 0.25-0.5 mg per kg, intravenously, and that for 10.0 3 3 diminution of the activity of the receptor systems 0.1 0 2 2.0 mg per kg; this again did not indicate that selec- 0.3 2 2 tive action on the receptors was involved. Further- 0.5 1 1 more, although many antitussives that have been 1.0 1 1 used clinically have also been shown to be local Procaine anaesthetics (e. g., isoaminile, clofedanol, pholcodine, hydrochloride 10.0 0 3 diphenhydramine, oxolamine, caramiphen, etc.) a 20.0 2 3 3' role for such action in promoting the antitussive 0.1 0 2 effect has not been proven experimentally or clini- 0.3 1 2 cally, however reasonable this may seem. 0.5 3 3 Afferent nerve fibres. Specific pharmacodynamic blockade of fibres transmitting cough-inducing An effect was obtained but large intravenous doses afferent action potentials has not been demonstrated. were required and probably had only a small margin In special circumstances, interruption of the afferent of safety. Such doses would stimulate respiration pathways could be effected by section (cough pro- and and affect the blood pressure cardiac rhythm. voked by the involvement of, or pressure from, a one- mean cat The intravenous lethal dose for the was only sided bronchial tumour (Bucher, 1958) or by conduc- 30-40 mg per kg for procaine and 15 mg per kg tion anaesthesia (Tomori et al., 1960)). for cocaine (Sollmann, 1957). Kas6 did not recom- mend this antitussive effect or ascribe a peripheral Cough centre. The truly specific site of action of site of action; the smallness of the effective doses cough-depressing drugs may be presumed to be the when given into the cisterna suggests indeed that the cough centre, or its neuronal system, in the medulla. action was central. " In this area all factors leading to cough are united " Ther & Lindner (1958) studied the antitussive po- (Bucher, 1958). However, even among the com- tency of local anaesthetics, and found no correlation pounds acting at such a site the specificity of action, between it and surface anaesthesia on the rabbit the details of the response and the breadth of other cornea or conduction anaesthesia in the frog sciatic surrounding actions vary. Dextromethorphan and nerve. They used the superior-laryngeal-nerve-stimu- noscapine, for example, have little of note other than lation technique (cat) and demonstrated cough antitussive action; other materials, such as codeine, depression most readily among substances related to pipazetate, 2-aminoindane, and hexobarbital act on tetracaine. They suggested that the presence of a broader scale. Hahn & Friebel (1966) studied the p-aminobenzoic acid in the molecule enabled the range of action of such a group, using unanaesthe- compound to gain access to the cough centre. tized guinea-pigs and producing cough by stimulation Sell, Lindner & Jahn (1958) compared the endo- ofelectrodes implanted in the trachea. Data recorded anaesthetic and antitussive properties of tetracaine, included the electric current at the cough threshold, benzonatate and related compounds in cats and the integration of the explosive expiration (the cough) into the breathing rhythm, a kymographic record of guinea-pigs. The former was determined from the transmission of stretch-receptor impulses arising in the cough reaction, and the sound of the cough before and treatment. the the lung, via the vagus nerve, and the latter from the during They reported following: transmission of mechanoreceptor impulses arising in (1) Hexobarbital-antitussive dose 70 mg per kg, the trachea, via the superior laryngeal nerve. The subcutaneously. The cough threshold was elevated antitussive activity was measured after electrical by 3.5 mA (the cough was somewhat modified when stimulation of the superior laryngeal nerve for central the stimulation was elevated above the threshold) evocation of cough. In these experiments benzona- and there was somnolence to narcosis, and retar- tate at 1.0 mg per kg, intravenously, depressed the dation of respiration. In contrast to normal the cough and the transmission of the impulses from the cough reaction began invariably with an inspiration. CODEINE AND ITS ALTERNATES FOR PAIN AND COU GH RELIEF. 3 437

The subsequent reduction in the cough response was mate the incoming stimuli and that this failure of slowed; in some cases animals coughed for three summation is the principal effect of the antitussive seconds after interruption of the cough-provoking compound, though it may also sometimes change current. A broadening of the kymographic V-record the character or form of the response. of the cough further illustrated the retardation of Some compounds affect the correlation between the response. Signs of closure of the glottis as a the respiratory cycle and cough. With hexobarbital, part of the event disappeared and the sound of the as was noted, the cough was initiated with an inspira- cough was weakened. tion; the same was true in barbiturate-treated cats (2) Aminoindane-antitussive dose 14.5 mg per kg, (Kroepfli, 1950). In untreated cats, 16 % of electric- subcutaneously. The cough threshold rose quickly ally-provoked cough reactions began without the by 1.5 mA and showed no tendency to recover during introductory inspiration. the next three hours. Respiratory frequency was If the central action of antitussives were directed increased and there was motor excitement, piloerec- primarily against the stimulus-summating function tion, secretion of tears and saliva. The explosive of the centre, this would explain the equal therapeutic expiration began during the expiratory phase of the effectiveness of these drugs against most forms of breathing cycle more frequently than is normal; it irritating cough. Codeine, dextromethorphan, pipa- was never integrated into the breathing cycle as zetate, and benzonatate seem to be equally satisfac- sometimes occurred before drugging. The sound of tory whether the cough arises from acute or chronic the cough increased and the kymographic record bronchitis, tuberculosis, emphysematous bronchitis, became irregular and broader in form. asthmatic bronchitis or bronchial carcinoma; the doses necessary for diminution of the cough are (3) Codeine phosphate-antitussive dose 15.4 mg about equal, and this is understandable only if they per kg, subcutaneously. The cough threshold was are acting uniformly on a central mechanism acti- elevated by 0.5 mA for about three hours (a larger vated by all kinds of cough. The clinical significance dose of 30.8 mg per kg elevated the threshold by of the other cough modifications described by Hahn 0.9 mA). The noise of the cough was weakened and & Friebel (1966) has not been clarified. its kymographic record was simplified but not broad- ened. Efferent activating structures. Inhibition of cough through an effect on responding structures has been (4) Pipazetate-antitussive dose 14.5 mg per kg, claimed for oxolamine (Silvestrini & Pozzatti, 1960) subcutaneously. The cough threshold was elevated and sodium dibunate (DeVleeschhouwer, 1954) as a by about 1.6 mA, and was still above normal after part of peripheral antitussive action. However, the three hours. The noise of the cough was decreased; respiratory centre is also in control of these structures the record was irregular and sometimes broader. and differentiation is not possible. (5) Dextromethorphan-antitussive dose 26.0 mg per kg, subcutaneously. The cough threshold was Effects other than specific cough-depressant action raised by 0.4 mA for three hours. Respiration was Cough-depressing substances, of course, exhibit unchanged. The explosive expiration began more other pharmacodynamic effects, which may be detect- frequently (18%) than is normal in the expiratory able at the antitussive dosage level and which may phase of the breathing cycle, and in 14% of cases at promote or even interfere with antitussive usefulness. the very end of expiration. The record showed numerous irregularities and broadening. The noise Respiratory effects, other than antitussive effects. of the cough was decreased only during the first B-nzonatate, tetracaine and related substances, if 20 minutes. injected intravenously, diminish the spontaneous activity of the pulmonary stretch receptors, though Thus these compounds produced several different the margin between effective and toxic doses is small effects, only one common to them all, an elevation (receptor-depressing doses in cats are between 0.5 and of the cough threshold, which could be overcome by 2.0 mg per kg whilst depression of blood pressure, increasing the strength of the stimulation. The sen- temporary arrest of respiration, and ozcasional death sitivity of the centre is decreased but it is not para- occurs with doses of 2.0 or 3.0 mg per kg). Depres- lysed. Based upon their experimental results Friebel sion of the pulmonary stretch receptors is probably & Hahn (1966) and Hahn & Friebel (1966) concluded not significant in antitussive action. Cutting off im- that the centre temporarily loses the ability to sum- pulses from these receptors from the vagus, below 438 N. B. EDDY AND OTHERS the entrance of the superior laryngeal nerve, did not mechanism for both types of relief, but antitussive affect the cough response to electrical stimulation of action without analgesia implies a fundamental dif- that nerve (Widdicombe, 1954; Sell, Lindner & Jahn, ference in the mechanisms involved. 1958). Pholcodine enhanced the Hering-Brauer pul- Friebel, Reichle & von Graevenitz (1955) and monary-inflation reflex, which May & Widdicombe Reichle & Friebel (1955) assessed the antitussive and (1954) thought was secondary to bronchoconstriction analgesic potencies of a number of compounds in and indicated increased activity of the pulmonary guinea-pigs and rats by uniform methods (inhalation stretch receptors, yet pholcodine was equivalent to ofsulfur dioxide and skin-twitch to radiant heat) with codeine as an antitussive. the following results: Many antitussives affect central regulation of res- Antitussive Analgesic Ratio: piration. Friebel & Kuhn (1962) and Hahn & Friebel Agent EDro ED,, analgesicl (1966) demonstrated that hexobarbital decreased (mg/kg) (mg/kg) antitussive respiratory volume and frequency at doses below Guinea-pig those which affected the cough reflex mechanism. Methadone 1.6 3.7 2.31 Pentoxyverine 1 citrate, pipazetate and the piperidine Morphine 4.0 11.5 2.87 Normethadone 4.6 12.3 2.66 derivative, L 2909, depressed the cough reflex at a Pethidine 8.0 12.3 1.53 dose level which did not seem to affect respiration, Codeine 8.0 48.5 6.62 but pipazetate would increase the respiratory depres- Ethylmorphine 13.0 43.0 3.30 sant effect of hexobarbital. 2-Aminoindane stimula- Dihydrocodeine 14.0 80.0 5.71 ted respiration even when the respiratory centre was Rat depressed by previous administration of hexobarbital Methadone 1.0 1.8 1.80 and its antitussive effectiveness equalled that of Morphine 3.0 2.3 0.76 codeine (Friebel & Kuhn, 1962; Kuhn & Friebel, Pethidine 5.6 8.2 1.46 1960a, 1960b). Again these results support the sug- Normethadone 7.0 5.8 0.82 gestion that the respiratory centre, in its reactivity to Codeine 10.0 13.0 1.30 drugs, is to a considerable extent independent of the cough centre, that depression or stimulation of res- The higher the ratio the greater the antitussive piration may be independent of antitussive effective- specificity, and this was greater for codeine than for ness (Eichler & Smiatek, 1940; Haas, 1955; La morphine in both species in spite of the greater anal- Barre & Plisnier, 1959), and that the degree ofrespira- gesic effectiveness of the latter. These data further tory depression produced by many antitussives does support the independence (or dissimilarity) of the not impair their antitussive usefulness at usual thera- two effects. A further illustration is pholcodine. neutic doses. Structurally closely related to codeine, it has com- parable antitussive action but no analgesic action Analgesic effect. The virtually exclusive use of the (Kelentey et al., 1958; Cahen & Boucherle, 1961). opiates, or of analgesics related to morphine, for Many other authors have claimed equivalence to antitussive effect was interrupted by the observations codeine in clinical antitussive effectiveness for com- of Pellmont & Bachtold (1954) and Winter & Flata- pounds which lacked ability to relieve pain (Capello ker (1954), which demonstrated, on the one hand, & DiPasquale, 1955; Maurer, 1955; Carter & Maley, that the structurally related dextromethorphan, and, 1957; Cass & Frederik, 1958; Haslreiter, 1959; on the other hand, that the unrelated noscapine both Bickerman, 1960, Prime, 1961). depressed experimentally produced cough, although Analgesic compounds, at least compounds with both lacked analgesic action. a morphine-like analgesic effect, will depress the Cough and pain are nociceptive warning signals cough response, but compounds which depress the which can evoke reflex protective responses and which cough response are not necessarily analgesic. are perceived as discomforts of greater or less degree. Antitussives and analgesics suppress the response Sedative and tranquillizing effect. Gravenstein, under experimental conditions and make the per- Devloo & Beecher (1954), failed to detect a decrease son feel better by decreasing his awareness of the in cough frequency and implied, at least, that the discomfort, apart from the suppression of the res- sedative effect of codeine made the patient feel better ponse. This would seem to bring into play a similar and he thought he coughed less. After subcutaneous injection of hexobarbital in I Also known as carbetapentane. guinea-pigs the sequence of symptoms with increas- CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 439 ing dosage was sedation at 50 mg per kg (to anaesthe- In man, as in laboratory animals, narcosis abo- sia), depression of respiration and depression of lishes the cough reflex; in deep narcosis the mecha- cough at 77 mg per kg (Friebel & Kuhn, 1962). Is nism is paralysed. Barbiturates and other sedatives, the depression of cough in hexobarbital-treated in sleep-inducing doses, are not expected to act as animals a specific antitussive action or a part of the cough depressants, though the sedation may be as general depression of reflexes? Hahn & Friebel useful an adjunct to cough therapy in some circum- (1966) observed, in guinea-pigs in which cough was stances as simultaneous analgesia may be in others. elicited by electrical stimulation, that hexobarbital at 70 mg per kg, subcutaneously, first caused sedation Neuroleptic effect. Chlorpromazine, a standard (sometimes with delirium) followed by decrease in substance among neuroleptic compounds, and thio- respiratory frequency and later, while in an anaesthe- ridazine act similarly, protecting against vomiting, tized, condition by a retardation and decrease in the having moderate anticholinergic, adrenolytic and cough response which, as anaesthesia declined, out- antihistaminic properties, and are both credited with lasted the respiratory depression. Throughout the strengthening the pharmacodynamic effects of opi- period under hexobarbital the cough was always ates, barbiturates and other sedatives and anaesthe- initiated by inspiration, which was not the case in tics. Both substances, according to Boissier & Pagny untreated animals. (1960a, 1960b), showed a good antitussive effect in Tedeschi et al. (1959) found an ED50 of 13.7 mg guinea-pigs subjected to sulfuric acid aerosol. Pro- per kg for pentobarbital sodium in unanaesthetized methazine, which is weaker in its neuroleptic potency, also the at dogs, for depression of cough due to mechanical depressed cough reflex higher doses. The stimulation. However, the barbiturate was effective authors believed that non-specific central depressant only in doses which produced signs of neurological properties were mainly responsible for the cough- deficit, marked ataxia and/or unconsciousness indi- depressing effect. Kase & Yuizono (1959) tested cating that " pentobarbital acts as a non-selective some compounds structurally related to chlorproma- antitussive agent ". The cough response of decere- zine and found chlorpromazine in dogs, at 20 mg per brate cats to electrical stimulation has been abolished kg intravenously, sometimes slightly effective, per- by thiopental, 4-12 mg per kg, intravenously (Matal- phenazine, prochlorperazine and other phenothia- lana & Borison, 1955). zines more definitely effective against cough at equal or smaller doses. Thus powerful neuroleptics affect Methaqualone, similar in its sedative effect to the the cough-reflex mechanism and act in this respect barbiturates, diminished by 50% the cough response like the opiate analgesics; absence of neuroleptic to electrical stimulation in narcotized cats at a dose action in substances related to chlorpromazine does of 4 mg per kg, intramuscularly (Boissier & Pagny, not exclude antitussive effectiveness. 1960a, 1960b). Up to 40 mg per kg, it also had an Tomori antitussive effect in unanaesthetized ex- et al. (1960) studied the effect of neurolep- guinea-pigs on and posed to sulfur dioxide, without sedation. tics tracheal bronchial reflexes in man. Ten producing patients under (1) ether-oxygen anaesthesia, (2) Certain hypnotics then may cause specific antitus- " sive effects. A broader examination of the antitussive autonomic blockade with the " lytic cocktail or (3) blockade were potency of sedatives (hypnotics) is desirable to deli- plus hypothermia, stimulated mechan- neate better the relation between sedation and cough ically in the lower trachea. Signs of positive reflex relief. activity were coughing and apnoea. A positive res- Lorenz the and antitussive ponse failed to occur in 12 % in (1), 26.5 % in (2) and (1962) studied sedative in The mechanism effects of dipropylbarbituric acid, codeine, and a mix- 46.9% (3). of the cough de- ture of the two in guinea-pigs and cats. In anaesthe- pressing effect is a matter for speculation, since the tized composition of the " lytic cocktail " was not dis- cats, stimulated electrically, the antitussive effect closed but of intravenous codeine was not increased by the addi- may have contained chlorpromazine or tion of the barbiturate. In unanaesthetized guinea- pethidine or both. There may have been a deepening pigs, however, the cough reflex was depressed by of the ether narcosis or strengthening of the antitus- smaller doses of codeine in the mixture than when sive action of pethidine. it was given alone, subcutaneously or orally; the Ganglionic blockade. Three observations may jus- sedative effect of the barbiturate in the mixture was tify a consideration of whether ganglionic blockade also enhanced. Whether this was potentiation ofanti- can influence the cough reaction. Tiffeneau (1955) tussive action or of sedation was not clarified. found that inhalation of hexamethonium instantly

8 440 N. B. EDDY AND OTHERS

stopped the excessive cough reactions produced showed similar properties. Oxeladin exerted papa- in man by inhalation of acetylcholine aerosol. Hillis verine-like spasmolytic action on the isolated ileum & Kelly (1951) suppressed the cough, produced in (David, Leith-Ross & Vallance, 1957) and on the human volunteers by intravenous injection of lobe- tracheal chain preparation (Ozawa et al., 1962). The line, also by using hexamethonium. Korpas, Tomori thiambutene derivative, AT-327, had little spasmo- & Ivanca (1957) studied the cough-depressing effect lytic effect (Kase et al., 1959). of pentamethonium in cats. Doses, which were under Epinephrine caused no depression of the cough some experimental conditions slightly effective reflex, mechanically stimulated in unanaesthetized against cough reactions, were considerably higher dogs (Kase, 1955), but ephedrine, less effective as a than those which paralyse the cervical ganglion. bronchodilator, depressed the cough in 6 out of 12 Tests with quatemary salts of antitussive agents gave dogs, at a dose of 5.0 mg per kg, intravenously. no convincing evidence that ganglionic blocking con- Increasing the dose to 10-20 mg per kg did not pro- tributes to the cough-depressing effect. duce an effect in the dogs which did not respond. Methamphetamine' was more centrally exciting than Local anaesthetic effect. The role of local anaesthe- ephedrine, and had antitussive (Kase, 1955) and sia in antitussive action has been discussed (see above, analgesic (Suter, 1951) properties. Part 3, section headed " Reflexogenic receptor area "). Codeine, as well as morphine, causes slight bron- Spasmolytic and spasmogenic effects. The evidence choconstriction (Krueger, Eddy & Sumwalt, 1943; in this connexion may be summarized briefly. Konzett, 1955; La Barre & Plisnier, 1959) and the Some antihistamines depress experimental cough codeine bronchospasm can be relieved partly by in laboratory animals and in man (see Part 4). How- isoprenaline (Friebel & Kuhn, 1962). Pholcodine ever, Boissier & Pagny (1960a, 1960b) concluded that also has a bronchoconstrictor effect comparable to antihistamines were not antitussives per se. They that of codeine (Plisnier, 1960). Antitussive potency could not demonstrate a parallelism of antihistaminic would not appear to be impaired, though warnings and antitussive potencies. with respect to impairment of respiratory function Atropine was tested in unanaesthetized dogs by have been expressed, especially if a bronchospastic Kase (1955). It had no cough-depressing action in condition already exists. doses of 0.1-5.0 mg per kg, intravenously. Silvestrini All of this does not suggest any significant relation- & Maffii (1959) observed no significant effect of ship of spasmolytic or spasmogenic effects to anti- atropine on the acrolein-induced cough in guinea- tussive action, though under some circumstances a pigs; the doses used were 1.0-5.0 mg per kg, sub- potent bronchodilator may help in the relief of an cutaneously. Bobb & Ellis (1951) treated with atro- associated condition. pine and curare unanaesthetized dogs, in which the Augmentation and reduction of bronchial secretion. vagus loop was electrically stimulated: both were Noscapine, 1.0-2.0 mg per kg, intravenously, in- ineffective. creased bronchial secretion in anaesthetized rabbits Domenjoz (1952), Kase (1955) and Hara & Yanau- (Konzett & Rothlin, 1954; La Barre & Plisnier, 1959; ra (1959) showed no definitive effect of papaverine Plisnier & Hernalsteen, 1959). The dose was sufficient in dogs and cats, though the doses were high in com- to depress the cough. Caramiphen ethane disulfonate parison with that of codeine. This is discussed more also increased bronchial secretion in the same type of fully in Part 4. experiment at 6.0 mg per kg, intravenously (Plisnier, Isoaminile relaxes other smooth muscle, isolated 1957). Silomat, 40 mg per kg intravenously, increased rabbit uterus and intestine, and inhibits the peristal- bronchial secretion in rabbits by 20-45% for 2-6 tic reflex (Krause, 1958; Graham, 1962). Clofedanol hours (Engelhorn, 1960). and nicodicodine relaxed the rabbit duodenum and Pholcodine, orally or subcutaneously in doses of relieved spasms induced by barium chloride or acetyl- 0.1-50 mg per kg, did not cause a significant change choline (Cahen & Boucherle, 1964, 1965). Oxolamine in output of respiratory-tract fluid; nor did diphen- was about as active as papaverine in its inhibitory hydramine (Bidwell & Boyd, 1952). effect on the tracheal chain preparation (Silvestrini Bedell & Seebohm (1959) warned against the use & Pozzatti, 1960; Balea & Buret, 1964). Caramiphen of codeine in patients with emphysema because of prevented or retarded bronchoconstriction provoked its drying effect. Boyd & Trainer (quoted by Boyd, in guinea-pigs by injection or inhalation of acetyl- choline or histamine (Plisnier, 1957). Pentoxyverine 1 Also known as methylpropamine. CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 441

1954) recorded a decrease in bronchial secretion after Perhaps with greater significance for the therapeu- dextromethorphan. Clofedanol (Boyd & Boyd, 1960) tic use of codeine in man, it would be interesting to and pentoxyverine (Levis et al., 1955) decreased the know if codeine could depress or abolish cough, in secretion of saliva and presumably of bronchial fluid. animals, which occurred spontaneously in bronchitis, There is no clinical evidence that antitussive ther- for example. Stefko, Denzel & Hickey (1961) report- apy is significantly affected by the changes in secre- ed suppression of cough due to upper respiratory tion described. infection, in a dog for 30 minutes with 1.0 mg per kg and for 38 minutes with 2.0 mg per kg, orally. Exper- ANTITUSSIVE ACTION OF CODEINE imentally produced cough in dogs has been abolished by codeine, if the conditions were adequate, for 40 Animal experiments minutes with 4.0 mg per kg, orally, for example, Many reports of attempts to evaluate the antitus- when the cough was elicited by electrical stimulation sive effectiveness of codeine in animals are summa- of the trachea. Chen, Biller & Montgomery (1960) rized in Table 27. These have been arranged by have reported complete suppression of cough pro- species and by the method of cough production, duced in dogs by chemical stimulation. though the latter has not been uniform in detail, Other authors who have attempted to establish even when the cough-provoking stimulus was said to dose-effect relationships have not been successful in be the same. Some of these variations have been obtaining 100% suppression of experimentally in- discussed previously (see Part 3, section headed duced cough (Kdllquist & Melander, 1957; Silves- " Measurement of antitussive potency in animal trini & Maffii, 1959; Engelhorn, 1960). The explana- experiments "). Also the effective doses for the most tion may lie in the experiments of Rosiere, Winder part are indicative only of reduction or depression of & Wax (1956) which showed that complete inhibition the cough of varying degree, an ED50 having been of cough could be overcome by intensifying the calculated only in rare instances (those doses are cough-provoking stimulation, but that stronger stim- marked in Table 27). Consequently, doses range ulation could be overcome only to a certain extent widely, even for similar methods and species, from by increasing the dose of codeine. 10 to 26 mg, subcutaneously or intraperitoneally, in Onset and duration of action of codeine in relation mice, rats and guinea-pigs with irritant vapour to experimentally induced cough in animals have inhalation, but calculated ED50s vary widely also, been studied by Winter & Flataker (1954), Kas6 10 mg per kg and 13.5 mg per kg, subcutaneously, in (1955), Chen, Biller & Montgomery (1959), Engel- rats, 25 mg per kg by the same route in guinea-pigs horn (1960), Stefko, Denzel & Hickey (1961), Kraus- with sulfur dioxide inhalation. If ammonia vapour haar, Schunk & Thym (1964), Hahn & Friebel (1966), was used in guinea-pigs the effective dose was only and others. In conscious dogs, for example, some 0.3 mg per kg, subcutaneously. Even so the dose had cough-depressing action has been observed 15 min- to be increased remarkably, to 12 mg per kg, if the utes after oral administration, peaks of action at drug was given orally. In rabbits also, with cough pro- 45-60 minutes after ingestion (Chen, Biller & Mont- duced by mechanical stimulation, there was a consid- gomery, 1959; Stefko, Denzel & Hickey, 1961) and erable ratio between oral and parenteral effective a duration of action mainly dependent upon the dose doses. In cats there was very wide variation in of codeine but usually not exceeding three hours. effective dosage. Electrical stimulation and intra- Repetition of codeine doses at short intervals was venous administration were commonly used in this followed by tolerance to antitussive action (Kase, species but mostly in anaesthetized animals. The 1955; Friebel & Kuhn, 1964). In the experiments of anaesthetic, usually a barbiturate, possessing in itself Kase, dogs were injected intravenously daily. The some cough-depressing action (see above, Part 3, sec- initial dose of 10 mg per kg depressed coughing for tion headed " Sedative and tranquillizing effect ") 20-30 minutes on the first day; the same dose was might be in part responsible for these variations, ineffective on the 14th day of administration. The because of differences in depth of anaesthesia, but cough-depressing effect, which reappeared when the there are many other variables in these experiments. dose was increased to 15 and 20 mg per kg, was The sensitivity of dogs to the cough-depressing action reduced again in 4-10 days. Friebel & Kuhn treated of codeine appears to be high and more uniform, guinea-pigs and rats with codeine, subcutaneously, with a dose range of only 0.5-7.5 mg per kg for three times a day for six to seven weeks. The daily various procedures and routes of administration. dose was increased from 20 to 110 mg per kg and 442 N. B. EDDY AND OTHERS

TABLE 27 ANTITUSSIVE ACTION OF CODEINE IN ANIMALS

Route Effective Cough-provoking agent Anaesthesia of admi- dose Reference nistration a (mg/kg)

Species: MOUSE

Sulfuric acid aerosol 0 SC 10.0 Kelenty et al. (1957)

Sulfuric acid aerosol 0 SC 1 Kelenty et al. (1958)

Species: RAT

Sulf ur dioxide vapour Friebel & Kuhn (1964) Sulfur dioxide vapour Reichle & Friebel (1955) Sulfur dioxide vapour Hengen & Kasparek (1958) Sulfur dioxide vapour Weidemeyer, Kramer & DeJongh (1960)

Species: GUINEA-PIG

Sulfur dioxide vapour 0 SC 15.4 b Friebel & Kuhn (1962)

Sulfur dioxide vapour 0 SC 20.0 Boissier & Pagny (1960) Sulfur dioxide vapour 0 SC 19.0 b Friebel & Kuhn (1964) Sulfur dioxide vapour 0 IP 30.0 Cahen & Boucherle (1964)

Sulfur dioxide vapour 0 IV 10.0 Green & Ward (1955)

Ammonia vapour 0 0 12.0 Salle & Brunaud (1960) Ammonia vapour 0 SC 0.3 b Graham (1962) Ammonia vapour 0 SC 0.5-2.0 Winter & Flataker (1954)

Ammonia vapour 0 SC 1.0 Silvestrini & Maffii (1959) Acroleine vapour 0 SC 3.5 Silvestrini & Maffli (1959) Acroleine vapour 0 IP 1.0-2.0 Silvestrini & Pozzatti (1961)

Species: RABBIT

Mechanical stimulation of the trachea 0 0 100.0 Furakawa & Okabe (1958)

Mechanical stimulation of the trachea 0 IV 3.0-4.0 Furakawa & Okabe (1958)

Mechanical stimulation of the trachea 0 IV 13.5 Graham (1962) -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Species: CAT

Ammonia vapour IV 1.5 Silvestrini & Maffli (1959) Ammonia vapour IV 3.5 Pellmont & Bachtold (1954) Ammonia vapour IV 10.0 Mulinos (1960) Ammonia vapour +) c SC 0.61 Engelhorn (1960) Ammonia vapour IP 15.0 Cahen & Boucherle (1964) Mechanical stimulation of the trachea +) c IV 0.74 Engelhorn (1960) Mechanical stimulation of the trachea IV 3.0 Silvestrini & Maffli (1959) CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 443

TABLE 27 (continued)

Route Effective Cough-provoking agent Anaesthesia of admi- dose Reference I nistration a (mg/kg)

Species: CAT (continued) Mechanical stimulation of the trachea IV 10.0 Mulinos (1960) Mechanical stimulation of the trachea IV 16.3 May & Widdicombe (1954) Sterile pleurisy + SC 2.5-3.0 Van Dongen (1956) Soap powder insufflation IV 1.5 Pellmont & BAchtold (1954) Electrical stimulation of the superior laryngeal nerve IV 1.0-3.0 Domenjoz (1952) Electrical stimulation of the superior laryngeal nerve 0 8.1 Engelhorn (1960) Electrical stimulation of the superior laryngeal nerve (+) c IV 0.52 Engelhorn (1960) Electrical stimulation of the superior laryngeal nerve IP 1.0 Cahen & Boucherle (1964) Electrical stimulation of the superior laryngeal nerve IV 1.0 Plisnier (1958) Electrical stimulation of the superior laryngeal nerve IV 1.5 b Kohli et al. (1960) Electrical stimulation of the superior laryngeal nerve IV 1.5 Levis, Preat & Moyersoons (1955) Electrical stimulation of the superior + laryngeal nerve IV 1.5 Klein (1958) Electrical stimulation of the superior laryngeal nerve IV 0.5-2.0 Boissier & Pagny (1960a, 1960b) Electrical stimulation of the superior + laryngeal nerve IV 0.6-3.0 Haas (1955) Electrical stimulation of the superior laryngeal nerve IV 4.0 b Green & Ward (1955) Electrical stimulation of the medulla IV 1.0-6.0 Matallana & Borrison (1955) Electrical stimulation of the medulla IV 4.0-10.0 Chakravarty et al. (1956) Electrical stimulation of the trachea IV 1.0 Toner & Macko (1952) Electrical stimulation of the trachea IV 1.5 Silvestrini & Maffli (1959) Electrical stimulation of the trachea IV 2.3 Green & Ward (1955)

Species: DOG Sulfuric acid aerosol 0 0 0.5 Winter & Flataker (1952) Sulfuric acid aerosol 0 SC i.0 Silvestrini & Maffli (1959) Sulfur dioxide vapour 0 1.0 Chen, Biller & Montgomery (1960) Sulfur dioxide vapour IV 1.0 Green & Ward (1955) Ammonia vapour SC 2.5-5.0 Winder & Rosiere 0 (Personal communication, 1954) Ammonia vapour SC 2.0 b Rosiere & Winder (1955) 0 Ammonia vapour SC 1.0-5.7 Winder & Rosiere (1955) 0 Ammonia vapour SC 2.8 Winder et al. (1961) 0 Ammonia vapour SC 1.0-5.7 Rosiere, Winder & Wax (1956) 444 N. B. EDDY AND OTHERS

TABLE 27 (concluded)

Route Effective Cough-provoking agent Anaesthesia of admi- dose Reference nistration a (mg/kg)

Species: DOG (continued)

Mechanical stimulation of the trachea 0 SC 3.31 b Tedeschi et al. (1959) Mechanical stimulation of the trachea 0 IP 7.5 Cahen & Boucherle (1964) Mechanical stimulation of the trachea + IV ca 1.0 Green & Ward (1955) Mechanical stimulation of the trachea 0 IV 2.0-4.0 Kas6 (1952) Tracheal electrodes 0 0 0.25-4.0 Stefko, Denzel & Hickey (1961) Tracheal electrodes SC 3.5 Silvestrini & Maffii (1959) Tracheal electrodes 0 SC 1.0-4.0 Stefko & Benson (1953) Tracheal electrodes 0 SC 1.04.0 Benson, Stefko & Randall (1953) Tracheal electrodes 0 IV 2.04.0 Granier-Doyeux et al. (1959) Tracheal electrodes 0 2.5-3.5 Hara & Yanaura (1959) Tracheal electrodes + 5.0 Graham (1962) Pathological 0 0 1.0-2.0 Stefko, Denzel & Hickey (1961)

a SC = subcutaneous, IP = intraperitoneal, IV = intravenous, and 0 = oral. b Dose so marked is the calculated EDso. c (+) = Light anaesthesia. from 20 to 150 mg per kg for the guinea-pigs and rats, decreased by a maximum of 5.2 in the normal sub- respectively. The antitussive ED50 for guinea-pigs jects and by 5.9 in the asthmatics. In a later study increased from 19 to 55 mg per kg, and for rats from Bickerman et al. (1957) used individually determined 10.9 to 32.5 mg per kg. Tolerance in guinea-pigs citric acid aerosol concentrations as the cough- decreased slowly over a period of 27 weeks; it could provoking stimulus and a pneumotachograph for be induced again but at a slower rate. The graph of cough registration. They demonstrated a statistically the onset and decline of tolerance was linear if log significant dose-related decrease in the cough res- ED50s were plotted. ponse after oral doses of 5, 15 and 30 mg of codeine as shown below. Experimentally induced cough in man Dose Initial Percentage decrease after The antitussive action of codeine on experimental- (mg) no. of (hours) (m) coughs 1 2 3 4 ly produced cough in man has been studied exten- Placebo sively by many authors with various methods. Most 2.67 - 7.5 +2.5 +5.0 +25.0 have found it more or less effective, only a few have 2.81 +10.6 +1.3 -4.3 +12.8 failed to demonstrate a cough-depressing effect. The 3.33 - 8.0 -12.0 -12.0 +2.0 results are summarized in Table 28. Codeine Bickerman & Barach (1954) subjected 21 selected 5 2.76 - 17.2 - 22.4* -22.4* -17.2 healthy adults and 17 patients with bronchial 15 2.22 - 32.4* - 32.5* - 27.5* -10.0 asthma to a 5-10% citric acid aerosol for five in- 30 3.22 -32.8* -43.1* -34.5* -22.4* halations and obtained a fairly consistent cough * Significant difference. response, with a mean of 4.8 coughs in the normal subjects and 6.6 in the asthmatics. During the three- Shane, Krzyski & Copp (1957), Rasch (1957), hour observation period following oral administra- Archibald, Slipp & Shane (1959) and Calesnick, tion of a placebo or 30 mg of codeine in random Christensen & Munch (1961) demonstrated the anti- order under double-blind conditions, the mean cough tussive effectiveness of 15 or 25 mg of codeine in response was not reduced after the placebo but was their experiments based in general on Bickerman's CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 445

TABLE 28 EFFECT OF CODEINE ON EXPERIMENTALLY INDUCED COUGH IN MAN

Route No. and kind Cough-inducing agent Dose(m)Iadmi-Iof Effect Criteria for evaluation Reference of subject (mg nistra- tion a

21 Normal, selected 10 % Citric acid aerosol 30 0 Decrease in no. of coughs Bickermann & Barach (1954) 17 Asthmatics 30 0 + + 13 Normal 5-15 % Citric acid aerosol 10-60 SC Decrease in no. of coughs Gravenstein, Devloo & Bucher (1954) 20 Patients 15 % Citric acid aerosol 32.5 0 Decrease in no. of coughs Shane, Krzyski & Copp (1957) 20 Normal 10 % Citric acid aerosol 25.0 0 Percentage reduction in Rasch (1957) (syrup) cough 19 Normal, trained 10 % Citric acid aerosol 5, 15, 30 0 Percentage reduction in Bickermann et al. (1957) cough 14 Normal, selected 10-15 % Citric acid aerosol 30 0 Decrease in no. of coughs Archibald, Slipp & Shane + (1959) 24 Mainly 10-15 % Citric acid aerosol 30 0 +b Decrease in no. of coughs Groszman, Birker & tuberculous + Casimir (1961) 70 Normal 25 % Citric acid aerosol 15 0 Decrease in no. of coughs Calesnick, Christensen & + Munch (1961) 6 Normal, selected 10 % Citric acid aerosol 15 0 Decrease in no. of coughs Tannenbaum (1965) 30 + 6 Normal 5-25 % Citric acid aerosol 15-30 0 0 Decrease in no. of coughs Sevelius, Lester & Colmore (1965) 9 Normal Citric acid aerosol 30 0 + Increase in threshold Silson (1965) 32 Normal, selected 25 % Citric acid aerosol 5-120 0 Increase in latency Calesnick & Christensen (1967) 15 Normal Ammonia vapour 10-30 0 Total rise in threshold Hoglund & Michaelsson (1950) Normal Ammonia vapour + Hahn & Wilbrand (1952) 6 Normal Ammonia vapour 10-40 0 0 Total rise in threshold Trendelenbarg (1950) 30 Normal Ammonia vapour 20 0 Increase in threshold Blix (personal communi- cation 1952) 24 Normal Ammonia vapour 10-60 SC 0 Increase in threshold Gravenstein, Devloo & Beecher (1954) Patients, respira- Acetylcholine aerosol, 1% 40 0 (+)b Decrease in no. of coughs Tiffeneau (1955) tory disease + 120 Patients, Acetylcholine aerosol, 30 0 Suppression of cough Hasireiter (1959) respiratory disease 0.1-1 % 12 Patients, Acetylcholine aerosol, I % 16 0 + Decrease in no. of coughs Prime (1961) respiratory disease 12 Smokers Acetylcholine aerosol, 1 % 30 0 + Decrease in no. of coughs Sevelius, Lester & Colmore (1965) 0 1 Normal Peppermint water spray, 40 SC Percentage reduction of Hillis (1952) 1 :40 180 0 0 cough Normal Lobeline High IV 0 Suppression of cough Hillis & Kelly (1951) 5 Normal Paraldehyde 10-60 IV Decrease in no. of coughs Gravenstein, Devloo & 0 Beecher (1954) Normal Paraldehyde Suppression of cough Konar (1959)

a 0 = oral, SC = subcutaneous, and IV - intravenous. b Slight effect. 446 N. B. EDDY AND OTHERS method. Tannenbaum (1965) obtained results which acetylcholine inhalation methods feasible for anti- were not significant for a 15-mg dose of codeine but tussive testing. which showed significant cough suppression with Hoglund & Michaelsson (1950), who were the 30 mg compared with a placebo. Groszman, Birker first to develop a method for induction of cough, or & Casimir (1961) found a moderate antitussive effect its equivalent, experimentally in man, added amounts with 30 mg of codeine, orally, which was greater in of ammonia to the inspired air with human volun- those tuberculous patients who responded only mod- teers. The closure of the glottis and the momentary erately to a citric acid aerosol. pause in respiration which followed was considered Silson (1965) made his subjects inhale a fixed con- a substitute for a cough response and the antitussive centration of citric acid aerosol continuously until a effectiveness of codeine was calculated from the cough response occurred. From the time for this to amount of codeine required to raise the threshold. occur and the concentration of the aerosol, a citric A dose-response curve was established for 10, 15, 20 acid threshold stimulus was calculated and effect was and 30 mg ofcodeine. Trendelenburg (1950) used the expressed as percentage increase of this threshold. same method and criteria and found a good dose- An oral dose of 30 mg of codeine raised the threshold response relationship for 10-40 mg ofcodeine, orally, in the nine subjects tested. as shown below. Calesnick & Christensen (1967) considered the Codeine Maximum Duration Total effect dose No. threshold of elevation (o% x time) latency of response (time elapsing between beginning (mg) of trials elevation (min) of inhalation of a constant citric acid concentration (%) and occurrence of cough response) to be a purely 10 3 62 60 2 310 objective index useful for measuring antitussive po- 20 15 119 72 4695 30 5 91 100 4780 tency. By this parameter a single oral dose of 5 mg 40 3 137 80 5 950 of codeine exerted a measurable antitussive effect, differing with statistical significance from the res- In nearly all cases the threshold returned to the ponse to a placebo. A dose-response curve was initial value within the 3-4 hour observation period. plotted for doses of 5-120 mg with the maximum Hahn & Wilbrand (1952) and Blix (personal com- effect at 40 mg of codeine. These authors also stud- munication, 1952) also reported a depressing action ied the influence of the citric acid concentration in of codeine on ammonia-induced cough. the aerosol on the cough-suppressant effect of a Tiffeneau (1955), who initiated the use ofan acetyl- constant dose of40 mg ofcodeine. When the amount choline aerosol for cough production, saw only a of citric acid provided supramaximal stimulation, a very moderate effect on the cough frequency by an gradual reduction in the antitussive effectiveness of oral dose of 40 mg of codeine. Haslreiter (1959) codeine occurred until a cough-suppressing action conducted a double-blind trial on 120 patients with could not be distinguished from the response to respiratory disease, who coughed in response to indi- placebo. vidually determined concentrations of acetylcholine Gravenstein, Devloo & Beecher (1954) were unable aerosol. After 30 mg of codeine, orally, coughing to detect a cough-suppressing effect of oral doses bouts could not be induced with the same concentra- of 30 mg of codeine, whether citric acid aerosols, tion ofacetylcholine in 58 % ofthe patients; coughing ammonia inhalations or intravenous paraldehyde bouts could not be induced in 6% of the patients injections were used as stimuli, because of the great after a placebo. Prime (1961) using a 1 % acetyl- variability in the reactions of the subjects. Other choline aerosol demonstrated a significant decrease antitussives used for comparison also failed to in the number of coughs in 12 patients, 20 minutes demonstrate effectiveness, except for heroin after after an oral dose of 20 mg of codeine. Comparison paraldehyde injection. was with a placebo in random order under double- Sevelius, Lester & Colmore (1965) said that their blind conditions. subjects became adapted to the citric acid aerosol Other attempts to demonstrate an antitussive effect and consequently an antitussive effect of 30 mg of of codeine on cough induced by other procedures codeine was not demonstrated. Using an acetyl- have failed. Hillis (1952) used a peppermint: water choline aerosol to induce cough, the effect of 30 mg spray (1: 40). Codeine, 60 mg subcutaneously, or ofcodeine was no greater than that of a placebo whe- 180 mg orally, did not reduce the number of coughs ther the post-drug test was made at 60 or 120 minutes. but morphine did. Hillis & Kelly (1951) injected Other authors have regarded the ammonia and lobeline intravenously. A substantial dose of mor- CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 447 phine or codeine did not block the resulting cough 87% experienced relief while only 14% of those on but hexamethonium was effective. According to placebo registered improvement. Di Gregorio (1960) Gravenstein, Devloo & Beecher (1954) and Konar demonstrated a good antitussive effect of 30 mg of (1959) the cough following intravenous injection of codeine, given three times a day, on acute as well as paraldehyde could not be influenced by codeine. on chronic cough. Gravenstein, Devloo & Beecher (1954) and Gra- Clinical trials and experience (Table 29) venstein & Beecher (1955) devised an objective meth- od of cough counting by means of a sound recorder The first report on a clinical trial of codeine as an and compared the results of therapy obtained by antitussive which might meet today's standards was subjective assessment of coughing with the actual by Davenport (1938). He compared, double-blind, cough counts. While a difference of statistical sig- oral doses four times a day of dihydroisocodeine and nificance in the patients' judgement of antitussive codeine in tuberculous patients. Evaluation was by effectiveness of codeine and heroin against placebo subjective effect following medication and the two was found, at least in one of the two groups of drugs were judged equally effective. There was no patients, the cough counts did not show a corre- comparison with placebo. The antitussive effective- sponding difference. From this the authors suggested ness ofcodeine was accepted without question at that that the activity of antitussives, especially of codeine, time, though it had not been established for patholo- consisted essentially in a modification of psycho- gical cough in man by placebo comparison. Him- logical factors so that patients felt better and thought melsbach et al. (1940) added placebo to a double- they coughed less when changes in the cough fre- blind comparison of morphine and codeine as quency did not occur. antitussives. The cough of the majority of the Proof to the contrary was obtained subsequently patients worsened when codeine was replaced by by Cavalieri et al. (1960), Nicolis & Pasquariello placebo and improved when the switch was to mor- (1962a, 1962b) and Pasquariello & Nicolis (1962). phine. No change occurred when a 2-mg dose of They counted the number of coughs for a three-hour morphine was replaced by 10 mg of codeine. These period following the oral administration of 60 mg of authors believed 10 mg of codeine to be an optimal codeine or placebo and demonstrated a significant dose and 30 mg unnecessarily large. difference. The same patients were evaluated for Cass & Frederik (1953) compared 17 mg ofcodeine their subjective response. They preferred codeine, 4 times a day with 4 mg of dextromethorphan and 15 patients to 5, but scores based on patients' a placebo in 65 patients with chronic cough under judgement of cough intensity showed no marked double-blind conditions, and clearly established the difference between codeine and placebo. antitussive effect ofcodeine. Their patients identified Woolf & Rosenberg (1962) compared, in a double- two of the three coded placebo medications correctly blind trial, codeine in syrup, 32.4 mg five times daily, as least effective and the three coded codeine prepa- with the syrup alone and found a definite antitussive rations as most effective. The authors transformed effect of codeine by a preference-ranking scheme. In subjective assessment of cough severity into an aver- a subsequent study (Woolf & Rosenberg 1964), in age cough-intensity score for each treatment. In which cough was sound-recorded over a 24-hour another trial Cass, Frederik & Andosca (1954) period, the syrup with codeine, 32.4 mg five times a demonstrated the ability of their patients to discern day, was significantly more effective compared with correctly differences in the antitussive effectiveness a control period and to a period of administration of three different doses of dextromethorphan. The of syrup alone. Thus they showed antitussive poten- placebo medication was least effective and the cough- cy for codeine by both subjective and objective intensity score of 15 mg of codeine lay between those methods of evaluation. of 12 and 18 mg of dextromethorphan. Sevelius & Colmore (1967) also found 15 mg of Abelmann, Gaensler & Badger (1954) found codeine, three times a day, more effective than 16.2 mg of codeine three times a day more effective placebo by subjective and objective methods. In than placebo, measured by days in which cough was 12 male hospitalized patients with chronic cough, improved. Haslreiter (1959) assessed the effect of the coughing pattern was determined for two days. 30 mg of codeine three times a day against placebo If the counts agreed within 20% the experiment was in a double-blind trial on 60 patients during a test continued with alternative administration of drug of pipazetate. Of the patients receiving codeine, and placebo. The reduction in the number of coughs 448 N. B. EDDY AND OTHERS

- LO0

(Uto U Q U _r_ ) An (U 0QA Qr Loc'0 ~~~~~~.. - ~~~~~~ 0 O .c ~~~~~g,t0QC) 0) U) 1L Q CO4 0) (c ( 0 VdC -C U. U 4) ._ c(U \coC] o cA~~~~~~~~~~~~~~~C0 * (~ 0- -(U -0 \- - I >30c ..(U(U (U0( *U o (Uc n.. -((w C.(U0( (U (UO) (U .2 C: 11 'a

-c 0 ~~~A 0. (U 0. (L 0. A 0.) A 0..00L.0.L A A Ct C> >Q a) c)A >LO_ _ >, Us>Wa) CDA A Ac AA 2 0) -~ 0X0 0 0XI (L0X0 00 00.X co cIIte XA A A A AA A AII A A AA m 0

:1 U') CIO 004 w .4 4 0))'nO)(- 00 0Ra R - i 0 co co -00 (0 co0o(o0(U' U,)Fn ._.. -4 (U ( 04 co 0 0(00o0 000DO 0 LO to H O cm 000 6 0-a E x x xxx x xx xx xxx x xxx xx z a coC3COCO CO co CO COCOco lt11, CO) COCOCO

z 0) C ccc c sc c (U (Uc( (U (U (U U) -a CD o 000 0 0 0 w c z a w s E E E E *E 2 *E E E 0 EE 0 0 X,.- (U (U(U(U (U (U(U (U(U (U*c x x x (Um U o DOD 0 DIs D 000o. 000 mm 0) 1 CO4 ll 0 0 c co~~~ 0 ~~~0 U 0 coO 0 0 ,i (U w Q >° > 0° > s (U cm (U CD (U cz 'C *.-E -4 E, <, ,c > > 0 c > cmw (U IL- U ( 0 M O 0 E\ \0n 00

z 0 (U- oCQ .- s1. 0°° av) a) - : ' cr). a O : C4. cn :3 U): -LJE _ -Q a E)_ 0@(~ C CMc 00 0 0. 0 c On0 0.00 0Oos 0 010 0Oo00.*0fL - O0(> U) eDU:> O U: Z 0 a)n 0 o C a() L-OOl ~0(U (0'ut - ( 0 0 o X cXo X X X

C C Xl tt co O CO CD LtO OLO 0 X CO 0 0 0 00 (c. >, 0n clncn u o Cn I- z 0 ~~~~~~ -aco~~~~~~~~~~~a C 4 C C CO cO

COO) OCOC7OCDO0CD U)oCOOC 0CO 0 0

> ~>~ >, n U U U U (U 0COUnu~~~co co0 UE co~~~0 00CDoCDO L CLO o o 0 0 0 0 00 (UUC) O (U (U( O .c_c, c 2 c.c) c cc. c2 .2E2 2 .22

0 0 0 0 0 0 .c 0ca 0 .c0 .)c0 ~cc0 0 H O 00

N1 0 00 CO cotD N E CD CID0 CIO m CY) LO ~~~~~~C,)( cn UU CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 449

0

U) 0 0 m Q Z0, 0) ° c U) -Q U)

U)

O a)O ' 7cO CL~~~~~~~~~~~~~~LO 5 U) D w z Z .2in n -w w - U)0

a. a cu a.a 0 a. 0 A A A A a. a ca. 0AA A A 0.0a. I 0 co0() IIIA AAA 11 A A fA A A liii A 0000000) a. 0. a. 000) c/) 00 0 0 co cb L- CD \\IX uf) CD 0 00C4 0! -: 0 C co

00 0 C') 0 o r- 0 _ 0 oCD 0 0 0 0 U 0 _Cli cOD co cm x x x x x x x Xx x Z IV C. (D XL M '); 760 . a~~~~~~~~~~~~~~~~~~~~~) s:~ C r-

c ~~~~~CL 0 U O ° 0 OuC CU~~~~~~~~~~~ q~~~~~U C 2co0 S E t)4~0) °* E - 2,t q) I ~ ~~~~~~O o -co -c a) 0)° Ia. a. ..Xi u3 cw

Q U)c -Jc a C 0 00 -a) ) CD a) 0 cD °° O C c L E E E 0U m) 0 0iE a E 0 0 u D - _ C

c, D 0U nC oU 000 Ua)O Oa, *_*_ Ma}0°*-co UW 0 (Dn) U) a D D B E ME C 0c a)00) C CD, Cl) -EU° 000 a.~3- -o a~ -a~.c a. ou0 E x > E \1 I CO3 \1 0 c CD 0 CO 0 a C (e cto CO _ sosoCRf CD 2a C.f- '- LO + C9^ --:!-zz C') 0- C) CO -c CQ t co Ca c0 C-cO C) CDCD - .0 C'o C') 04CO o t 1- LO r m~~ 04 C) CXO C( oo 750 W Q.>) CUC r-D Ul) .0CD0

u 0 0 0*.C C 0~ 0 04~01C. LO In UC D 0oCO _ Cl' COC CO- 0 CD xx xxx x x x x x x x x x x 0 0 x COCO LO "W U).~ ° E C') .? C a. a

c/)C) 000 C/ cn un cn un un O cn cn u oO c C O ' ~ E cn ._ ° 1 cn II t o : Ucu U)OOc CU 0 CU 0 n H C VS Vt C _ D C') CO t Zo tQ _c- a) lr o o oo o Ua) Q CU, C_ ~ C_ C_ .C csc cs c (I) V C C S 0 I- - 000 0 0 0 (1 Cl) 0. 0 C.)CD):- co Cl o) CO U)co) N UD ItLO) COC) C4c co LO c4 E

TABLE 30 CODEINE DOSAGE REPORTED OR RECOMMENDED

Kind of patient Dosage (mg single or daily) Reference

Cough of various origin 64.5 Saundby (1879) Tuberculous 30 Schr6der (1899) Cough of various origin 25-30 Behrens (1932) Cough of various origin 16.2-32.4 Pottenger (1934) Severe cough 30-GO Fantus (1936) Tuberculous I or2x20 Davenport (1938) Acute influenza 3x10, with papaverine Moll (1953) Cough due to neurosis or habit 64.8 every 3 h for 2 days Douglass (1954) Irritative cough in postoperative 3-4x 15-20 drops of 3 % solution Prussing (1960) Children with bronchial asthma Glaser (1961) 1-3 years of age 3-5 3-5 years of age 5-6 5-10 years of age 6-9 >10 years of age 9-12 Acute and chronic cough 4 x 12-30 Anonymous (1962) Cough due to cold 16.2 Glick (1963)

(-107) paralleled the subjective judgement of im- A few investigators for various, and often un- provement (seven of 12 patients improved a lot). known reasons have failed to demonstrate codeine Placebo and 50 mg of ethyldibunate included in the superiority over placebo subjectively or objectively comparisons showed no effect by the objective meth- but the majority have found it superior in reducing od and only two of 12 claimed marked improvement cough frequency and intensity, and in giving relief with ethyldibunate or placebo. from cough as a symptom; in one-third of the trials Renovanz & Liebrich (1965) reported significant the difference has been proven statistically. The trials effectiveness of codeine in a controlled clinical trial have shown that a single dose of codeine of 30-60 mg, on 196 patients with acute cough. The trial was or a daily dosage of 45-160 mg, exerts an antitussive designed as a cross-over comparison of 40 mg of action on pathological cough in man which can be codeine, 60 mg of Silomat and placebo, each three demonstrated, taking into account certain precau- times a day. The patients compared, by question- tions. It is feasible, therefore, to use codeine as the naire, frequency and intensity of cough, effects on standard of comparison in testing other antitussives. mood and breathing, and sedation. Of the 62 pa- Codeine has been referred to in many of the tients who received both codeine and placebo, a reports on clinical experience with new antitussives statistically significant majority rated codeine more and stated ratios of effectiveness are cited in the effective on frequency (67.7 %) and intensity (64.5 %) descriptions of these compounds in Part 4. In addi- ofcough. Placebo was rated more effective in respect tion, in many reports on the use of codeine in respi- to these two parameters in 29.0 and 32.3%. The ratory disease in adults and children, various recom- differences for the other points compared were not mendations have been made in regard to dosage. A significant. few of these are noted in Table 30. CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 451

RESUME

LA CODEINE ET SES SUCCEDANES SEDATIFS DE LA DOULEUR ET DE LA TOUX: 3. LA SEDATION DE LA TOUX PAR LA CODIINE - MECANISME METHODOLOGIE ET EVALUATION

Cette troisieme partie du rapport d'ensemble consacre caine, 1'ethoforme, le benzonatate, le caramiphene, le aux propriet6s sedatives de la codeine et de ses succedanes clof6danol, la diphenhydramine, l'isoamimile, l'oxola- expose en detail les donnees experimentales relatives aux mine et la pholcodine sont capables d'agir sur les recep- voies qu'emprunte le r6flexe de la toux et au mecanisme teurs peripheriques reflexogenes. Bien que le site sp6ci- de ce moyen de protection contre les irritations nocives fique d'action des drogues sedatives de la toux soit et les lesions des voies aeriennes par des agents m6ca- probablement le ((centre de la toux *, situe dans la moelle, niques ou chimiques. Les etudes les plus r6centes de sa on note entre les composes qui agissent a ce niveau nature et de son fonctionnement supportent l'hypothese (aminoindane, codeine, dextromethorphane, hexobarbi- selon laquelle le reflexe est en grande partie ind6pendant tal, pipaz6tate) des variations portant sur la sp.cificitM de du mecanisme de la respiration. La toux est un symptome l'effet, les modalites de la reponse et l'ampleur des que l'on observe dans un tres grand nombre de maladies reactions accessoires. (inflammatoires, infectieuses, allergiques, circulatoires, Les substances qui apaisent la toux ont parfois d'autres psychiques). effets pharmacodynamiques qui accroissent ou, au On a mis au point toute une serie de methodes en vue contraire, diminuent l'efficacite de leur action sp6cifique d'evaluer chez l'animal les effets s6datifs sur la toux de sur le reflexe. On examine a cet egard la fagon dont la codeine ou d'autres drogues. On peut recourir pour l'aminoindane, le benzonatate, l'hexobarbital, la pen- provoquer la toux a des stimuli chimiques, mecaniques toxyverine, la pholcodine et le pipazetate agissent sur la ou 6lectriques. Ces divers procedes font l'objet d'une respiration. Les analg6siques - du moins ceux dont description minutieuse et leurs avantages et desavantages l'action s'apparente a celle de la morphine - attenuent respectifs sont examines. la toux, mais des composes comme le dextrom6thor- Pour juger de la valeur therapeutique d'une drogue, il phane, la noscapine et la pholcodine, qui ont un effet est particulierement utile d'apprecier son action sedative sedatif sur la toux, ne sont pas necessairement dou6s de sur la toux directement chez l'homme. A cet effet, la proprietes analgesiques. Pour mieux comprendre les rela- toux peut etre provoqu6e experimentalement au moyen tions entre l'att6nuation de la douleur et la sedation de d'inhalations d'aerosols contenant de l'acetylcholine, de la toux, il faudra, par une 6tude experimentale plus pous- l'ammoniac ou de l'acide citrique ou par l'injection de see, rechercher dans quelle mesure les sedatifs gen6raux lobeline ou de parald6hyde. Ces methodes ont rendu (hypnotiques) sont capables de combattre la toux. Les possibles de nombreuses experiences qui sont relatees en neuroleptiques comme la chlorpromazine, la perph6na- detail. Bien que la toux artificiellement induite chez un zine et la prochlorp6razine modifient le m6canisme du sujet sain differe par maints caracteres de la toux d'origine reflexe de toux. On a montr6 que l'inhalation d'hexa- pathologique, les etudes de ce genre permettent cependant methonium supprime la toux provoqu6e experimentale- de definir les analogies entre les resultats obtenus chez ment chez l'homme par l'inhalation d'ac6tylcholine, l'animal et les donnees recueillies chez l'homme. II mais des tests utilisant des sels quaternaires de substances importe dans tous les cas de leur adjoindre des recherches sedatives de la toux n'ont pas permis d'apporter la concernant l'absorption, la distribution dans l'organisme preuve definitive du r6le du blocage ganglionnaire dans et le metabolisme des drogues exp6riment6es afin de la suppression du r6flexe. L'examen des donn6es experi- reunir davantage d'informations sur les similitudes entre mentales concernant un certain nombre de substances les especes. Ces donnees fourniront la base de l'inter- douees de propri6tes spasmolytiques ou spasmogenes ne pretation des resultats de laboratoire et des applications fait ressortir aucune relation precise entre ces effets cliniques. sp6cifiques et l'action sedative sur la toux. Du point de L'evaluation des propriet6s sedatives des drogues chez vue clinique, 1'efficacite des substances actives contre la des malades presentant une toux de nature pathologique toux ne parait pas alt6ree lorsque leur emploi a pour re- due a l'une ou l'autre cause doit etre men6e en se confor- sultat d'accroitre ou de diminuer la s6cretion bronchique. mant aux exigences fondamentales de la pharmacologie L'action de la codeine sur la toux provoqu6e, tant chez clinique. Diverses methodes utilisees 'a cet effet sont 1'homme que chez l'animal, a fourni matiere a de tres decrites et discutees. nombreuses recherches. La plupart des exp6rimentateurs Le cas d'un certain nombre de substances sedatives de assignent a la codeine une efficacit6 plus ou moins la toux est examin6 sous l'angle de leur lieu d'action dans grande, et rares sont ceux qui n'ont pas constate une l'organisme: des produits douds de proprietes analge- sedation de la toux apres administration de la substance. siques locales, comme la cocaine, la t6tracalne, la lido- Pour la majorite des chercheurs qui ont procd6 is des 452 N. B. EDDY AND OTHERS essais cliniques, les effets calmants de la codeine, sur la raison de 30 a 60 mg en dose unique ou de 45 a 160 mg frequence comme sur l'intensite de la toux, sont supe- quotidiennement, combat efficacement la toux patholo- rieurs a ceux des placebos; dans un tiers des cas, la gique chez l'homme. II est des lors possible de l'utiliser difference d'action est statistiquement significative. Ces comme substance de reference pour les essais d'evaluation memes essais ont montr6 que la codeine, administree a d'autres composes sedatifs de la toux.

REFERENCES

Abelmann, W. H., Gaensler, E. A. & Badger, T. L. (1954) Cahen, R. & Boucherle, A. (1961) C. R. Soc. Biol. (Paris), Dis. Chest, 25, 532 155, 1034 Abruzzi, W. A. (1962) J. New Drugs, 2, 310 Cahen, R. & Boucherle, A. (1964) C. R. Soc. Biol. (Paris), Amdur, M. 0. (1957) Amer. industr. Hyg. Ass. Quart., 158, 1021 18, 149 Cahen, R. & Boucherle, A. (1965) Fed. Proc., 24, 549 Amler, A. B. & Rothman, C. B. (1963) J. New Drugs, 3, Calesnick, B. & Christensen, J. A. (1967) Clin. Pharmacol. 362 Ther., 8, 374 Anonymous (1962) Practitioner, 188, 249 Calesnick, B., Christensen, J. A. & Munch, J. C. (1961) Archibald, D. W., Slipp, L. B. & Shane, S. J. (1959) Amer. J. med. Sci., 242, 560 Canad. med. Ass. J., 80, 734 Capello, G. & DiPasquale, S. (1955) Schweiz. Z. Tuberk., Balea, T. & Buret, J. (1964) ThJrapie, 19, 631 12, 80 Banyai, A. L. & Johannides, M., Jr (1956) Dis. Chest, Carter, C. H. & Maley, M. C. (1957) Amer. J. med. Sci., 29, 52 233, 77 Bedell, G. N. & Seebohm, P. M. (1959) J. Amer. med. Cass, L. J. & Frederik, W. S. (1953) New Engl. J. Med., Ass., 169, 1699 249, 139 Beecher, H. K. (1955) J. Amer. med. Ass., 159, 1602 Cass. L. J. & Frederik, W. S. (1956) J. Lab. clin. Med., Behrens, H. 0. (1932) Ther. d. Gegenw., 73, 399 48, 879 Bein, H. J. & Bucher, K. (1957) Helv. physiol. pharmacol. Cass, L. J. & Frederik, W. S. (1958) Ann. intern. Med., 49, Acta, 15, 55 151 Benson, W. M., Stefko, P. L. & Randall, L. 0. (1953) Cass, L. J. & Frederik, W. S. (1964) Curr. ther. Res., 6, 14 J. Pharmacol. exp. Ther., 109, 189 Cass, L. J., Frederik, W. S. & Andosca, J. B. (1954) Amer. Bickerman, H. A. (1960a) In: Modell, W., ed., Drugs J. med. Sci., 227, 291 of choice, St. Louis, USA, Mosby Co., p. 493 Cavalieri, U., et al. (1960) J. Geront., 8, 981 Bickerman, H. A. (1960b) In: Gordon, B. L., ed., Clinical Chabrier, B., Giudicelli, R. & Thuillier, J. (1950) Ann. cardiopulmonary physiology, New York, Grune and pharm. franc., 8, 261 Stratton, p. 495 Chakravarty, N. K., et al. (1956) J. Pharmacol. exp. Ther., Bickerman, H. A. & Barach, A. L. (1954) Amer. J. med. 117, 127 Sci., 228, 156 Chen, J. Y. P., Biller, H. F. & Montgomery, E. G. (1960) Bickerman, H. A. & Itkin, S. E. (1960) Clin. Pharmacol. J. Pharmacol. exp. Ther., 128, 384 Ther., 1, 180 Dalhamn, T. (1960) Acta physiol. scand., 49, 242 Bickerman, H. A., et al. (1956) Amer. J. med. Sci., 232, Davenport, L. F. (1938) J. Pharmacol. exp. Ther., 64, 236 57 David, A., Leith-Ross, F. & Vallance, D. K. (1957) Bickerman, H. A., et al. (1957) Amer. J. med. Sci., 234, J. Pharm. Pharmacol., 9, 446 191 DeVleeschhauwer (1954) Arch. int. Pharmacodyn., 97, 34 Bidwell, S. M. R. & Boyd, E. M. (1952) J. Pharmacol. DiGregorio, M. (1960) Minerva med., 51, 4086 exp. Ther., 104, 224 Domenjoz, R. (1952) Arch. exp. Path. Pharmak., 215, 19 Bobb, J. R. & Ellis, S. (1951) J. Physiol., 167, 768 Douglass, B. E. (1954) Med. Clin. N. Amer., 38, 949 Boissier, J.-R. & Pagny, J. (1960a) Therapie, 15, 93 Eichler, 0. & Smiatek, A. (1940) Arch. exp. Path. Boissier, J.-R. & Pagny, J. (1960b) Therapie, 15, 97 Pharmak., 194, 621 Borison, H. L. (1948) Amer. J. Physiol., 154, 55 Engelhorn, R. (1960) Arzneimittel-Forsch., 10, 785 Boyd, E. M. (1954) Pharmacol. Rev., 6, 521 Engelhorn, R. & Weller, E. (1961) Pflugers Arch. ges. Boyd, E. M. & Boyd, C. E. (1960) Canad. med. Ass. J., Physiol., 273, 614 83, 1298 Engelhorn, R. & Weller, E. (1965) Pfluigers Arch. ges. Bucher, K. (1958) Pharmacol. Rev., 10, 43 Physiol., 284, 224 Bucher, K. & Jacot, C. (1951) Helv. physiol. pharmacol. Ernst, A. M. (1938) Arch. int. Pharmacodyn., 58, 363 Acta, 9, 454 Ernst, A. M. (1939) Arch. int. Pharmacodyn., 61, 73 CODEINE AND ITS ALTERNATES FOR PAIN AND COUGH RELIEF. 3 453

Fantus, B. (1936) J. Amer. med. Ass., 106, 375 Kase, Y., et al. (1959) Chem. pharm. Bull., 7, 372 Floersheim, G. L. (1959) Helv. physiol. pharmacol. Acta, Kelentey, B., et al. (1957) Arzneimittel-Forsch., 7, 594 17, 153 Kelentey, B., et al. (1958) Arzneimittel-Forsch., 8, 325 Frey, R., Hugin, W. & Mayrhofer, 0. (1955) Lehrbuch Kleibel, F. (1964) Med. Welt (Stuttg.) 24, 1317 der Anaesthesiologie, Bcrlin, Springer-Verlag Kleibel, F., Steinhoff, E. & Wilde, W. (1965) Therapie- Friebel, H. (1960) Klin. Wschr., 38, 621 woche, 15, 1115 Friebel, H. (1963) In: Muller, W. A., ed., Klinische Physio- Klein, B. (1958) Antibiot. Med., 5, 462 logie, Stuttgart, Thieme, vol. 1, p. 301 Kohli, R. P., et al. (1960) Ind. J. med. Res., 48, 193 Friebel, H. & Hahn, K.-J. (1966) Med. pharmacol. exp. Konar, N. R. & Dasgupta, S. (1959) J. Ind. med. Ass., (Basel), 14, 78 32, 189 Friebel, H. & Kuhn, H.-F. (1962) Arch. exp. Path. Konzett, H. (1955) Wien. klin. Wschr., 67, 306 Pharmak., 243, 162 Konzett, H. & Rothlin, E. (1954) Experientia, 10, 472 Friebel, H. & Kuhn, H.-F. (1964) Arch. exp. Path. Korpas, J., Tomori, Z. & Ivanca, 1. (1957) Physiol. Pharmak., 246, 527 bohemoslov., 6, 111 Friebel, H., Reichle, C. & von Graevenitz (1955) Arch. Krause, D. (1958) Arzneimittel-Forsch., 8, 553 exp. Path. Pharmak., 224, 384 Kraushaar, A. E., Schunk, R. W. & Thym, H. F. (1964) Furakawa, F. & Okabe, M. (1958) Ig. Kenkyyu, 28, 1643 Arzneimittel-Forsch., 14, 986 Glaser, J. (1961) Practitioner, 12, 715 Kroepfli, P. (1950) Helv. physiol. pharmacol. Acta, 8, 33 Glick, J. (1963) Delaware med. J., 35, 180 Krueger, H. M., Eddy, N. B. & Sumwalt, M. (1943) Publ. Gordonoff, T. (1938) Ergebn. Physiol., 40, 53 Hlth Rep. (Wash.), Suppl. 165 Gosswald, R. (1958) Arzneimittel-Forsch., 8, 550 Krug, F. (1964) Med. Welt (Stuttg.), 18, 1026 Graham, J. D. P. (1962) J. New Drugs, 2, 43 Kuhn, H.-F. & Friebel, H. (1960a) Arch. exp. Path. Granier-Doyeux, M. et al. (1959) Arch. int. Pharmacodyn. Pharmak., 240, 18 121, 287 Kuhn, H.-F. & Friebel, H. (1960b) Med. exp. (Basel), 3, Gravenstein, J. S. & Beecher, H. K. (1955) Arzneitmittel- 329 Forsch., 5, 364 LaBarre, J. & Plisnier, H. (1959) Arch. int. Pharmacodyn., Gravenstein, J. S., Devloo, R. A. & Beecher, H. K. (1954) 119, 205 J. appl. Physiol., 7, 119 Larsell, 0. & Burget, G. E. (1924) Amer. J. Physiol., 70, Green, A. F. & Ward, N. B. (1955) Brit. J. Pharmacol., 311 10, 418 Levis, S., Preat, S. & Moyersoons, F. (1955) Arch. int. Groszman, M., Birker, I. 0. & Casimir, F. (1961) App!. Pharmacodyn., 103, 200 Ther., 3, 95 Lorenz, D. (1962) Arzneimittel-Forsch., 12, 1183 Haas, H. (1955) Arch. exp. Path. Pharmak., 225, 442 Martin (1834) Arch. gen. Mid., 4, 2s, 692 Hahn, K.-J. & Friebel, H. (1966) Med. Pharmacol. exp. Martini, H. (1957) In: Nuckel, H. Aerosol-Therapie, Stutt- (Basel), 14, 87 gart, Schattauer, p. 120 Hahn, L. A. & Wilbrand, H. (1952) Arch. int. Pharmaco- Matallana, A. & Borison, H. L. (1955) Fed. Proc., 14, 367 dyn., 91, 144 Maurer, H. (1955) Deutsch. med. Wschr., 80, 351 Hara, S. & Yanaura, S. (1959) Jap. J. Pharmacol., 9, 46 May, A. J. &Widdicombe, J. G.(1954)Brit. J. Pharmacol., Haslreiter, E. (1959) Arzneimittel-Forsch., 9, 769 9, 335 Hengen, 0. and Kasparek, H. (1958) Arzneimittel-Forsch., Moll, E. L. (1953) Practitioner, 170, 76 8, 620 Morris, D. J. & Shane, S. J. (1960) Canad. med. Ass. J., Herzog, H. (1967) Therapiewoche, 17, 532 83, 1093 Hillis, B. R. (1952) Lancet, 1, 1230 Mulinos, M. G. (1960) Toxicol. appl. Pharmacol., 2, 635 Hillis, B. R. & Kelly, J. C. C. (1951) Glasgow med. J., 32, Mulinos, M. G., Nair, K. G. & Epstein, J. G. (1962) N. Y. 72 Med. J., 62, 2373 Himmelsbach, C. K., et al. (1940) Publ. Hlth Rep. (Wash.) Nicolis, F. B. & Pasquariello, G. (1962a) Clin. ter., Suppl. 158 23, 215 Hoglund, N. J. & Michaelsson, M. (1950) Acta physiol. Nicolis, F. B. & Pasquariello, G. (1962b) J. Pharmacol. scand., 21, 168 exp. Ther., 136, 183 Isbell, H. & Fraser, H. F. (1953) J. Pharmacol. exp. Ther., Ozawa, H., et al. (1962) J. pharmacol. Soc. Jap., 82, 1314 107, 524 Parish, F. A. (1959) Med. Tms (N.Y.), 87, 1488 Kallquist, I. & Melander, B. (1957) Arzneimittel-Forsch., Pasquariello, G. & Nicolis, F. B. (1962) Clin. ter., 23, 226 7, 301 Pellmont, B. & Bachtold, H. (1954) Schweiz. med. Wschr., Kase, Y. (1952) Jap. J. Pharmacol., 2, 7 84, 1368 Kase, Y. (1954) Chem. pharm. Bull., 2, 298 Perry, W. P. & Boyd, E. M. (1941) J. Pharmacol. exp. Kase, Y. (1955) Jap. J. Pharmacol., 4, 130 Ther., 73, 65 Kase, Y. & Yuizono, T. (1959) Chem. pharm. Bull., Phillips, A. M., Phillips, R. W. & Thompson, J. L. (1956) 7, 378 Ann. intern. Med., 45, 216 454 N. B. EDDY AND OTHERS

Plisnier, H. (1957) Etude comparative de l'action des deri- Silvestrini, B. & Pozzatti, C. (1961) Brit. J. Pharmacol., ves de l'opium et de diverses substances synthetiques a 16, 209 effet non morphinique sur le mecanisme de la toux, sur le Snell, E. S. & Armitage, P. (1957) Lancet, 1, 860 bronchospasme et sur la secretion bronchique, Thesis, Sollmann, T. (1957) Manual of pharmacology, 8th ed., Brussels Philadelphia, Saunders Plisnier, H. (1958) C. R. Soc. Biol. (Paris), 152, 1267 Staehelin, R. (1914) Jahreskurse drztl. Fortbild, 5, 32 Plisnier, H. (1960) C. R. Soc. Biol. (Paris), 154, 451 Stefko, P. L. and Benson, W. M. (1953) J. Pharmacol. Plisnier, H. & Hernalsteen, L. (1953) C. R. Soc. Biol. exp. Ther., 108. 217 (Paris), 153, 363 Stefko, P. L., Denzel, J. & Hickey, I. (1961) J. pharm. Pottenger, F. M. (1934) Clinical Tuberculosis, 2nd ed., St. Sci., 50, 216 Louis, Mosby Co., p. 485 Suter, C. M. (1951) Medical chemistry, New York, J. Wiley Prime, F. J. (1961) Brit. med. J., 1, 1149 & Sons, p. 405 Priissing, G. (1960) Deutsch. med. J., 11, 95 Tannenbaum, F. J. (1965) Rasch, M. (1957) Nord. Med., 57, 629 Pharmacologist, 7, 77 Reichle, C. & Friebel, H. (1955) Arch. exp. Path. Pharmak., Tedeschi, R. E. et al. (1959) J. Pharmacol. exp. Ther., 126, 226, 558 338 Renovanz, H. D. & Liebrich, K. G. (1965) Arzneimittel- Ther, L. & Lindner, E. (1958) In: Medizin und Chemie, Forsch., 15, 1393 Berlin, I. G. Farbenindustrie, vol. 6, p. 351 Rosiere, C. E. & Winder, C. V. (1955) J. Pharmacol. exp. Tiffeneau, R. (1955) Z. Aerosol-Forsch., 4, 116 Ther., 113, 46 Tiffeneau, R. (1957) Dis. Chest., 31, 404 Rosiere, C. E., Winder, C. V. & Wax, J. (1956) J. Pharma- Tomori, Z. et al. (1960) Anesthetist, 9, 169 col. exp. Ther., 116, 296 Toner, J. J. & Macko, E. (1952) J. Pharmacol. exp. Ther., Ross, B. B., Gramia, R. & Rahn, H. (1955) J. appl. 106, 246 Physiol., 8, 264 Trendelenburg, U. (1950) Acta physiol. scand., 21, 174 Salem, H. & Aviado, D. M. (1964) Amer. J. med. Sci., Van Dongen, K. (1956) Acta physiol. pharmacol. neerl., 247, 585 4, 500 Salle, J. & Brunaud, M. (1960) Arch. int. Pharmacodyn., Watt, J. (1959) Pharm. Industrie, 21, 386 126, 120 Weidemeyer, J. C., Kremer, H. W. & DeJongh, D. K. Saundby, R. (1879) Brit. med. J., 1, 545 (1960) Acta physiol. pharmacol. neerl., 9, 501 Schlez, K. (1965) Ober die Verwendung des Vagusschlingen- Widdicombe, J. G. (1954a) J. Physiol., 123, 55 hundes far die Wertbestimmung hustenstillender Sub- stanzen, Thesis, Heidelberg Widdicombe, J. G. (1954b) J. Physiol., 123, 71 Schofield, P. B. (1963) Brit. J. Clin. Pract., 17, 583 Widdicombe, J. G. (1961) Arch. exp. Path. Pharmak., 241, Schroder, G. (1899) Ther. d. Gegenew., 1, 113 415 Schroeder, W. (1951) Arch. exp. Path. Pharmak., 212,433 Widdicombe, J. G. (1964) In: Fenn, W. 0. & Rahn, H., Sell, R., Lindner, E. & Jahn, H. (1958) Arch. exp. Path. ed., Handbook of Physiology, Baltimore, Williams & Pharmak., 234, 164 Wilkins, vol. 1, sect. 3, Respiration, p. 585 Sevelius, H. D. and Colmore, J. P. (1967) Clin. Pharmacol. Winder, C. V. & Rosiere, C. E. (1955) J. Pharmacol. exp. Ther., 8, 381 Ther., 113, 55 Sevelius, H., Lester, P. D. & Colmore, J. P. (1965) Clin. Winder, C. V., et al. (1961) J. Pharmacol. exp. Ther., 133, Pharmacol. Ther., 6, 146 117 Shane, S. J., Krzyski, T. K. & Copp, S. E. (1957) Canad. Winter, C. A. & Flataker, L. (1952) Proc. Soc. exp. Biol. med. Ass. J., 77, 600 (N.Y.), 81, 463 Silson, J. S. (1965) J. New Drugs, 5, 94 Winter, C. A. & Flataker, L. (1954) J. Pharmacol. exp. Silvestrini, B. & Maffli, G. (1959) Farmaco, Ed. Sci., 14, Ther., 112, 99 440 Woolf, C. R. & Rosenberg, A. (1962) Canad. med. Ass. J., Silvestrini, B. & Pozzatti, C. (1960) Arch. int. Pharmacodyn., 86, 810 129, 249 Woolf, C. R. & Rosenberg, A. (1964) Thorax, 19, 125