The relationship between the biotype of Klebsiella

Transcription

1 J. clin. Path., 1973, 26, The relationship between the biotype of Klebsiella species an their pathogenicity R. J. FALLON From the Department oflaboratory Meicine, Ruchill Hospital, Glasgow SYNOPSIS Using the biochemical scheme of Cowan, Steel, Shaw, an Dugui (1960) 209 strains of the genus Klebsiella isolate from sputum were classifie an their relationship to isease in the patients from whom they were isolate was examine. K. aerogenes preominate an was selom of proven pathogenicity. K. ozaenae was the next most common species isolate an was of pathological significance in 38 % of cases. K. ewarsii var ewarsii, K. ewarsii var atlantae, an K. pneumoniae were less commonly isolate but were more frequently associate with severe isease than K. aerogenes an K. ozaenae. Foster an Bragg (1962) observe that certain of the biochemical types of Klebsiella as classifie by Cowan et al (1960) appeare to be associate with severe chest isease more frequently than others. Their observations were extene by Darrell an Hurle (1964) in their stuy of Klebsiella in chest infections but neither series containe many organisms classifie as Klebsiella ozaenae. This organism has been isolate from many sputa submitte to this laboratory for examination an in view of this fining it was thought to be of interest to exten the observations of these workers. A preliminary stuy was reporte briefly in 1968 (Fallon, 1968). Further points of interest were to see how the biochemical classification correlate with capsular serotype an also whether any significant biochemical variations occurre in strains isolate from the same patient over a perio of two or more years. Materials an Methos CASES The recors of cases seen at Ruchill Hospital from 1962 to 1967 either as inpatients or outpatients, from whose sputum Klebsiella sp were isolate, were reviewe in orer to efine what relationship the organism bore to the patient's clinical conition. Nearly all patients ha either suspecte or establishe respiratory infection so that, except in a very few cases, it was impossible to state that the organism Receive for publication 18 May isolate from their sputum bore no relationship to their conition, ie, the Klebsiella may have been seconary invaers even although they were not the primary bacterial pathogens. It was possible, however, to try an assess, by the persistence of an organism or its isappearance from the sputum, what relationship it bore to an acute infection where one occurre. The criteria on which the significance of the organisms isolate from sputum was assesse are set out in table I. EXAMINATION OF SPUTUM Sputum was treate with buffere pancreatin (Oxoi) by the metho of Rawlins (1953) an the centrifuge eposit from the resulting homogenize specimen was culture on 5 % horse bloo agar an MacConkey agar. In aition a bloo agar plate was inoculate for irect antibiotic sensitivity tests an a Gram film was mae. Klebsiella sp were initially recognize as non-motile, Gram-negative ros growing as moist an often mucoi non-pigmente colonies. They were catalase positive, oxiase negative an fermente glucose in the meium of Hugh an Leifson (Cowan an Steel, 1965). Organisms fulfilling these criteria were then examine using the tests shown in table H with the exception of examination for the presence of fimbriae. In aition, strains examine for capsular serotype (vie infra) were examine for the prouction of aci from salicin, xylose, aonitol, cellobiose, an inositol. They were also examine for eamination of phenylalanine an reuction of nitrate. Strains were classifie biochemically as escribe 523

2 524 Associate with Disease Isolate from lung at necropsy Present in acute but not convalescent specimen Present in acute specimen. Patient recovere after appropriate antibiotic Table I Assessment of significance of isolation of Klebsiella from sputum Not Primarily Responsible for Disease Persistently present espite recovery of patient (especially if other pathogens isappeare coincient with recovery) Appeare transiently uring illness without clinical eterioration. Only one sputum specimen obtaine but patient recovere on penicillin therapy. Not isolate at necropsy but other pathogens present No attributable illness K. aerogenes K. pneutnoniae K. ewarsii- K. ewarsii- K. rhino K. o:aenae ewarsii atlantae scleromatis Glucose (gas) Mannitol (aci) Dulcitol (aci) Lactose (aci) ( ) (-) Sucrose () (aci) Inole Gelatin liquefaction - H,S - MR - VP Urease _ Lysine ecarboxylase Arginine ecarboxylase Ornithine ecarboxvlase KCN Gluconate Malonate Fimbriae' MacConkey (growth) Citrate Table IL Diagnostic scheme of Cowan an Steel (1961) 'Not teste for by Cowan et al (1960) with the moifications introuce by Cowan an Steel (1961) (table II). BIOCHEMICAL TESTS The methos use were the same as those escribe by Cowan an Steel (1965) with the following exceptions or where relevant, alternatives. Unless state cultures were incubate at 37 C for 48 hours. Fermentation of carbohyrates Strains examine for capsular serotype or for reproucibility of fermentation pattern (see table IV) were incubate for up to three weeks at 370C before a test was accepte as negative for aci or gas prouction. H,Sprouction Organisms for this test were grown in peptone water. Voges-Proskauer reaction Barritt's metho was use. Urease prouction Christensen's urea meium was use. Decarboxylase reactions The metho of M0ller was use. Strains were not examine for fimbriae nor for /-galactosiase activity. Serological stuies The capsular antigens of a proportion of the strains stuie were kinly etermine by Dr Ia 0rskov of the State Serum Institute, Copenhagen. Results R. J. Fallont BIOTYPES OF KLEBSIELLA ISOLATED Two hunre an nine patients were inclue in this stuy. The etails of the biochemical types of Klebsiella isolate from their sputa are summarize in table III. The majority of strains were classifie either as K. aerogenes or K. ozaenae. Strains of K. ozaenae were of a fairly constant colonial morphology on bloo agar forming colourless or slightly creamy, low, convex, mucoi colonies which were quite istinct from the large cream-coloure, high-convex, colonies of K. ewarsii

3 The relationship between the biotype of Klebsiella species an their pathogenicity Type of Klebsiella Number ofpatients Total M. F. K. aerogenes K. ozaenae K. ewarsii var ewarsii K. ewarsii var atlantae K. pneumoniae Table III Klebsiella sp. examine an the meium-size moist, cream-colourecolonies of K. aerogenes. In some patients with purulent chronic bronchitis K. ozaenae persiste in the sputum for over three years. This was not seen with other Klebsiella sp. In one patient the colonial morphology of the strain of K. ozaenae isolate repeately from her sputum was atypical in that this strain consistently prouce small mucoi colonies which woul often have been misse on inspection of the bloo agar plate, on the occasions when they were present in relatively small numbers, ha not either the McConkey agar shown a mucoi smear of growth in the well of the plate, or the patient's name been note an closer re-examination of the plate been carrie out. BIOCHEMICAL STABILITY OF STRAINS OF K. ozaenae. Biochemical investigation There was a close correlation between the results Patient obtaine by the late K. Patricia Carpenter at the Dysentery Reference Laboratory, Colinale, an this laboratory using the various biochemical tests escribe by Cowan et al (1960) for the classification of strains of Klebsiella except for the prouction of gas from glucose. When teste at Colinale five strains prouce aci only an 11 aci an gas from glucose compare with 13 strains pro - ucing aci only an three aci an gas at Ruchill. It was consiere that this ifference coul have been ue to the ifferent makes of peptone use by our laboratories. Evans' peptone was use at Colinale an Oxoi peptone at Ruchill. However, when peptone water sugars from both laboratories were inoculate in parallel with a wie range of stock cultures of Enterobacteriaceae ientical results were obtaine. Hence variation between the two laboratories may have reflecte an intrinsic variability in the reactions of K. ozaenae. This possibility was examine further with strains of K. ozaenae isolate on more than one occasion from iniviual patients. It was foun that gas prouction from glucose was variable as occasionally was that from other sugars. It coul well be argue that we were examining ifferent strains of K. ozaenae from a particular patient, but, as note above, the atypical colonial appearance was note in every strain isolate an stuie from one of the patients (patient A in table IV). Table IV shows the variability of biochemical reaction obtaine with strains isolate from five patients. (The following were negative an never varie: ulcitol fermentation, inole prouction, gelatin liquefaction, hyrogen sulphie prouction A B C D E No. of strains Test result. - + Test MR VP Urease Lysine KCN' Citrate A&G A A&G A A&G A A&G A A&G A Glucose Mannitol Lactose Sucrose Table IV Biochemical reactions of K. ozaenae strains isolate from patients with the organism persistently in the sputum 'Not all strains were teste A = aci prouction G = gas prouction 525

4 526 from peptone water, ornithine ecarboxylase, malonate utilization, an gluconate oxiation.) The variabitity in results in the KCN an citrate tests may have been influence by technique but those in sugar fermentation, especially with respect to gas prouction, coul not have been influence in this way. All of the strains submitte for capsular serotyping were examine using the extene series of biochemical tests note previously. All strains prouce aci from aonitol, salicin, xylose, an cellobiose. Two of 16 strains of K. ozaenae faile to prouce aci from inositol. The lysine-positive strains from patient C woul be ifficult to ifferentiate from K. ewarsii var atlantae using the criteria of Cowan et al (1960) an even those of Cowan an Steel (1961). SEROTYPES Twenty strains of Klebsiella were submitte to Dr I. 0rskov for serotyping. The 16 strains which were biochemically type as K. ozaenae were all foun to belong to capsular serotype 4. The other four strains, three K. ewarsii an one K. ewarsii var atlantae, belonge to serotype 1. These results agree with the observations of Cowan et al (1960). RELATIONSHIP TO DISEASE The relationship between the biochemical type of Klebsiella an the isease in the patient from whom they were isolate is shown in table V. The biotype most commonly isolate, K. aerogenes, was foun in the sputum of 92 patients but in the nm,jority of these was not of primary significance although it appeare to be significant in seven cases of pneumonia. K. ozaenae was the next commonest biotype an was foun most frequently in patients with a history of chronic bronchitis. In about half of these cases it seeme to be relate to the patient's isease. In most others its presence seeme to be of R. J. Falloll little significance except in five cases of primary pneumonia where, on the criteria shown in table 1, it appeare to be of importance. K. ewarsii *ar ewarsii was isolate from 19 patients an K. ewarsii var atlantae from seven. K. ewarsii strains were more often associate with isease than the other Klebsiella sp but even so in only seven of I1 cases of primary pneumonia i they seem to be of significance. Nevertheless all three cases of primary pneumonia ue to K. ewarsii i*ar atlantae were fatal as were two ue to K. ewarcsii var ewarsii. Three of the five patients who ie were age 60 or over an the other two age 38 an 41 were moribun on amission to hospital. One patient with K. ozaenae an four with K. aerogenes, where these were juge to be significant, ie. The patient with K. ozaenae infection an two of those with K. aerogenes infection were age over 70 but K. aerogenes was also isolate at necropsy from the lung of a man age 57 an from a lung abscess in a chil of 42 years. As in the other series quote K. pneumoniae was rare but one female age 69 who ie of bronchopneumonia yiele the organism at necropsy. Discussion It is probable that the criteria as shown in table I ten to uner- rather than overestimate an organism's significance. Even so the results with respect to K. ewarsii var ewarsii agree with the observations of Foster an Bragg (1962). In the present series K. ewarsii var atlantae was also associate with severe isease but more obviously than in Foster an Bragg's series. The iscrepancy is oubtless ue to the small numbers in each series. Hence K. ewarsii was associate more frequently with severe isease than other types whereas K. aerogenes was infrequently associate with severe isease. Organismi Relationship A cute Chronic Prinmar.y Acuite Other Total to Disease Eracerbation Bronchitis an Pneunionic Bronciitis (Seconary of Chronic Pneunionia Infection, eg, Bronchitis in Tuberculosis Lung Cancer, Tracheostomy) K. aerogenes ' K. ozaenae K. ewarsii var ewarsii K. ewarsii ear atlantae K. pneunmoniae I Table V Relationship of Kiebsiella to isease

5 The relationship between the biotype of Klebsiella species an their pathogenicity Darrell an Hurle (1964) also were of the opinion that K. aerogenes was not a angerous pathogen but ha a significance in sputum similar to that of Escherichia coli. One interesting fining in the present survey is that from 1962 to 1964 only 19 strains of K. aerogenes were isolate from the 101 cases reviewe but in strains appeare in 108 cases reviewe. This woul correlate well with the increasing use of ampicillin in the treatment of chest infections because K. aerogenes is frequently foun in the sputum of patients who have been treate with this antibiotic. The prevalence of K. aerogenes in the sputum of patients previously treate with antibiotics was also note by Darrell an Hurle (1964). Apart from the observations of Burns (1968) there have been no assessments mae of the pathological significance of K. ozaenae since Cowan et al (1960) publishe their scheme of classification. Hence the present series is of interest in that full clinical ata have been available on 86 patients from whose sputum K. ozaenae has been isolate. Perhaps the most funamental observation has been that K. ozaenae has selom (both uring the perio uner review an subsequently) been isolate from patients who have not got chronic lung isease. Hence 47 of the 87 patients ha chronic bronchitis an almost all of the 22 with 'other iseases' ha either pulmonary tuberculosis or bronchial carcinoma. It woul seem that a amage respiratory tract provies a situation where K. ozaenae can colonize but, in over half these patients, not give rise to clinical isease. However, in 33 of the 86 cases, K. ozaenae was acting as a pathogen, incluing five cases of primary pneumonia where no other pathogen coul be emonstrate. The proportion of patients in whom K. ozaenae was associate with isease (38 %) was higher than the proportion (19 5 %) of patients significantly infecte with K. aerogenes but consierably less than the proportion of significant infections with K. ewarsii or K. pneumoniae. Burns (1968) foun Enterobacteria-specific precipitins in serum in relation to capsular type 4 K. ozaenae in the sputum of patients with chronic estructive isorers an suggeste a particular association of this species of Klebsiella with chronic seconary infection in these cases. It woul seem that although K. ozaenae has not the pathological significance of K. ewarsii its presence requires careful clinico-bacteriological assessment in the iniviual patient an may point to the nee for further investigation to efine the nature an extent of the pulmonary lesion in a particular patient. The classification of any group of organisms epens on the observation of reactions which are usually reproucible so that in iagnostic tables '+' 527 usually inicates that 75 % or more of strains perform that reaction an conversely '-' that 25 % or less o not perform that reaction. For this reason it was thought to be of interest to see how reproucible the reactions characteristic of K. ozaenae weie in strains repeately isolate from one patient. The results in table IV suggest that gas prouction from glucose is an unreliable characteristic as was recognize by Cowan an Steel (1961) who change this characteristic from + (Cowan et al, 1960) to. This variability of gas prouction (amongst other reactions stuie) between ifferent strains of K. ozaenae has been note also by glopek an Durlakowa (1967) but such variation between isolates from the same patient has not been recore. The three negative MR results with the strains from patient A were possibly ue to the fact that these strains grew poorly in the meium on these occasions. Growth in the presence of KCN was also an unreliable test in this laboratory as only 69 % of strains were positive. However, with the exception of growth in the presence of KCN the strains, although showing consierable variations in some characteristics still fell clearly, on most occasions, within the biochemical classification of K. ozaenae. Biotype correlate well with capsular serotype in the strains examine by Dr 0rskov. In the present series the biochemical classification of Klebsiella has been foun helpful in the assessment of the significance of the presence of a Klebsiella in sputum. At a primary iagnostic level the presence of the three colonial forms associate with the ifferent biotypes-the highly convex, large, sticky colonies of K. ewarsii, meium-size, moist, creamy colonies of K. aerogenes (which may be ifficult to ifferentiate from mucoi strains of E. coli), an the colourless, mucoi, an sometimes small colonies of K. ozaenae give the clinical bacteriologist early information which he can then correlate with the clinical state of the patient. The presence of K. ewarsii (which shoul be ifferentiate from the occasional mucoi strain of Pseuomonas aeruginosa by the oxiase test) shoul always prompt urgent enquiry into the patient's conition, whereas with K. ozaenae the situation is less urgent an with K. aerogenes the colonial similarity to some strains of E. coli prompts the same sort of caution in correlating its presence with significant clinical isease as woul be accore to E. coli. This relationship between colonial form an pathogenicity may well assume greater importance if the classification scheme propose by Bascomb, Lapage, Willcox, an Curtis (1971) gains acceptance in view of the fact that they groupe K. aerogenes, K. oxytoca, an K. ewarsii together in one taxon. It has been suggeste (Darrell an Hurle, 1964)

6 528 that K. ewarsii, K. pneumoniae, an possibly K. ozaenae shoul be reporte as 'Frieliners' bacillus'. This woul ten to overemphasize the importance of K. ozaenae but woul err on the safe sie in terms of arousing clinical suspicion as to the true nature of a patient's illness. It is, however, an unsatisfactory substitute for the irect consultation between bacteriologist an clinician which the isolation of such organisms shoul initiate. The work was starte as the result ofiscussions with the late Dr K. Patricia Carpenter who was to have written a section on the biotyping of Klebsiella sp. Her help an interest were greatly appreciate as was also the kinness of Dr I 0rskov in serotyping 20 strains of Klebsiella forware to her by Dr Carpenter. I am also grateful to Mr W. M. Brown, FIMLT, for careful technical assistance, to Mrs A. C. Ramage for secretarial assistance, an to my clinical colleagues for permission to review the recors of cases amitte uner their care. References R. J. Falloni Bascomb, S., Lapage, S. P., Willcox, W. R., an Curtis, M. A. (1971). Numerical classification of the tribe Klebsiellae. J. gen. Microbiol., Burns, M. W. (1968). 'Precipitins to Klebsiella an other enterobacteria in the serum of.patients with chronic respiratory isorers. Lancet. 1, Cowan, S. T., an Steel, K. J. (1961). Diagnostic tables for the common meical bacteria. J. Hyg. (Lon.), 59, Cowan, S. T., an Steel, K. J. (1965). Manualfor the Ientificationi of Meical Bacteria. Cambrige University Press, Cambrige. Cowan, S. T., Steel, K. J., Shaw, C., an Dugui, J. P. (1960). A classification of the Klebsiella group. J. gent. Microbiol., Darrell, J. H., an Hurle, A. D. F. (1964). Ientification an clinical significance of Klebsiella species in chest infections. J. clin. Path., 17, Fallon, R. J. (1968). Klebsiellae in chronic respiratory isease (Letter). Lancet, 1, Foster, W. D., an Bragg, J. (1962). Biochemical classification of Klebsiella correlate with the severity of the associate isease. J. clin. Path., 15, Rawlins, G. A. (1953). Liquefaction of sputum for bacteriological examination. Lancet, 2, glopek, S., an Durlakowa, 1. (1967). Stuies on the taxonomy of Klebsiella bacilli. Arch. Imnmunol. Ther. evp., 15, J Clin Pathol: first publishe as /jcp on 1 July Downloae from on 26 March 2019 by guest. Protecte by copyright.