You are on page 1of 3


19, No 4
Copyright C 1984, American Society for Microbiology

Distribution and Relationship to Serotype of Haemophilus influenzae

Biotypes Isolated from Upper Respiratory Tracts of Children and
Adults in Papua New Guinea
Papua New Guinea Institute of Medical Research, Goroka, Eastern Highlands Province, Papua New Guinea
Received 13 October 1983/Accepted 16 January 1984

The relationship between serotypes and biotypes of 505 carriage strains of Haemophilus influenzae
isolated from the upper respiratory tracts of well children, children with pneumonia, and healthy adults was
studied. All except serotype c were significantly associated with one or two specific biotypes (P < 0.001).
No encapsulated organisms belonging to biotypes V, VI, or VII were encountered. No significant difference
in the interaction of biotypes and serotypes isolated from well and sick children was present. Both
encapsulated and nonserotypable biotype I H. influenzae strains were commonly carried in the upper
respiratory tracts of healthy Melanesian children. The distribution of nonserotypable H. influenzae strains
occurred throughout all biotypes, and the frequency of nonencapsulated biotype III and IV strains differed
significantly from serotypable organisms with the same biotype (P < 0.001).
Few data are available on either the relationship between Primary cultures were made on chocolate (horse blood)
biotypes and serotypes of Haemophilus influenzae isolated agar with and without bacitracin (300 pLg/ml) and incubated
from the upper respiratory tracts of children and adults or at 36°C in a 10% carbon dioxide-enriched atmosphere. H.
the distribution of biotypes among nonserotypable strains by influenzae isolates were identified by Gram stain, colony
source of isolation. In reports in which biochemical profiles morphology on chocolate agar, and dependence for growth
and capsular antigens have been compared, the sample size, on hemin and NAD. Encapsulated strains were serotyped by
apart from serotype b, has been too small to provide slide agglutination with H. influenzae antisera a through f
statistically meaningful results (1, 9, 15, 18, 19). Kilian (10) (Wellcome Reagents Ltd., Beckenham, England). The sub-
and Kilian and Frederiksen (11) have biotyped 296 encapsu- strates urea, ornithine, and tryptophane, in conjunction with
lated clinical isolates, but type b strains predominated and the Minitek system (BBL Microbiology Systems, Cockeys-
constituted over 80% of the total sample. ville, Md.) were used to biochemically type all strains by the
Systemic infections caused by nonencapsulated H. in- method of Kilian (10) as modified by Back and Oberhofer
fluenzae strains and serotypes other than b are common in (2).
both children and adults (3, 4, 6, 8, 13, 14, 17, 20, 21). Statistical analysis was carried out with the biomedical
Biochemical profiling developed by Kilian (10) has provided data processing software package developed at the Universi-
a useful epidemiological marker which has enabled invasive, ty of California, Los Angeles in conjunction with a PDP
nonserotypable H. influenzae strains to be both differentiat- 11/34A minicomputer (Digital Equipment Corp., Marlboro,
ed and related to coexisting carriage in the upper respiratory Mass.).
tract and elsewhere (3, 6, 13, 14, 20, 21).
The object of the present study was to clarify the relation- RESULTS
ships among H. influenzae biotypes and serotypes and the All isolates were biotypable as types I through VII, and all
distribution of biotypes among nonencapsulated carriage encapsulated strains were serotype specific. For statistical
organisms isolated from the upper respiratory tract. Because analysis, encapsulated biotype III strains were omitted be-
beta-lactamase-producing variants are uncommon in Papua cause of their infrequency. No encapsulated biotype V, VI,
New Guinea (16), no attempt could be made to relate or VII organisms were isolated. Of the serotypable strains,
penicillin resistance to antigenic or biochemical properties. 97% belonged to biotypes I, II, and IV. A comparison of
serotypes a through f with biotypes I, II, and IV (Table 1)
MATERIALS AND METHODS revealed a significant association of all serotypes except c
A total of 505 H. influenzae strains isolated from the noses with one or two specific biotypes (X2 = 181.384; 10 df;
or nasopharynxes of 410 children aged 2 weeks to 5 years P < 0.001); i.e., serotype f strains were biotype I specific,
and 95 healthy adults were studied between March 1980 and 98% of type b strains belonged to biotypes I and II, biotype
January 1983. A total of 343 isolates, including 169 serotyp- IV contained most (89%) serotype d isolates, serotype e
able organisms, were cultured from well children, and a organisms were distributed exclusively in biotypes I and IV,
further 66 isolates, of which 35 were type specific, were and 58% of serotype a strains were found in biotype II.
recovered from children with mild or moderate pneumonia. Nonserotypable H. influenzae isolates from children and
(The classification of acute lower respiratory tract infection adults were distributed throughout all biotypes, with those of
within the Pneumonia Research Programme in Goroka is I, II, and III containing more than 70% of all isolates from
based on International Classification of Diseases, 9th revi- either group. No significant age-related difference in the
sion, World Health Organization, Geneva, 1977.) All strains number of strains in any biotype was present in either group.
from adults were nonserotypable. However, the distribution of encapsulated biotypable organ-
isms differed significantly from the biotype distribution of
Corresponding author. nonserotypable organisms within biotypes III through VI (X2

TABLE 1. Relationship between biotypes and serotypes of H. types II, III. and IV. None of 17 biotype VI isolates were
inflienzae isolated from the upper respiratory tracts of Melanesian encapsulated.
children Frederiksen and Kilian (5) have reported the close associ-
No. of strains with following biotype: ation of biotype I H. influienzae strains with serotype b
Serotype No. of strains and pathogenicity. Long et al. (12) have also found
isolates I II 1II IV V VI vii that biotype I strains, regardless of encapsulation, are signif-
a 43 10 25 8 icantly related to disease. In well carriers, most H. infliueni-
b 45 17 27 1 zae strains belong to biotypes other than I, with those of
c 16 8 4 3 1 biotypes II and III being predominant (11). In tCle current
d 36 1 3 32 study, however, the carriage of nonserotypable biotype II
e 36 13 23 and III organisms in well and sick children does not differ
f 28 28 significantly from biotype I strains (X2 = 0.715, Yates
NT" 301 72 88 62 15 46 17 1 correction) and the latter, regardless of capsules, are present
" NT, Non-serotypable. in the upper respiratory tracts of 30% of well children in
Papua New Guinea.
Although the findings in this study differ in part from those
= 130.85: 5 df; P < 0.001). Types IV, VI, and VII were the published elsewhere. a number of reports of H. inflienzaie
least common biotypes among nonencapsulated strains. Bio- biotype and serotype interactions have either studied few
type VII has only recently been described (7). non-type b encapsulated strains or have failed to clearly
Table 2 shows the relationship between serotypes and define the source of clinical isolates or relate biotype distri-
biotypes isolated from well and sick children. The distribu- bution to serotypable isolates. Different regional or ethno-
tion of nonserotypable organisms in each group did not differ graphic serotype-biotype patterns of carriage and invasion
significantly. This was also true for the biotype and serotype by H. iniflutenizae may exist, particularly between developed
interaction among type-specific strains in the same groups and developing countries. Further studies are required to
(X2 = 24.74; 15 df; 0.10 > P > 0.05), although here the chi- clarify these differences.
square value borders on significance.
The assistance of the field staff of the Pneumonia Unit, Papua
This study has shown that the biotypes of encapsulated New Guinea Institute of Medical Research. in specimen collection is
strains of H. iniflienzae isolated from the upper respiratory acknowledged. We are grateful also to Damien Jolly for advice in
tracts of children in Papua New Guinea are confined to four statistical analysis and to Heather Johnson for clerical help.
of seven biotypes within which a marked biotype specificity
is apparent. The report of Kilian and Frederiksen (11) which LITERATURE CITED
describes the serotype-biotype relationship of 423 clinical 1. Albritton, W. L., S. Penner, L. Slaney, and J. Brunton. 1978.
isolates of H. infllenzae is the most substantive to date. Biochemical characteristics of Haetnophliluis infliuenzae in rela-
Although the source of their strains is not clear, several tionship to source of isolation and antibiotic resistance. J. Clin.
findings that closely parallel those in the current investiga- Microbiol. 7:519-523.
tion include the rarity of encapsulated biotype III and V 2. Back, A. E., and T. R. Oberhofer. 1978. Use of the Minitek
system for biotyping Haetinoplhilius species. J. Clin. Microbiol.
organisms and the biotype specificity of serotypes d and f. 7:312-313.
The association of type e isolates with biotype IV appears to 3. Controni, G., M. J. Chang, B. G. Gold, and W. J. Rodriguez.
be exclusive, according to Kilian and Frederiksen (11) and 1981. Haemnophlillus in fluen-ae biotype III infections in children
Kamme (9), but respiratory strains of this serotype in Papua and report of three unusual cases. Am. J. Clin. Pathol. 76:718-
New Guinea are found in both biotypes I and IV. The 720.
biotypes of smaller numbers of types d, e, and f strains 4. Eng, R. H. K., M. L. Corrado, D. Cleri, and M. F. Sierra. 1980.
published earlier by Kilian (10) are similar to more recent Non-type b Haernophilits influenZae infections in adults with
findings. Oberhofer and Back (15) have noted the specificity reference to biotype. J. Clin. Microbiol. 11:669-671.
of type f variants for biotype I. Type c H. inflienzae strains 5. Frederiksen, W., and M. Kilian. 1981. Haeemophilus-Pasteur-
ella-Ac tinobacillius: their significance in human medicine, p. 39-
are unusual elsewhere (11) but comparatively common in
55. In M. Kilian, W. Frederiksen, and E. L. Biberstein (ed.).
Papua New Guinea, although less so than other serotypes. Haemophilius Pasteuirella and Actinobacillius. Academic Press,
Of 16 serotype c strains reported in this study, 50% were Inc., London.
biotype I, and the remainder were distributed among bio- 6. Granato, P. A., E. A. Jurek, and L. B. Weiner. 1983. Biotypes of

TABLE 2. Relationship between biotypes and serotypes of H. infliuenzae isolated from well (ni = 344) and sick (n = 66) children
No. of isolates from: No. of strains from well (no. from sick) children with following biotype:
Serotype Well Sick
children children I II 111 IV V VI VII
a 38 5 10 22 (3) 6 (2)
b 35 10 11 (6) 23 (4) 1
c 11 5 4(4) 4 3 (1)
d 31 5 (1) 2(1) 29 (3)
e 30 6 11 (2) 19 (4)
f 24 4 24(4)
NT" 175 31 45 (6) 53 (5) 32 (7) 8 (1) 26 (9) 10 (3) 1
a NT, Non-serotypable.

Haemophilus influenzae: relationship to clinical source of isola- systemic infections in children: 3 years of experience. J. Clin.
tion, serotype, and antibiotic susceptibility. Am. J. Clin. Pathol. Microbiol. 15:543-546.
79:73-77. 15. Oberhofer, T. R., and A. E. Back. 1979. Biotypes of Haemophi-
7. Gratten, M. 1983. Haemophilus influenzae, biotype VII. J. Clin. lus encountered in clinical laboratories. J. Clin. Microbiol.
Microbiol. 18:1015-1016. 10:168-174.
8. Greene, G. R. 1978. Meningitis due to Haetnophilus influenzae 16. Shann, F., M. Gratten, J. Montgomery, T. Lupiwa, and H.
other than type b: case report and review. Pediatrics 64:1021- Polume. 1982. Haemophilus influenzae resistant to penicillin in
1025. Goroka. Papua New Guinea Med. J. 25:23-25.
9. Kamme, C. 1980. Biotypes of capsulated and non-capsulated 17. Shishido, H., and K. Matsumoto. 1979. Meningitis due to
Haemophilus influenzae. Acta Pathol. Microbiol. Scand. Sect. Haemophilus influenzae type e biotype 4. J. Clin. Microbiol.
B 88:261-264. 10:926-927.
10. Kilian, M. 1976. A taxonomic study of the genus Haemnophilus, 18. Tiller, F. W. 1980. Haemophilus influenzae biochemotyping. J.
with the proposal of a new species. J. Gen. Microbiol. 93:9-62. Hyg. Epidemiol. Microbiol. Immunol. 24:478-481.
11. Kilian, M., and W. Frederiksen. 1981. Ecology of Haemnophilus, 19. Tsai, W. C., and J. J. Wu. 1979. Serotypes and biotypes and
Pasteurella and Actinobacillus, p. 11-38. In M. Kilian, W. antibiotic susceptibility of Haemophilus influenzae encountered
Frederiksen, and E. L. Bibersein (ed.), Haemophilus, Pasteur- in a clinical laboratory in Taiwan. Chin. J. Microbiol. 12:140-
ella and Actinobacillus. Academic Press, Inc., London. 148.
12. Long, S. S., M. J. Teter, and P. H. Gilligan. 1983. Biotype of 20. Wallace, R. J., C. J. Baker, F. J. Quinones, D. G. Hollis, R. E.
Haemophilus influenzae: correlation with virulence and ampi- Weaver, and K. Wiss. 1983. Nontypable Haemophilus influen-
cillin resistance. J. Infect. Dis. 147:800-806. zae (biotype 4) as a neonatal, maternal, and genital pathogen.
13. Marraro, R. V., F. K. McCleskey, and J. L. Mitchell. 1977. Rev. Infect. Dis. 5:123-136.
Pneumonia due to Haemophilus influenzae (H. aegyptius) bio- 21. Wallace, R. J., D. M. Musher, E. J. Septimus, J. E. McGowan,
type 3. J. Clin. Microbiol. 6:172-173. F. J. Quinones, K. Wiss, P. H. Vance, and P. A. Trier. 1981.
14. Mason, E. O., S. L. Kaplan, L. B. Lamberth, D. B. Hinds, S. J. Haemophilus influenzae infections in adults: characterization of
Kvernland, E. M. Loiselle, and R. D. Feigin. 1982. Serotype and strains by serotypes, biotypes, and P-lactamase production. J.
ampicillin susceptibility of Haemophiluis influenzae causing Infect. Dis. 144:101-106.