Variables Influencing the the Protein C Pathway Response in Children with ALL.

The protein C pathway is believed to be a crucial regulator of coagulation. Similar to the adult population, it is of utmost importance in pediatric patients to maintain the balance between pro- and anticoagulatory signals (Petaja and Manco-Johnson 2003). Cancer patients have been shown to have cancer and cancer-therapy associated pathologies of the protein C pathway. (Nijziel, van Oerle et al. 2003; Woodley-Cook, Shin et al. 2006).

A test evaluating the response of the protein C pathway after activation of endogenous protein C has been shown to be a sensitive tool to screen for protein C deficiency and APC resistance whereas protein S deficiency is detected less reliably (Toulon, Halbmeyer et al. 2000). This is in line with findings suggesting that any pathology of this assay is related to an increased risk of VTE, independent of the pathology related to the abnormality of the test result (Toulon, Perez et al. 2007). Similarly, patients with idiopathic pregnancy loss were found to have a pathological response independent of the pathology related (Sarig, Lanir et al. 2002).

We therefore evaluated the influence of different variables on the results of the global protein C assay (Pro C Global on a BCS analyzer, Siemens, Marburg, Germany) in 431 children enrolled into the Thrombotect study. The Thrombotect study is a prospective, randomized trial evaluating unfractionated heparin vs low molecular weight heparin vs antithrombin replacement for prevention of thromboembolism in children undergoing asparaginase containing chemotherapy for acute lymphoblasic leukaemia within the BFM protocol. A multivariate linear regression analysis on results obtained before treatment was performed (MedCalc 10.0.1).

To date (as of 8/09), 720 children have been randomized in the Thrombotect trial. At the time, we had complete data for laboratory evaluation available for 431 children on day 0 of the treatment protocol.

Free protein S, IgG antibodies against cardiolipin or phosphatidlyserine, dRVVT, fibrinogen, the presence of F II G20210A, homocystein concentration, Lp(a) concentration and PAI-1 activity had no predictive value towards the protein C pathway assay result in this multivariate analysis.

In contrast, the protein C pathway activity was expectedly shown to be dependent on the presence of the F V Leiden mutation (p<0.0001), protein C activity (p<0.0001) and F VIIIc (p=0.0001).

Unexpectedly, the protein C pathway activity was also dependent on antithrombin activity (p=0.0166), F XIIc (p=0.0137) and IgM (but not IgG) antibodies against cardiolipin (p=0.0304) and phosphatidylserine (p=0.0006) in this multivariate analysis.

Importantly, the association between protein C pathway activity and the “unexpected” predictors were independent of whether or not the results were in the reference range or not (median [95% CI] antithrombin activity 0.95 [0.93-0.97], median F XIIc 0.94 [0.90-0.98], median ACL IgM 2 [2-3] U/ml, median APS IgM 2 [2-3] U/ml).

The protein C pathway is believed to be a crucial regulatory element of haemostasis. Our results from a large cohort (n = 431) of children with ALL and thus at risk for thromboembolism indicate that the protein C pathway is -expectedly (such as F V Leiden mutation, protein C activity, F VIIIc) and unexpectedly (such as antithrombin activity and the concentration of antiphospholipid IgM antibodies) - interdependent with many other variables of the coagulation system.

These data provide evidence that the protein C pathway ex vivo in children with ALL is targeted by various elements of other haemostasis pathways in a significant manner in multivariate analysis.

No relevant conflicts of interest to declare.