Unregulated Contaminant Monitoring Rule

How are microcystins being measured under the fourth cycle of EPA’s Unregulated Contaminant Monitoring Rule (UCMR 4)?

• As a first step, EPA Method 546 (“Determination of Total Microcystins and Nodularins in Drinking Water and Ambient Water by ADDA Enzyme-Linked Immunosorbent Assay”) is being used to measure “total microcystins” in finished (treated) drinking water under UCMR 4, as described in the December 2016 final rule.
• If “total microcystins” are detected in finished water at or above 0.3 µg/L by EPA Method 546, UCMR 4 requires that a sample also be analyzed for six microcystin “congeners” and nodularin-R using EPA Method 544 (“Determination of Microcystins and Nodularin in Drinking Water by Solid Phase Extraction and Liquid Chromatography/ Tandem Mass Spectrometry (LC/MS/MS)”).
• The purpose of performing analyses with Method 544 for UCMR 4 samples that yielded positive results by Method 546 is to understand the occurrence of the six microcystin congeners measurable by Method 544 and determine how they generally contribute and compare to the “total” microcystin measurement. This phased approach (i.e., employing the lower-cost ADDA-ELISA technique as a first step) is expected to significantly reduce the cost of monitoring as compared to analyzing every sample by Method 544.

Why is EPA using ADDA-ELISA to support UCMR 4 monitoring?

• Method performance data collected during the development of EPA Method 546 showed that quantitative microcystin results were within typical QC thresholds (for both precision and accuracy) used in drinking water methods.
• UCMR 4 specifies a minimum reporting level (MRL) of 0.3 µg/L for Method 546.

Why is a second UCMR 4 sample analyzed using Method 544 after a positive result (i.e., a result above the minimum reporting level) from a Method 546 test? Is this a confirmation sample and should this be considered standard for other monitoring programs?

• The subsequent analysis by Method 544 under UCMR 4 is not a confirmation of the initial (Method 546) result. See above for the rationale behind the phased UCMR 4 approach to microcystin analyses, and see below for additional discussion of how the two methods measure different analytes.
• Other programs (e.g., states) may very well chose to follow a different approach for monitoring beyond UCMR 4.

Comparing and Contrasting Microcystin Results by Different Methods

What are the advantages and limitations associated with EPA’s two methods (ADDA-ELISA Method 546 and LC/MS/MS Method 544) when measuring microcystins in drinking water?

• LC/MS/MS analysis provides the most selective, sensitive and accurate analysis of individual microcystin congeners. EPA Method 544 currently measures the concentration of six individual microcystin congeners (microcystin-LR, - YR, -RR, -LA, -LF, -LY).  While LC/MS/MS analysis is well suited for those six congeners (and possibly additional congeners if analytical standards are available), it does not identify and accurately quantify other microcystins that may be present in the sample; it may therefore produce lower values than measurement by ADDA-ELISA.
• The ADDA-ELISA technique has higher quantitation limits than LC/MS/MS and does not identify concentrations of individual microcystin congeners.  It can also be subject to high or low bias depending on the congeners present in the sample. However, unlike LC/MS/MS, analysis by ADDA-ELISA is designed to measure the “total microcystin” concentration. ADDA-ELISA also represents a significantly simpler, cheaper, and faster approach than LC/MS/MS and does not require laboratories equipped with sophisticated and expensive LC/MS/MS instrumentation.

Where microcystin congeners are detected by LC/MS/MS, will measurement by ADDA-ELISA always yield a positive result? Will “total microcystin” concentrations measured by ADDA-ELISA always be higher than the sum of congener concentrations measured by LC/MS/MS?

• Not necessarily. Because ELISA has a higher quantitation limit than LC/MS/MS, it is possible that water samples with very low concentrations may be quantifiable by LC/MS/MS but not by the ADDA-ELISA method.
• If the water sample only contains those congeners that can be measured by LC/MS/MS, and presuming that the concentrations are above the quantitation limits for both methods, one would generally expect the ADDA-ELISA and LC/MS/MS methods to provide comparable results (subject to the earlier points about differences in ELISA cross-reactivity for different congeners).
• If the water sample contains congeners that cannot be measured by LC/MS/MS, and presuming that the concentrations are above the method quantitation limits, one would generally expect the ADDA-ELISA result to be greater than the LC/MS/MS result.

Does EPA consider one method to be better than the other?

• EPA does not view Method 544 or 546 as better than the other. They measure different analytes and each method has advantages and limitations.  EPA recognizes that each approach may be better suited for different applications, dependent on the goals of the entity studying microcystin occurrence.  

Why isn’t there better agreement between EPA Method 544 (LC/MS/MS) and 546 (ADDA-ELISA)?

• EPA Method 544 (LC/MS/MS) and 546 (ADDA-ELISA) measure different analytes. Method 544 is an LC/MS/MS analytical technique that measures the concentration of six individual microcystin congeners (microcystin-LR, - YR, -RR, -LA, -LF, -LY) and nodularin. ADDA-ELISA measures an aggregate “total microcystin” concentration based on the affinities of antibodies to the ADDA portion of most microcystin and nodularin congeners. In short, ADDA-ELISA is detecting any congener that contains the ADDA functional group (currently associated with more than 100 verified congeners), while Method 544 only detects six microcystin congeners and nodularin-R that were commercially available as analytical standards when Method 544 was developed. The microcystins addressed in Method 544 may or may not be the dominant congeners in a particular water sample.

Method Performance

Is EPA satisfied with the method variability associated with EPA Method 546, the ADDA-ELISA technique?

• Yes. Method performance data collected during the development of EPA Method 546 showed that quantitative microcystin results were within typical QC thresholds (for both precision and accuracy) used in drinking water methods.

How do you reconcile potential quantitation bias associated with measuring non-LR microcystin congeners using an ADDA-ELISA technique that relies on calibration with the microcystin-LR standard?

• The technique associated with ADDA-ELISA (and that supports its ability to measure microcystins as a group in a simple, cost-effective manner) requires that a single congener be used to establish calibration curves. The antibodies used in ADDA-ELISA have different cross-reactivities for different microcystin congeners, such that when microcystin-LR is being used to calibrate the method (as is typically the case), there can be some variability in the measurements for the other (non-LR) microcystin congeners. Recognizing this limitation, and considering the limitations of the LC/MS/MS techniques, EPA views ADDA-ELISA as an appropriate approach for generating a “total microcystin” result for a sample with an unknown (and potentially wide) array of congeners.

Are ADDA-ELISA results as low as 0.3 µg/L reliable?

• Yes. Method performance data collected during the development of EPA Method 546 shows that results were reliable (with appropriate accuracy and precision) at 0.3 µg/L. The method underwent extensive validation by its developers as well as validation by three external labs and peer review. Participants in EPA’s UCMR 4 Laboratory Approval program have demonstrated the ability to perform analyses at this level using Method 546.

Why isn’t analysis by “MMPB” a good alternative for analyzing drinking water samples?

• “MMPB” is an LC/MS/MS technique that, like ADDA-ELISA, is designed to quantify microcystins by detecting a portion of the molecule common to most congeners. Recent research has shown that the MMPB analysis can create high bias or lead to false positives due to the presence of chlorination byproducts resulting from drinking water treatment. Rosenblum, et. al., demonstrated that oxidation (by chlorine) can result in destruction of microcystins while producing chlorinated byproducts that MMPB may erroneously identify as microcystins. Since untreated source water or ambient water should be less likely to exhibit this bias, MMPB may be an appropriate method for analyzing these waters. EPA has not published an MMPB-based analytical method, nor has the Agency fully evaluated others’ MMPB methods.

Federal Requirements for Methods and Laboratories

Does EPA have federal requirements for using particular methods or laboratories to analyze drinking water samples for microcystins?

• EPA Methods 544 and 546 are specified for the analysis of samples for UCMR 4, and those samples must be analyzed by laboratories approved by EPA.
• For monitoring that is not being performed for UCMR4, one should consider the following to ensure their monitoring programs are yielding high-quality results: establish clear data quality objectives; evaluate the degree to which those objectives can be met by different methods; and review initial demonstration of capability (IDC) data and ongoing quality control data from supporting laboratories to ensure that they are following the method (as written), meeting required quality control criteria, and generating reliable data.