VTE IN CHRONIC LYMPHOCYTIC LEUKEMIA

The etiology of VTE in CLL patients has not been thoroughly investigated, however a few studies indicated that the risk of VTE could be noticeably higher than in the background population. Some studies indicated that CLL patients had a considerable risk of second primary cancers, but it was not investigated if this was associated with the risk of VTE. In study 4,²⁴⁷ we investigated the epidemiology of VTE in a Danish, national CLL cohort diagnosed in 2008-2015. Information concerning second primary cancers was included. The VTE diagnoses were validated in a local dataset covering 10% of the total study population.

0 1 2 3 4 5 6 7 8 9

All cancers

combined Breast Prostate Colorectal Other solid tumors Hematological

cancers

IR of VTE per 1000 PY in cancer survivors and references

Reference subjects >2 years Cancer survivors > 2 years Reference subjects > 5 years Cancer survivors >5 years

VALIDITY OF THE VTE DIAGNOSES IN CLL PATIENTS

Medical journals from 469 CLL patients were systematically reviewed in the valida-tion study. Thirty-two VTEs were coded in the DNPR. During the validavalida-tion process, six additional VTEs were identified. Fifteen percent of the registered VTEs were not confirmed after review of medical journals and diagnostic tests. In five cases, VTE was classified as probable and in 28 cases, VTEs were confirmed. Fourteen were pulmo-nary embolisms and 22 deep vein thromboses, and in three of the cases, the deep vein thrombosis and pulmonary embolism were concurrent. Four (12.1%) of the validated VTEs were asymptomatic (i.e. diagnosed coincidentally because a CT scan was per-formed), three of which were registered in the DNPR. The positive predictive value of VTE diagnosis codes in DNPR was 84.4 % (95% CI, 67.2 - 97.4) while the sensitivity was 81.8% (95% CI, 64.5 - 93.0). In order to assess if a diagnosis of CLL influenced the validity of VTE diagnoses, we calculated the positive predictive value and sensi-tivity of VTEs that occurred after the CLL diagnosis. Twenty-two of the 33 validated VTEs occurred after the CLL diagnosis, the positive predictive value was 84.2% (95%

CI, 60.4 – 96.6) and the sensitivity was 76.2% (95% CI, 52.8 - 91.8).

VTE IN THE DANISH CLL PATIENTS

We followed 3609 CLL patients for a median of 2.6 years. In total, 102 of the CLL patients had a VTE diagnosis registered after the CLL diagnosis. For ten of these, however, no subsequent ambulatory or ward coding of VTE was done, and no antico-agulation treatment was prescribed. These ten events were not included in the analysis due to questionable validity of the diagnosis. A considerable proportion of the CLL patients died during follow-up (20%) and 13% were diagnosed with a second primary cancer after the CLL diagnosis.

Sixteen of the 92 VTEs were diagnosed at or after a diagnosis of second primary cancer, and six of the 92 VTEs were diagnosed before a diagnosis of second primary cancer. Fifteen of the 92 CLL patients with VTE after the CLL diagnosis had a history of VTE. Three of the 15 VTEs preceding the CLL diagnosis could be the same as the one after the diagnosis of CLL because the time interval between the first registered VTE and the second was less than 30 days.

Exposure to VTE before the CLL diagnosis was associated with the highest risk of VTE followed by exposure to second primary cancer. A second primary cancer among CLL patients exposed to VTE before the CLL confounded this association, HR of VTE in previous VTE compared with no previous VTE however remained the highest obser-ved in the study (Table 6).

Table 6. Hazard ratios of VTE by patients related factors and CLL specific markers. Modified from ^247.

In interaction analysis, we found markedly higher effect of previous VTE on the risk of VTE after the CLL diagnosis in age groups >60 - ≤70 years, >70 - ≤80 and >80 years compared with ≤ 60 years. In analysis of previous VTE stratified by age, HR for VTE after CLL was 0.83 (0.10 – 6.69) in subjects ≤65 years of age exposed to previous VTE compared with subjects free of previous VTE, while for those > 65 years the HR was 6.34 ( 3.37 – 11.91) if exposed to previous VTE (Table 7). In other words, age was not strongly related to the risk of VTE unless exposed to VTE before the CLL diagnosis.

HR, crude (95% CI) HR, model 1 (95% CI) HR, model 2 (95% CI)

Model 1: Adjusted for anti-coagulation treatment, WHO-PF, previous VTE, sex and age Model 2: Adjusted for anti-coagulation treatment, WHO-PF, previous VTE, sex age and second primary cancer

*Adjusted for previous VTE, age and sex in model 1 and previous VTE, age, sex and second primary cancer in model 2

Table 7. Age stratified analysis of the effect of previous VTE on VTE after the CLL diagnosis.

Regarding CLL prognostic markers, the risk of VTE was highest in CLL patients with β₂-microglobulin levels above 4 mg/L and unmutated IgHV genes (Figure 7). These associations were not confounded by second primary cancer. (Table 6). Trisomy 12, Del11q and del17p were also correlated to the risk of VTE, while normal FISH did not increase the risk of VTE significantly compared with Del 13, which is a favorable prognostic marker in CLL²⁴⁸ (Table 6, Figure 7).

PY IR (95% CI) HR, crude (95% CI) HR, adjusted* (95% CI)

≤ 65 years, previous VTE

No 4 257 6.3 (4.3 – 9.2) Reference Reference

Yes 71 14.0 (2.0 – 99.3) 2.03 (0.27 – 14.95) 0.97 (0.12 – 7.88)

> 65 years, previous VTE

No 6 637 7.5 (5.7 – 9.9) Reference Reference

Yes 234 59.7 (35.4 – 100.9) 6.62 (3.59 – 12.20) 6.68 (3.57 – 12.49)

*Adjusted for anti-coagulation treatment, WHO-PF, previous VTE, sex and second primary cancer.

Figure 7. Cumulative incidences of VTE according to CLL prognostic factors, modified from ²⁴⁷. Exposure to CLL treatment increased the risk of VTE (HR 2.40, 95% CI, 0.91-6.30).

Several types of chemotherapy and immunotherapy and combinations hereof were represented among the CLL patients who were diagnosed with VTE during CLL treatment.

02%4%6%8%Cumulative incidence VTE

0 1 2 3 4

Time since CLL diagnosis (years)

Second primary cancer after CLL No additional cancer β2-microglobulin> 4mg/L β2-microglobulin< 4mg/L

Binet B Binet A

IgHV unmutated IgHV mutated

The mortality of CLL patients with VTE was marginally higher than in CLL patients without VTE after the CLL diagnosis (Figure 8). WHO performance status, age, sex, CLL risk profile and second primary cancer did not influence this association (cru-de and adjusted HRs 2.20 [95% CI, 1.43 – 3.37] and 2.13 [95% CI, 1.39 – 3.27], respectively).

Figure 8. Cumulative mortality and associated 95% confidence intervals of CLL patients accor-ding to VTE after the CLL diagnosis. Modified from ²⁴⁷.

020%40%60%80%Cumulative mortality risk

0 2 4 6

Time since CLL diagnosis (years) With VTE after CLL

No VTE