Pediatric Pulmonology 12:90-94 (1992)

Sputum Cysteinyl-Leukotriene Levels Correlate With the Severity of Pulmonary Disease in Children With Cystic Fibrosis D.A. Spencer, MRCP,' A.P. Sampson, PhD,' C.P. Green, MRCP,~ J.F. Costello, M D , ~ P.J. Piper, pm,' and J.F. Price, M D ~ Summary. Sputum samples from 13 children with cystic fibrosis (CF) were analyzed for leukotrienes (LTs) LTB,, LTC,, LTD,, and LTE,. Distributionof LTB, appeared to be normal, and of cysteinyl-LTs log normal. Total cysteinyl-LT levels, of which on average 75% was LTE, were nearly 10 times higher than in earlier studies. Log LTE, and total cysteinyl-LT levels correlated with the overall severity of pulmonary disease assessed by Chrispin-Norman chest radiograph score (Log LTE,: r = 0.701, P = 49.1%, P = 0.008. Log total cysteinyl-LTs: r = 0.715, P = 51.1%, P = 0.006).There was no apparent relationship between LTB, levels and ChrispinNorman chest radiograph score, nor between the level of any of the LTs and age or organism cultured from sputum. These findings suggest that the cysteinyl-LTs may be involved in the pathophysiology of pulmonary disease in CF. Pediatr Pulmonoll992; 12:90-94. Key words: Reversephase high-performance liquid chromatography; radioimmunoassay; Chrispin-Norman chest radiograph score; Minitab statistical software package; leukotrienes.

INTRODUCTION

The potential importance of LTE, as a mediator of inflammation has only recently been recognized. Although on a molar basis LTE, is a less potent bronchoconstrictor than LTC, or LTD,, it has a more prolonged duration of action, and is the only member of the series known to induce a prolonged increase in bronchial responsiveness when inhaled by patients with asthma.' We have recently improved the methodology for the measurement of LTB,, LTC,, LTD, in sputum and incorporated a radioimmunoassay for LTE,." In this study we have also assessed the relationships between sputum LT levels and the overall severity of pulmonary disease assessed by the Chrispin-Norman chest radiograph score. I ' , I 2

Inflammation of the airways is an important feature which may contribute to the destructive pathological changes and clinical findings of airways obstruction, mucus hypersecretion and increased bronchial responsiveness observed in cystic fibrosis (CF). These inflammatory changes may be related to the known actions of the LTs, a group of 5-lipoxygenase products of arachidonic acid derived from a variety of resident and circulating inflammatory cells within the lung.3 The LTs can be divided into the cysteinyl-LTs (LTC,, LTD,, LTE,), and LTB, which is devoid of an amino acid residue. Both LTB, and LTC, are formed by enzymatic conversion from LTA,; LTC, is then further metabolized in a series of enzymatic steps to LTD, and LTE,. LTC, and LTD, are potent bronchoc~nstrictors,~ they stimulate secretion of mucus from human bronchus in vitro5 and cause intense vasoconstriction and plasma From the Departments of Paediatrics and Thoracic Medicine, King's leakage from postcapillary venules in vitro.6 In contrast, College Hospital, Denmark Hill' and the Department of PharmacolLTB, has no activity as a bronchoconstrictor, but is a ogy, Hunterian Institute, Royal College of Surgeons, Lincoln's Inn potent chemotactic agent and activator of inflammatory Fields,* London, U.K. cells, it potentiates edema formation and impairs mucoReceivedDecember 14, 1990; (revision) accepted for publicationSepciliary clearance. tember 26, 1991. Previous studies from this department* and other centers* have reported finding LTB,, LTC,, and LTD, in the Address correspondence and reprint requests to Dr. D.A. Spencer: sputum of adults with CF. However, LTE, levels were until April I, 1992 at the Department of Paediatrics, Ward B5. not measured in these studies, and no attempts were made Walsgrdve Hospital, Clifford Bridge Road, Coventry CV2 2DX, U.K.;subsequently at the Children's Hospital, Ladywood Middleway, to relate LT levels to the clinical severity of the disease. Birmingham 816 8ET, U.K.

',*

377

0 1992 Wiley-Liss, Inc.

Sputum LT Levels in Children With CF

91

MATERIALS AND METHODS TABLE 1-Sputum LT Levels (pg/mL Sputum) Patients Total LTD4 LTE4 cysteinyl-LTs LTC4 Sputum specimens were collected with parental con- Subject LTB4 sent from 13 outpatients (mean age, 10.6 years; range 1 62 5.6 0 24.3 29.9 2 61 5.0 0 87.2 92.2 6.7-14.3 years) with positive sweat tests who attended 3 108 16.2 7.0 49.7 72.9 CF clinics at King’s College Hospital and the Hospital for 4 186 2.4 9.8 11.7 23.9 Sick Children, Great Ormond St, London. The first 13 5 81 6.5 17.4 174.0 197.9 patients attending the clinics who could produce sputum 6 52 10.0 0 19.6 29.6 were enrolled. There was no formal age minimum, but 7 94 14.8 0 200.0 214.8 8 109 5.6 17.9 36.2 59.7 obviously the very young would be effectively excluded 9 3.2 1.7 0 25.0 26.7 by their inability to expectorate. The patients included 10 2.4 2.8 14.6 109.0 126.4 ( n = 13) were not formally compared to those attending 11 143 5.6 0 47.3 52.9 these clinics who were not included (n > 400), as we had 12 22 1.1 0 14.0 15.1 no reason to suspect a selection bias. Ten of the 13 13 88 27.0 0 1.o 28.0 children were prescribed a regular inhaled (3-agonist, but none was receiving methylxanthines or any form of stethe Pearson product-moment coefficient and linear reroid medication. gression analyses were performed using the least-squares Methods method. Data was analyzed using the “Minitab” statistiSputum (minimum 1 mL) was expectorated freely or cal software package (Minitab Inc., State College, PA). after physiotherapy into a dry cup on ice. Part of each sample was sent for bacteriological examination and the RESULTS remainder stored at -20°C for subsequent analysis. LT levels are given in Table I . The mean recovery of Extraction and Assay of LTs LTs was 24% (range, 20-28%). LTB, and cysteinyl-LTs LT assays were performed as previously described’” were detected in all samples, although LTD, was undeby an investigator who was unaware of the individual tectable in eight specimens. The mean (range) level of clinical details. Briefly, sputum samples were weighed LTB, was 77.7 (2.4-186) pmol/g sputum. The geometric and to each were added internal standards of tritium- mean (range) level of LTC, (pmol/g sputum) was 5.5 labeled LTB,, LTC,, LTD, and LTE,. Ice cold ethanol (1.1-26.9); of LTD, 2.6 (0-17.9); of LTE, 33.0 (1-200). (4 vol) was added to the samples, which were homoge- This gave a geometric mean total cysteinyl-LT level of nized and left at 4°C for 30 min to denature the protein. 53.2 (12.2-215). Distribution of LTB, appeared to be Precipitated protein and cell debris were removed by normal, and of cysteinyl-LTs log normal. There were no centrifugation at 12,OOOg for 6 min at 4°C and superna- significant correlations between the levels of LTB, and tants evaporated to dryness in vacuo. Samples were re- either individual or total cysteinyl-LTs. Clinical details are given in Table 2. The Chrispinconstituted with 5 mL distilled water and partially puriNorman score correlated with both log LTE, ( r = 0.701, fied using successive washes of distilled water, r2 = 49.l%, P = 0.008) (Fig. 1) and log total cyspetroleum ether, and chloroform through Waters C 18 ( r = 0.715, r2 = 51.1%, P = 0.006) teinyl-LT levels “Sep-PAK’ cartridges. Methanol-eluted specimens were separated by reverse-phase high-performance liquid (Fig. 2). There was no apparent relationship between LTB,, chromatography and radioimmunoassay was performed for LTB,, LTC,, LTD,, and LTE, using a double anti- LTC,, or LTD, levels and Chrispin-Norman score, or body technique. LT immunoreactivity was corrected for between the Chrispin-Norman score and age. Nor was the percentage of the respective tritium-labeled internal there a correlation between any individual LT or total standard recovered, and levels were expressed as pmoYg cysteinyl-LT levels and age. Sputum from eight children yielded a pure growth of wet weight of sputum. P . aeruginosa, one a pure growth of Staph. aureus, and Clinical Measurements two cultures were positive for both organisms. There was Chrispin-Norman scores were allocated by two one pure culture of H . influenzue and two negative culblinded independent assessors (one radiologist, one pedi- tures. atrician) on recent posteroanterior and lateral chest films.

’’

Statistical Analysis The distribution of data was assessed by plotting normal probability scores. l 3 Correlation was measured using

DISCUSSION

We have found that levels of cysteinyl-LTs and LTB, in the sputum of children with CF are much higher than

Spencer et al.

92

TABLE 2-Clinical Subject 1

2 3 4

Bacterial culture

Crispin-Norman score

7.7 13.9 11.0

P. aeruginosa P. aerugbsa Staph. aureus P. aeruginosa, Staph. aureus P. aeruginosa P. aeruginosa P. aeruginosa P. aeruginosa, Staph. aureus No growth P. aeruginosa P. aeruginosa No growth H. injluenzae

12

5

9.0

6 7 8

12.0 9.9 9.2

9 10 I1

13.6 11.3

13.0 6.7 6.I

13

30 -

Age (years)

14.3

12

Details

14

27

r = 0.71

12

26 21 19

22

0

II 27 16 8 6

I

10

I

100 1000 Total cysteinyl-leukotrienes (pmollg sputum)

Fig. 2. The relationship between Chrispin-Norman score and log total sputum cysteinyl-LT levels.

30 r

0 1 1

I

10

I

1

100

1000

LTE4 (pmollg sputum)

Fig. 1. The relationship between Chrispin-Norman score and log sputum LTE., levels.

previously reported. In addition we have observed a highly significant statistical correlation between the Chrispin-Norman chest radiograph score and both log LTE, and log total sputum cysteinyl-LT levels in children with CF. As the Chrispin-Norman score in turn correlates very well with the severity of airway obstruction and reduction in forced vital capacity assessed by spirometry in children with CF,'* this suggests that there is a relationship between sputum cysteinyl-LT levels and impaired pulmonary function in this condition. The very high total levels of cysteinyl-LTs in the sputum of this group of patients, compared to earlier studies,***can be largely explained by the inclusion of an LTE, radioimmunoassay, as LTE, accounted for 75% of the total cysteinyl-LTs. Improvements in analytic techniques will also have contributed to these findings, as the levels of LTB,, LTC,, and LTD, are all somewhat higher than previously described. The total sputum cysteinyl-LT levels in this study would be sufficient to cause biological effects of smooth

muscle contraction, mucus production and edema formation in ~ i t r o . ~ -Smooth ~ * ' ~ muscle spasm and bronchial hyperresponsiveness are common in CF; inhalation of low concentrations of either LTC,, LTD,, or LTE, by man leads to marked bronchoconstriction, which in the case of LTE, inhalation is accompanied by a 2.5-fold increase in bronchial responsiveness persisting for up to 1 week in patients with asthma.' In addition, bronchoconstriction induced by allergen challenge in asthmatics is accompanied by a significant increase in urinary LTE, excretion suggesting involvement of the cysteinyl-LTs in this process. The relatively low sputum levels of LTC, and LTD, compared to those of LTE, probably reflect their short half lives prior to sequential enzymatic conversion to LTE,. This may account for the lack of correlation between the levels of LTC, and LTD, and the chest radiograph score. The sputum levels of LTB, in this group of patients would also be sufficiently high to cause biological effects in vitro. It is perhaps surprising that these levels did not correlate with the seventy of pulmonary disease, as LTB, has a wide range of actions relevant to the pathophysiology of lung disease in CF.337*'6 These negative findings may reflect a large indirect element in the biological effects of LTB,, perhaps dependent on the secondary release of proinflammatory products from cells attracted to the region by this mediator. In addition, LTB, may undergo extensive local degradation by o-oxidation in activated neutrophils, possibly limiting its biological availability. LTB, levels did not correlate with either individual or total cysteinyl-LT levels, which could be explained by the different cellular origins of these substances. Although the number of patients in each subgroup was small, there was no apparent relationship between the nature of the organism cultured from sputum and LT

''

Sputum LT Levels in Children With CF

93

levels; LTs were also recovered from sputum from which ACKNOWLEDGMENTS no organism was recovered. This would suggest that bacD.A.S. was funded by the King's College Hospital terial infection is not a critical factor for LT production in Joint Research Trust and A.P.S. by the Children Nationthese patients. This conclusion is supported by the studies wide Medical Research Fund. We thank Dr. Dinwiddie of Zakrzewski et al. , 2 who found that antibiotic treatment and his colleagues at the Hospital for Sick Children, was not associated with decreases in eicosanoid levels in Great Ormond Street, for allowing their patients to particCF patients. ipate in this study. The mechanisms of LT generation and their precise cellular origins in CF are not known. Interactions between various cells may also be important in the process- REFERENCES ing of LT precursors in vivo. Isolated neutrophils and I . Phelan PD, Landau L1, Olinsky A. Cystic fibrosis. In: Respiratory alveolar macrophages are able to generate large amounts Illness in Children. Oxford: Blackwell Scientific Publications, of LTB,. LTC, can be generated by e~sinophils,'~ 1982: 239-293. and also by pulmonary mast cells3 which have been ob2. Zakrzewski JT, Barnes NC, Costello JF, Piper PJ. Lipid mediators in cystic fibrosis and chronic obstructive pulmonary disease. served in great numbers around the bronchial mucous Am Rev Respir Dis 1987; 136:77!&782. glands in CF (JA Nadel, personal communication). 3. Piper PJ. Formation and actions of leukotrienes. Physiol Rev One important stimulus for LT production may be viral 1984; 64:744-761. respiratory tract infections, which are associated with 4. Barnes NC, Piper PJ, Costello JF. Comparative effects of inhaled deterioration of pulmonary function in children with leukotriene C4, leukotriene D4 and histamine in normal human subjects. Thorax 1984; 39:50&504. CF. l9 High LTC, levels have been found in the nasopha5 . Coles SJ,Neil1 KH, Reid LM, et al. Effects of leukotrienes C, and ryngeal secretions of normal children with viral upper D, on glycoprotein and lysozyme secretion by human bronchial respiratory tract infections.20 Abnormal mucociliary mucosa. Prostaglandins 1983; 25: 155-170. transport and tenacious sputum in CF will delay clearance 6. Dahltn S-E, Bjork J, Hedqvist P, et al. Leukotrienes promote of mediators and inflammatory cells from the respiratory plasma leakage and leukocyte adhesion in postcapillary venules: in vivo effects with relevance to the acute inflammatory response. tract, possibly prolonging their biological effects. A sinProc Natl Acad Sci USA 1981; 78:3887-3891. gle infection might then contribute to a vicious circle of 7. Ford-Hutchinson AW, Bray MA, b i g MV, Shipley ME, Smith persistent inflammation with increased mucus producMJH. Leukotriene B,, a potent chemokinetic and aggregating tion, persistent cellular chemotaxis and activation, and substance released from polymorphonuclear leukocytes. Nature further mediator production. An alternative explanation (London) 1980; 286:264265. 8. Cromwell 0, Walport MJ, Moms HR, et al. Identification of for our findings is that increased LT formation is directly leukotrienes D and B in sputum from cystic fibrosis patients. related to the abnormal function of the cystic fibrosis Lancet 1981; ii:164-165. transmembrane conductance regulator (CFTR) responsi9. Arm JP, Spur BW, Lee TH. The effects of inhaled leukotriene E, ble for the fundamental biochemical abnormalities of celon the airway responsiveness to histamine in subjects with asthma lular chloride excretion in this condition.*' This speculaand normal subjects. J Allergy Clin lmmunol 1988; 82:654-660. tion is indirectly supported by the findings of Carlstedt- 10. Sampson AP, Spencer DA, Green CP, Piper PJ, Price JF. Leukotrienes in the sputum and urine of children with cystic fibrosis. Br Duke et a1.22 who described an abnormality in the J Clin Pharmacol 1990; 30:861-869. regulation of arachidonic acid release from lymphocyte 11. Chrispin AR, Norman AP. The systematic evaluation of the chest membranes of CF patients. radiograph in cystic fibrosis. Pediat Radiol 1974; 2:lOl-106. Oral steroids have been reported to reduce morbidity 12. Mathew DJ, Warner JO, Chrispin AR, Norman AP. The relationship between chest radiographic scores and respiratory function and improve pulmonary function in children with CF.23 tests in children with cystic fibrosis. Pediat Radiol 1977; 5:198These beneficial effects could possibly be related to inhi200. bition of phospholipase A,, with a consequent reduction 13. Ryan BF, Joiner BL, Ryan TA. Minitab Handbook, 2nd ed. in the amount of free arachidonic acid available for the Boston, MA: PWS-Kent Publishers, 1976: 177-179. synthesis of LTs and other lipid mediators.', Alterna- 14. Piper PJ, Samhoun MN. Leukotrienes. Br Med Bull 1987; 43~287-31 1. tively, steroids may act principally by inhibition of cytokine p r o d ~ c t i o n resulting ,~~ in an indirect reduction in 15. Taylor GW, Black P, Turner N, et al. Urinary leukotriene E, after antigen challenge and in allergic rhinitis. Lancet 1989; i584588. lipid mediator generation, secondary to decreased inflam16. Bray MA. The pharmacology and pathophysiology of leukotriene matory cell infiltration and activation within the lungs. B,. Br Med Bull 1983; 393249-254. Selective modification of the biological effects of LTs 17. Fels AOS, Pawlowski NA, Cramer EB, King TKC, Cohn ZA, Scott WA. Human alveolar macrophages produce leukotriene B,. using 5-lipoxygenase inhibitors and cysteinyl-LT recepRoc Natl Acad Sci USA 1982; 79:78667870. tor antagonists could represent a novel therapeutic approach in CF. Such compounds are now becoming avail- 18. Weller PF, Lee CW, Foster DW, Corey El, Austen KF, Lewis RA. Generation and metabolism of 5-lipoxygenasepathway leukoable for clinical which could help to trienes by human eosinophils: Predominant production of LTC,. determine the precise role of LTs in the pathophysiology Roc Natl Acad Sci USA 1983; 8076267630. 19. Wang EEL, Prober CG, Manson B, Corey M, Levison H.Associof lung disease in CF.

94

20. 21. 22. 23.

Spencer et al. ation of respiratory viral infections with pulmonary deterioration in patients with cystic fibrosis. N Engl J Med 1984; 311:16531658. Volovitz B, Faden H, Ogra PL. Release of leukotriene C, in respiratory tract during acute viral infection. J Paediatr 1988; 112:218-222. Riordan JR, Rommens JM, Bat-Sheva K , et al. Identification of the cystic fibrosis gene: Cloning and characterization of complementary DNA. Science 1989; 245:1066-1073. Carlstedt-DukeJ, Bronnegard M, Strandvik B. Pathological regulation of arachidonic acid release in cystic fibrosis: The putative basic defect. Proc Natl Acad Sci USA 1986; 83:9202-9206. Auerbach HS, Williams M, Kirkpatrick JA, Colten HR. Alter-

24. 25. 26.

27.

nate-day prednisolone reduces morbidity and improves pulmonary function in cystic fibrosis. Lancet 1985; ii:686-688. Morris HG. Mechanism of action and therapeutic role of corticosteroids in asthma. J Allergy Clin Immunol 1985; 75:l-13. Kelley J. Cytokines of the lung. Am Rev Respir Dis 1990; 141:765-788. Salmon JA, Jackson WP, Garland LG. Development and in vivo evaluation of 5-lipoxygenase inhibitors: Potential drugs for asthma and inflammation. In: Lewis AJ, Doherty NS, Ackerman NR, eds. Therapeutic Approaches to Inflammatory Diseases. Amsterdam: Elsevier, 1989: 137-146. Krell RD. The emergence of potent and selective peptide leukotriene receptor antagonists. Pulmon Pharmacol 1989; 2:27-31.