Calculating BeLPT Metrics

DOE-SPEC-1142-2001

7.4.6. If cell yield is low, the minimum acceptable assay is:

4 wells of each of the following: (Be concentration 1, be concentration 2, be

concentration 3, and control--16 total wells)

4 wells of PHA with at least 2 x 106 cells are required

Only an abnormal response would be acceptable in this circumstance.

8. CALCULATING BeLPT METRICS

The measurement of lymphocyte proliferation is based on beta particle counts from the

tritiated thymidine. The data is analyzed by comparing counts from BeSO4 exposed wells and

mitogen exposed wells to the counts from unexposed control wells. The laboratory must

calculate the metrics it uses to determine if a test is acceptable, and whether the result is

normal, abnormal, or borderline. Electronic copies of count data and all calculations must be

provided to the purchaser upon request. Appendix B describes the least absolute value

(LAV) method of performing the required calculations. Steps in the calculations are

described in detail with a numerical example. It also describes assumptions and possible

alternative methods of analysis. A laboratory that chooses an alternative method must include

a description of the method. The steps in the calculation based on the LAV method are

summarized here.

8.1.

For each treatment group, calculate a metric called a "stimulation index" (SI) that is a

ratio of the response in treated wells to the response in unexposed control wells (see

Appendix B.3).

8.2.

For each treatment group, calculate a metric called a standardized natural log SI

[standardized Ln(SI)] that uses the amount of well-to-well variance in the test to

estimate the standard error of the Ln(SI). The standardized Ln(SI) is obtained by

dividing the Ln(SI) by its standard error (see Appendix B.4-B.6). This standardized

statistic indicates the extent to which the Ln(SI) differs from the reference value of

zero. Large (i.e., greater than 2.5) positive values indicate a response to beryllium.

8.3.

Establish a reference data set for each serum, and determine the Ln of the maximum

SI for each BeLPT. Calculate the median (M) and standard deviation (SD) of the

Ln(maximum SI) for the reference data set (see Appendix B section B.10).

8.4.

Calculate a metric called the "standardized maximum Ln(SI)." First, find the

treatment group with the largest Ln(SI), i.e., the maximum Ln(SI). Next, subtract the

value of M for the reference data set and divide by the standard deviation, i.e.,

[maximum Ln(SI)-M]/SD. This metric compares the strongest response for each

BeLPT with the strongest responses from normal individuals in the reference data set.

9. RESULTS. The results of BeLPTs shall be reported as unacceptable, abnormal, normal, or

borderline (see Appendix B and C for illustrations.) If a test is unacceptable, it will be