Seasonality of Cat-Scratch Disease, France, 1999–2009

Cat-scratch disease is seasonal in the United States and Japan; but no data are available from Europe. To assess the seasonality of the disease in France, we analyzed lymph node biopsy specimens collected during 1999–2009. Most (87.5%) cases occurred during September–April and peaked in December.

B artonella henselae is the causative agent of catscratch disease (CSD), the most common cause of lymphadenopathy in adults and children (1). Cats are the main reservoir of B. henselae, which is transmitted among cats by the Ctenocephalides felis fl ea (2). Bartonella organisms remain viable in fl ea feces, and transmission to humans results in inoculation of B. henselae-contaminated fl ea feces into the skin through a scratch (3). However, transmission of B. henselae from cats to humans through scratches is rare (4). In classic CSD, gradual regional lymph node enlargement is accompanied by a papule that develops in the scratch line after 3-10 days and persists from a few days to 2-3 weeks (4).
The link between seasons and CSD incidence has been described in the United States (5,6) and in Japan (7). However, because no data are available on seasonal variations of CSD in Europe, or in France, we studied lymph node biopsy specimens obtained January 1999-December 2009 from patients throughout France with suspected CSD.

The Study
Tissue specimens were sent to the National Reference Center (Marseilles, France) either frozen or in transport media. CSD was defi nitively diagnosed when a specimen was positive for 2 genes of Bartonella spp. (1). Total genomic DNA was extracted from samples by using a QIAamp tissue kit (QIAGEN, Hilden, Germany). Before 2005, PCR amplifi cation and sequencing of the internal transcribed spacer (ITS) region and pap31 gene were used for detecting B. henselae and thus confi rming CSD (1). Beginning in 2005, real-time PCR to amplify the ITS region and pap31 gene was used (8). For all assays, 2 sets of negative controls were used. DNA from B. elizabethae and B. henselae Houston-I was used as the positive control for the ITS region and the pap31 gene, respectively (1). To exclude false-positive results, we performed a second independent extraction when false-positive or unexpected results were obtained. Results were confi rmed by using PCR amplifi cation and sequencing aimed at 16S rRNA gene (8).
Epi Info version 6.0 software (Centers for Disease Control and Prevention, Atlanta, GA, USA) was used for signifi cance variations in the number of positive specimens between 2 consecutive months, nonconsecutive months, and seasons (p<0.05). The Mantel-Haenszel test or the Fisher exact test was used to test for signifi cance.
We tested 1,849 lymph node biopsy specimens and identifi ed B. henselae in 493. Positive and negative controls yielded the expected results in all tests. Positive CSD cases were plotted for each month ( Figure 1) to identify seasonal distributions of CSD from 1999 through 2009. Monthly mean numbers of CSD were lowest from May through August, followed by signifi cant increases during August-September (p = 0.002) and during November-December (p = 0.01). During December-January, the mean number of CSD cases decreased signifi cantly (p = 0.005), then plateaued from January through March (Table). Cases decreased slightly in April, then decreased signifi cantly during April-May (p = 0.002).
The odds of a CSD diagnosis based on lymph node biopsy was 9.2× higher in December, the month with the highest number of cases (92/493), than in July, the month with the lowest number of cases (12/493). The number of CSD cases was signifi cantly higher in autumn (October-December) than in summer (July-September) (p<0.0001). Fewer cases were identifi ed in winter (January-March) than autumn (p = 0.02), and cases decreased signifi cantly from winter to spring (April-June) (p = 0.0001). The number of cases did not differ signifi cantly from spring to summer (p = 0.06).

Conclusions
Our fi ndings that the number CSD cases in France varies by season are similar to fi ndings in Japan and the United States. In Japan, 64% of CSD occurred during September-December and peaked in November (7). In the United States, most CSD have occurred during the last 6 months of the year, with a peak in September (9). Moreover, the analysis of 3 US national databases indicated that most CSD cases have occurred during September-January, with peaks in November and December (5). On the other hand, 60% of admissions for CSD in children in the United States have occurred during July-October (6). The fact that the United States is a large country with diverse climates, whereas continental France has a more homogeneous climate, may explain the differences in seasonality.
The presence of Ct. felis fl eas is essential for maintaining B. henselae infection within the cat population (2). Flea infestation is more frequent in bacteremic than in nonbacteremic cats, particularly in pet cats (10). After adult cat fl eas parasitize a host cat, they feed on its blood and transmit B. henselae. Fleas go through 4 life cycle stages: egg, larva, pupa, and imago (adult) (Figure 2). Temperature and relative humidity are the 2 most essential factors for the successful reproduction, development, and survival of fl eas (11). Cats reported to have been infested with fl eas during the preceding 6 months were more likely than cats without fl eas to be seropositive (12), and the seroprevalence of B. henselae is higher in the pet cat population in warm, humid climates than in cold, dry climates because Ct. felis fl eas are more common in warmer climates (10). As a result, cats have more fl eas during the summer and autumn months than in the other 2 seasons (13).
In Nancy, France, 53% of 94 stray cats were infected with either B. henselae or B. clarridgeiae (14). In Paris, Chomel et al. reported a B. henselae seroprevalence of 36% in 64 pet cats, of which 11% were B. henselae infected (12). Gurfi eld et al. determined that 16.5% of cats tested were Bartonella infected, and 41% were seropositive for B. henselae or B. clarridgeiae (15). Risk for Bartonella infection or seropositivity was higher in cats from multicat households and in cats adopted from the pound or from the street (15).
Feline sexual activity also may infl uence the seasonality of CSD. In the Northern Hemisphere, cat reproduction increases during spring and summer, and kittens stay with their mothers until they are 12-16 weeks of age. In addition, humans are more likely to acquire kittens during the autumn months. B. henselae infection appears to be more common in young cats (10), and infection decreases with the length of cat ownership (15  encounter more fl eas during summer and autumn, and transmission of B. henselae from cat to cat is facilitated during this period (13).
In conclusion, CSD is a seasonal disease in France with increased incidence in autumn, with peaks in December, and a decrease in spring. This pattern may be explained by seasonality in cat reproductive behavior, their fl eas' activities, and the fact that during summer cats spend most time outside the house, whereas during autumn they stay indoors.
Ms Sanguinetti-Morelli is a resident at the biological laboratory of Timone Hospital in Marseille. Her research interest is zoonotic pathogens.