Prophylaxis with enoxaparin and antithrombin III in drug-induced coagulation alterations in childhood leukemia: a retrospective experience of 20 years

The retrospectively evaluated principal cohort included all children and adolescents (age > 1 year to < 18 years) with newly diagnosed ALL at our institution (Department of Pediatrics I, Medical University of Innsbruck) between June 2001 and July 2021, included in the randomized controlled trials ALL-BFM 2000, AIEOP-BFM ALL 2009 or 2017. Originally, 97 patients from the ALL-BFM 2000, 94 patients from the AIEOP-BFM ALL 2009, and 26 patients from the AIEOP-BFM ALL 2017 study were evaluated. However, due to lack of data sets or lack of comparability in the case of randomization with increased L-asparaginase administration (ASP +), 25 patients from ALL-BFM 2000, 16 patients from AIEOP-BFM ALL 2009, and three patients from AIEOP-BFM ALL 2017 were excluded. Thus, the effective study cohort included 173 patients, 74 (42.8%) of them female and 99 (57.2%) male. Mean age at diagnosis was 6.66 years (± 4.49 standard deviation; range 0.22 – 17.92 years; Table 1). All patients received chemotherapy according to the AIEOP-BFM ALL study protocols (2000, 2009 or 2017), including prednisone 60 mg/m²/d (7 days prephase, 21 days normal dose, tapering; AIEOP-BFM 2000: randomization prednisone versus dexamethasone), vincristine 1.5 mg/m²/d (days 8, 15, 22, 29), daunorubicin 30 mg/m²/d (days 8, 15, 22, 29), E.coli asparaginase 5000 IU/m²/d for ALL-BFM 2000 (8 times during induction from day 12) or PEG L-asparaginase 2500 IU/m²/d for AIEOP-BFM ALL 2009/2017 (days 12 and 26). Prior to chemotherapy, a central catheter (Broviac/Hickman) was implanted in the subclavian vein of each patient. Relevant for this study is the comparable induction phase in all three studies with i.v. administration of asparaginase (eight times E. coli asparaginase for ALL-BFM 2000 and twice PEG L-asparaginase for AIEOP-BFM ALL 2009/2017) and identical cortisone therapy. All patients included in this retrospective evaluation received daily prophylaxis with enoxaparin subcutaneously from the day of first asparaginase administration (dose 1 mg/kg) to normalization of coagulation parameters. Antithrombin III (AT III) was additionally substituted intravenously below a limit of 50%. This prophylactic combination therapy with enoxaparin and AT III was discontinued for platelets below 30G/L, but was otherwise administered until the coagulation values normalized (especially AT III levels above 50% without substitution and fibrinogen stable above 100 mg/dL). The diagnosis of deep vein thrombosis was made by ultrasound in all cases, except in the patient who developed sinus vein thrombosis, where the diagnosis was made using computed tomography.

All our patients undergo a basic assessment of coagulation parameters at the start of intensive therapy.

Routine blood specimens for coagulation parameters were drawn at least twice a week during induction therapy (first 40 days). If abnormalities such as a reduced AT III level were detected, blood specimens were drawn daily to ensure timely substitution. AT III was determined using a chromogenic substrate from citrated plasma in our central laboratory (Innsbruck Medical University Hospital). Substitution of AT III was performed below a level of 50% according to the following formula: required units = body weight [kg] x (100 – current antithrombin activity [%]) × 2/3.

Study data were analyzed with descriptive and inferential statistics using Python as programming language along with several special Python modules (pandas, numpy, scipy). Statistical differences of mean values were calculated by means of customary T tests – with preceding F tests, as usually required. Resulting p values and confidence intervals are based on a confidence level of 95%. Consequently, p values below α = 0.05 are deemed statistically significant. Confidence intervals of regression slopes or mean value differences that are completely located in the positive or negative range are considered statistically significant as well as corresponding with p values below 0.05.

The Ethics Committee of the Medical University of Innsbruck approved retrospective evaluation (EC No. 1478/2020). All data were obtained from medical records. This study was performed in accordance with the Declaration of Helsinki.

As has already been shown several times in the literature, we, too, can demonstrate a significant drop in the relevant coagulation parameters during therapy with asparaginase and cortisone in our ALL cohort. In the process, fibrinogen dropped below a level of 100 mg/dL at day 13 of induction therapy (Fig. 1B; regression curve based on the mean values of the measured fibrinogen). AT III was continuously substituted in our patient cohort for values below 50%. Therefore, the drop observed here is not as sharp as described in the literature and the minimum mean values in our figure settle at about 50% (Fig. 1A).

Prophylaxis with enoxaparin was performed in 167 of the 173 patients in the ALL cohort examined, i.e. the compliance for this supportive measure was very high as compared to other studies that administered this therapy [8] (96.5%; Table 2). The 6/173 (3.5%) patients who did not receive enoxaparin are all in the ALL 2000 study, which could possibly be due to the moderate experience with this prophylaxis at that time and the lack of staff training. Substitution of AT III was performed in 122 (70.5%) of the 173 patients; only 51 of the 173 patients did not receive any AT III (values always measured above 50%, 29.5%; Table 2). In the 122 patients substituted with AT III, we were able to evaluate a total of 173 AT III doses. The frequency of doses showed a mean value of 2.06 doses per patient during the induction phase of therapy (median 2.0; standard deviation ± 2.0; range 0—9 doses). There was no significant difference between the sexes or the frequency of AT III administration (p = 0.076; mean doses female/male: 1.76/2.28; range female/male: 0–6 versus 0–9). Duration of the required AT III substitutions was on average 7.84 days after start of AT III substitution (range 0–45 days). Again, there was no significant difference in gender (p = 0.248; mean value female/male: 6.9/8.56 days). In terms of age, we found a high statistical significance regarding the need for AT III substitution (total population p < 0.0001; female p < 0.0001; male p = 0.003). This is also demonstrated in Table 2: In the age group of patients < 10 years of age, 65.2% required AT III substitution, whereas in the age group > 10 years of age 91.4% of patients needed AT III administration, calculated for the total population (Table 2). Likewise, the duration of the necessary AT III substitution in female patients was statistically significantly dependent on age (total population p = 0.064; female p = 0.022; male p = 0.567). That means adolescent patients have a greater need for AT III substitution, and female adolescent patients need these substitutions for a longer time.

We analyzed the incidence of both mild thromboembolic complications (e.g. blockage of the central venous catheter during induction therapy) and severe deep thrombosis and came to the conclusion that the incidence under prophylaxis with enoxaparin and AT III is significantly lower than in the described literature:

Blockade of central venous catheter during induction therapy (first 50 days) was seen in 14 of the 173 patients (8.1%; Table 2) who needed lysis. Deep vein thrombosis was observed in only five of the 173 patients (2.9%) in our ALL cohort (Table 2; 3/72 patients (4.2%) in the ALL-BFM 2000 study, 1/78 patients (1.3%) in the AIEOP-BFM ALL 2009 study, and 1/23 patients (4.3%) in the AIEOP-BFM ALL 2017 study), compared to the for the most part much higher prevalence reported in the literature. The patients who had deep vein thrombosis in our collective were on average 9.13 years old (range 3.27–17.9 years; Table 3). Interestingly, four (80%) out of the five patients with deep thrombosis were male. Risk factors for the development of thrombotic complications, such as hypertriglyceridemia, hyperleukocytosis, sepsis or adipositas, are not relevant here. Patient #4 in Table 3 had a heterozygous mutation of Factor II (G20210A) and a homozygous MTHFR mutation (C677T) as risk factors for developing thrombosis. Interestingly, only one of the patients with deep vein thrombosis did not require AT III administration, suggesting that there were no severe thromboses in the group of patients who did not require AT III administration.