Abstract

The purpose of lupus anticoagulant (LA) testing is to predict increased tendency for thrombosis or fetal loss or to explain prolonged PTTs found on routine testing. Detection rate varies between laboratories with poor interlaboratory agreement when samples are exchanged. This reflects a lack of standardization of the tests and of the reagents. Few labs adhere to the recommendations on LA testing of the Scientific Subcommittee (SSC) of the International Society of Thrombosis and Haemostasis (ISTH) (

Brandt et al.
Thromb Haemost
1995
;
74
:
465
) which are: prolonged phospholipid dependent coagulation test, failure to correct the prolonged test by mixing with normal platelet poor plasma, shortening or correction of the prolonged test by addition of excess phospholipid, and exclusion of other coagulopathies as appropriate. For example, 60% of laboratories do not confirm that an abnormal mixing test is shortened by excess phospholipid (46th Annual SSC meeting, ISTH, 2000).

To evaluate the effect of complexity of testing on diagnosis, we prospectively studied 311 patient samples sent to the laboratory for LA testing over a 3 month period. The following tests are routinely performed in this lab: PTT with mixing study when prolonged using a moderately sensitive reagent (PTT-A, Diagnostica Stago), and dRVVT with mixing study when prolonged (Gradipore). Based on the results in Table 1, samples in groups 1– 4 would be considered negative for LA and those in group 5 a-d, positive.

Table 1
GroupNumberPTT-APTT mixdRVVTdRVVT mix
N=normal, N/A=not applicable, ↑ =increased 
182 N/A N/A 
23 ↑ N/A 
18 ↑ ↑ 
24 N/A ↑ 
5a ↑ N/A ↑ ↑ 
5b 16 ↑ ↑ N/A 
5c 13 ↑ ↑ ↑ 
5d 27 ↑ ↑ ↑ ↑ 
GroupNumberPTT-APTT mixdRVVTdRVVT mix
N=normal, N/A=not applicable, ↑ =increased 
182 N/A N/A 
23 ↑ N/A 
18 ↑ ↑ 
24 N/A ↑ 
5a ↑ N/A ↑ ↑ 
5b 16 ↑ ↑ N/A 
5c 13 ↑ ↑ ↑ 
5d 27 ↑ ↑ ↑ ↑ 

The following additional study was done: PTT with STACLOT-LA (Diagnostica Stago) using reagent 1 (lupus sensitive). If this PTT was prolonged, a mix was performed. If this mix was prolonged, the PTT was repeated with reagent 2 (lupus insensitive). Shortening of the PTT by reagent 2 (8 or more seconds) was considered confirmatory and the test positive. In group 1, one percent was weakly positive. In groups 2 to 4 fourteen percent was positive (11% strongly), and in groups 5a to 5d 79% was positive. In samples from groups 5a and 5d (where dRVVT was prolonged) RVVT was performed with excess phospholipid. The test was confirmatory according to the manufacturer’s criteria in 75% and 79%, respectively.

We conclude that when the PTT with a moderately sensitive reagent and the dRVVT are both normal (group1) these could be considered negative and no further testing is necessary (58% in this study). This is probably also true of samples in which both PTT mix and dRVVT mix are prolonged (5d, 9% in this study). These could be considered positive. This would significantly reduce the cost of testing; for example, the reagent costs alone for STACLOT-LA in this laboratory would increase the price of testing approximately 5-fold. In addition, the STACLOT-LA is cumbersome and labor-intensive. In the future, if more specific tests such as β2-glycoprotein1 antibody detection prove to be accurate predictors of thrombosis, coagulation tests for LA could be greatly simplified.

Author notes

Corresponding author