The pharmacokinetic assessment of sodium cromoglycate

Flip-flop pharmacokinetics--delivering a reversal of disposition: challenges and opportunities during drug development

Flip-flop pharmacokinetics is a phenomenon often encountered with extravascularly administered drugs. Occurrence of flip-flop spans preclinical to human studies. The purpose of this article is to analyze both the pharmacokinetic interpretation errors and opportunities underlying the presence of flip-flop pharmacokinetics during drug development. Flip-flop occurs when the rate of absorption is slower than the rate of elimination. If it is not recognized, it can create difficulties in the acquisition and interpretation of pharmacokinetic parameters. When flip-flop is expected or discovered, a longer duration of sampling may be necessary in order to avoid overestimation of fraction of dose absorbed. Common culprits of flip-flop disposition are modified dosage formulations; however, formulation characteristics such as the drug chemical entities themselves or the incorporated excipients can also cause the phenomenon. Yet another contributing factor is the physiological makeup of the extravascular site of administration. In this article, these causes of flip-flop pharmacokinetics are discussed with incorporation of relevant examples and the implications for drug development outlined.

A simple pharmacokinetic method to evaluate the pulmonary dose in clinical practice—analyses of inhaled sodium cromoglycate

When the expected effect of an inhaled drug is not achieved, the cause could be poor inhalation technique and consequently a low pulmonary dose. A simple in vivo test to evaluate the pulmonary dose would be a benefit. This study evaluates the relative and systemic bioavailability following inhalation of nebulized sodium cromoglycate (SCG) in healthy subjects. Blood samples were collected during 240 min and urine was collected in two portions, up to 6 h post-inhalation. Two exposures were performed and comparisons based on the quantification of drug in plasma and urine by a high-performance liquid chromatography (HPLC) procedure were done. In one of the exposures, a pulmonary function test was performed to study if an expected effect of increased absorption could be detected. There was a good correlation between the two exposures shown in the plasma concentrations, but not in the urine analyses. The forced exhaled volume manoeuvres were associated with a higher Cmax and plasma concentrations up to 60 min post-inhalation (P<0.01). This effect was not detected in the urine analyses. We conclude that this pharmacokinetic method with inhaled SCG and plasma analyses could be used to evaluate individual inhalation technique. The HPLC method used was rapid and had adequate sensitivity.

Lung deposition in cystic fibrosis patients using an ultrasonic or a jet nebulizer

This study was conducted to compare pulmonary deposition following inhalation with an ultrasonic and a jet nebulizer in CF patients under conditions relevant to practice. The marker substance used to estimate the relative lung bioavailability was sodium cromoglycate (SCG), which is poorly absorbed from the gastrointestinal tract, but is completely absorbed from the lungs. Ten CF patients (aged 9-21 years) used an ultrasonic nebulizer (Multisonic compact 2.4 MHz) and a jet nebulizer (Parimaster, LC Plus Turbo) in a crossover design to inhale a solution containing 20 mg of SCG and a beta(2)-agonist. Urine was collected in five fractions until 12 h p. a., and the excreted SCG was determined by means of HPLC. Prior to each inhalation, the patients' pulmonary function was measured employing a Pneumoscope. Using the ultrasonic nebulizer, the amount of SCG excreted in urine was significantly greater than that after inhalation with the jet nebulizer (1.43 +/- 0.47 mg vs. 1.04 +/- 0.47 mg; p = 0.002), despite the larger residual volume in the ultrasonic nebulizer. The absorption half-life for SCG following ultrasonic nebulization was significantly shorter when compared with jet nebulization (84 +/- 14 min vs. 101 +/- 19 min; p = 0.005), being suggestive of a more peripheral deposition. Furthermore, an inverse relationship was found between absorption half-life and FEV(1) (% pred.) (r = -0.655, p = 0.04) or MMEF(75/25) (% pred.) (r = -0.844, p = 0.031), but only with the ultrasonic nebulizer. In conclusion, the ultrasonic nebulizer tested when used for inhalation in CF patients was found to be at least equivalent to the jet nebulizer.

Determination of sodium cromoglycate in human plasma by liquid chromatography-mass spectrometry in the turbo ion spray mode

A highly sensitivity liquid chromatography-tandem mass spectrometry method has been developed for the quantitation of sodium cromoglycate (SCG) in human plasma. The method was validated over a linear range of 0.100-50.0 ng/ml, using 13C4 sodium cromoglycate as the internal standard. Compounds were extracted from 1.0 ml of lithium heparin plasma by methanol elution of C18 solid-phase extraction cartridges. The dried residue was reconstituted with 100 microl of 0.01 N HCl. and 30 microl was injected onto the LC-MS-MS system. Chromatographic separation was achieved on a C8 (3.5 microm) column with an isocratic mobile phase of methanol-water-0.5 M ammonium acetate (35:64.8:0.2, v/v/v). The analytes were detected with a PE Sciex API 3000 mass spectrometer using turbo ion spray with positive ionization. Ions monitored in the multiple reaction monitoring (MRM) mode were m/z 469.2 (precursor ion) to m/z 245.1 (product ion) for SCG and m/z 473.2 (precursor ion) to m/z 247.1 (product ion) for 13C4 SCG (I.S.). The average recoveries of SCG and the I.S. from human plasma were 91 and 87%, respectively. The low limit of quantitation was 0.100 ng/ml. Results from a 4-day validation study demonstrated excellent precision (C.V.% values were between 1.9 and 6.5%) and accuracy (-5.4 to - 1.2%) across the calibration range of 0.100-50.0 ng/ml.

Plasma concentrations of disodium cromoglycate after various inhalation methods in healthy subjects

AIMS

To compare the plasma concentrations of disodium cromoglycate (DSCG) following various inhalation procedures in healthy volunteers.

METHODS

Nine healthy subjects inhaled 2 mg of aerosol, 20 mg of nebuliser solution only, 20 mg of nebuliser solution mixed with isotonic saline, or 20 mg of nebuliser solution mixed with saline and procaterol, a beta2-adenoceptor agonist, on separate occasions 2-3 weeks apart. Plasma concentrations of DSCG were determined by high-performance liquid chromatography (h.p.l.c.).

RESULTS

The peak plasma concentrations of DSCG were 1.5+/-0.7 (range 0.4-2.4) ng ml-1 in the aerosol group, 8.8+/-6.2 (range 5.3-19.9) ng ml-1 in the nebuliser solution only group, 17.2+/-16.3 (range 5.0-38.6) ng ml-1 in the nebuliser solution plus isotonic saline group, and 24.5+/-11. 9 (range 10.2-44.9) ng ml-1 in the nebuliser solution plus saline and procaterol group. Thus subjects who used the nebuliser had markedly higher plasma concentrations of DSCG than subjects who used the aerosol inhaler.

CONCLUSIONS

These findings may have important implications for the evaluation of inhalation treatment with DSCG for bronchial asthma.

Inhalers and nebulizers: which to choose and why

It is obvious that many factors should be considered when an inhaler is prescribed. Based upon the information discussed above, a rational inhaler strategy could be as follows: (1) Children < or = 5 years and elderly patients are prescribed a spacer with a valve system (and a face mask for the children) for the delivery of all drugs. When they are severely obstructed, some may need a nebulizer. If the patient cannot be taught the correct use of a spacer, a nebulizer should be prescribed. (2) Children > or = 5 years and adults are prescribed a spacer or a Turbuhaler for the administration of inhaled corticosteroids and a dry powder inhaler (preferably multiple dose) or a breath-actuated MDI for other drugs. If these alternatives are not available or the patient prefers, a conventional MDI can be used (preferably not for other corticosteroids than fluticasone propionate) provided that careful tuition is given. Fluticasone dipropionate may be given by DPI, Spacer or MDI. (3) Nebulizers are mainly reserved for severe acute attacks of bronchoconstriction. With this approach, most patients can be taught effective inhaler use with a minimum of instructional time. Finally, it must always be remembered to consider the patient's wish, since prescription of an inhaler which the physician likes but the patient does not is likely to reduce compliance.

Comparison of gamma scintigraphy and a pharmacokinetic technique for assessing pulmonary deposition of terbutaline sulphate delivered by pressurized metered dose inhaler

A comparison has been made of pulmonary deposition of terbutaline sulphate from a pressurized metered dose inhaler (pMDI), measured in 8 healthy male subjects by gamma scintigraphy and by a pharmacokinetic (charcoal-block) method, involving drug recovery in urine. Measurements were carried out with a pMDI at slow (27 l/min) and fast (151 l/min) inhaled flows and with Nebuhaler large volume spacer device (average inhaled flow 17 l/min). Overall, the two methods did not differ significantly in their estimates of whole lung deposition, although values obtained by gamma scintigraphy exceeded those from the charcoal-block method for the pMDI with fast inhalation. The regional distribution of drug within the lungs and deposition in the oropharynx could be assessed by gamma scintigraphy, but not by the charcoal-block method. It is concluded that either method may be used to assess whole lung deposition of terbutaline sulphate from pMDIs, both with and without a spacer, although each method has its own inherent advantages and disadvantages.

Pharmacokinetic optimisation of asthma treatment

Asthma is generally managed with bronchodilator therapy and/or anti-inflammatory drugs. Guidelines now advocate selection of drugs and pharmaceutical formulations (long-acting vs short-acting, inhaled vs systemic) on the basis of disease severity. Theophylline has a narrow therapeutic margin. Clearance is highly variable and plasma concentrations should be monitored to avoid the occurrence of plasma concentration-related adverse effects. The rate of absorption of theophylline differs depending on the sustained release formulation administered. Some products do not provide sufficient plasma drug concentrations for therapeutic efficacy over a 12-hour period, particularly in patients with high clearance rates (e.g. children and patients who smoke). Administration of drugs via inhalation offers several advantages over systemic routes of administration (e.g. adverse effects are decreased). Inhalation is now advocated as first-line therapy. Aerosol medications available for the treatment of asthma are beta 2-agonist (including the newer long-acting agents such as salmeterol), corticosteroids, anticholinergic drugs, sodium cromoglycate (cromolyn sodium) and nedocromil. To reach the airways, aerosolised particles should be 1 to 5 microns in diameter. Particles of this size can be produced by nebuliser for continuous administration or by metered-dose inhaler and drug powder inhaler for unit dose medication. For efficient use of the metered-dose inhaler, slow inhalation and actuation must be coordinated. However, efficacy and convenience can be improved when spacer devices are used. Furthermore, spacer devices lessen the oropharyngeal adverse effects of inhaled corticosteroids. Dry powder inhalers are more easily used by children and elderly patients than metered-dose inhalers. Regardless of the device used, a maximum of 10% of the inhaled dose reaches the airways. The rest of the dose is swallowed and absorbed through the gastrointestinal tract. Most inhaled drugs have low oral bioavailability, either because of a high first-pass metabolism (beta 2-agonists and glucocorticoids) or because of lack of absorption (sodium cromoglycate). Sulphation of beta 2-agonists occurs in the wall of the gastrointestinal tract and extensive metabolism of inhaled corticosteroids occurs in the liver. Low bioavailability of the swallowed fraction contributes to reduced adverse effects. The pharmacokinetic properties of an inhaled drug are of interest. The fraction of the dose absorbed through the lung has the same disposition characteristics as an intravenous dose, and the swallowed fraction has the same disposition as an orally administered dose. However, for many drugs, pharmacokinetic data after inhalation are limited and cannot be used as a criteria for selection of therapy.(ABSTRACT TRUNCATED AT 400 WORDS)

Inhaled histamine increases the rate of absorption of sodium cromoglycate from the lung

Since many factors may alter lung epithelial permeability (LEP) to water soluble molecules, the effect of histamine on the absorption and clearance of inhaled sodium cromoglycate was examined in seven mildly asthmatic patients with hyperresponsive airways and eight normal subjects. The subjects underwent histamine challenge to determine the provocative concentration of histamine required to reduce the forced expiratory volume in one second (FEV1) by 20% (PC20) from baseline. On two further visits they inhaled either saline placebo or histamine and 5 min later inhaled an aerosol containing sodium cromoglycate. Measurements of FEV1 were made and blood samples taken for analysis of plasma sodium cromoglycate concentration at intervals for 3 h. In the asthmatic group histamine inhalation led to a 24 +/- 4% reduction in FEV1 but had no effect on the normal subjects. When compared with inhaled saline, histamine increased the initial pulmonary absorption of SCG without influencing the total amount of drug absorbed in both asthmatics and normals. These observations suggest that the pharmacokinetics of inhaled sodium cromoglycate may be altered significantly by inflammatory mediators present at the site of drug absorption from the airways.

The effect of respiratory disorders on clinical pharmacokinetic variables

Respiratory disorders induce several pathophysiological changes involving gas exchange and acid-base balance, regional haemodynamics, and alterations of the alveolocapillary membrane. The consequences for the absorption, distribution and elimination of drugs are evaluated. Drug absorption after inhalation is not significantly impaired in patients. With drugs administered by this route, an average of 10% of the dose reaches the lungs. It is not completely clear whether changes in pulmonary endothelium in respiratory failure enhance lung absorption. The effects of changes in blood pH on plasma protein binding and volume of distribution are discussed, but relevant data are not available to explain the distribution changes observed in acutely ill patients. Lung diffusion of some antimicrobial agents is enhanced in patients with pulmonary infections. Decreased cardiac output and hepatic blood flow in patients under mechanical ventilation cause an increase in the plasma concentration of drugs with a high hepatic extraction ratio, such as lidocaine (lignocaine). On a theoretical basis, hypoxia should lead to decreased biotransformation of drugs with a low hepatic extraction ratio, but in vivo data with phenazone (antipyrine) or theophylline are conflicting. The effects of disease on the lung clearance of drugs are discussed but clinically relevant data are lacking. The pharmacokinetics of drugs in patients with asthma or chronic obstructive pulmonary disease are reviewed. Stable asthma and chronic obstructive pulmonary disease do not appear to affect the disposition of theophylline or beta 2-agonists such as salbutamol (albuterol) or terbutaline. Important variations in theophylline pharmacokinetics have been reported in critically ill patients, the causes of which are more likely to be linked to the poor condition of the patients than to a direct effect of hypoxia or hypercapnia. Little is known regarding the pharmacokinetics of cromoglycate, ipratropium, corticoids or antimicrobial agents in pulmonary disease. In patients under mechanical ventilation, the half-life of midazolam, a new benzodiazepine used as a sedative, has been found to be lengthened but the underlying mechanism is not well understood. Pulmonary absorption of pentamidine was found to be increased in patients under mechanical ventilation. Pharmacokinetic impairment does occur in patients with severe pulmonary disease but more work is needed to understand the exact mechanisms and to propose proper dosage regimens.

The influence of liposomal encapsulation on sodium cromoglycate pharmacokinetics in man

The pharmacokinetics of pulmonary-administered sodium cromoglycate (SCG) has been studied in five healthy volunteers. SCG, 20 mg, was inhaled as a solution and encapsulated in dipalmitoyl phosphatidylcholine/cholesterol (1:1) liposomes. Liposomal SCG produced detectable drug levels in plasma from four volunteers taken 24 and 25 hr after inhalation. Inhaled SCG solution, although producing peak plasma levels more than sevenfold greater than liposomal drug, was not detectable in 24-hr samples from any volunteer. The decline in plasma levels following inhalation of liposomal SCG (reflecting the absorption phase) was best described by a biexponential equation. The two absorption rate constants differed by more than an order of magnitude. The rapid absorption phase was probably due to free or surface-adsorbed SCG in the liposomal formulation, since the absorption rate constant for this phase did not differ significantly from the absorption rate constant for SCG in solution. The phase of slow drug absorption may then be attributed to absorption of drug released from vesicles. The data indicate that encapsulation of SCG prior to pulmonary administration prolonged drug retention within the lungs and altered its pharmacokinetics.

Deep inspiration increases the absorption of inhaled sodium cromoglycate

The plasma concentrations of sodium cromoglycate were measured for 4 h following a single dose of 20 mg given by inhalation to six normal volunteers. A series of forced expiratory manoeuvres was performed 2 h after the dose, which resulted in a rapid and marked increase in the plasma concentrations of the drug. A similar increase was found in three volunteers who undertook a single deep inspiration at 4 h. These data indicate that the absorption of cromoglycate from the airways can be affected by manoeuvres used to assess lung function.

Effect of methacholine induced bronchoconstriction on the pulmonary distribution and plasma pharmacokinetics of inhaled sodium cromoglycate in subjects with normal and hyperreactive airways

Inhalation treatment may be less effective in the presence of bronchoconstriction because of the reduced penetration of drugs into the airways. The effect of bronchoconstriction on the lung deposition and plasma pharmacokinetics of inhaled sodium cromoglycate was examined. Ten subjects attended the laboratory on three occasions. On the first occasion a bronchial provocation test was performed to determine the concentration of methacholine required to reduce the forced expiratory volume in one second (FEV1) by 20% (PC20). On the two subsequent occasions subjects inhaled either saline or their PC20 methacholine, followed five minutes later by an aerosol containing sodium cromoglycate and stannous phytate labelled with technetium-99m. Twenty minutes later a gamma emission lung scan was performed to determine the intrathoracic deposition of the nebulised aerosol. The central:peripheral (C:P) ratio of lung deposition was then calculated. Measurements of FEV1 were made and blood samples taken for analysis of plasma sodium cromoglycate concentration at intervals for four hours. Methacholine led to a 23.4% (SEM 0.6%) lower FEV1 and a 2.8 times higher C:P ratio than those observed after saline. There was a direct correlation between log PC20 methacholine and the increase in the C:P ratio (r = 0.81). Despite these changes with methacholine, the plasma pharmacokinetics of inhaled sodium cromoglycate were not significantly different after methacholine and after saline, except that the maximum concentration achieved (Cmax) was increased. These observations suggest that the area of cromoglycate deposition and the anatomical site are less important in determining the plasma pharmacokinetics of cromoglycate than is the total dose delivered to the lung.

Sodium cromoglycate in asthma: correlation between response and serum concentrations

The clinical response to sodium cromoglycate treatment was compared with its concentration in serum. Twenty five children with asthma entered a 10 week trial of two baseline weeks followed by eight weeks of treatment by the inhalation of 20 mg of sodium cromoglycate spincaps four times a day. Individual clinical response was determined by the differences between baseline and treatment periods of: (a) percentage of symptomless days (delta score 0); (b) diary derived daily score for four symptoms (delta DS); and (c) peak expiratory flow rate (delta PEFR). At the end of the treatment period, patients inhaled a 20 mg spincap of sodium cromoglycate and the technique of inhalation was graded. Concentrations of sodium cromoglycate in serum were measured by radioimmunoassay in samples withdrawn 5 to 120 minutes after inhalation. Delta Score 0, delta DS, and delta PEFR correlated significantly with the area under the concentration time curve. Both the area under the sodium cromoglycate concentration time curve and clinical response correlated significantly with inhalation technique score. We suggest that response of children with asthma to inhalation treatment with sodium cromoglycate is correlated to its serum concentrations.

Cromolyn sodium: a review of mechanisms and clinical use in asthma

The cellular and clinical pharmacology of cromolyn sodium are reviewed. Cromolyn sodium inhibits the release of mediators of inflammation, induced by specific antigens as well as nonspecific mechanisms, such as exercise, from mast cells. Cromolyn may also inhibit the activity of other cell types that produce inflammation. It is the only antiasthmatic that blocks both early and late asthmatic responses induced by allergen inhalation and exercise. Consequently, cromolyn therapy can block the increase in bronchial hyperreactivity induced by chronic allergen exposure. Cromolyn is effective for controlling the symptoms of mild to moderate chronic asthma in 60 to 70 percent of patients. In comparative studies, cromolyn has been as effective as theophylline for controlling symptoms of chronic asthma with fewer side effects, and may be particularly suited for asthmatic patients with learning or behavioral problems. The combination of theophylline and cromolyn is often more effective than either agent alone. Because of the convenience of administration, cromolyn is preferred over theophylline for exercise-induced asthma. Cromolyn may allow the corticosteroid dosage to be reduced in severe steroid-dependent asthmatics, but it is primarily indicated for the prophylaxis of mild to moderate disease. Adverse reactions to cromolyn are uncommon to rare. Cromolyn is now available in a Spinhaler, a metered-dose aerosol, and a nebulizer solution. If taken properly all preparations appear to be equally effective. Attention to proper inhalation technique and appropriate education of the patient is essential to ensure a good outcome with cromolyn.

The pharmacokinetics of sodium cromoglycate in man after intravenous and inhalation administration

The pharmacokinetics of sodium cromoglycate in four healthy volunteers after slow intravenous infusion have been evaluated following measurement of plasma concentrations by radioimmunoassay. The results confirm earlier findings that sodium cromoglycate is rapidly eliminated from the body and that the data can be fitted to a two compartment open model. The pharmacokinetic parameters derived from the intravenous administration were used to evaluate the pharmacokinetics after inhalation administration via the Spinhaler. A model for absorption from the lungs is described which involves absorption at two different rates; this gives a better fit to the observed data than a single absorption rate. A fast absorption rate constant with a mean value of 0.54 min-1 and a slower rate constant with a mean value of 0.0097 min-1 were found. Of a mean total of 2.84 mg absorbed from a 20 mg inhaled dose, 0.68 +/- 0.15 (s.e. mean) mg were absorbed at the fast rate and 2.17 +/- 0.37 mg at the slower rate. These rates probably reflect absorption from different sites within the lungs. The results may have important implications for interpretation of clinical findings.

Pulmonary absorption of carboxyfluorescein in the rat

The pulmonary absorption of the fluorescent marker 6-carboxyfluorescein (CF) has been characterized. CF was administered intratracheally (i.t.) as a fluid instillate to pentobarbitone-anaesthetized rats at doses of 0.5 and 2 mg kg-1. The absorption was characterized by both model-independent and model-dependent pharmacokinetic analyses of blood concentration data with reference to previous intravenous (i.v.) studies. The mean fraction available (F) of CF was 90 and 112% with a mean absorption time of 107 and 109 min for the lower and higher doses, respectively. The terminal half-life for the i.t. administered CF (73 and 83 min for the 0.5 and 2 mg kg-1 doses, respectively) was significantly longer (P less than 0.001) than after i.v. dosing (18 min). This indicates a slow pulmonary absorption of CF. Blood concentration-time profiles could not be adequately described by models involving a simple first-order absorption process; a model incorporating two simultaneous first-order inputs gave a much better description, its absorption rate constants differing by almost two orders of magnitude.

Sodium cromoglycate and atropine block the fall in FEV1 but not the cough induced by hypotonic mist

In a group of patients with mild asthma the inhalation of mist derived from ultrasonically nebulised distilled water caused an increase in cough and a fall in FEV1. Double blind administration for five minutes of sodium cromoglycate (from an original solution containing 30 mg/ml) or atropine (2 mg/ml) by inhalation from a Minineb nebuliser, 30 minutes before the mist challenge, caused a significant reduction in the fall in FEV1 (p less than 0.05), but not in cough, by comparison with the protection afforded by placebo (saline). In a second study the fall in FEV1 caused by the inhalation of distilled water was not significantly different from that seen in response to hypotonic sodium chloride (1.7 g/l, 58 mmol/l), but both produced a significantly greater fall than did a similar mist containing sodium cromoglycate at an original concentration of 10 mg/ml (58 mmol/l). The results show that both atropine and sodium cromoglycate can block the fall in FEV1 due to mist and that protection by sodium cromoglycate is immediate. These results suggest that sodium cromoglycate blocks the nervous reflexes concerned in the response to mist, probably in the afferent limb of the reflex.

Capsaicin inhalation in man and the effects of sodium cromoglycate

The inhalation of capsaicin for 1 min, delivered as an aerosol by nebulising solutions of capsaicin at concentrations of 2-65 mumol 1(-1), caused dose-dependent coughing in normal volunteers and subjects with mild asthma. Capsaicin did not cause a feeling of breathlessness, and had no effect on forced expiratory volume in 1 s (FEV1) measured at the 1st, 5th and 9th min after the challenge was completed. Coughing started within seconds of applying the face mask, continued throughout the minute of capsaicin inhalation, and stopped within seconds of the mask being removed. In any one subject the number of coughs was reproducible when repeated on the same day or after an interval of several days. Experiments using local anaesthesia applied to the buccal mucosa or larynx indicated that the cough was caused by the stimulation of capsaicin-sensitive nerve terminals situated in the larynx. Cough response was not altered by the prior inhalation of sodium cromoglycate.

Evidence for an effect of sodium cromoglycate on sensory nerves in man

Sodium cromoglycate was given by both intravenous injection and local intra-arterial infusion to healthy volunteers. Intravenous injection of a dose of 4 mg in four subjects caused a statistically significant rise in blood pressure and pulse rate associated with a feeling of warmth in the perineum and blush areas of the face and chest. Brachial artery infusion of sodium cromoglycate at doses of 100-1000 microgram/min caused a feeling of warmth in the limb during 26 out of 30 infusions and this sensation was subject to tachyphylaxis. During eight infusions in which there was a sensation of warmth there was no change in local blood flow as measured by strain-gauge plethysmography. In a further six studies involving 12 infusions of sodium cromoglycate the feeling of warmth was not accompanied by a rise in local skin temperature. The results suggest that sodium cromoglycate may stimulate afferent nerves in man.