Outdoor environmental levels of Aspergillus spp. conidia over a wide geographical area

Transcription

1 Medical Mycology June 2006, 44, Outdoor environmental levels of Aspergillus spp. conidia over a wide geographical area JESÚS GUINEA, TERESA PELÁEZ, LUIS ALCALÁ & EMILIO BOUZA Clinical Microbiology and Infectious Diseases Department, Hospital General Universitario Gregorio Marañón, Universidad Complutense, Madrid, Spain Introduction Unfortunately, little information is available on the normal air and water load of Aspergillus spores and their seasonal changes. We describe the air and water load of Aspergillus spores across the province of Madrid (Spain). We collected samples of air and water (332 samples of air and 148 of water) from selected points and taps (urban and rural environment) in summer, autumn, winter and spring. Temperature, wind speed and humidity at each point were obtained. We collected a total of 369 Aspergillus spp. isolates: 200 A. fumigatus, 94 A. niger, 40 A. flavus, 16 A. nidulans, two A. terreus, and 17 Aspergillus spp. We always found more isolates in the urban environment than in the rural environment (P/0.11). This was also true of A. fumigatus (P /0.014). The autumn collection yielded more isolates than that of the other seasons. The level of conidia in air varied from 0 to 85 c.f.u./m 3 : 6.4 c.f.u./m 3 in summer, 12 in autumn, 2.5 in winter and 1.3 in spring. Changes in atmospheric parameters influenced the levels in air. None of the water samples were positive. Counts of Aspergillus spp. spores were always under 85 c.f.u./m 3. Public water does not seem to contain Aspergillus spores. Keywords Aspergillus spp. is a ubiquitous mould whose infections, especially invasive aspergillosis (IA), are increasing in hospitalized patients, although there is a substantial variation from centre to centre [1,2]. The development of IA requires the exposure of a susceptible patient to a relevant inoculum [2], but the size of the inoculum required is unknown. The incubation period of the disease is also unknown but could range from hours to months [35] and depend on the inoculum size. Received 26 September 2005; Accepted 23 November 2005 This study was partially presented at the 13th European Congress of Clinical Microbiology and Infectious Diseases (ECCMID), Glasgow, UK, 2003 (P696); and at the 43rd Annual Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), Chicago, USA, 2003 (M-998). Correspondence: Luis Alcalá, Servicio de Microbiología, Hospital General Universitario Gregorio Marañón, Avda. Dr. Esquerdo, 46, Madrid, Spain. Tel: / ; fax: / ; E- mail: luislacala@efd.net Aspergillus, conidia levels, outdoor, environmental, water Although air seems to be the main vehicle of the conidia involved in primary infection in the lungs, other potential vehicles, e.g., water, have been described during recent years [6 11]. Unfortunately, the normal air and water load of Aspergillus spp. conidia, their seasonal changes and the levels of conidia over wide geographic areas are mostly unknown [12 14]. The aim of our study was to describe the Aspergillus spp. conidia load in outdoor air and drinking water during different seasons in the province of Madrid (Spain) in order to determine the load that may be considered as normal in the outdoor air of a wide geographic area, and the potential role of drinking water as a vehicle. We also aimed to correlate these variations in levels with meteorological parameters. Material and methods The province of Madrid is located in the centre of Spain. It is 8,028 km 2 in area and divided into 11 healthcare areas, depending on the population in 2006 ISHAM DOI: /

2 350 Guinea et al. each one. We collected air and water samples from each area. Air samples Samples from the urban (different points of the provincial capital, Madrid, and towns throughout the province) and rural environment were collected. The number of samples from each area was determined according to several criteria: (1) more urban than rural samples; (2) wide distribution of the sample points; (3) the wider areas had more sampling points; (4) the collection point in each area was near a meteorological station if possible. In order to satisfy these four criteria, we collected at least one air sample per each square of the topographic map of the province of Madrid, but we were careful not to leave wide areas without sampling. The result can be seen in Fig. 1, which shows the distribution of sampling in the province. We collected samples of air from 83 selected points across the province of Madrid (62 in an urban environment and 21 in a rural environment). Every point was tested four times: in August (summer) and November (autumn) of 2002, and in February (winter) and May (spring) of 2003 to obtain a total of 332 air samples. Fig. 1 shows the final distribution of all the sampling points. Temperature, wind speed, rainfall and humidity at each point and season were obtained from the National Institute of Meteorology. The samples were collected using the Merck Air Sampler MAS 100, which was positioned 1.5 m above the ground and oriented vertically towards the sky. The sampler drew a final air volume of 200 l per sample, with a flow rate of 100 l per minute (two minutes), and was protected from the wind in order to minimize the effect of wind speed on the efficacy of the sampler. Czapeck agar (Oxoid ) plates were used as culture medium. This was the only way to make it difficult for other non-aspergillus fungi to grow, as we demonstrated in a recent study [15]. The samples in the urban environment were collected from a wide variety of points, such as squares, streets, underground station entrances and gardens. We did this in order to minimize the effect of a confounding factor, such as the presence of a conidia source that would make our data unreliable. Water samples One-hundred-and-forty-eight samples of water (140 drinking and eight non-drinking) from 37 selected taps were also collected and filtered. They were all located in urban areas of the province as follows: 22 in private residences (bathrooms and kitchens); 10 fountains (eight drinking water and two ornamental) located in city squares; three public toilets (two in Fig. 1 Distribution of points of water and air samples across the province of Madrid.

3 Environmental Aspergillus conidia levels 351 pubs and one in a petrol station) and the last two in two different laboratories. The samples were collected from all over the province in order to obtain as wide a distribution as possible. Every tap was tested four times during the study period, at the same times as the air samples. The samples were taken after letting the water run for five minutes to clear stagnant water from the pipe. Each sample contained 500 ml of water and was transported in sterile vacuums. These were kept at 48C until processing. Each sample was filtered using Millipore filters (Microfil TM V, 0.45 mm pore diameter). The filters were removed from the filtration structure, and placed over the surface of Sabouraud-Chloramphenicol dextrose agar (biomérieux ) plates. A sterile filter was put over the surface of the plate and used as a negative control. A suspension of conidia of Aspergillus niger was performed and filtered and used as a positive control. Culture processing The Czapeck agar plates were incubated for 5 days. We reported the results of each sampled place as c.f.u of total fungi, Aspergillus spp. per m 3 of air and the distribution by species. The Sabouraud-Chloramphenicol agar plates were incubated at 358C for7daysand checked daily to observe the growth of fungi. The results were reported as c.f.u. per ml of water. We isolated and sub-cultured every Aspergillus strain that was identified by conventional methods [16], and the only remaining filamentous fungi c.f.u. were counted. The Aspergillus strains were stored at /708. Statistical analysis The results at each sample point were recorded on an individual form and transferred to a database produced using Microsoft ACCESS. The conidia levels at each point were expressed as c.f.u./m 3. The statistical analysis was performed using the SPSS 11.5 software package. We studied the normality of the variable using the Kolmogorov-Smirnov test. Due to the asymmetric data Table 1 distribution, we used median and range as descriptive statistics to express the air levels of conidia; the interquartile range displays the dispersion of the values. On several occasions, the median level of conidia by species was 0 in the four collections because it was the most repeated value in the samples, despite the fact that the range had changed. We used the non-parametric Mann-Whitney test to establish differences between both environments (rural and urban). To determine whether atmospheric parameters influenced the levels of conidia in air, we used the mean instead of the median, because we grouped all the cases of the four collections (332) and the distribution was normal. We looked for a correlation between these parameters and the levels of conidia in air, using the Pearson correlation coefficient. A predictor equation of the conidia levels, depending on the atmospheric parameters, was designed from these data. The parameters analyzed were Daily Maximum Temperature, Daily Minimum Temperature, Monthly Mean Maximum Temperature, Monthly Mean Minimum Temperature, Monthly Mean Temperature, Maximum Wind Speed, Monthly Wind Speed, Monthly Precipitation, Daily Precipitation (day of sampling), Number of Days with Rainfall (month of sampling), Relative Humidity. We built a predictive model with a step-wise method of variable selection. The dependent variable chosen was Log 10 conidia per m 3 of air to achieve better results, and different combinations with the atmospheric parameters were examined using an ANOVA. Results Air samples Seasonal distribution of the isolates of Aspergillus collected from air samples. Table 1 shows the total number of isolates from each collection and Fig. 2 shows the percentage of the isolates of each species. Of the total strains isolated, most were A. fumigatus (54%), followed by A. niger (25%) and A. flavus (11%); some species of purportedly clinical relevance, such as A. terreus and others, were Collection A. fumigatus A. niger A. flavus A. nidulans A. terreus Aspergillus spp. Total Isolates per collection Summer Autumn Winter Spring Total Isolates

4 352 Guinea et al. Fig. 2 Distribution of the isolates of Aspergillus in air samples in the province of Madrid. scarce (10%). A. fumigatus was the most frequently isolated species in the four seasonal collections, with the highest number of isolates in autumn. A. niger was the second most common species. The autumn collection of samples yielded more isolates than the others: 106 in summer, 200 in autumn, 41 in winter and 22 in spring. We summarize the median conidia levels of each Aspergillus species in each seasonal collection with its range of values in Table 2. We observed that the median load of total filamentous fungi was higher in the autumn (105), followed by spring (75), summer (55) and winter (35). The total Aspergillus load was also higher in autumn. The maximum peak of total Aspergillus conidia was 85 (in autumn) and the minimum was 0 (all four seasons). The level of A. fumigatus conidia in air varied from 070 c.f.u./m 3, with the highest level in autumn. We observed that the mean level of Aspergillus spp. conidia in air varied with increased statistical significance (P B/ 0.050): total Aspergillus c.f.u./m 3 from winter spring ( ) to summer (6.4) and from summer to autumn (12). Table 3 shows the distribution of the results of conidia levels in urban and rural air for the four collections, with the median of these levels and the range of maximum and minimum level. We always found more isolates of filamentous fungi in the urban environment than in the rural environment, but this difference was only statistically significant in the summer collection (P / 0.001). In the case of total Aspergillus conidia, we found more Aspergillus isolates in the urban environment in all collections but winter; these differences were statistically significant for Table 2 Median, range and interquartile range of conidia levels of Aspergillus species in air (c.f.u./m 3 ) in each collection. Interquartile range (Spring) range (Spring) Interquartile range (Winter) range (Winter) Interquartile range (Autumn) range (Autumn) Interquartile range (Summer) Kind of fungus range (Summer) Total filamentous fungi 55 (445-5) (31015) (3150) (6855) 120 Total Aspergillus 0 (450) 10 5 (85 0) (30 0) 5 0 (300) 0 Aspergillus niger 0 (200) 5 0 (20 0) 5 0 (15 0) 0 0 (5 0) 0 Aspergillus fumigatus 0 (250) 0 0 (70 0) 10 0 (30 0) 0 0 (300) 0 Aspergillus flavus 0 (350) 0 0 (20 0) 0 0 (0) 0 0 (150) 0 Aspergillus terreus 0 (0) 0 0 (50) 0 0 (0) 0 0 (0) 0 Aspergillus nidulans 0 (100) 0 0 (10 0) 0 0 (50) 0 0 (0) 0 Aspergillus spp. 0 (150) 0 0 (50 0) 0 0 (0) 0 0 (0) 0

5 Environmental Aspergillus conidia levels 353 Table 3 range of conidia levels of Aspergillus species in the air of urban and rural environments (c.f.u./m 3 ) in each collection. range Rural (Spring) range Urban (Spring) range Rural (Winter) range Urban (Winter) range Rural (Autumn) range Urban Environment (Autumn) range Rural (Summer) Kind of fungus range Urban (Summer) Total filamentous fungi 60 (10 365) 30 (5445) 115 (15310) 85 (40250) 35 (0 315) ) 75 (5 685) 70 (15 420) Total Aspergillus 0 (1545) 0 (015) 5 (085) 0 (0 35) 0 (0 30) 0 (010) 0 (0 30) 0 (010) Aspergillus niger 0 (1020) 0 (010) 0 (020) 0 (0 20) 0 (0 15) 0 (05) 0 (0 5) 0 (05) Aspergillus fumigatus 0(0 25) 0 (0) 5 (070) 0 (0 25) 0 (0 30) 0 (05) 0 (0 30) 0 (05) Aspergillus flavus 0(035) 0 (05) 0 (020) 0 (0 5) 0 (0) 0 (0) 0 (0 15) 0 (05) Aspergillus terreus 0 (0) 0 (0) 0 (05) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) Aspergillus nidulans 0(010) 0 (0) 0 (010) 0 (0 5) 0 (0 5) 0 (0) 0 (0) 0 (0) Aspergillus spp. 0 (0 15) 0 (0) 0 (050) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) the summer and autumn collections (P/0.011 and P/0.030). This was also true of A. fumigatus (P /0.014), but the remaining species varied without statistical significance in each collection. Table 4 shows the results throughout the year. We can see that throughout the year, there were statistically significant differences only with total Aspergillus conidia and A. fumigatus: their levels varied from winter spring summer ( ) to autumn (7.7). The remaining species varied but without statistical significance: A. niger levels varied from winter spring ( ) to summerautumn ( ). Table 5 shows the correlation between the conidia levels and the meteorological parameters. Only those cases where the correlation was found (P B/ 0.050) are contained in the table showing the Spearman coefficient. We observed that the levels of Aspergillus conidia increase when temperature increases and decrease when wind speed is high and when it rains on the collection day. The analysis by species shows that A. fumigatus is not heavily influenced by these parameters and that A. niger and A. flavus are more dependent on them. In general terms, we could say that when temperature increases, the conidia load also increases, but when the wind speed of the collection day is high, the conidia are removed from the atmosphere and the load decreases. The meteorological factors that best predicted the levels in each season were: maximum mean temperature of the month (TMMax), minimum mean temperature of the month (TMMin) and monthly wind speed (MWS). The equation changed depending on the season of the year: Summer predictions: *TMMax0.040*TMMin0.016*MWS Winter predictions: *TMMax0.040*TMMin0.016*MWS Autumn predictions: *TMMax0.040*TMMin0.016*MWS Spring predictions: T*MMax0.040*TMMin0.016*MWS The model presented an R square of Water samples None of the drinking water samples were positive for Aspergillus spp. in any season collection. Only one sample from an ornamental fountain yielded a colony of A. niger but it probably came from the air due to contact between the water in the fountain and the air. At 10 points, dematiaceous fungi were isolated in at least one of the four samples (four private dwellings, two ornamental fountains, three drinking water

6 354 Guinea et al. Table 4 range of conidia levels of Aspergillus species in the air of urban and rural environments (c.f.u./ m 3 ) throughout the year. Kind of fungus range Urban range Rural Statistical significance Total filamentous fungi 80 ( ) 62.5 ( ) No (P/0.319) Total Aspergillus 5(031.25) 1.25 (010) Yes (P/0.003) Aspergillus niger 1.25 (06.25) 1.25 (06.25) No (P/0.498) Aspergillus fumigatus 2.5 (022.5) 0 (06.25) Yes (P/0.009) Aspergillus flavus 0(08.75) 0 (01.25) No (P/0.238) Aspergillus terreus 0(01.25) 0 (0) No (P/0.400) Aspergillus nidulans 0(02.5) 0 (01.25) No (P/0.323) Aspergillus spp. 0 (012.5) 0 (0) No (P/0.135) fountains, one laboratory), and mucorals were also present in these cases in the ornamental fountains. The remaining 27 points were always negative for fungi in the four collections. Water does not seem to be a common vehicle of Aspergillus spp. conidia. Discussion The most important disease caused by microorganisms of the genus Aspergillus is IA, with A. fumigatus being the principal species involved. It is primarily acquired by inhaling airborne conidia. The conidia, commonly found in indoor and outdoor air, can withstand extreme atmospheric conditions [17] and their very small size (35 mm) allows them to reach the deepest parts of the lungs. The primary ecological niche of Aspergillus spp. is decomposing vegetable matter but its presence is particularly marked near human habitation, including hospitals [18 20]. We observed that Aspergillus is found more frequently in the urban environment, and that this was especially important in the case of A. fumigatus, whose levels changed with statistical significance. This could be a risk factor for people who are susceptible to IA, because urban environments have higher levels of conidia. We analysed the changes of these levels during the different seasons of the year and the effect of weather on these changes. It is worth noting that A. fumigatus was the species most frequently found in every season, but we know that the high flow rate selected in our study may affect the spores that cannot withstand the collection/impaction process, and the survival under such conditions can possibly vary according to species. The maximum levels in all the species were observed in autumn, when atmospheric conditions are appropriate for fungal sporulation. All Aspergillus conidia were affected by temperature: when the temperature increased, conidia levels increased. Nevertheless, other factors such as monthly precipitation and number of days with rainfall also had a positive effect. On the other hand, when the wind speed was high, we observed a decrease in the conidia levels, which may be explained as resistance to the high volume collection process. A. fumigatus does not seem to be very affected by these parameters, but A. niger and A. flavus do. These correlations and the predictive equation were valid, but to obtain an exact correlation effect we would have to measure the levels every day for a year at each point, and this was methodologically impossible. In conclusion, while recognizing the inherent variability associated with the aerodynamics of fungal aerosol sampling, we could say that Aspergillus is a ubiquitous fungus found mainly in the urban environment and near human habitation and its levels change with meteorological parameters. As levels were higher in the urban than in the rural environment, patients at risk are usually in contact with these levels in the city. Autumn and summer were the seasons when the levels of conidia were higher and the inhalation of conidia by patients at risk of developing IA is higher than in other seasons. Counts of Aspergillus spp. conidia were always under 85 c.f.u./m 3 and for A. fumigatus the levels were always under 70 c.f.u./m 3. The fact that A. fumigatus was the most frequently isolated species could explain its more frequent recovery from clinical samples, regardless of its pathogenic factors. Although future studies must correlate Aspergillus air counts and the risk of IA, we already know that when the levels increase, the probability of inhaling conidia is higher. The public water of our province does not seem to contain Aspergillus conidia and air continues to be the main vehicle of conidia. Acknowledgements This study was supported by a grant from Red Española de Investigación en Patología Infecciosa C/ 03/75 (RESITRA: Red de Infección en el Transplante) and by grant number 08.2/0026/ from the

7 Environmental Aspergillus conidia levels 355 Table 5 Correlation between meteorological parameters and total filamentous fungi and Aspergillus conidia levels. Kind of fungí DMaxT a DMinT b DMeanT c MMeanMaxT d MMeanMinT e MMeanMinT f MaxWS g MWS h MWS i DP j NDR k RH l Total filamentous /0.179 m ñ Total Aspergillus /0.179 / A. fumigatus /0.109 / A. niger /0.107 A. flavus /0.106 o /0.113 A. terreus / A. nidulans Aspergillus spp. a DMaxT: Daily Maximum Temperature. b DMinT: Daily Minimum Temperature. c DMeanT: Daily Mean Temperature. d MMeanMaxT: Monthly Mean Maximum Temperature. e MMeanMinT: Monthly Mean Minimum Temperature. f MMeanT: Monthly Mean Temperature. g MaxWS: Maximum Wind Speed. h MWS: Monthly Wind Speed. i MP: Monthly precipitation. j DP: Daily precipitation (day of sampling). k NDR: Number of days with rainfall (month of sampling). l RH: Relative humidity. m Spearman negative coefficient means that an increase in the variable meteorological parameters correlates a decrease in the conidia levels. ñ Spearman positive coefficient means that an increase in the variable meteorological parameters correlates an increase in the conidia levels. o P/ Comunidad de Madrid Research Fund. Jesús Guinea receives a pre-doctoral grant from Universidad Complutense de Madrid. We would like to thank Thomas O?Boyle for his help in the translation of the article and José María Bellón, from the Preventive Medicine Department of Gregorio Marañón Hospital, for his help in the statistical analysis of the study. This study does not present any conflict of interest for its authors. References 1 Groll AH, Shah PM, Mentzel C, Schneider M, et al. Trends in the postmortem epidemiology of invasive fungal infections at a university hospital. J Infect 1996; 33: Denning DW. Invasive aspergillosis. Clin Infect Dis 1998; 26: Gerson SL, Talbot GH, Hurwitz S, et al. Prolonged granulocytopenia: the major risk factor for invasive pulmonary aspergillosis in patients with acute leukemia. Ann Intern Med 1984; 100: Birch M, Nolard N, Shankland GS, Denning DW. DNA typing of epidemiologically-related isolates of Aspergillus fumigatus. Epidemiol Infect 1995; 114: Carlson GL, Mughal MM, Birch M, Denning DW. Aspergillus wound infection following laparostomy. J Infect 1996; 33: VandenBergh MF, Verweij PE, Voss A. Epidemiology of nosocomial fungal infections: invasive aspergillosis and the environment. Diagn Microbiol Infect Dis 1999; 34: Gottlich E, van der Lubbe W, Lange B, et al. Fungal flora in groundwater-derived public drinking water. Int J Hyg Environ Health 2002; 205: Warris A, Gaustad P, Meis JF, et al. Recovery of filamentous fungi from water in a paediatric bone marrow transplantation unit. J Hosp Infect 2001; 47: Warris A, Voss A, Abrahamsen TG, Verweij PE. Contamination of hospital water with Aspergillus fumigatus and other molds. Clin Infect Dis 2002; 34: Anaissie EJ, Stratton SL, Dignani MC, et al. Pathogenic Aspergillus species recovered from a hospital water system: a 3- year prospective study. Clin Infect Dis 2002; 34: Anaissie EJ, Stratton SL, Dignani MC, et al. Cleaning patient shower facilities: a novel approach to reducing patient exposure to aerosolized Aspergillus species and other opportunistic molds. Clin Infect Dis 2002; 35: E De-Wei li BK. A year-round study on functional relationships of airborne fungi with meteorological factor. Int J Biometeorol 1995; 39: Stephen E, Raftery AE, Dowding P. Forecasting spore concentrations: A time series approach. Int J Biometeorol 1990; 34: Rohit K, Katial YZ, Jones RH, Dyer PD. Atmospheric mold spore counts in relation to meteorological parameter. Int J Biometeorol 1997; 41: Guinea J, Pelaez T, Alcala L, Bouza E. Evaluation of Czapeck agar and Sabouraud-Dextrose agar for the culture of airborne Aspergillus conidia. Diagn Microbiol Infect Dis 2005; 53: Murray P, Baron E, Pfaller M, Tenover F, Yolken R. Aspergillus, Fusarium and other oppurtunistic moniliaceous fungi. In: Manual

8 356 Guinea et al. of Clinical Microbiology. American Society for Microbiology 1999, pp Parta M, Chang Y, Rulong S, Pinto-DaSilva P, Kwon-Chung KJ. HYP1, a hydrophobin gene from Aspergillus fumigatus, complements the rodletless phenotype in Aspergillus nidulans. Infect Immun 1994; 62: Mullins J, Harvey R, Seaton A. Sources and incidence of airborne Aspergillus fumigatus (Fres). Clin Allergy 1976; 6: Staib F. Ecological and epidemiological aspects of aspergilli pathogenic for man and animal in Berlin (West). Zentralbl Bakteriol Mikrobiol Hyg [A] 1984; 257: Rainer J, Peintner U, Poder R. Biodiversity and concentration of airborne fungi in a hospital environment. Mycopathologia 2001; 149: 8797.