Incremental Lifetime Cancer Risk of PM 2 . 5 Bound Polycyclic Aromatic Hydrocarbons ( PAHs ) Before and After the Wildland Fire Episode

In Northern Thailand, wildland fire during cold period releases large amounts of smoke and fine particles into the atmosphere. The fine particles include several persistent organic compounds such as PAHs. In this study, PM2.5-bound PAH concentrations in the air of nine administrative provinces, namely Chiang-Mai, Chiang-Rai, Nan, Phayao, Mae Hong Son, Phrae, Lampang, Lamphun, Uttaradit (N Thailand) were determined during the wildland fire and non-wildland fire seasons. The monitoring strategy comprised two campaigns in each season. PM2.5 was collected using MiniVolTM portable air samplers (Airmetrics) with quartz fibre filters. Both PAHs and their B[a]PEquivalent concentrations of other urban cities around the world were significantly higher than those of northern provinces for both seasons. The average cancer risks observed at nine administrative provinces were 8.525 × 10 ± 3.493 × 10 and 2.558 × 10 ± 6.986 × 10 for ingestion rate of 50 and 100 mg day, respectively. The excess cancer risks of world cities for ingestion rate of 50 and 100 mg day were much higher than those of Northern Thailand for 851 and 567 times in that order. Dust ingestion was exceedingly critical to non-dietary PAH exposure in comparison with PM2.5 inhalation. These results are in good agreement with those of previous studies, underlining the significance of indoor air quality on long-term adverse respiratory diseases in Asian cities.

Previous epidemiological studies indicate that lung cancer evidences were substantially higher among residents in Northern Thailand, particularly in women (Pisani et al., 2006;Bumroongkit et al., 2008;Kamnerdsupaphon et al., 2008).For instance, a retrospective study used patient's records at the Chiang Mai Cancer registry and Maharaj Nakorn Chiang Mai Hospital from January 2001 to December 2005 revealed a comparatively high lung and breast cancer incidences (Kamnerdsupaphon et al., 2008).Another epidemiological study raised concerns over pesticide exposure or habitation in poorly ventilated houses may instead be related to the tumorigenesis of lung cancer patients in upper Northern Thailand (Bumroongkit et al., 2008).On the contrary, a previous study highlighted "tobacco smoking" as a main cause responsible for 96% of male and 64% of female lung cancer incidences found in rural area of Northern Thailand (Pisani et al., 2006).
In spite of some discrepancies over the causes of high lung cancer incidences, there has been no researches aiming on neither chemical characterization nor human exposure risk assessment of PM 2.5 bound PAHs before and after the wildland-burning episode in Northern part of Thailand.Overall, the major objectives of this study are to (i) use three different toxicity equivalent concentration (TEQ) equations for assessing the risk of exposure to PAHs, (ii) calculate daily exposure of PM 2.5 bounded PAHs, and (iii) conduct incremental lifetime cancer risk of PM 2.5 bounded PAHs via ingestion, inhalation, and dermal contact with dust particles at nine administrative provinces in northern part of Thailand.

Sampling of PM 2.5
Samplings of PM 2.5 were conducted and collected every week on pre-baked (550°C for 12 h) quartz-fiber filters (QFFs; Whatman 47 mm; Article No., 28418542 (US reference)) with the assistance of MiniVolTM portable air samplers (Airmetrics) using the flow rate of 5 litres minute -1 through a particle size separator (impactor) and consequently through a 47 mm filter.More details of the PM 2.5 sampling method was provided in "EPA Quality Assurance Guidance Document: Method Compendium, Field Standard Operating Procedures for the PM 2.5 Performance Evaluation Program, United States Environmental Protection Agency Office of Air Quality Planning and Standards" (U.S. EPA, 2002).The weight measurement of PM 2.5 was strictly followed the instructions of US-EPA Quality Assurance Document: Method Compendium, PM 2.5 Mass Weighing Laboratory Standard Operating Procedures for the Performance Evaluation Program, United States Environmental Protection Agency Office of Air Quality Planning and Standards (U.S. EPA, 1998) with the employment of micro balances (Mettler Toledo, New Classic MF, MS205DU, Switzerland).The sampling campaigns can be separated into two groups according to the monitoring period during which they were conducted: Campaign-І was conducted before the "wildland fire episode" in the winter of 2012 (i.e., from the 7 th to 22 nd of December 2012), while air sampling during Campaign-П was done in March 2013 (i.e., from the 4 th to 19 th of March 2013).

Campaign-I (Nine Northern Provinces-I (NNP-I); Sampling before the Haze Episode)
Observations of PM 2.5 at cluster-I sampling stations (CROS, PYOS and NPOS), cluster-II sampling stations (LMOS, PHOS and UTOS) and cluster-III sampling stations (MHOS, CMOS and LPOS) was conducted synchronously daily from the 28 th of November to the 4 th of December 2012, from the 7 th to 13 th of December 2012 and from the 16 th to 22 nd of December 2012, respectively.

Campaign-II (Nine Northern Provinces-II (NNP-II); Sampling after the Haze Episode)
Air quality observations at cluster-I monitoring stations (CROS, PYOS and NPOS), cluster-II monitoring stations (LMOS, PHOS and UTOS) and cluster-III monitoring stations (MHOS, CMOS and LPOS) was performed synchronously daily from the 23 th of February to the 2 nd of March 2013, from the 4 th to 11 th of March 2013 and from the 13 th to 20 th of March 2013, respectively.
Since the two campaigns consist of only three weeks durations, it is crucial to carefully investigate seasonal and daily variations, which can be effected by meteorological conditions.The seasonal representativeness of these campaigns should be validated and discussed.As displayed in Fig. 3, the hotspot numbers in northern Thailand provided by the Centre for Remote Imaging, Sensing and Processing (CRISP), the National University of Singapore, were classified into types namely Campaign-I and Campaign-II.The average hotspot number during Campaign-II (i.e., 532 ± 543) was significantly (p < 0.05) higher than those of Campaign-I (i.e., 96 ± 92), highlighting the importance of biomass burnings in March 2013 and thus the selected sampling period can be used to justify the seasonal representativeness of monitoring campaigns.

Polycyclic Aromatic Hydrocarbons (PAHs)
A The analytical procedure used in this study and relevant information has been explained in previous publications (Pongpiachan, 2009(Pongpiachan, , 2011(Pongpiachan, , 2013a, b;, b;Pongpiachan et al., 2013a, b).Briefly, all QFFs samples were chemically extracted with dichloromethane (Fisher Scientific, HPLC grade) by employing the Soxhlet apparatus.The QFFs samples were spiked with known amount of deuterated internal standards (i.e., d 10 -fluorene and d 12perylene) prior to the Soxhlet extraction.The fractionization protocol was started with the elution of PAH fraction with15 mL of 4:6 (v/v) toluene:hexane by using silica gel.Subsequently, the organic extract was rotary evaporated and reduced the volume with the assistance of gentle nitrogen stream to almost dryness until the final volume was 100 µL.More details of GC/MS method were provided in previous studies and will not be discussed here (Pongpiachan et al., 2009(Pongpiachan et al., , 2011)).Both precisions and accuracies were quarantined by applying the standard reference material (SRM 1941b: National Institute of Standard and Technology).

Health Risk Assessment of PAHs Toxicity Equivalent Concentration (TEQ)
The toxicity equivalent concentration (TEQ) equation is broadly used for assessing the risk of exposure to PAHs, which can be calculated as described bellows (Yang et al., 2007;Yu et al., 2008;Pongpiachan et al., 2013b): here, C i and TEF i are acronyms for the concentrations of individual PAH and the toxic equivalency factors, respectively.Since the poisonousness of a single PAH congener can differ by orders of degree, it is essential to show the toxicity of PAHs in terms of the most toxic form of PAHs, B[a]P.By employing TEF, the toxicity of a combination of PAH congeners can be evaluated in a single number, which is generally referred as the toxic equivalency, TEQ.It is a specific number causing from the product of the concentration and individual TEF values of each PAH, as described in Eq. ( 1).In this study, three TEQ equations were used with the citations of Nisbet and Lagoy (1992) (Eq.( 2)), US EPA (1993) (Eq. ( 3)), and Cecinato (1997) (Eq.( 4)).In these three TEQ formulas, the acronyms for the PAH congeners represent their concentrations.

Daily Exposure of PM 2.5 Bounded PAHs
The calculations of daily exposure of PM 2.5 bounded PAHs were conducted by using three different TEQ values as previously described in section 2.3.1.PM 2.5 13 1, 000 (5) M DID (mg day -1 ) represents the daily intake of dust by either adults or children.In this study, the M DID values of 4.16-100 and 55-200 mg day -1 were applied for adults and children, respectively (U.S. EPA, 1997).As a consequence, 4.16 and 55 mg day -1 for adults and children were assumed as comparatively low exposure situation whilst 100 and 200 mg day -1 were taken into account as relatively high exposure situation, correspondingly.A similar concept was applied to the calculation of PM 2.5 bounded TEQs daily exposure via inhalation (M PDI-inhalation ; ng day -1 ), which is based on Eq. ( 6): PM2.5 13 where, ∑ 13 TEQ PM2.5 (ng m -3 ) is the sum of the TEQ contents of the 13 PAHs measured in PM 2.5 .The inhalation rate (IR inhalation ) used in this research was 10 and 20 m 3 day -1 for children and adults respectively (SFT, 1999).

Incremental Lifetime Cancer Risk of PM 2.5 Bounded PAHs
Assuming that (i) there are only ingestion, inhalation, and dermal contact with dust particles, which can be considered as three major pathways for human exposure; (ii) the total carcinogenic risk could be evaluated by using TEQ values as previously mentioned in section 2.3.1;(iii) intake rates and particle emission can be estimated by those developed for soil particles, the exposure risk for environmental PAHs can be quantitatively evaluated by using the concept of incremental lifetime cancer risk (ILCR) (U.S. EPA, 1991;Chen and Liao, 2006;Peng et al., 2011).The subsequent equations were extensively applied to estimate the ILCR in terms of ingestion, dermal contact and inhalation:   (2)), US EPA (1993) (Eq. ( 3)), and Cecinato (1997) (Eq.( 4)).
It is obvious that the measured PAH congeners in the present study fell within relatively low range, in comparison with other studies around the world (Table 2).Several factors such as unique tropical wet and dry climate (Köppen Aw) characterized by hot and rainy weather coupled with low industrial emission sources of northern provinces might have influenced the comparatively low PAH contents observed in Table 1.Statistical descriptions of PAHs [pg m -3 ] in PM 2.5 collected from nine administrative provinces before (NNP-I) and in the middle (NNP-II) of biomass burning episode coupled with the average of world cities.  (Slezakova et al., 2011;Martellini et al., 2012;Slezakova et al., 2013).These findings underline the prominence of traffic emissions as the main contributors of PM 2.5 bounded PAHs, and it might have been exceeded numerous emission sources involving wild land fire and biomass burnings.As there were no significant differences detected at NNP-I and NNP-II (Table 1), it seems plausible to mention that the wild land fire episode plays a minor role on variations of PM 2.5 bounded PAHs in nine administrative provinces of Northern Thailand.

Health Risk Assessment
In this study, non-dietary exposure is defined as human exposure to PM 2.5 bounded PAHs via both household air and dust (Wilford et al., 2005).Table 3 displays the excess cancer risks associated with house dust exposure in NNP-I, NNP-II, and WCs, ranged from 3.732 × 10 -4 ± 7.929 × 10 -4 (NNP-II-TEQ US-EPA with ingestion rate of 50 mg day -1 ) to 1.451 ± 1.741 (WCs-TEQ Nisbet-Lagoy with ingestion rate of 100 mg day -1 ), depending on TEQ equations and ingestion rate (IR) as displayed in Eq (5).The mean cancer risks of NNP were 8.525 × 10 -4 ± 3.493 × 10 -3 and 2.558 × 10 -3 ± 6.986 × 10 -3 for ingestion rate of 50 and 100 mg day -1 , respectively.The cancer risks of WCs for ingestion rate of 50 and 100 mg day -1 were greatly higher than those of NNP for 851 and 567 times in that order.Although the cancer risks of NNP were much lower than those of WCs, its average values ranged from 10 -4 to 10 -3 (i.e., one cancer incidence case per million), which can be considered as unacceptable cancer risks (Maertens et al., 2008).These values are comparable to those of foundry workers (9.06 × 10 -4 and 1.09 × 10 -3 ) in Taiwan (Liu et al., 2010), but higher than those predicted as occupational exposure for sinter metal workers (3.18 × 10 -5 and 4.98 × 10 -5 ) (Lin et al., 2008;Chiang et al., 2009) and the Canadian maximum acceptable level of risk (1 × 10 -5 ) (Maertens et al., 2008).The current study indicates there is some concerns on adverse health impact related with PM 2.5 bounded PAHs exposure for preschool children via non-dietary exposure in home environment, based on the calculation of Eqs. ( 1)-( 5).
The highest average values of TEQ daily exposure doses via house dust (M PDI-ingestion ; ng day -1 ) for adults and children was 7.345 × 10 4 ± 8.813 × 10 4 and 1.469 × 10 5 ± 1.763 × 10 5 ng day -1 under high exposure condition (with ingestion rate of 100 and 200 mg day -1 for adults and children) with the application of WCs-TEQ Nisbet-Lagoy , while 4.715 ± 12.08 and 62.33 ± 159.8 ng day -1 under low exposure condition (4.16 and 55 mg day -1 for adults and children) with the employment of NNP-II-TEQ US-EPA , respectively (Table 4).It is also important to note that M PDI-ingestion values of NNP-II-TEQ US-EPA were almost similar to those of Guangzhou for adults and children with the values of 4.31 and 57.0 ng day -1 , respectively (Wang et al., 2013).For TEQs daily exposure via inhalation (M PDI-inhalation ; ng day -1 ), WCs-M-PDI-inhalation-Nisbet-Lagoy shows the highest values for both adults and children with the values of 147 ± 176 and 735 ± 881 ng day -1 , in that order (Table 5).Since the values of M PDI-ingestion were two to three orders of magnitude exceed those of M PDI-inhalation , it would be safe to assume that dust ingestion was extremely crucial to non-dietary PAH exposure in comparison with PM 2.5 inhalation.These findings in NNP Table 3.The cancer risks associated with non-dietary ingestion of PAHs in house dust for preschool children as estimated by using Eq. ( 5) (U.S. EPA, 1997;Maertens et al., 2008)  are consistent with those of Guangzhou (Wang et al., 2013), highlighting the importance of indoor air quality on longterm adverse respiratory diseases in Asian cities.

Incremental Lifetime Cancer Risk
This is the first attempt that evaluated the potential cancer risk of human exposure to PM 2.5 bounded PAHs of different exposure routes before and after the haze episode by using three different TEQ methods.The potential cancer risks of adults and children can be assessed by using ILCR for exposure routes of ingestion, dermal contact and inhalation as displayed in Eq. ( 7), Eq. ( 8), and Eq. ( 9), respectively (Table 6) (Chen and Liao, 2006;Peng et al., 2011;Wang et al., 2011).It is important to note that ILCR Dermal and ILCR Ingestion varied from 10 -10 to 10 -8 in both NNP-I and NNP-II samples, while those of WCs was 10 -7 to 10 -6 ; approximately two to three orders of magnitude higher than those of NNP.This indicates that human activities in urban world cities play an important role in governing potential cancer risk of human exposure to PM 2.5 bounded PAHs.It is also worth mentioning that both and ILCR Ingestion were significantly higher than ILCR Inhalation , denoting that respiration of PM 2.5 bounded PAHs through mouth and nose was practically unimportant, when compared with the other exposure pathways.These findings were in good agreement with studies of exposure to heavy metals in street dust of Angola (Ferreira-Baptista and De Miguel, 2005) and dust samples collected from Guangzhou, a typical urban center in South China (Wang et al., 2011).In the case of children, the ILCR Ingestion was significantly eight times higher (p < 0.05) than those of ILCR Dermal , suggesting that the hand-to-mouth activity prominently contributed to the cancer risk for children.This finding is also in consistent with the investigation of potential risk of children exposure on heavy metal distribution in dust from elementary schools in Hermosillo, Sonora, Mexico (Meza-Figueroa et al., 2007) and potential cancer risk of children exposure to urban dust bounded PAHs in Guangzhou, China (Wang et al., 2011).Therefore, it appears reasonable to conclude that the hazard health risk for children exposed to PM 2.5 bounded PAHs through direct ingestion is considered to be noticeably greater than that of dermal contact and inhalation.
Under most guidelines for carcinogen risk assessment, the values of ILCR range from 10 -6 to 10 -4 implied potential cancer risk, while the practical safety was expressed with an ILCR of 10 -6 or less and a potential high risk was evaluated by an ILCR of higher than 10 -4 (Chen and Liao, 2006;Wang et al., 2011).Furthermore, U.S. EPA Risk Assessment Guidance for Superfund Program adopts a cancer risk level of 10 -6 to suggest the point at which risk management decisions should be taken.By applying the same guidance, the chance of adverse cancer-causing effects in northern provinces of Thailand was considerably lower in comparison to world urban cities.In this study, the average values of ILCR of children as estimated by three different TEQs for NNP for both monitoring campaigns were 5.371 × 10 -9 ± 1.518 × 10 -8 , 6.696 × 10 -10 ± 1.893 × 10 -9 and 1.042 × 10 -14 ± 2.944 × 10 -14 , for ingestion, dermal contact, and inhalation, respectively; these risk levels are 1.862 × 10 2 , 1.493 × 10 3 , Table 6.Calculations of ILCR in terms of ingestion (Eq.( 7)), dermal contact (Eq.( 8)) and inhalation (Eq.( 9)) due to human exposure to PAHs via PM Nisbet-Lagoy 1.064E-13 3.126E-13 1.330E-14 3.907E-14 8.357E-14 2.431E-13 1.045E-14 3.039E-14 3.348E-11 4.017E-11 4.185E-12 5.022E-12 and 9.602 × 10 7 times lower than that of US EPA (1991) baseline in that order.In addition, the average ILCR of adults living in NNP as estimated by three different TEQs for three various exposure pathways was 256 and 237 times lower than the WCs and US EPA baseline, further indicating that the cancer risk of NNP falls into the ''acceptable level'' range.

CONCLUSIONS
In this study, PAHs analysis in PM 2.5 collected in nine administrative provinces of Northern Thailand from December 2012 to March 2013 did not show any significant differences between wildland fire and non-wildland fire seasons.These findings underline the influences of traffic emissions as regular contributors of PM 2.5 -bound PAHs.While the excess cancer risks of PM 2.5 -bound PAHs in Northern Thailand were greatly lesser than those of other world cities, its average values ranged from 10 -4 to 10 -3 , which was higher than the Canadian maximum acceptable level of risk (1 × 10 -5 ) and other previous occupation exposure studies.The average ILCR values of adults living in Northern Thailand as predicted by three different TEQs for ingestion, dermal contact, and inhalation exposure pathways were much lower than the US EPA baseline, further denoting that the cancer risk of NNP falls into the ''acceptable level'' range.
All nine administrative provinces of Northern Thailand are classified as tropical wet and dry climate (Köppen climate classification: Aw).The weather is fairly dry and warm in winters while the temperatures rise until April with the hottest average daily maximum at 34.9°C (94.8°F).The monsoon season runs from late April through October, with heavy rain and somewhat cooler temperatures during the day, although nights remain warm.The geographical positions of each monitoring station were written as below: Chiang-Mai Observatory Site (CMOS): (E: 498805, N: 2077713), Chiang-Rai Province Observatory Site (E: 593783, N: 2258302), Nan Province Observatory Site (NPOS; Thewarat Hotel; E: 687123, N: 2077209), Phayao Province Observatory Site (PYOS; Arunothai Coffee House Homestay; E: 594420, N: 2119215), Mae Hong Son Province Observatory Site (MHOS; Mae Hong Son Provincial Forestry Office; E: 391834, N: 2134869), Phrae Province Observatory Site (PHOS; Nana Charoenmuang Hotel; E: 620935, N: 2006155), Lampang Province Observatory Site (LMOS; Maemoh Training Center; E: 568200, N: 2020017), Lamphun Province Observatory Site (LPOS; Lamphun Provincial Administration Organization Stadium; E: 500441, N: 2052987), Uttaradit Province Observatory Site (UTOS; OUM Hotel; E: 615923, N: 1948269) (See Fig. 1).All air quality observatory sites were situated in the city center of each administrative province of Northern Thailand.The population density vary widely, from the relatively urbanized Chiang-Mai and Chiang-Rai Provinces with the range of 80 to 100 people km -2 , to the mostly-forest area of Mae Hon Song Province with less than 40 people km -2 (See Fig. 2).It is also worth mentioning that Northern Thailand contrasts remarkably from other regions of Thailand.All monitoring stations are geologically categorized by several mountainous belts, which continue from the Shan Hills in bordering RUM and LPDR, and climatically highlighted by a tropical savanna.

Fig. 1 .
Fig. 1.PM 2.5 monitoring stations in nine administrative provinces of Northern Thailand.

Fig. 2 .
Fig. 2. Population density in nine administrative provinces of Northern Thailand.

Fig. 3 .
Fig. 3. Hot spot numbers before and after burning episodes in nine administrative provinces of Northern Thailand.

Table 2 .
Aerosol observatory sites from cities around the world (WCs) for PAHs analysis. .