United States
             Environmental Protection
             Agency
                    Office of Science and Technology
                    Office of Water
                    Washington, DC 20460
EPA-822-B-01-007
August 1999
vwwv.epa.gov
&EPA
Mycobacteria:
Health Advisory


                                                          v*-. -
                                                          -%£&&




-------

-------
I.      INTRODUCTION

       The Health Advisory Program, sponsored by the Office of Water (OW), provides
information on the health effects, analytical methodology and treatment technology that would
be useful in dealing with the contamination of drinking water. Most of the Health Advisories
(HAs) prepared by the Office of Water are for chemical substances.  This Health Advisory
however, addresses contamination of drinking water by a microbial pathogen, examines
pathogen control,  and addresses risk factors for exposure and infection.

       Health Advisories serve as informal technical guidance to assist Federal, State and local
officials responsible for protecting public health when emergency spills or contamination
situations occur. They are not to be construed as legally enforceable Federal standards. The
HAs are subject to change as new information becomes available.

       This Health Advisory is based on information presented in the Office of Water's Criteria
Document (CD) forMycobacteria. Individuals desiring further information should consult the
CD.  This document will be available from the U.S. Environmental Protection Agency,  OW
Resource Center, Room M6099; Mail Code: PC-4100, 401 M Street, S.W., Washington, D.C.
20460; the telephone number is (202) 260-7786.  The document can also be obtained by calling
the Safe Drinking Water Hotline at 1-800-426-4791.

-------
II.     GENERAL INFORMATION

History

•      Mycobacteria were one of the first types of bacteria recognized to cause disease
       (tuberculosis and leprosy).

•      The name Mycobacterium, which means fungus-bacterium, was introduced in 1896. The
       name does not imply that Mycobacterium are fungi; rather it describes the way that the
       tubercle bacillus grows on the surface of liquid media as mold-like pellicles
       (Gangadharam & Jenkins, 1998).

•      In the  1940s and 1950s, it became apparent that there were many other species of
       mycobacteria (in addition to those which cause tuberculosis and leprosy) which could
       contribute to disease, based on their isolation from clinical patients.

Taxonomy

•      Mycobacteria belong to the Order Actinomycetales, Family Mycobacteriaceae and
       Genus Mycobacterium.

•      One of the early techniques used to classify mycobacteria was the system of Adansonian
       taxonomy (Grange, 1996b). This system was based on the use of cultural and
       biochemical properties to group related strains.  Although this system is rarely used
       today, this approach facilitated the use of discriminative identification tests for use in
       diagnostic mycobacteriology.

•      Dr. Ernest Runyon performed the initial work and grouping classifications on the
       taxonomy of Mycobacterium in the mid-1950s.  In Runyon's classification,
       mycobacteria, excluding those in the M. tuberculosis complex and noncultivable taxa
       (e.g., M. leprae), were divided into four groups based on growth rates and pigmentational
       properties.

•      In 1980, 41 species of mycobacteria were included in the Approved Lists of Bacterial
       Names (Grange, 1996b).  Today, there are 71 recognized or proposed species of
       mycobacteria (Shinnick and Good, 1994).  Over 20 of these species are known to cause
       disease in humans.

•      Mycobacteria not identified as tuberculosis or leprosy complex, have been addressed by a
       variety of nomenclature including; 'atypical mycobacteria', 'mycobacteria other than
       tubercle' bacilli (MOTT), 'environmental mycobacteria' or 'non-tuberculous
       mycobacteria' (NTM) (Wolinsky, 1979).    This document uses the NTM nomenclature.
                                          -2-

-------
Morphologic and Taxonomic Characteristics

•      Aerobic, asporogenous rods, which tend to be acid-fast at some stage of their growth
       cycle.

•      Have been referred to as the 'ducks of the microbial world' due to their thick, waxy,
       outer coating which enables them to thrive in aquatic environments.

•      Have cell walls with very low permeability, contributing to their resistance to therapeutic
       agents. The mycobacterial cell wall is highly complex and has a lipid content that
       approximates 60% of the structure (Brennan and Nikaido, 1995). The cell wall
       characteristics allow the mycobacterial  species to survive in different environments (e.g.,
       in biofilms in water habitats or particulate matter in soils and water) and resist
       disinfection procedures.
Lifecycle
       Most studies have revealed that mycobacteria exist as non-sporing acid-fast rods and
       replicate by binary fission (Grange, 1996b). However, some investigators have
       hypothesized that a more complex life cycle (possibly involving cell-wall-free forms or
       microspores) exists. Several authors have produced evidence for unusual forms and life
       cycles of mycobacteria.

       Unusual life cycles may explain the phenomenon of mycobacterial dormancy and
       persistence (Grange, 1996b).  Wayne (1994) reviewed the mechanism of dormancy seen
       inM tuberculosis. There is ample evidence that these organisms  are capable of adapting
       to prolonged periods of dormancy in tissues, and that this dormancy is responsible for the
       latency of disease.  Wayne (1994) indicates that there may be two or more stages
       involved in the process leading from active replication to dormancy.  One of these steps
       involves a shift from rapid to  slow replication, the second involves complete shutdown of
       replication, but does not result in death of the cell.
Host Range
       In addition to human, tuberculosis infection also occurs in a wide range of domesticated
       and wild animals.  M. bovis, which causes tuberculosis in animals, has one of the widest
       host ranges of all pathogens.  Hosts include the African buffalo, baboon, badger, bison,
       opossum, cat, elk, Fallow deer, goat, horse, Leche antelope, maral, pig and wild boar,
       Rock hyrax, and seal (Morris et al., 1994; Grange, 1996b).

       Leprosy is a disease that is primarily transmitted from person to person.  Humans were
       thought to serve as the only reservoir of M leprae until 1974, when scientists discovered
       nine-banded armadillos in Louisiana with leprosy in an advanced stage.  It is now

-------
       thought that the nine-banded armadillo is the primary animal reservoir forM leprae and
       may serve as a vector for transmission of leprosy to humans either by direct or indirect
       contact (Jenkins,  1991). M. leprae has also been found to infect chimpanzees and sooty
       mangabey monkeys in Africa (Grange, 1996a).

       Several of the NTM species have been isolated from animals, birds and fish. These
       include: M. avium complex, M. marinum., M. ulcerans, M. paratuberculosis, M. simiae,
       M.fortuitum andM smegmatis. Most of the NTM agents that cause human and animal
       diseases have demonstrated very little person-to-person contagiousness (Wayne &
       Sramek, 1992).
III.    OCCURRENCE

No evidence was found that water serves as a source of infection for tuberculosis or leprosy.
Because the bacterial species that cause these diseases have not been recovered from water
sources, the remaining sections of this health advisory will focus only on NTM species.

NTM are ubiquitous in nature and consist of a large number of species which vary in
pathogenicity. These infections are more likely transmitted from environmental sources by
ingestion, inhalation and inoculation of Mycobacterium bacilli.  These environmental sources
may include aerosols, water, soil, dust, food products and contaminated medical equipment
(Gangadharam & Jenkins, 1998).

Because routine environmental monitoring for NTM is not a common practice, the occurrence of
these bacteria is often only indicated by outbreaks or sporadic cases of mycobacterial infection.
Therefore, this section considers the worldwide occurrence or incidence of mycobacterial
infections as well as the occurrence of NTM in water, soil, and air.  Environmental factors
influencing NTM survival also are discussed.
Occurrence in Water

Environmental mycobacterial species have been repeatedly isolated from natural and municipal
waters. They occur in surface water, notably ponds, streams, and estuaries. Piped water supplies
are readily colonized by mycobacteria which can thereby lead to more frequent exposure of
humans.  Mycobacterial characteristics such as surface hydrophobicity and charge, as well as
certain physiochemical factors like salinity, temperature, humidity and wind currents can
influence the distribution of mycobacteria in water systems (Falkinham, 1996).

       Waste Water

•      In 1980, a survey conducted in Korea found that both slow- and rapid-growing
       environmental mycobacteria (EM) were isolated from 67% of the sewage samples

                                           -4-

-------
collected. Slow growers constituted 49% of sewage isolates. The isolates from sewage
samples included: M gordonae, M. scrofulaceum, M.flavescens, M. phlei, M. terrae (,
M.fortuitum, M.  chelonei andM smegmatis (Won Jin et al., 1984).

In Czechoslovakia, environmental mycobacteria have been isolated from waste water in
the North-Movarian Region (Kaustova, 1981).  Between 1985 and 1991, the reference
laboratory in the  Czech Republic reported that 8.2% (102) of 1244 waste water samples
tested were positive for mycobacterial species (Slosarek et al., 1994).
Surface Waters

Mycobacterium-avium-intracellulare-scrofulaceum (MAIS) organisms were recovered
from so-called acid-brown swamps and lake water samples collected in Georgia,
Virginia, and West Virginia.  The high rate of recovery of MAIS organisms in these
regions was attributed to the combination of higher temperatures, low oxygenated waters,
low pH soils, higher zinc, and fulvic and humic acids (Kirschner et al., 1992).

In a study in Finland, MAIS complex organisms were detected in 40% of surface water
samples collected from streams.  Concentrations of these organisms were found to range
from 50 to 1,400 CFU/L (mean 370 CFU/L). Additionally, M. malmoense was detected
in two stream waters at concentrations of 320 and 750 CFU/L. In all surface water
samples (not limited to streams), mycobacteria were detected at a mean of 1,500 CFU/L
(Katila et al., 1995).

In Valencia, Spain, 15 strains ofM gordonae and 10 strains of MAC were identified in a
variety of surface water samples  (Sabater and Zaragoza, 1993).
Swimming Pools/Hot Tubs

Mycobacterial species including, M. marinum, M.  chelonei, M. scrofulaceum andM
gordonae were isolated from swimming pool samples collected in Araraquara, Sao
Paulo, Brazil (Falcao et al., 1993).  The number of isolates ranged from 1-3 per site.

A survey was made of water quality of swimming pools in semi-public areas (hotels,
recreational parks and camping grounds) and whirlpools (in sauna institutes and fitness
clubs) that were not yet subject to water quality standards under Dutch legislation
(Havelaar et al., 1985). Mycobacteria were detected in all water samples.  Water
temperatures in pools and whirlpools ranged from 18-25'C and 35-40'C, respectively.
The mycobacterial numbers in whirlpools on the average were about ten times higher
than those in swimming pools, with M. gordonae being the predominant species
recovered. M. marinum was not detected in any of the samples and M. kansasii was
recovered three times from whirlpools. Only M avium andM fortuitum-M. chelonei

                                    -5-

-------
complex were found in both types of pool samples, but the highest densities were seen in
whirlpools.

Species in theM avium complex have been recovered from hot tubs (Embil et al., 1997;
Kahana et al., 1997). Aerosols generated by the hot tub action are linked to case reports
of disease in humans.
 Ground Water

 An investigation was conducted to determine whether ground water could be a natural
 source of NTM, particularly the MAIS group, in order to explain the geographic bias for
 infected people in the southeastern U.S. The samples that were tested originated in three
 geographic regions characterized by different human reactivity to PPD-B.  These regions
 included the Georgia coastal plain, the Virginia coastal plain,  and the Valley and Ridge
 Region of Montgomery Co. Virginia.  Relatively low numbers of mycobacterial isolates
 and very low numbers of MAIS isolates were recovered in the selected regions and were
 not found to correlate with the distribution of PPD-B reactivity or the incidence of MAIS
 infection. The data strongly suggest that clean ground waters are not sources of human
 infection by MAIS or other mycobacteria (Martin et al., 1987).

 Environmental mycobacteria were isolated from 27 out of 63 well water samples
 examined during a nationwide tuberculosis prevalence survey  in Korea in 1980. M.
fortuitum, M. terrae,  andM gordonae were the predominant species. (Won Jin et al.,
 1984).

 Water reservoirs constructed of concrete or steel tanks, located below and above ground
 in a small town in Texas were reported to contain M. kansasii  and M. gordonae
 (Steadham, 1980). The investigators postulated that the NTM species could have entered
 the reservoirs and passed into the water distribution  system from deep water wells or
 through seepage of water and soil which accumulate in both concrete and steel  reservoirs.
Drinking Water
von Reyn et al. (1993) investigated mycobacterial recovery from water supply samples
from wells, hot and cold municipal water supplies, showers and stand pipes. M. avium
complex organisms were recovered from 1/6 and 2/8 of samples collected in New
Hampshire and Boston, respectively.
Tap water samples from both hot and cold outlets in a French hospital contained NTM.
The predominant species included M. kansasii, M. gordonae, and M. fortuitum (Dailloux,
1991).
                                    -6-

-------
       Scarlata et al. (1985) reported detection of NTM in 37% of tap water samples collected in
       Palermo, Italy.

       Kubalek and Mysak (1995) investigated the prevalence of environmental mycobacteria
       collected from a drinking water supply system in the North Moravian region of the Czech
       Republic during 1984 to 1989. Samples were obtained from 16 localities by tap  swabs,
       tap scrapings or collecting running water.  The most commonly identified species of
       NTM wereM gordonae (20.4%), M.flavescem (13.8%), rapidly growing mycobacteria
       (5%) and occasionally M. fortuitum, M. terrae andM. scrofulaceum.

       von Reyn et al. (1994) investigated  mycobacteria in water collected from hospital taps.
       They found concentrations ofM  avium ranging from none to 5.2 CFU/mL.

       Falsely reported outbreaks of nosocomial infections by NTM species have been traced to
       the use of tap water, used on medical devices and equipment during certain surgical and
       lab procedures, which is contaminated withM xenopi (Sniadack et al., 1993), M
       gordonae (Stine et al., 1987) and M. fortuitum (Jacobsen et al., 1996).
Occurrence in Other Media

       Soil

•      Mycobacterial species that have been shown to be normal inhabitants of soil include M.
       kansasii (Jones, 1965), MAIS complex (Brooks et al., 1984a; Kirschner et al., 1992) M.
       malmoense (Saito et al., 1994) and rapidly growing species M. smegmatis and M.
      fortuitum (Jones, 1965; Wolinsky and Rynearson, 1968).

•      Katila et al. (1995) investigated the occurrence of mycobacteria in soils from unpolluted
       areas of Finland. Mycobacteria were detected in all soil samples with a mean of 3.6 x 105
       CFU/g dry weight.

•      The presence of M avium of the same serotype in tuberculous lesions in pigs and in sputa
       of swine workers led Reznikov and Robinson (1970) to postulate that dusts from M.
       avium contaminated soil generated in the swine facility were the source of transmission
       for both pigs and man.

•      Potting soil samples collected from the homes of HIV patients were found to contain
       serotypes of MAC that were similar to the isolates from the HIV patients of the study
       group. Although a relationship of exposure to potting soil and acquisition of MAC could
       not be demonstrated, the data  suggest that potting soil may be a potential reservoir of
       organisms causing MAC infection in San Francisco (Yajko et al.,  1995).
                                          -7-

-------
Air

Shelton et al. (1999) recently investigated aerosolized NTM as a possible cause of
hypersensitivity pneumonitis in three machine workers.  Water-based coolants are
typically recycled and may frequently become colonized by microorganisms.  In one case
study, mycobacteria from theM chelonae complex were identified in concentrations
ranging from non-detectable to 6.6 x 106 CFU/mL in bulk coolant samples.  Air samples
collected in the area around the colonized machines revealed concentrations ranging from
56 to 2,256 CFU/m3. In the second case study, Shelton et al. (1999) found mycobacterial
counts of 102 to 107 CFU/mL in machine fluid samples. Air samples collected near the
colonized machines yielded concentrations exceeding the limits of the sampler (greater
than 9,424 CFU/m3). The mycobacteria detected in this case were rapid-growers which
were identical to a newly proposed mycobacterial species M. immunogen. In the third
case study, concentrations of mycobacteria in metal removal fluids were found to range
from non-detectable to greater than 106 CFU/mL. Again, the mycobacteria  were identical
toM immunogen.

Some evidence suggests that  some forms of NTM infection may be associated with heavy
occupational exposures to dust (Oldham et al., 1975). An apparent association was found
between infection withM kansasii and exposure to dusts in the metal grinding trades.
The association between dust exposure and mycobacterial infection may have two
complementary mechanisms: (1) dust may  act as a pulmonary irritant or toxin, resulting
in increased susceptibility to mycobacterial infection; and (2) dust serves as a vehicle for
mycobacterial exposure.
Surfaces

In a model system, the accumulation of NTM in biofilms resulted in mycobacterial
densities of more than 106 colony-forming units per square centimeter (CFU/cm2) within
10 weeks (Schulze-Robbecke et al., 1991) providing evidence that biofilms may serve as
reservoirs for these organisms.

In a study conducted in Bonn, Germany, 90% of biofilm samples from piped water
systems contained mycobacteria at average densities ranging between 103 andlO4
CFU/cm2 with a maximum density of 5.6 x 106 CFU/cm2. Samples collected from
organic substances such as plastic and rubber typically had higher concentrations of
mycobacteria than did inorganic substances such as copper or glass. The opportunistic
pathogens identified in this study included M. chelonae, M. fortuitum, M. gordonae and
M. kansasii (Schulze-Robbecke et al., 1992).

Food

-------
       Studies of risk factors for MAC infection among AIDS patients in San Francisco did not
       support the hypothesis that food was a likely source of organisms that cause MAC
       infection in AIDS patients (Yajko et al., 1995).
Worldwide Distribution of Mycobacterial Infection

•      NTM are generally found in the environment as free-living organisms, but many are also
       known opportunistic human pathogens.  NTM diseases are not notifiable or required to
       be reported; therefore, information on the occurrence of disease is likely to be
       underestimated.  However, human infections due to NTM appear to be increasing at a
       significant rate across the United States.

•      With the emergence of the AIDS pandemic, it is estimated that 25% - 50% of HIV-
       patients in the United States and Europe are infected with NTM, the primary species of
       which isM avium (Horsburg, 1991).  Presently, the use of highly active anti-retroviral
       therapy (HAART) has led to a decrease in this estimated risk and rate of infection of
       NTM in AIDS patients.

•      In non-AIDS patients, the occurrence of NTM diseases is also on the rise. This is
       suggested by the frequency with which MAC has been isolated from clinical samples in
       regional reference laboratories in the state of Massachusetts from 1972 to 1983 (du
       Moulin et al.,  1985). Reports from Milwaukee, Philadelphia and Portland indicate that
       the prevalence of pulmonary disease due to MAC in the general population has begun to
       approach and exceed that of tuberculosis (Iseman, 1998). The Centers for Disease
       Control and Prevention (CDC) estimates that rates for non-AIDS NTM diseases in the
       United States are 1.8 per 100,000 per year, of which 1.3 are due to the M. avium complex
       (MAC) (O'Brien, 1989).

•      The incidence of mycobacterial disease may be influenced by exposure to dusty
       conditions. Among 154 patients in Great Britain with pulmonary M kansasii disease, 33
       had pneumoconiosis and another 31 were coal miners, steel workers or worked in dusty
       conditions (Jenkins, 1981). In a study of 12 patients withM kansasii pulmonary
       infections in southern California, seven had pre-existing pulmonary disease and three
       reported previous exposures to dust (Gorse et al., 1983 ).

•      M. marinum has been well established as a human pathogen, based on clusters of cases
       observed between 1930 and 1970 (Dobos et al., 1999). Infections in humans have been
       reported in coastal areas of the Middle East (Evan-Paz et al., 1976), in the Far East
       (Iredell et al.,  1992) and in several countries in Europe (Collins et al., 1984) as well as in
       the United States (Zeligman, 1972).
                                          -9-

-------
       In some developing countries, M. ulcerous disease is a  problem, with hundreds of
       individuals afflicted with disabling lesions and only limited therapy available (Marston et
       al., 1995). Although endemic occurrence of the disease has been reported mostly in
       tropical and subtropical areas of the world (Hayman, 1991), a large outbreak ofM
       ulcercms infection on a temperate southern Australian island was reported by Veitch et al.
       (1997).

       From the time thatM haemophilum was first described as a human pathogen in 1978 up
       until 1989, only 18 cases of infection had been reported: seven patients were from the
       United States, and 11 were from Australia, Canada and France (MMWR, 1991). From
       1989 to 1991, CDC identified M haemophilum in eight patients from Connecticut,
       Florida, Georgia, Pennsylvania, Texas and Virginia.

       There  is a lack of information on the occurrence of NTM in animals; however, reports
       raise the possibility that certain animals may represent a natural reservoir for these
       mycobacteria (Gangadharam and Jenkins, 1998).  Strains of disease-causing M avium
       have been recovered from ducks, geese and swans in an English wildlife reserve. NTM
       species such asM marinum andM scrofulaceum which are normally associated with
       water  as a habitat, have been found in fish.
Epidemiology and Disease Outbreaks

•      NTM diseases are not reportable; therefore, information regarding the occurrence of
       disease outbreaks is likely to be underestimated. Many infections caused by these
       organisms are episodic and sporadic (e.g.,M kamasii, M. marinum, andM ulcerans
       infections).

•      Waterborne NTM have been associated with a large number of nosocomial and pseudo-
       outbreaks worldwide. Nosocomial disease outbreaks usually involve sternal wound
       infections, plastic surgery wound infections or postinjection abscesses. Falsely reported
       outbreaks are usually related to contaminated hospital equipment and water supplies.

•      Outbreaks of mycobacterial disease have been reported after exposures in public
       swimming areas. In 1954, an outbreak of 80 cases of M marinum were reported due to
       exposure in a contaminated swimming pool. An additional outbreak involving at least
       290 cases occurred in children who swam in a warm mineral water pool in Glenwood
       Springs, Colorado (Wolinsky, 1979).
Factors Affecting Environmental Survival
                                          -10-

-------
•      The survival of environmental mycobacteria in habitats that are potential reservoirs or
       sources of infection may be influenced by certain physiochemical factors which include
       temperature, pH, organic matter salinity and humidity.

•      du Moulin et al. (1988) reported that temperatures between 52° and 57°C encouraged the
       proliferation of M. avium in hospital water supplies and recommended raising the
       temperature of hot-water systems to reduce exposure of patients to organisms of the
       MAIS complex.

•      Other investigators have found thatM kansasii colonized cold water systems and mixer
       taps, whereas M. xenopi predominated in hot water systems and mixer taps (McSwiggan
       and Collins, 1974; Wright et al., 1985).

•      An investigation was conducted to measure the heat susceptibility of opportunistic
       mycobacteria frequently isolated from domestic water supply systems versus the heat
       susceptibility of L. pneumophila (Schulze-Robbecke and Buchholtz, 1992). It was
       concluded thatM kansasii is more susceptible to heat than L. pneumophila, whereas M
      fortuitum,M. intracellulare andM  marinum are equally  susceptible to temperatures
       between 55° and 60° C.  However, M. avium, M.  chelonae, M. phlei, M. scrofulaceum
       and M. xenopi were found to be more heat resistant than L. pneumophila.

•      A study designed to evaluate the efficiency of ozone as a disinfectant in secondary
       wastewater effluent under different environmental conditions, showed that the survival of
       M.fortuitum increased with an increase in pH from 5.7 to 10.1 (Farooq et al., 1977)

•      Humic and fulvic acids are known to affect the survival of environmental mycobacteria
       either directly (Brooks et al., 1984b) or indirectly in combination with other
       physiochemical variables such as temperature, oxygen content, pH and inorganic
       substances (Kirschner et al.,  1992).  The combination of higher temperatures, low
       oxygenated  waters low soil pH, and waters high in zinc, humic and fulvic acids from
       swamp waters most likely favor the growth and survival of MAIS organisms (Flaig et al.,
       1975; Schnitzer, 1982).

IV.    HEALTH EFFECTS IN HUMANS

       Clinical Symptoms

•      The clinical symptoms seen following infection with NTM depend greatly on the
       mycobacterial species and site of the infection. NTM diseases in immunocompetent
       hosts are relatively rare even though exposure to organisms is common based on their
       ubiquitous nature in the environment.

•      Common clinical syndromes include:
       • •    Pulmonary infection

                                          -11-

-------
• •     Lymph infection (lymphadenitis)
• •     Ear infection
• •     Skin & soft tissue infection
• •     Catheter-associated infection
• •     Disseminated Infection

Pulmonary Infection

In adults, pulmonary  infection is the most commonly recognized form of NTM
infections.  These infections often present clinically as chronic cough, sputum production
and fatigue. Older adults are generally the population in which chronic lung disease due
to NTM is observed (ATS, 1997).

Two forms of pulmonary disease are recognized: apical cavitary disease that is best
recognized with MAC and M. kansasii infection in middle aged male smokers, and
modular bronchiectasis that is best recognized in MAC and M. abscessus infection and is
most commonly seen in elderly non-smoking women (ATS, 1997).

Members of the M avium complex orM kansasii are the NTM species most commonly
associated with pulmonary infection in the U.S.  However, other species known to
occasionally cause pulmonary disease include; M. xenopi, M. fortuitum, M. abscessus, M.
szulgai, M. malmoense, M. simiae, M. celatum, M. asiaticum and M. shimodii (ATS,
1997).
Lymph Infection (Lymphadenitis)

Lymphadenitis occurs predominantly in young children, between 1 and 5 years old, and
typically affects the cervical, submaxillary, submandibular and preauricular lymph nodes
(ATS, 1997; Jenkins, 1991).  In the absence of HIV infection, this disease rarely affects
adults. Historically, the classical cause of cervical lymphadenitis wasM scrofulaceum
whereas today, the species most commonly involved is theM avium complex. M.
malmoense., M. scrofulaceum, M. kansasii and M. fortuitum have been implicated to a
lesser extent.
Ear Infection

M. abscessus has recently been implicated in causing sporadic ear (otologic) infections
after placement of tympanotomy tubes. This infection is characterized by nonspecific
otorrhea and abundant granulation tissue which has lasted over 3 months and is
unresponsive to standard antibiotic therapy (Correa and Starke, 1996).
                                   -12-

-------
Skin & Soft Tissue Infection

Multiple species of mycobacteria have been identified as causative agents of skin and
soft tissue infections.  These include; M. marinum, M. ulcerans, M. haemophilum, M.
fortuitum, M. abscessus, M. chelonae and species within the M. avium complex.  These
infections can be either community acquired or nosocomial infections.

M. marinum is the species of Mycobacterium most commonly associated with skin
infections.  Most often, infection from this species occurs following an exposure of cut or
abraded skin to organisms present in aquariums,  pools, natural water supplies and salt
water (Feldman, 1974; Wolinsky,  1979; ATS, 1997). The incubation period can range
from two weeks to several months (Dobos et al.,  1999).  The typical outcome of infection
is the development of a localized skin lesion on the arms or legs.  Occasionally, synovial
involvement or development of subcutaneous nodules along lymph channels will occur.

M. ulcerans causes distinctive, often severe, skin lesions. It is thought that the primary
mode of infection with this species is through cuts from vegetation (e.g., grass) which
allow the organisms to enter the skin. Lesions develop as small, palpable, painless,
subcutaneous swellings approximately 4 to 10 weeks after infection.   The growing
nodule, which is firm  and attached to the skin, remains superficial and extends laterally
involving fat and fascia around muscle bundles or the muscles themselves. The skin
overlying the lesion loses pigmentation, becomes filled with fluid and necrotic and often
ulcerates. The ulceration typically has undermined edges and enlarges over many months
(Feldman, 1974). M. ulcerans has not been  found in the United States, but is well
recognized in Australia and Africa.

M. haemophilum has been observed to cause joint and skin infections in
immunocompromised patients and lymphadenitis and skin lesions in healthy children.
The lesions frequently occur on the arms or  legs  as raised violaceus nodules, which often
may become erythematous and ulcerated (Dobos et al., 1999). Recurrence of ulcers may
occur in patients who  have not undergone a  complete excision of the lesion.
The rapidly growing mycobacterial species M. abscessus, M. fortuitum and M. chelonae
are also a common cause of skin and soft tissue infections following local trauma.
 Catheter-Associated Infection

 Although infrequent, catheter-associated mycobacterial infections have most often been
 associated with long-term central venous catheters and are linked to rapidly growing
 mycobacterial species. The disease may include exit site infections, tunnel infections or
 catheter-related bacteremia (Correa and Starke, 1996). Exit site infection is characterized
 by white to green purulent drainage. Tunnel tract infection is accompanied by erythema
 and induration of the surrounding tissues.
                                    -13-

-------
The most common NTM species associated with catheter-associated infections are M.
fortuitum, M. chelonae, M. abscessus andM mucogenicum (Wallace et al., 1993).
Rarely, other species may be seen, including M avium complex. Infections may also
follow use of other types of catheters, including peritoneal catheters, ventriculostomy
tubes and nasolacrimal duct tubes.
Disseminated Infection
Disseminated NTM infection in HIV patients appears to originate from a primary
infection of either the respiratory or gastrointestinal tracts (Correa and Starke, 1996).
These infections may involve any organ, but most commonly occur in the lungs, liver,
spleen, lymph nodes or bone marrow (Correa and Starke, 1996).  Common symptoms
include prolonged fevers (often accompanied by night sweats), weight loss and
occasional abdominal pain or diarrhea. This disease is most commonly seen in patients
with less than 50 CD4 cells (ATS, 1997). The primary Mycobacterium species
associated with disseminated infections in HIV infected patients isM avium.  However,
M. kansasii, M. haemophilum and M. genavense have also been implicated.

Prior to the HIV epidemic, disseminated infection caused by MAC was rare, occurring
primarily in patients with underlying malignancy or immunodeficiency. M. avium
complex, M kansasii, M. chelonae, M. scrofulaceum, M. abscessus andM haemophilum
have  all been observed to cause disease in individuals without HIV infection.  The typical
symptom of disseminated infection withM avium complex is a fever of unknown origin,
whereas  symptoms caused by the other species consist of multiple subcutaneous nodules
or abscesses that drain spontaneously (ATS, 1997).

In immune-suppressed individuals other than those with HIV, dissemination of disease
from  a cutaneous infection is the most common form of NTM disease. These infections
are usually due toM chelonae, M. abscessus, M. haemophilum and rarely other species
such as M. kansasii.
Dose Response

An organism is pathogenic for a specific host if it can infect that host and produce signs
and symptoms of disease. Some pathogens, such asM tuberculosis, can successfully
infect and cause disease in a completely normal host in whom all defenses are intact.
Other organisms have a more limited pathogenicity and cause disease almost exclusively
in hosts with one or more defense defects. Such organisms are thus termed
"opportunistic".  Almost all of the NTM fit into this latter category.
                                   -14-

-------
       The dose-response relationship is dependent on the host's immune status, presence of
       predisposing factors (e.g., broken skin), and the species of NTM.  Therefore, a dose of
       organisms that results in infection in an immunocompromised individual, may have no
       apparent effect in a healthy host.

       At least one study has shown that inhalation of MAC is more likely to cause infection
       than oral exposure (Gangadharam et al. 1989; cited in Rusin et al., 1997). In this study,
       beige mice were exposed once via either the intranasal or oral route to MAC.  Via the
       intranasal route, lung infections were seen within one day and lymph node, liver and
       spleen infections were seen within 8 weeks. The oral exposure did not result in systemic
       infection; however, MAC organisms were isolated from spleen and lung tissue samples 6
       to 8 weeks post-exposure.

       Bermudez et al. (1992) (cited in Rusin et al., 1997) found that when beige mice were
       challenged to five oral doses of 104 CPU of M. avium given on alternate days of
       exposure, bacteremia was detected in 26.9% and mortality occurred in 11.5% of mice
       within 4 weeks.  By 8 weeks, all of the animals had disseminated  disease and as many as
       70% had bacteremia caused by MAC strains. After exposure to five doses of 108 CPU of
       MAC, bacteremia occurred in 45% of the dosed animals and mortality was seen in 25%.
Immunity
       The stages of immune response seen in mycobacterial disease are similar to those in other
       infections and consist of recognition, response and reaction. The first stage consists of
       'recognizing' the invading mycobacteria as 'foreign'. In the second stage, the host
       defense mechanisms are triggered.  The last stage represents the interaction between the
       Mycobacterium and the host.

       Cell-mediated immunity is thought to be the host immune response which is mainly
       responsible for protection against NTM infection. However, some mycobacteria can
       survive within macrophages.
Treatment

Treatment of NTM infection depends on the location and extent of disease involvement, status of
the host's immune system, and the mycobacterial species. Since there are a variety of disease
manifestations for the NTM species, a variety of treatment options exist.  A statement reviewing
treatment of NTM disease has been issued by the American Thoracic Society (ATS, 1997).

       Pulmonary Infection
                                          -15-

-------
Recent studies have shown that the macrolides clarithromycin and azithromycin have
strong activity against MAC (ATS, 1997). According to a recent statement by the
American Thoracic Society (ATS, 1997), initial therapy for this MAC-associated
pulmonary disease in adult HIV-negative patients should consist of a minimum three
drug regimen, consisting of clarithromycin or azithromycin, and rifabutin or rifampin,
and ethambutol.  On this regimen, patients should show clinical improvement within 3 to
6 months. For patients whose disease has failed to respond to the regimen, an alternative
four drug regimen, consisting of isoniazid, rifampin, ethambutol and streptomycin, has
been recommended (ATS, 1997).

In adult patients with pulmonary infection caused by M. kansasii, a three drug treatment
regimen consisting of isoniazid, rifampin and ethambutol is recommended (ATS, 1997).
In cases where no treatment is provided, there is progression of clinical and radiographic
disease.

Although there are multiple other species of NTM that are known to cause pulmonary
infection, treatment recommendations are only available for four species; M. malmoense,
M. simiae, M. szulgai andM xenopi.  Treatment durations typically last between 18 to 24
months  (ATS, 1997). MostM malmoense isolates are susceptible to ethambutol, and
many are also susceptible to rifampin and streptomycin. In most cases, favorable
responses have been seen using the alternative four-drug treatment regimen for MAC.
For patients requiring treatment of M simiae infection, initial therapy may be
implemented using the alternative four-drug treatment regimen for MAC. Based on
susceptibility tests, this regimen can be modified as needed (ATS, 1997). M. szulgai is
typically susceptible to rifampin and higher concentrations of isoniazid, streptomycin and
ethambutol (ATS, 1997).  Most patients treated with these drugs in combination respond
favorably to therapy. In vitro tests have shown that the susceptibility of M xenopi to
antituberculosis drugs is variable (ATS, 1997).  It is recommended that for initial therapy,
patients should be administered a macrolide, rifampin or rifabutin, and ethambutol with
or without initial streptomycin. If treatment fails or patients show signs of relapse,
surgery  may be a consideration.

Skin and Soft Tissue Infection

Cutaneous lesions caused by NTM may spontaneously regress over a few week period,
but the deeper lesions can persist or advance,  often requiring treatment  (Rosenzweig,
1996). Guidelines for drug therapy of disease caused by rapidly growing mycobacteria
have been developed; however, susceptibility testing of isolates is recommended prior to
beginning a treatment regimen.

For serious infection caused by M. fortuitum and M. abscessus,  an initial therapy of
amikacin combined with high-dose cefoxitin given intravenously is recommended (ATS,
1997).  Surgery may be required in cases of extensive disease, abscess  formation or
where drug therapy is difficult.

                                   -16-

-------
       Several treatment strategies have been identified for cutaneous infection caused by M.
       marinum.  These include observation, surgical excision, use of antituberculosis agents
       and use of single antibiotic agents. The American Thoracic Society (ATS, 1997) has
       indicated that acceptable treatment regimens for adults include clarithromycin,
       minocycline or doxycycline, trimethoprim-sulfamethoxazole, or rifampin plus
       ethambutol daily. Each of these regimens should be administered for a minimum of three
       months.

       Drugs for treatment of infection by M ulceram are considered ineffective.  Surgical
       removal is generally considered the treatment of choice for this disease.
       Disseminated Infection

       Treatment of disseminated M avium infection requires multi-drug therapy due to
       problems with drug resistance.  Recommendations for treatment, based on currently
       available data, advise a minimum of three drugs, one of which should be clarithromycin
       or azithromycin (ATS, 1997). Ethambutol is often used as the second agent and rifabutin
       as the third.  In patients with AIDS, adverse effects resulting from NTM treatment drugs
       may necessitate frequent changes to the therapeutic regimen.
Sensitive Subpopulations

•      In many patients with mycobacterial disease, there are predisposing factors present.
       These factors include, traumatic breaches of the skin, pre-existing pulmonary disease or
       damage, lung architectural defects, bronchiectasis and generalized congenital and
       acquired immunosuppressive disorders (such as HIV) (Grange, 1996a; Dawson, 1990).

•      NTM has been recovered frequently in patients with cystic fibrosis, particularly in the
       southeastern United States.

       Several species of NTM are generally seen only or primarily in patients with suppressed
       immune status.  These species include M genavense andM haemophilum.
Children
       It is difficult to estimate the true incidence of disease caused by NTM in children. These
       infections are typically underestimated due to a lack of mandatory reporting and the fact
       that NTM disease is seldom a cause of death in children.
                                          -17-

-------
•      In general, symptoms seen in children are similar to those presented in adults.  The
       superficial lymph nodes of the head and neck are the most common sites of clinically
       significant NTM infection in children, with onset occurring more rapidly in younger
       children.  Of the mycobacteria isolated from children under 12 years of age with
       lymphadenitis, 65% to 80% are MAC, and 10% to 20% areM scrofulaceum (Correa and
       Starke, 1996).  Pulmonary disease is relatively rare in the pediatric age group.

•      Symptoms of disseminated NTM infection in pediatric patients include recurrent fever,
       failure to thrive, neutropenia, night sweats, abdominal pain, anemia and anorexia (Correa
       and Starke, 1996).

•      Mycobacterial  lymphadenitis in children most commonly occurs as a lump in the neck of
       a child who is otherwise healthy (Colville, 1993). The involved nodes are typically high
       in the neck, just under or near the jaw. There is little or no associated pain or tenderness
       (Wolinsky, 1979). The nodes typically progress rather rapidly to softening, rupture, sinus
       formation and prolonged drainage.  The most common treatment for lymphadenitis
       caused by NTM is complete surgical excision. This treatment has a greater than 95%
       cure rate.

•      Therapy for otologic infections consists of surgical debridement, removal of foreign
       material, and long-term antimicrobial therapy determined by susceptibility testing.  This
       disease is almost invariably due toM abscessus.  Clarithromycin is the treatment drug of
       choice for this infection.

•      The primary treatment for skin and soft tissue infection in children is antimicrobial drug
       therapy, often accompanied by surgical debridement if necessary  (ATS,  1997).

V.     HEALTH EFFECTS IN ANIMALS
       Several of the NTM species have been known to cause disease in animals. These
       include: M. avium complex, M. marinum, M. ulcemns, M. paratuberculosis, M. simiae,
       M. fortuitum and M. smegmatis.

       Strains of the M avium complex have been recovered from numerous animals including
       pigs, ducks, geese, swans, and monkeys (Gangadharam and Jenkins, 1998). Most are
       presumed to beM avium. Most strains of MAC are non-pathogenic in guinea pigs and
       mice, although some have been shown to be virulent in chickens (Gangadharam and
       Jenkins, 1998). Progressive disease in laboratory mice has been seen after infection with
       several strains of M avium-intracellulare.

       The causative agent of Johne's disease isM paratuberculosis. This disease is a slow,
       progressive infection of the intestine which can result in diarrhea and wasting of the
       infected animal. The disease is usually associated with cattle, sheep and goats.

                                          -18-

-------
However, other ruminants, both domestic and wild, have been affected (Power et al.,
1993).

It is estimated that approximately 10% of dairy cattle and 33% of dairy herds in the state
of Wisconsin are infected with M. paratuberculosis (Falkinham,  1996).  Since the
principal routes of infection are from infected to uninfected cattle and from infected
mothers to their young, removal of infected animals from herds or of newborns calves
from their mothers has been successful in reducing the incidence of the disease.

Some investigators have observed that paratuberculosis in cattle may be transmitted in
utero.  In fetuses from cows with clinical signs of paratuberculosis, prevalence of fetal
paratuberculosis ranged from 21% to 37%.

Manning et al. (1998) investigated an epizootic of paratuberculosis in farmed elk. Elk
infected withM paratuberculosis have clinical signs (e.g., weight loss and diarrhea) and
histologic lesions which are similar to those in cattle.

M.  marinum is known to infect fish and serve as an important cause of morbidity,
mortality and economic loss.  Some reports indicate that the prevalence of M marinum
infection in fish may be as high as 15% (Gangadharam and  Jenkins,  1998). Clinical signs
of M marinum infection in fish include emaciation, poor growth and retarded sexual
maturation. Infection rates reportedly range from 10% to 100% when there is an infected
fish present in the population. Mycobacterial infections in fish are considered non-
treatable. It is recommended that infected stocks be destroyed and that disinfection occur
prior to restocking (Reed and Francis-Floyd,  1995).

Cases of disease from infection withM ulcerous have been reported in koala bears
inhabiting Australia (Grange, 1996a).

In 1965, M. simiae was described as a new species of mycobacteria upon its isolation
from monkeys. Data suggest that transmission occurs from animal to animal  as shown by
a case study in which 26% of healthy monkeys caged in groups with infected monkeys
developed M simiae infection over a 13 to 90 day contact period (Falkinham, 1996).

M.fortuitum andM smegmatis are known to cause mastitis in sheep and cattle and skin
and soft tissue infection in domestic house cats. M.fortuitum has also been noted to be
pathogenic for mice, producing distinctive abscesses in the kidneys and the syndrome of
spinning disease as a result of lesions in the cochlea (Wolinsky, 1979). Additionally, M
fortuitum may produce localized abscesses, but not progressive disease, when inoculated
subcutaneously or intraperitoneally in rabbits and guinea pigs (Wolinsky,  1979).
                                    -19-

-------
VI.    RISK ASSESSMENT
       Pathogen Characterization and Occurrence

       NTM species have been recovered from waste waters, surface waters and ground waters
       as well as from aerosols associated with water sources. Mycobacteria can multiply in
       water that is essentially devoid of nutrients, and they are relatively resistant to
       disinfection with many water treatment chemicals, including chlorine. Although there
       are data on the occurrence of NTM species in water sources, the data are often not
       quantitative.
       Exposure

       The Centers for Disease Control and Prevention (CDC) conducted a surveillance study of
       NTM disease from 1981 to 1983. Based on this study, the annual case rate estimates for
       non-AIDS related NTM diseases was 1.8 per 100,000 individuals, of which
       approximately 1.3 were attributable to MAC (O'Brien, 1989). In 1996, the rate of NTM
       infection in the United States population based on reported data was 7.7 per 100,000
       individuals (CDC, 1999a).

       There are three primary pathways of exposure to mycobacteria in water by which humans
       are known or suspected to become infected: ingestion, inhalation and entry through
       abraded skin. For water ingestion, USEPA (1993) recommends a value of 2 L/day be
       used as an estimate of a reasonable maximum exposure in adults.  Based on this ingestion
       rate of 2 L/day and concentrations of mycobacteria ranging from 0.01  to 5.2 CFU/mL in
       municipal water systems, adults could be expected to ingest from <20 to 10,400 CPU on
       a daily basis.  No information was located on the amount of aerosolized water which is
       inhaled daily, nor the amount of water which can potentially enter through broken or
       abraded skin, but the exposure rate is presumably much lower than for ingestion.
       Human Health Effects

       In many patients with mycobacterial disease, there are often predisposing factors present,
       including pre-existing pulmonary disease or damage, and generalized congenital and
       acquired immunosuppressive disorders (Grange, 1996a).

       The most common clinical syndromes are pulmonary infection, lymphadenitis, otologic
       infection, skin and soft tissue infection, catheter-associated infection and disseminated
       infection.

       Dose Response
                                          -20-

-------
Little is known regarding the Dose-Response relationship between mycobacterial
exposure and infection in humans by either ingestion, inhalation or dermal exposure
pathways. Some oral ingestion studies have been conducted in animal models (Rusin et
al., 1997), however, these studies are limited in that they did not test responses at a range
of doses, but instead tested the health effects after oral ingestion of a single
concentration.
Risk Characterization

NTM are opportunistic pathogens with widespread distribution in the environment but a
very low rate of infection in the general population.

Sufficient information is not available to support a quantitative characterization of the
threshold infective dose (i.e., the  dose required to produce infection) of NTM species.

Despite the lack of data on factors necessary to perform a risk assessment, there are direct
observations which can be utilized to provide a rough overview of risks in the United
States.  As mentioned above, NTM infections are not required to be reported, therefore
the following rates may be underestimated. The annual case rate estimates from 1981 to
1983 for non-AIDS related NTM diseases are 1.8 per 100,000 individuals. This is
equivalent to an annual risk of 1.8 x  10"5 for developing NTM disease as a result of
exposure to NTM organisms from all sources (including non-water sources) in the United
States.

The goal of the Surface Water Treatment Rule (SWTR) is to reduce risk to less than one
infection per year per 10,000 people (risk = 1 x  10"4) based on exposure to a
microorganism in drinking water. Assuming that case rates have not increased
dramatically over the last two decades, the population average risk of developing disease
associated with NTM organisms from any exposure source is probably below the goal
identified  by the SWTR.  Risks to individuals may be higher.

It has been estimated that in the United States, 25% to 50% of individuals with AIDS
will develop NTM diseases (Pozniak et al., 1996; Falkinham,  1996).  In 1998, the annual
rate of AIDS  in the United States was 17.1 cases per 100,000 population. However,
some states have case rates as high as 189 per 100,000 population, as seen in 1998 in the
District of Columbia. If it is expected that 25% to 50% of AIDS patients will develop
NTM disease, this would  be equivalent to a prevalence of approximately 4.3 to 8.6 per
100,000 across the United States  and 47.3 to 95 per 100,000 in the District of Columbia.
The recent use of highly active anti-retroviral therapy (HAART) has led to a decrease in
risk/rate of NTM infiction in AIDA patients.
                                    -21-

-------
VII.   METHODOLOGY

       Sampling and Recovery

•      Because the majority of mycobacterial species grow slowly, some method of
       decontamination is necessary to kill the other bacteria and fungi present in the water.  If
       this decontamination step does not occur, these other microorganisms will overgrow the
       culture medium and often cause its breakdown by proteolysis.  Acids, alkalis and
       detergents are often used during the decontamination process since mycobacteria are
       generally more resistant to these chemicals than are other bacteria (Jenkins, 1991).

•      Carson et al. (1988) found that sodium hypochlorite (0.2 ug/mL of free chlorine)
       appeared to be the most effective in reducing gram-negative bacterial populations
       stemming from background contamination while still allowing good recovery of NTM in
       test samples.

•      livanainen et al.  (1997) compared two decontamination methods for the isolation of
       mycobacteria from brook waters. The decontaminates were 0.7 mol/L NaOH followed
       by 50 g/L oxalic acid and 0.9 mol/L H2SO4 combined with 0.5  g/L cycloheximide.  The
       authors reported that in general, the NaOH-oxalic acid method resulted in lower
       contamination and higher isolation of mycobacteria than the H2SO4-cycloheximide
       method.

•      In addition to decontamination methods, recovery of mycobacteria is often dependent on
       the culture medium. Some media favor rapid-growing mycobacteria and may result in
       obscured growth of the slower growing species. One of the most common media used in
       culturing mycobacteria, the Lowenstein-Jensen medium, contains eggs,  asparagine,
       glycerol and some mineral salts.
       Detection, Quantification and Identification

       Numerous methods have been developed for the detection of mycobacteria in samples.
       Although the majority of these methods have been developed for the analysis of clinical
       specimens (e.g., blood, sputum), they can also be applied to detection in water sources.
       These methods include:
             • •     Polymerase chain reaction (PCR)
                    Radiometric methods (BACTEC)
                    GC/MS
             • •     Nucleic acid probes
             • •     Cultural and biochemical tests
       Overall,  the most common and reliable method for the detection and identification of
       mycobacteria is culture isolation.

                                         -22-

-------
       Polymerase chain reaction (PCR) techniques take a small quantity of bacterial DNA and
       enzymatically produce multiple (as many as a billion) copies of the target DNA segment.
       These segments can then be separated via gel electrophoresis. This technique can be
       used to identify, the disease-causing bacteria. One drawback of PCR techniques is that
       they can result in a positive reading even if the bacteria have been inactivated, due to the
       presence of residual genetic material  in the water source. This may lead to inappropriate
       assertions that a drinking water source is contaminated, when the detected genetic
       material may in fact be from non-viable cells (Cormican et al., 1992).

       Radiometric detection of mycobacteria, also known as BACTEC methods, is becoming
       widely used.  The BACTEC TB460 system can distinguish between M tuberculosis and
       NTM via the use of a selective growth inhibitor called NAP (ATS, 1997).  The average
       time for performing this  test is five days. Although the BACTEC  system can detect
       growth of mycobacteria; it is not quantitative. Additionally, this system can not be used
       for species identification.

       Gas chromatography-mass spectrometry (GC/MS) has been used for the detection and
       quantificiation of various mycobacterial strains in drinking water.

       Species of mycobacteria can also be identified via High Performance Liquid
       Chromatography (HPLC) analysis of species-specific mycolic acids (Butler et al., 1991,
       1992; CDC, 1999b). Most commonly reported species can be rapidly identified using
       this method.  CDC has developed a document entitled "Mycolic Acid Pattern Standards
       for HPLC Identification  of Mycobacteria" (CDC, 1999b) to serve as a resource in species
       identification.

       Nucleic acid probes are a tool  which  can be used for  rapid identification of pure cultures
       of mycobacteria (Gangadharam and Jenkins,  1998).  Probes have been developed for the
       following species:M tuberculosis complex, M avium,M. intracellulare^M. avium
       complex, M kansasii,  andM gordonae.

       In addition to the above methods, identification of mycobacteria is often performed by
       evaluating both cultural and biochemical characteristics (Grange, 1996; Gangadharam
       and Jenkins, 1998). Until more methods are developed for species identification, the
       conventional methods  based on morphology, growth rates and biochemical parameters
       will continue to be the most utilized.
VIII.  ANALYSIS AND TREATMENT
       Drinking Water Treatment Methods
                                         -23-

-------
In general, drinking water treatment methods can be separated into two modes of action:
removal and disinfection. Removal methods typically are physical rather than chemical
techniques. Disinfection is defined as the destruction or inactivation of pathogenic
microorganisms including bacteria, amoebic cysts, algae, spores and viruses
(Montgomery, 1985).

The quality of the raw or source water and the measures that are taken to improve water
quality prior to  disinfection will have a significant effect on the efficacy of any treatment
method.

Water quality factors that influence disinfection efficiency include particulates or
aggregates (suspended solids or turbidity), dissolved organic matter, inorganic
constituents, pH and temperature (Sobsey, 1989).

Physicochemical treatment methods are generally used prior to disinfection to 'clarify'
the source or raw waters for improvement of water quality. These treatment methods are
also used to improve the efficiency of disinfection (called disinfection demand) by
reducing the inorganic and organic loads present in  source water prior to drinking water
treatment. These methods include:
       • •     Sedimentation
       • •    Flocculation and coagulation
       • •    Filtration
       • •    Adsorption
Disinfection/Inactivation

Disinfection treatment methods consist of a number of different processes that are used to
destroy or inactivate pathogenic microorganisms.  These include treatment with: chlorine
(free chlorine), chloramines, chlorine dioxide, bromine, iodine, ozone and ultraviolet
radiation.

Many species of mycobacteria have proven resistant to chlorine treatment. This is of
concern, since most waters that are intended for human use are treated with chlorine. In
fact, it is estimated that 98% of municipal water suppliers in the United States use
chlorine for water treatment.
The susceptibility of mycobacterial strains to chloramines has not been fully evaluated
(Pelletier et al., 1988).

Preliminary studies have shown that M. fortuitum is more resistant to ozonation than is E.
coli.  Farooq et al. (1977) concluded that ozone residual is the controlling factor in the
inactivation of M. fortuitum.
                                    -24-

-------
       No information was located regarding the disinfection efficacy of bromine, iodine or
       hydrogen peroxide against NTM species.

       Inactivation by UV radiation is believed to act through direct absorption of UV energy by
       the microorganism, causing molecular rearrangement and disruption of unsaturated
       chemical bonds.  Studies have shown that UV radiation can be effective for disinfecting
       water contaminated with pathogenic mycobacteria (Kubin et al., 1982).
IX.    RESEARCH NEEDS

       NTM are an important cause of community- and hospital-acquired illness, and they can
be associated with morbidity and mortality when an infection is not rapidly diagnosed and
treated. In addition, NTM are widely distributed in the environment, including treated water
supplies. Additional information is needed to institute optimal prevention and control measures
and to minimize the morbidity and mortality associated with these organisms. Specific
information gaps include the following:

•      More information is required on the latency and dormancy periods of NTM diseases.
       This information will facilitate the identification of environmental sources which lead to
       disease outbreaks and may help identify species-specific transmission factors.

•      More comprehensive data on the concentration of NTM species in water sources is
       needed,  especially as it relates to potable water supplies in order to accurately estimate
       exposures to these organisms.

•      Further information is needed regarding the nature of the dose-response relationship for
       NTM species, particularly for exposures from potable water. More specifically, research
       is needed to establish the minimal infectious dose for high-risk and general populations.
       This information will help support development of acceptable levels of these organisms
       in water supplies.

•      Identification of the most effective (and most cost-effective) biocidal treatments for
       NTM species in water sources.
X.     REFERENCES

ATS (American Thoracic Society).  1997. Diagnosis and Treatment of Disease Caused by
       Nontuberculous Mycobacteria.  Am J Respir Crit Care Med v!56 pp.51-525.

Brennan, PJ. and Nikaido, H. 1995.  The envelope of mycobacteria. Annu Rev Biochem 64:29-
       63.
                                          -25-

-------
Brooks, R.W., Parker, B.C., Gruft, H. et al.  1984a. Epidemiology of infection by
       nontuberculous mycobacteria. V. Numbers in eastern United States soils and correlation
       with soil characteristics.  Am Rev Respir Dis 130 (4) : 630-633.  (As cited in
       Gangadharam and Jenkins, 1998)

Brooks, R.W., George, K.L., Parker, B.C. et al. 1984b.  Recovery and survival  of
       nontuberculous mycobacteria under various growth and decontamination conditions.
       Can J Microbiol 30 (9) :  1112-1117.

Butler, W.R., lost Jr., K.C. and Kilburn, J.O. 1991. Identification of Mycobacteria by High-
       Performance Liquid Chromatography. J Clin Microbiol 29(11):2468-2472.

Butler, W.R., Thibert, L. and Kilburn, J.O.  1992.  Identification of Mycobacterium avium
       Complex Strains and Some Similar Species by High-Performance Liquid
       Chromatography.  J Clin  Microbiol 30(10):2698-2704.

Carson, L.A., Petersen, N.J., Favero, M.S. et al.  1978.  Growth characteristics of atypical
       mycobacteria in water and their comparative resistance to disinfectants.  Appl Environ
       Microbiol 36 (6) : 839-846.

Carson, L.A., Cusick, L.B., Bland, L.A. et al.  1988. Efficacy of chemical dosing methods for
       isolating nontuberculous  mycobacteria from water supplies of dialysis centers. Appl
       Environ Microbiol  54(7) : 1756-1760.

CDC.  1999a. Nontuberculous Mycobacteria Reported to the Public Health Laboratory
       Information System by State Public  Health Laboratories United States, 1993-1996.
       Centers for Disease Control and Prevention. July 1999.

CDC.  1999b. Mycolic Acid Pattern Standards for HPLC Identification of Mycobacteria.
       February 1999.

Collins, C.H., Grange, J.M., and Yates, M.D.  1984. Mycobacteria in water.  J Appl Bacteriol
       57(2): 193-211.
Colville, A.  1993.  Retrospective review of culture-positive mycobacterial lymphadenitis cases
       in children in Nottingham, 1979-1990. Eur J Clin Microbiol Infect Dis 12 (3) :  192-5.

Cormican, M.G., Barry, T., Gannon, F. et al. 1992. Use of polymerase chain reaction for early
       identification of Mycobacterium tuberculosis in positive cultures [see comments]. J Clin
       Pathol  45 (7) : 601-604.


                                          -26-

-------
Correa, A.G. and Starke, J.R.  1996. Nontuberculous mycobacterial disease in children.  Semin
      Respir Infect  11 (4) : 262-271.

Dailloux, M. and Blech, M.F.  1992.  Occurrence of water associated mycobacteria in
      immunosuppressed patients. Aggressologie 33 (2) : 84-86.

Dawson, DJ. 1990. Infection with Mycobacterium avium complex in Australian patients with
      AIDS. Med J Aust 153 (8) : 466-8.

Dobos, K.M., Quinn, F.D., Ashford, D.A., Horsburgh, C.R., King, C.H. et al.  1999. Emergence
      of a Unique Group of Necrotizing Mycobacterial Diseases.  Emerg Infect Dis.  1991.  5
      (3) : 367-378.

du Moulin, G.C., Sherman, I.H., Hoaglin, D.C. et al. 1985. Mycobacterium avium complex, an
      emerging pathogen in Massachusetts. J Clin Microbiol 22 (1) :  9-12.

du Moulin, G.C., Stottmeier, K.D., Pelletier, P. A. et al. 1988. Concentration of Mycobacterium
      avium by hospital hot water systems. JAMA 260(11):  1599-1601.

Embil, J., Warren, P., Yakrus, M., Start, R., Corne,  S., Forrest, D., Hershfield, E. et al. 1997.
      Pulmonary illness associated with exposure  to Mycobacterium-avium  complex in hot tub
      water. Hypersensitivity pneumonitis or infection?  Chest.  Ill (3):  813-6.

Evan-Paz, Z., Haas, H., Sacks, T., et al. 1976.  Mycobacterium mariunum skin infections
      mimicking cutaneous leishmaniasis. Br J Dermatol 94 : 435-442. (As cited in Joe and
      Hall, 1995)

Falcao, D.P., Leite, C.Q., Simoes, M.J.  et al. 1993.  Microbiological quality of recreational
      waters in Araraquara, SP, Brazil.  Sci Total Environ 128 (1) : 37-49.

Falkinham, J.O.  1996. Epidemiology of infection by nontuberculous mycobacteria.  Clin
      Microbiol Rev 9 (2) :  177-215.

Farooq,  S., Chian, E.S., and Engelbrecht,  R.S.  1977. Basic concepts in disinfection with ozone.
      J Water Pollut Control Fed 49 (8) : 1818-1831.

Feldman, R.A.  1974. Primary mycobacterial skin infection: A summary.  IntJ Dermatol  13(6)
       : 353-356.
                                          -27-

-------
Flaig, W., Beatelspacher, H., and Rietz, E.  1975.  Chemical composition and physical properties
       of humic substances. In: Gieseking, I.E., ed. Soil components. Vol 1. New York:
       Springer-Verlag. 1-211.  (As cited in Kirschner et al., 1992)

Gangadharam, P.R.J. and Jenkins, P.A.  1998.  Mycobacteria. I. Basic Aspects. New York, NY:
       Chapman & Hall.

Gorse, G.J., Fairshter, R.D., Friedly, G. et al.  1983.  Nontuberculous mycobacterial disease.
       Experience in a southern California hospital. Arch Intern Med 143 (2) : 225-228. (As
       cited in Falkinham,  1996)

Grange, J.M.  1996a. Mycobacteria and Human Disease, Second eds. New York, NY:  Oxford
       University Press, Inc.

Grange, J.M.  1996b.  The biology of the genus Mycobacterium.  Soc Appl Bacteriol Symp Ser
       25 :  1S-9S.

Havelaar, A.H., Berwald, L.G., Groothuis, D.G. et al. 1985. Mycobacteria in semi-public
       swimming-pools and whirlpools.  Zentralbl Bakteriol Mikrobiol Hyg [B]  180 (5-6) :
       505-514.

Hayman, J.  1991. Postulated epidemiology of Mycobacterium ulcerans infection.  Int J
       Epidemiol 20 (4) : 1093-1098. (As cited in Veitch et al., 1997)

Horsburgh,C.R., Jr. 1991.  Mycobacterium avium complex infection in the acquired
       immunodeficiency syndrome. N Engl J Med 324 (19) :  1332-1338.  (As cited in Yajko
       etal., 1995)

livanainen, E., Martikainen, P.J., and Katila, M.L. 1997. Comparison of some decontamination
       methods and growth media for isolation of mycobacteria from northern brook waters.  J
       Appl Microbiol 82  (1) :  121-127.

Iredell, J., Whitby, M., and  Blacklock, Z. 1992. Mycobacterium marinum infection:
       epidemiology and presentation in Queensland 1971-1990.  Med J Aust 157 (9) : 596-
       598.

Iseman, M.D. 1998. The Theodore E. Woodward Award. Mycobacterium avium and slender
       women: an unrequited affair. Trans Am Clin Climatol Assoc 109 : 199-202.
                                         -28-

-------
Jacobsen, E., Gurevich, I, Schoch, P. et al.  1996. Pseudoepidemic of nontuberculous
       mycobacteria in a community hospital [letter; comment]. Infect Control Hosp Epidemiol
       17 (6) : 348.

Jenkins, P. A.  1991.  Mycobacteria in the environment. Soc Appl Bacteriol Symp Ser 20 :
       137S-141S.

Jenkins, P. A.  1981.  The epidemiology of opportunist mycobacterial infections in Wales, 1952-
       -1978.  Rev Infect Dis  3 (5) : 1021-1023. (As cited in Falkinham, 1996)

Jones, W.D., Jr. and  Kubica, G.P. 1965. Differential colonial characteristics of Mycobacteria on
       oleic acid- albumin and modified corn meal agars. II. Investigation of rapidly growing
       Mycobacteria.  Zentralbl Bakteriol [Orig. ] 196 (1) : 68-81. (Secondary reference)

Kahana, L.M., Kay,  J.M., Yakrus, M.A., Waserman, S. et al.  1997. Mycobacterium avium
       complex infection in an immunocompetent young adult related to hot tub exposure.
       Chest  111 (1): 242-5.

Katila, M.L., livanainen, E., Torkko, P. et al. 1995. Isolation of potentially pathogenic
       mycobacteria in the Finnish environment. Scand J Infect Dis Suppl 98 : 9-11.

Kaustova, J., Olsovsky, Z., Kubin, M., Zatloukal, O., Pelikan, M., and Hradil, V.  1981.
       Endemic occurrence of Mycobacterium kansasii in water-supply systems. J Hyg
       Epidemiol Microbiol Immunol 25 : 24-30.

Kirschner, R.A.J., Parker, B.C., and Falkinham, J.O.  1992. Epidemiology of infection by
       nontuberculous mycobacteria. Mycobacterium avium, Mycobacterium intracellulare, and
       Mycobacterium scrofulaceum in acid, brown-water swamps of the southeastern United
       States and their association with environmental variables. Am Rev Respir Dis 145 (2 Pt
       1) : 271-275.

Kubalek, I. and Mysak, J.  1995.  The prevalence of environmental mycobacteria in drinking
       water supply systems in Olomouc County, north Moravia, Czech Republic, in the period
       1984-1989. Cent Eur J Public Health 3 (1) : 39-41.

Kubin, M., Sedlackova, J., and Vacek, K. 1982. Ionizing radiation in the disinfection of water
       contaminated with potentially pathogenic mycobacteria. J Hyg Epidemiol Microbiol
       Immunol 26(1): 31-36.

Manning, E.J., Steinberg, H., Rossow, K. et al.  1998.  Epizootic of paratuberculosis in farmed
       elk.  J Am Vet Med Assoc 213(9):  1320-1321.


                                         -29-

-------
Marston, B.J., Diallo, M.O., Horsburgh, C.R., Jr. et al. 1995. Emergence of Buruli ulcer disease
       in the Daloa region of Cote d'lvoire.  Am J Trop Med Hyg 52 (3) : 219-224. (As cited in
       Veitchetal., 1997)

Martin, E.G., Parker, B.C., and Falkinham, J.O.  1987. Epidemiology of infection by
       nontuberculous mycobacteria. VII. Absence of mycobacteria in southeastern
       groundwaters. Am Rev Respir Dis 136 (2) : 344-348.

McSwiggan, D.A. and Collins, C.H. 1974. The isolation of M. kansasii and M. xenopi from
       water systems. Tubercle  55 (4) : 291-297.

Morbidity and Mortality Weekly Reports (MMWR).  1991.  Epidemiologic Notes and Reports,
       Mycobacterium haemophilium infections - New York City metropolitan area, 1990-1991.
       MMWR 40 (37): 636-637, 643.

Montgomery, J.H. 1985.  Water  Treatment Principles and DesignJohn Wiley & Sons Press.

Morris, R.S., Pfeiffer, D.U., and Jackson, R. 1994. The epidemiology of Mycobacterium bovis
       infections. VetMicrobiol 40(1-2): 153-177.

National Academy Science (NAS). 1977. Drinking Water and Health. National Academy of
       Sciences, Washington, D.C. 939 pp.

O'Brien, R.J. 1989.  The epidemiology of nontuberculous mycobacterial disease.  Clin Chest
       Med 10 (3): 407-418.

Oldham, P.D., Selkon, J.H. and Thomas, H.E.  1975.  Opportunistic Mycobacterial Pulmonary
       Infection and Occupational Dust Exposure: An Investigation in England and Wales.
       Tubercle 56:295-310.

Papapetropoulou, M., Tsintzou, A., and Vantarakis, A. 1997. Environmental mycobacteria in
       bottled table waters in Greece. Can J Microbiol 43 (5) : 499-502.

Pelletier, P.A., du, M.G., and  Stottmeier, K.D.  1988. Mycobacteria in public water supplies:
       comparative resistance to chlorine. Microbiol Sci 5  (5) : 147-148.

Power, S.B., Haagsma, J.,  and Smyth, D.P. 1993. Paratuberculosis in farmed red deer (Cervus
       elaphus) in Ireland. VetRec 132 (9): 213-216.

Pozniak, A.L., Uttley, A.H., and Kent, R.J. 1996. Mycobacterium avium complex in AIDS:
       who, when, where, why and how?  Soc Appl Bacteriol Symp Ser  25 : 40S-46S.

                                         -30-

-------
Reed, P. A. and Francis-Floyd, R.  1995. University of Florida Cooperative Extension Service
       Institute of Food and Agricultural Sciences. Mycobacteriosis in Fish. VM96.
       Gainesville, FL.

Reznikov, M. and Robinson, E. 1970.  Serologically identical Battey mycobacteria from sputa
       of healthy piggery workers and lesions of pigs.  Aust Vet J 46 (12) : 606-607. (As cited
       in Goslee and Wolinsky, 1976)

Rosenzweig, D.Y.  1996.  Nontuberculous mycobacterial disease in the immunocompetent adult.
       Semin Respir Infect 11 (4) : 252-261.

Rusin, P. A., Rose, J.B., Haas, C.N. et al.  1997. Risk assessment of opportunistic bacterial
       pathogens in drinking water.  Rev Environ  Contam Toxicol 152 : 57-83.

Sabater, J.F. and Zaragoza, J.M. 1993.  A simple identification system for slowly growing
       mycobacteria. II. Identification of 25  strains isolated from surface water in Valencia
       (Spain).  Acta Microbiol Hung 40 (4) : 343-349.

Saito, H., Tomioka, H., Sato, K. et al. 1994.  Mycobacterium malmoense isolated from soil.
       Microbiol Immunol 38 (4) : 313-315. (As  cited in Falkinham,  1996)

Scarlata,  G., Pellerito, A.M., Di Benedetto, M. et al.  1985.  Isolation of Mycobacteria from
       drinking water in Palermo.  Boll 1st Sieroter Milan 64 (6) : 479-482.

Schnitzer, M. 1982.  Organic matter characterization. In: Page, A.L., Miller, R.H., Keeney,
       D.R., eds. Methods of soil analysis. Part 2. 2nd ed. Madison, WI: American Society of
       Agronomy, Inc. 581-594. (As cited in Kirschner et al., 1992)

Schulze-Robbecke, R., Weber, A., and Fischeder, R. 1991. Comparison of decontamination
       methods for the isolation of mycobacteria from drinking water samples. J Microbiol
       Methods  14 : 177-183.  (As cited in Falkinham, 1996)

Schulze-Robbecke, R., Tanning, B.,  and Fischeder, R. 1992. Occurrence of mycobacteria in
       biofilm samples. Tuber Lung Dis 73 (3) :  141-144.

Schulze-Robbecke, R. and Buchholtz, K. 1992.  Heat susceptibility of aquatic mycobacteria.
       Appl Environ Microbiol 58  (6)  : 1869-1873.

Schulze-Robbecke, R., Feldmann, C., Fischeder, R. et al.  1995. Dental units: an environmental
       study of sources of potentially pathogenic mycobacteria. Tuber Lung Dis 76 (4) : 318-
       323.

                                         -31-

-------
Shelton, E.G., Flanders, W.D. and Morris, G.K. 1999. Mycobacterium sp. as a Possible Cause
       of Hypersensitivity Pneumonitis in Machine Workers. Emerg Infect Dis 5(2):270-273.

Shinnick, T.M and Good, R.C. 1994.  Mycobacterial Taxonomy.  Eur J Clin Microbiol Infect
       Dis 13(11):884-901.

Slosarek, M. Kubin, M. Pokorny, J. 1994. Water as a possible factor of transmission in
       mycobacterial infections.  Cent Eur J Public Health 2 (2) :  103-105.

Sniadack, D.H., Ostroff, S.M., Karlix, M.A. et al. 1993.  A nosocomial pseudo-outbreak of
       Mycobacterium xenopi due to a contaminated potable water supply: lessons in
       prevention. Infect Control Hosp Epidemiol  14 (11) : 636-641.

Sobsey, M.D. 1989.  Inactivation of health-related microorganisms in water by disinfection
       processes. Water Science and Technology 21 (3): 179-195.

Steadham, I.E.  1980.  High-catalase strains of Mycobacterium kansasii isolated from water in
       Texas. J Clin Microbiol  11 (5) : 496-498.

Stine, T.M., Harris, A.A., Levin, S. et al.  1987. A pseudoepidemic due to atypical mycobacteria
       in a hospital water supply. JAMA 258 (6) : 809-811.

USEPA.  1993.  U.S. Environmental Protection Agency.  Superfund's Standard Default Exposure
       Factors for the Central Tendency and Reasonable Maximum Exposure. Draft, dated
       11/04/93.  Washington, D.C.

Veitch, M.G., Johnson, P.D., Flood, P.E. et al.  1997. A large localized outbreak of
       Mycobacterium ulcerans infection on a temperate southern Australian island.  Epidemiol
       Infect 119(3): 313-318.

von Reyn, C.F., Maslow, J.N., Barber, T.W. et al.  1994.  Persistent colonisation of potable water
       as a source of Mycobacterium avium infection in AIDS [see comments]. Lancet 343
       (8906):  1137-1141.

Wallace R.J.J, Silcox, V.A.,  Tsukamura, M., Brown, B.A., Kilburn, J.O., Butler, W.R., Onyi, G.
       1993.  Clinical significance, biochemical features, and susceptibility patterns of sporadic
       isolates of the Mycobacterium chelonae-like organism.  J Clin Microbiol. 31 (21):3231-9.

Wayne, L.G. and Sramek, H.A.  1992.  Agents of newly recognized or infrequently encountered
       mycobacterial diseases. Clin Microbiol Rev 5 (1) : 1-25.
                                          -32-

-------
Wayne, L.G.  1994.  Dormancy of Mycobacterium tuberculosis and Latency of Disease. Eur J
       Clin Microbiol Infect Dis 13(11):908-914.

Wolinsky, E.  1979. Nontuberculous mycobacteria and associated diseases. Am Rev Respir Dis
       119(1): 107-59.

Wolinsky,E. and Ry near son, T.K.  1968.  Mycobacteria in soil and their relation to
       disease-associated strains.  Am Rev Respir Dis 97 (6) :  1032-1037. (As cited in
       Wolinsky,  1979)

Won Jin, B., Saito, H., and Yoshii, Z.  1984. Environmental Mycobacteria in Korea. I.
       Distribution of the Organisms.  Microbiol Immunol 28 (6) : 667-677'.

Wright, E.P.,  Collins, C.H., and Yates, M.D.  1985.  Mycobacterium xenopi and Mycobacterium
       kansasii in a hospital water supply.  J Hosp Infect  6 (2)  : 175-178.  (As cited in
       Gangadharam and Jenkins, 1998)

Yajko, D.M.,  Chin, D.P., Gonzalez, P.C. et al. 1995. Mycobacterium avium complex in water,
       food, and soil samples collected from the environment of HIV-infected individuals. J
       Acquir Immune  Defic Syndr Hum Retrovirol 9 (2) : 176-182.

Zeligman, I.  1972. Mycobacterium mariunum granuloma. Arch Dermatol 106 : 26-31. (As
       cited in Joe and Hall, 1995)
                                         -33-

-------