Focus
ANALYZING FOR EPA scrambles to refine an analytical method CRYPTOSPORIDIUM
S
ince 1993 the U.S. Environmental Protection Agency (EPA) has been under intense pressure to regulate the waterborne pathogen Cryptosporidium in drinking water. That was the year that Cryptosporidium slipped through Milwaukee's water purification system, sickened an estimated 400,000 people, and contributed to 104 deaths. Yet regulations on Cryptosporidium probably will not be finalized until late 1997 at the earliest, says Stig Regli of EPA's Office of Ground Water and Drinking Water. And, admits Regli, the delay is largely the result of problems with EPA's analytical procedure to detect Cryptosporidium and another troublesome drinking-water pathogen, Giardia lamblia. The delay has important public health
As concern increases, no method satisfactorily detects the waterborne pathogen implications. There is no medical cure for Cryptosporidium infection. Those stricken develop flulike symptoms and diarrhea that can last for one to two weeks. Moreover, the very young and the elderly along with immunocompromised or immunosuppressed individuals can develop life-
threatening cholera-like symptoms that can linger for months. In June, EPA and the Centers for Disease Control issued drinking-water guidelines for people with weakened immune systems that recommend regularly boiling water or fitting water taps with filters. At the same time, concerns over Cryptosporidium are being played out in the political arena as Congress and the Clinton administration spar over new versions of the Clean Water Act and the Safe Drinking Water Act. Members of the administration, including the president, regularly invoke the Milwaukee Cryptosporidium outbreak in speeches attacking what they see as congressional efforts to weaken these environmental laws. Others cite EPA's failure to regulate Cryptosporidium
Analytical Chemistry, December 1, 1995 731 A
Focus as evidence that the agency has squandered its resources on developing regulations and analytical methods for what they say are less risky chemical contaminants. Since 1984 EPA has documented several Cryptosporidium outbreaks in the United States, including a 1994 event linked to the deaths of 35 HIV-infected individuals in Las Vegas, NV. Each outbreak has occurred at a utility that complied with EPA's drinking-water regulations. European waters are also infected with the pathogen; more than 300 people were stricken this August in Devon, U.K. Preliminary studies suggest that Cryptosporidium is already widespread in U.S. surface waters. EPA has estimated that more than 155 million U.S. citizens are at risk for infection from their drinking water. Yet the scale of the problem is unknown because of the lack of analytical data. "Only a few water quality and public health laboratories have the capability to evaluate any microbial contaminant, and most can perform only the simplest test for fecal coliform," says microbiologist Joan Rose of the University of South Florida. Problems and disappointments Cryptosporidium survives in environmental waters as ~ 5-pm sized oocysts—spherical or egg-shaped particles surrounded by a tough protective coat. Ingestion of the oocysts starts a complex life cycle that leads to infection of the small intestine by the protozoa and eventually the excretion of a new generation of oocysts. Unfortunately, standard drinking-water disinfection practices such as chlorination are ineffective against Cryptosporidium oocysts. Instead, water utilities reduce the turbidity of their treated water by flocculation, sedimentation, and filtration as a means of eliminating the micron-sized Cryptosporidium. Last year the American Water Works Association urged its members to lower the turbidity values in their finished drinking water from EPA's regulated value of 0.5 NTU (nephelometric turbidity units) to 0.1 NTU. EPA had initially planned to issue similar values as a short-term regulation this year, but the Republican congressional victories and the backlash against federal controls halted that approach. Instead, EPA has launched a cooperative, voluntary safe drinking-
water initiative to aid utilities in lowering their turbidity values. Meanwhile, EPA has had to work hard to solve problems with its analytical method for protozoa. EPA's method (which is based in turn on an American Society for Testing and Materials [ASTM] method) collects Cryptosporidium oocysts and Giardia cysts by passing large volumes of water, as much as 100 L of raw water or 1000 L of finished water, through yarn-wound polypropylene cartridge filters. Retained particles are eluted, concentrated by centrifugation, and further separated from unwanted debris by flotation on a Percoll-sucrose cushion. The concentrated organisms are collected on a cellulose-acetate membrane and stained with monoclonal antibodies tagged with fluorescein. The labeled pathogens are viewed
EPA specified everything in the methodfrom brand-name reagents to when slides should be cleaned. under a microscope equipped for epifluorescence and differential interference contrast or Hoffman modulation optics. A trained analyst classifies the cysts and oocysts according to size, shape, and internal morphologies. These characteristics must be identified under the microscope in green fluorescing pathogens scattered among interferences such as algae, which also fluoresce, or turbidity. Even at its best, the method known as indirect fluorescent antibody (IFA) fails to determine whether the oocysts or cysts are alive or not (viable or nonviable) or differentiate among different species of Cryptosporidium. (Not all species lead to illness in humans.) Tests sponsored by ASTM with commercial laboratories in 1991 and 1992 found that the procedure leads to widely varying results. A second study run by the Erie County Water Authority and re-
732 A Analytical Chemistry, December 1, 1995
ported in 1994 revealed that 6 of the 12 participating laboratories failed to detect any oocysts, and the rest of the labs recovered an average of 2.8% of the oocysts from the spiked samples (I). These results led to the development of a more detailed procedure, specifying everything from brand names of reagents to when slides should be cleaned with a Kimwipe. In addition, workshops were instituted to train analysts to recognize oocysts. EPA also plans to institute a laboratory approval program. However, an EPA-run round-robin performance evaluation this January of the revised method involving nine "expert" commercial laboratories yielded recoveries that ranged from 0 to 130%. These results fell short of EPA's average recovery goal of 20% for Cryptosporidium in spiked samples of environmental waters. Results for Giardia proved better, with a "passing" grade of more than 50% average recovery, although precision was also poor. Recent analyses of the overall analytical procedure differ as to where most of the oocysts are lost; one report cites the clarification and centrifugation steps (2), whereas another study blames the flotation step (3). The basic message, says research microbiologist Frank Schaefer of EPA's Office of Research and Development in Cincinnati, OH, is that the more you manipulate the sample, the more likely you'll lose analyte. In response to the disappointing Cryptosporidium results, EPA microbiologists along with experts from academia and industry held a series of meetings and workshops this year, including a major meeting on May 18 to "refine" the method. According to Schaefer, participants at the meeting agreed to remove the remaining options in the IFA procedure to reduce variations among laboratories. For example, analysts will no longer have a choice on how to prepare the Percoll-sucrose cushion. Participants also agreed to test some comparisons on small variations in the method such as using well slides instead of the cellulose membrane in the microscopic analysis. With the changes in place, a second round-robin study using 12 laboratories began in September, and additional tests requiring field collection and spiking of water
ο ζ
ώ ο
>
Œ.
if)
