Household Transmission of Streptococcus pneumoniae among Siblings with Acute Otitis Media

Abstract

Nasopharyngeal transmission of Streptococcus pneumoniae was evaluated among 23 siblings with acute otitis media (AOM). Restriction fragment length polymorphism revealed that the nasopharyngeal strains were identical between siblings in 12 of 13 clusters of AOM experienced in 11 families. This study demonstrated person-to-person transmission of S. pneumoniae, especially drug-resistant strains, among siblings with AOM.

Acute otitis media (AOM) is the most common infectious disease during childhood. Approximately one-third of children experience three or more episodes in the first 3 years of life (2). Nasopharyngeal colonization with potential middle ear pathogens is an important first step in the development of AOM. Especially heavy growth of a particular pathogen in nasopharyngeal specimens may correlate well with the organism causing otitis media (6). Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis are considered normal flora of the nasopharynx in children; however, these microbes are very likely to be causative pathogens of AOM. Moreover, there has been an alarming increase in the antimicrobial resistance of S. pneumoniae that may be responsible for treatment failures. The standard method for differentiating strains of S. pneumoniae is serotyping based on capsular antigens. However, this approach is limited because of the potential for horizontal transfer of capsular genes (1). Pulsed-field gel electrophoresis (PFGE) is currently considered the most reliable and practical tool for epidemiological analysis of bacterial infections. The present study was designed to investigate household transmission of S. pneumoniae among siblings with AOM utilizing restriction fragment length polymorphism analysis by PFGE.

Twenty-three siblings in 11 families (A through K) with concurrent or closely related episodes of AOM were enrolled in this study (Table 1). Nasopharyngeal isolates were obtained with a transnasal swab (Carry Mate calcium alginate swab; Osaka Seiyaku, Osaka, Japan) inserted into the nasopharynx through the nose. Specimens were cultured on sheep blood agar (Nippon Becton Dickinson, Tokyo, Japan) at 37°C with CO₂ for 24 to 48 h. S. pneumoniae was identified by standard methods (8). Potential middle ear pathogens were identified in the nasopharyngeal culture. The growth of pathogens in culture was recorded semiquantitatively from 0 to 3+, and growth at 2+ or 3+ was considered significant. MICs of antibiotics for S. pneumoniae isolates were determined by the NCCLS-recommended broth microdilution method (10). The antibiotics employed in this study were penicillin G, ampicillin (Meiji Seika Kaisha., Ltd., Tokyo, Japan), cefotaxime (Nippon Hoechst Marion Roussel, Tokyo, Japan), clindamycin (Pharmacia Co., Ltd., Tokyo, Japan), cefaclor, erythromycin (Shionogi & Co., Ltd., Osaka, Japan), and levofloxacin (Daiichi Pharmaceutical Co., Ltd., Tokyo, Japan). The definition of the susceptibility of S. pneumoniae to penicillin G was based on the criteria established by the NCCLS (10). Capsular serotypes of S. pneumoniae were determined by the Quellung reaction with Danish typing sera (Statens Seruminstitut, Copenhagen, Denmark) (9). Genomic DNA was prepared by modification of the procedure described by Mitsuda et al. (7). The isolate was grown at 37°C for 6 h in Todd-Hewitt broth under 10% CO₂. The cells were harvested and resuspended in washing buffer (50 mM Tris-HCl [pH 7.5]). Fifty microliters of each cell suspension was used for plug preparation. An equal volume of melted 2% agarose was added to this suspension. Sample plug molds were incubated with lysis buffer (0.25 M EDTA, 1% sodium dodecyl sulfate, 10 mM Tris-HCl [pH 9.5], proteinase K at 0.5 mg/ml) overnight at 50°C and then rinsed with washing buffer. Restriction of genomic DNA was carried out with 20 U of SmaI for 20 h at 30°C or ApaI at 37°C for 20 h. Electrophoresis was performed on a GenePath PFGE apparatus (Bio-Rad Laboratories, Hercules, Calif.) with program 12 (18.5 h at 6 V/cm and 14°C in 0.5× TBE buffer [1× TBE buffer is 0.1 M Tris-HCl + 0.1 M boric acid + 2 mM EDTA {pH 8.0}]). After gel electrophoresis, the gels were stained with ethidium bromide and photographed under UV light. Isolates were assigned a major pattern designation (S1 or A1) when they showed the same PFGE pattern.

TABLE 1.

Clinical and laboratory features of S. pneumoniae recovered from siblings with acute otitis media

Family	Siblinga (strain no.)	Age at onset	Day care attendance	Date of onset	Interval between episodes (days)	PFGE typeb		PFGE type	Sero- type	MICc (μg/ml) of:
						SmaI	ApaI			PcG	ABPC	CCL	CDTR	EM	CLDM	CTX	LVFX
A	A1 (384)	2 yr 7 mo	No	10/6/95	8	S1	A1	A	14	0.06	0.06	2	0.25	4	≦0.125	0.25	1
	A2 (389)	1 yr 8 mo	No	10/14/95		S1	A1	A	14	0.06	0.06	2	0.125	4	≦0.125	0.25	1
	A1 (478)	2 yr 11 mo	No	1/29/96	2	S2	A2	B	19	1	1	>16	0.25	4	≦0.125	0.25	1
	A2 (481)	1 yr 11 mo	No	1/27/96		S2	A2	B	19	1	1	>16	0.25	4	≦0.125	0.25	1
B	B1 (221)	8 mo	No	4/5/95	21	S3	A3	C	23	1	1	>16	0.5	>16	>16	1	0.5
	B2 (226)	3 yr	No	4/26/95		S3	A3	C	23	1	1	>16	0.5	16	>16	0.25	0.5
C	C1 (049)	4 yr 8 mo	Yes	3/16/94	6	S4	A4	A	19	0.06	0.06	1	0.25	2	≦0.125	0.25	1
	C2 (056)	3 yr 4 mo	Yes	3/22/94		S5	A5	D	6	≦0.03	≦0.03	1	0.125	>16	>16	0.125	1
D	D1 (400)	4 yr 7 mo	Yes	10/27/95	1	S6	A6	E	23	0.125	0.125	1	0.125	≦0.125	≦0.125	0.06	1
	D2 (401)	2 yr 4 mo	No	10/28/95		S6	A6	E	23	0.06	0.125	1	0.06	≦0.125	≦0.125	0.06	1
E	E1 (160)	2 yr 11 mo	No	1/3/95	8	S7	A7	F	19	≦0.03	≦0.03	1	0.06	≦0.125	≦0.125	≦0.015	1
	E2 (169)	4 yr 9 mo	Yes	1/11/95		S7	A7	F	19	≦0.03	≦0.03	0.5	0.06	≦0.125	≦0.125	≦0.015	1
F	F1 (293)	1 yr	No	7/21/95	26	S8	A8	G	19	1	1	>16	0.25	4	≦0.125	0.25	1
	F2 (328)	4 yr 3 mo	Yes	8/16/95		S8	A8a	G	19	1	2	>16	0.5	4	≦0.125	0.25	1
G	G1 (374)	1 yr 2 mo	No	9/26/95	0	S9	A9	H	19	1	2	>16	0.5	4	≦0.125	0.25	1
	G2 (372)	2 yr 4 mo	No	9/26/95		S9	A9	H	19	1	1	>16	0.5	4	≦0.125	0.25	1
	G1 (490)	1 yr 6 mo	No	1/31/96	0	S10	A10	I	6	0.06	0.06	1	0.125	>16	>16	0.25	1
	G2 (495)	2 yr 8 mo	No	1/31/96		S10	A10a	I	6	0.06	0.06	1	0.125	>16	>16	0.125	1
H	H1 (099)	1 yr 3 mo	Yes	5/24/94	23	S9	A9	H	19	1	1	>16	0.25	4	≦0.125	0.25	1
	H2 (085)	3 yr 3 mo	Yes	6/16/94		S9	A9	H	19	1	1	>16	0.25	2	≦0.125	0.25	1
I	I1 (244)	3 yr 8 mo	Yes	5/25/95	0	S11	A11	J	23	1	1	>16	0.5	≦0.125	≦0.125	0.25	0.5
	I2 (243)	3 yr 8 mo	Yes	5/25/95		S11	A11	J	23	1	1	>16	0.5	≦0.125	≦0.125	0.25	0.5
	I3 (253)	3 yr 8 mo	Yes	6/1/95	7	S12	A12	K	23	≦0.03	≦0.03	0.5	≦0.03	2	≦0.125	≦0.25	1
J	J1 (130)	3 mo	No	11/13/94	0	S9	A9	H	19	1	1	>16	0.25	4	≦0.125	0.25	1
	J2 (129)	2 yr 6 mo	Yes	11/13/94		S9	A9a	H	19	1	1	>16	0.25	4	≦0.125	0.25	1
K	K1 (378)	1 yr 7 mo	No	10/1/95	24	S13	A13	L	23	1	0.5	8	0.25	>16	>16	0.25	0.5
	K2 (403)	3 yr 7 mo	Yes	10/25/95		S13	A13	L	23	0.5	0.5	16	0.25	16	>16	0.25	0.5

^aEach sibling was assigned a letter corresponding to a family and a number designation, e.g., B1.\

^bEach SmaI and ApaI PFGE pattern was assigned a letter-and-number designation (S1 to S12, A1 to A12). Isolates having patterns that differed from these patterns by one to three bands were considered closely related and were clustered into subtypes, e.g., A7a.\

^cAntibiotic abbreviations: PcG, penicillin G; ABPC, ampicillin; CCL, cefaclor; CDTR, cefditren; EM, erythromycin; CLDM, clindamycin; CTX, cefotaxime; LVFX, levofloxacin.

Thirteen clusters of AOM in siblings were observed in 11 families. Two clusters of AOM among siblings occurred in families A and G. (Table 1). PFGE analysis with SmaI revealed 13 different patterns. The strain polymorphism found by ApaI testing was similar to that found by SmaI testing (Fig. 1). Comparison of the PFGE patterns of pneumococcal isolates with 13 clusters in 11 families revealed 100% homology between the siblings in 12 clusters (92%). One common PFGE pattern (S9, A9) was seen in siblings of three different families (families G, H, and J). Although these three families lived in different districts of the same town and the children attended different day care centers (families H and J), they might have had some contacts or there might be endemic dissemination of the specific pathogen in such a small town. Four different serotypes were detected. Serotype 19 was the most prevalent, and serotypes 23 and 19 accounted for the majority of the isolates. Antimicrobial susceptibility tests showed that 10 (37%) strains were penicillin-susceptible S. pneumoniae. Seventeen strains (63%) displayed reduced susceptibility to penicillin (penicillin-intermediately resistant S. pneumoniae). Resistance to two or more antibiotics was observed among 20 strains (74%).

FIG. 1.

PFGE patterns of SmaI (top) and ApaI (bottom) digests of genomic DNA from nasopharyngeal isolates of S. pneumoniae from siblings with AOM. S. pneumoniae isolates from 11 families were analyzed by PFGE. Families A to K are the same as those shown in Table 1. Lanes M contained a molecular size marker (48.5-kb lambda DNA ladder).

Typing by PFGE, capsular serotype determination, and antibiotic susceptibility testing were equally able to demonstrate homology between pairs of organisms from siblings; however, only PFGE was able to distinguish pairs in the population of 27 organisms since there were 12 PFGE types, four capsular serotypes, and overlapping antibiotic susceptibility patterns. Our study provides evidence for the household transmission of genetically similar strains of S. pneumoniae between siblings. Hendley et al. (4) first described intrafamily spread of S. pneumoniae more than 25 years ago by simply documenting pharyngeal colonization among family members. This crude technique demonstrated the high rate of colonization in families with young children as opposed to families without children. Subsequent studies used capsular typing to demonstrate the epidemiology of colonization within individuals over time (3, 5). More recent studies have utilized DNA fingerprinting to study the spread of resistant strains in a day care setting (11). Sluijter et al. (12) compared restriction fragment end labeling, capsular typing, and penicillin binding protein genotyping for effectiveness in the characterization of S. pneumoniae colonization in children from birth to 2 years of age. They concluded that restrictions fragment end labeling was superior to capsular typing in distinguishing strains, and they observed horizontal transfer of capsular genes between strains. Chromosomal analysis by PFGE in the present study allowed the identification of one endemic multi-drug-resistant strain (S9, A9) in three unrelated families, which were otherwise indistinguishable by serotyping or antibiogram. In conclusion, the PFGE analysis clearly demonstrated person-to-person transmission of S. pneumoniae among siblings with AOM and clinicians should take account of the likelihood of a single pathogen causing AOM among siblings.