The third generation tyrosine kinase inhibitor (TKI) ponatinib extends the therapeutic armamentarium for the management of nilotinib- and dasatinib-refractory chronic myeloid leukemia (CML).1 Vascular complications, especially arterial events, have been associated with nilotinib and ponatinib only and are an emerging issue when using these two TKIs in CML.2, 3, 4, 5

A 52-year-old male patient was diagnosed in June 2011 with high-risk (Hasford and Sokal score) Ph-positive CML in early chronic phase (CP) with additional trisomy 8 and monosomy 7. At diagnosis, leukocytes exceeded 400 G/l and hemoglobin was 7.9 g/dl, whereas platelets were within the normal range. The patient’s history did not show any relevant metabolic and/or vascular risk factors or comorbidities. Treatment with nilotinib (300 mg b.i.d.) was initiated within a clinical trial and the patient rapidly achieved a complete hematological response (CHR) followed by a complete cytogenetic response at month 8 (limitation of this analysis was the evaluation of 11 metaphases only). The patient never reached major molecular response, the minimum BCR-ABLIS he achieved was 7% at month 6. After 1 year of therapy BCR-ABLIS increased by 40% followed by subsequent loss of CHR with increasing leukocytes, rising up to 30 G/l (that is, treatment failure according to European LeukemiaNet (ELN)).6 BCR-ABL mutations were not detected and the patient’s compliance was not an issue. Five weeks after switching to dasatinib 100 mg/day, leukocytes remained elevated and dasatinib was subsequently escalated to 140 mg/day for 4 weeks, again without response (bone marrow smear and differential blood counts still showed CP-CML). The third generation TKI ponatinib (45 mg/day) was then initiated to further reduce leukemic burden before the planned allogeneic peripheral blood stem cell transplantation from an available 10/10 unrelated donor.

Three weeks after starting ponatinib the patient developed complete visual loss. Ponatinib was immediately stopped and radiological imaging revealed multiple bilateral infarctions in the posterior area of the middle and posterior cerebral artery territories (Figure 1a, arrows). Magnetic resonance imaging (MRI) angiography showed severly narrowed vessels in the circle of Willis bilaterally (Figure 1b). Digital substraction angiography (DSA) showed caliber changes with high-grade stenosis resembling ‘Moyamoya’ disease and suggesting vasculitis (Figure 1c). Infectious problems as well as meningeal leukemia were excluded. Despite a 3-day course of high-dose corticosteroid therapy (1 g per 24 h) together with low-dose aspirin and heparin, the patient’s condition deteriorated and had to be mechanically ventilated. Other reasons (for example, microangiopathy, posterior reversible leukencephalopathy syndrome and so on) for the vascular pathology were excluded. The patient subsequently died due to brain death including generalized brain edema and demarcation of bilateral infarctions (Figure 1d) resulting in compression of the brain stem. Postmortem examination revealed only minimal signs of generalized arteriosclerosis of the large vessels but severe brain edema leading to compression of the brain stem. Histologically, multiple both fresh and old ischemic cerebral insults were seen, without characteristics of cerebral vasculitis.

Figure 1
figure 1

Radiological changes during ponatinib-associated cerebral ischemia in a CML patient. (a) Cerebral MRI revealed multiple bilateral infarcts predominantly in the posterior area of the middle and posterior cerebral artery territories (arrows). (b) MR angiography showed severely narrowed vessels in the circle of Willis bilaterally. (c) DSA depicted pronounced caliber changes with high-grade stenosis and a ‘Moyamoya’-like pattern suggesting vasculitis. (d) Cranial computed tomography documented a generalized brain edema and demarcation of bilateral infarcts resulting in compression of the brain stem.

Full size image

Recent data from the phase II PACE trial reported serious arterial thrombotic events occurring in 8% of ponatinib-exposed patients including 2% experiencing serious cerebrovascular events.7 Moreover, the first-line EPIC trial was very recently stopped because of high rates of thrombotic events in ponatinib-exposed patients as well as the FDA asked the manufactorer to suspend marketing and sales of the drug. The exact mechanisms of TKI-induced arterial thrombotic events are still elusive, especially when classical risk factors and hallmarks of generalized arteriosclerosis are missing. In line with our case, recent data from the MDACC described several cases of non-arteriosclerotic vascular events.8 Moreover, in most cases of vascular complications described so far, clinical events more frequently occurred long after nilotinib treatment initiation. In the case presented here, the total exposure time to nilotinib and ponatinib was 13 months. This is a rather short exposure time and emphasizes the hypothesis of an eventual cross-toxicity between nilotinib and ponatinib. Thus, we are not able to clearly identify which TKI, if any, was responsible for the fatal course. However, our report supports the necessity to carefully evaluate patients receiving either nilotinib or ponatinib for vascular and metabolic changes even in the absence of cardiovascular risk factors.