Abstract
This study demonstrates that R-crizotinib, a frontline therapy for lung cancer, predisposes to and exacerbates PH in animal models. Caution and regular follow-up should be exercised in lung cancer patients treated with the compound. http://bit.ly/39s6stp
To the Editor:
Pulmonary hypertension (PH) is a life-threatening disease of multiple aetiologies. Regardless of the underlying cause, PH is characterised by vasoconstriction and progressive thickening of the pulmonary vessel wall, all of which is initiated by the loss of pulmonary artery endothelial cells (PAECs) [1]. Indeed, a large body of work has shown that damaged or apoptotic PAECs initiate the remodelling process through the release of growth, fibrogenic and pro-inflammatory factors that directly induce contraction and enhance survival and proliferation of adjacent pulmonary artery smooth muscle cells (PASMCs) and fibroblasts [1, 2]. Over the past decade, intense research efforts have been directed at deciphering how PH cells acquire their “cancer-like” properties. As a consequence, the therapeutic potential of numerous anti-neoplastic drugs has been tested in preclinical models, with some of them reaching clinical assays [3]. Considering the biphasic pattern of apoptosis that characterises the disease (i.e. PAEC apoptosis that triggers the disease is followed by an apoptosis-resistant state allowing vascular remodelling [4]), it is not surprising that some anticancer agents can both predispose to and treat pulmonary arterial hypertension (PAH). This is exemplified by studies showing that dasatinib, a second-generation tyrosine kinase inhibitor (TKI) approved for Philadelphia chromosome positive chronic myeloid leukaemia, improves established PAH in multiple animal models [5], while its administration before exposure to PH inducers exacerbates pulmonary vascular remodelling and pulmonary artery pressures; histological and haemodynamic changes not observed in rats exposed to dasatinib alone [6].
Recently, several observational studies have highlighted the development of PAH in patients with metastatic nonsmall cell lung cancer with anaplastic lymphocyte kinase (ALK) rearrangement who received ALK/cMET TKIs (including R-crizotinib, ceritinib, brigatinib and lorlatinib) [7, 8]. Since PH can be associated with multiple diseases including lung cancer [9], the question remains whether development of PAH in lung cancer patients receiving c-MET/ALK TKI represents an adverse drug event or disease spread. To clarify this point and potentially improve our understanding of the pathogenesis underlying PH development, we investigated in different animal models whether R-crizotinib (also known as Xalkori), a standard frontline therapy for c-MET-positive and ALK-rearranged lung cancer [10], exacerbates existing PH and/or predisposes to PH in well-established animal models (ethics approval #VRR-19-018).
We first explored the influence of R-crizotinib therapy on existing PH in the Sugen/Hypoxia (Su/Hx) rat model (figure 1a; protocol A). We found that one-third of R-crizotinib-treated Su/Hx rats died between weeks 4 and 5, whereas none of the Su/Hx rats receiving vehicle died. In agreement with this, treatment with R-crizotinib resulted in a significant increase in right ventricular (RV) systolic pressure (RVSP) and mean pulmonary artery pressure (mPAP) compared to the injured vehicle-treated group, as assessed by RV catheterisation in closed-chest animals (figure 1a). Although RV hypertrophy, as measured by Fulton index, and natriuretic peptide A (Nppa) and B (Nppb) transcript levels were not significantly different between groups (data not shown), stroke volume (SV) and cardiac output (CO) were more significantly declined in R-crizotinib-treated rats. Total pulmonary resistance (TPR, calculated by dividing the mPAP by the CO) was augmented and this was reflected by an increase in medial wall thickness of distal PAs (figure 1b). We next investigated whether R-crizotinib-induced adverse cardiopulmonary effects could be reproduced in a second PAH model; namely the monocrotaline (MCT) rat. In view of the high mortality rate seen in R-crizotinib-treated Su/Hx rats, the duration of R-crizotinib treatment was reduced to 1 week, initiated 2 weeks after MCT injection (figure 1a; protocol B). In comparison with vehicle-treated MCT rats, R-crizotinib treatment significantly increased RVSP and mPAP. No significant differences were observed with regard to SV, CO, TPR and vascular remodelling (figure 1a and b) nor with RV hypertrophy (data not shown), possibly due to the shorter duration of treatment.
Having demonstrated that treatment with R-crizotinib aggravates existing PH in rodent models, we next adopted a reverse reasoning and investigated whether the drug given to rats prior to exposure to a PH inducer potentiates the development of the disease by exacerbating haemodynamic and structural changes (figure 1c). Because a single dose of 60 mg·kg−1 of MCT produces severe PH that could mask the potential worsening effects of R-crizotinib pretreatment, mild PH was induced by a low dose of MCT (40 mg·kg−1). In this protocol, R-crizotinib-treated rats exhibited an exaggerated pulmonary hypertensive response compared to vehicle-pretreated animals, as demonstrated by a significant increase in RVSP and mPAP, lower SV and reduced CO (figure 1c) without any impact on the degree of RV hypertrophy (data not shown). Accordingly, TPR was significantly augmented and medial wall thickness of distal pulmonary arteries was increased approximately two-fold in rats who received R-crizotinib (figure 1c and d). It must be noticed that for each protocol, a left heart catheterisation was not performed. However, measurement of Nppa and Nppb expression in left ventricles revealed no significant changes between groups (data not shown). Finally, effects of chronic administration of R-crizotinib alone (100 mg·kg−1·day−1) for 21 consecutive days were investigated. R-Crizotinib-treated rats did not show any haemodynamic difference (i.e. RVSP, mPAP, SV, CO and TPR) when compared with vehicle-treated animals (figure 1c), indicating that the drug by itself is not sufficient to elicit PH.
Since injury-induced death of PAECs is recognised as a critical initiating event in PAH, we next investigated whether in vitro exposure of control human PAECs to a clinically relevant dose of R-crizotinib influences their survival and proliferative capacities. We first verified its capacity to inhibit basal phosphorylation of its primary target, c-Met, in cultured cells. As expected, R-crizotinib drastically diminished phosphorylation of c-Met and its downstream pro-survival signal AKT (figure 1e). As revealed by Annexin V staining, terminal transferase-mediated DNA end labelling (TUNEL) assay and immunoblot for cleaved Caspase-3, apoptosis of PAECs was markedly increased upon exposure to R-crizotinib (figure 1e and f). In agreement with this, R-crizotinib elicited anti-proliferative effects, as illustrated by a significant diminution in the proportion of Ki67-positive cells and reduced expression levels of proliferating cell nuclear antigen (figure 1e and f). Surprisingly, exposure to R-crizotinib was associated with an increase in the number of large, flat PAECs containing multiple nuclei or aberrant nuclei clusters (figure 1f), a feature usually seen in cells that undergo a form of cell death called mitotic catastrophe induced by ionising radiation and certain anticancer drugs. Accordingly, expression levels of the mitotic regulator Polo-like kinase 1 (PLK1) was nearly abolished in PAECs treated with R-crizotinib (figure 1e).
Interestingly, HGF/c-MET signalling was documented to elicit pro-survival effects on endothelial cells [11] and activation of cMET signalling, via supplementation in HGF 2 weeks after MCT injection, was shown to improve vascular remodelling [12]. These data may explain in part why R-crizotinib both predisposes to and aggravates established PH. Whether the deleterious impact of R-crizotinib on PAECs is mediated by on- or off-target effects remains to be explored. Several studies have evidenced that R-crizotinib elicits anti-tumour activity via off-target effects [13, 14], including the inhibition of multiple kinases, such as Src (figure 1e), Lck, Abl and Yes, all critically involved in PAH development [15]. This suggests that the adverse outcome induced by R-crizotinib on the pulmonary vasculature is likely the consequence of its cumulative effects on multiple targets.
In conclusion, our study shows for the first time that the anticancer agent R-crizotinib may cause endothelial cell injury, and by doing so, amplify the response to well-established PH inducers. Although it remains unclear whether a similar relationship between R-crizotinib-induced endothelial cell injury and PH development in lung cancer patients exists, further study of the affected signalling pathways may provide important information into the pathophysiology of PAH, and potentially new targets to combat this serious condition. In addition, our findings suggest that clinicians should consider further evaluation for PH in lung cancer patients treated with R-crizotinib who develop worsening dyspnoea or heart failure symptoms.
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-03271-2020.Shareable
Footnotes
Conflict of interest: C. Awada has nothing to disclose.
Conflict of interest: Y. Grobs has nothing to disclose.
Conflict of interest: W-H. Wu has nothing to disclose.
Conflict of interest: K. Habbout has nothing to disclose.
Conflict of interest: C. Romanet has nothing to disclose.
Conflict of interest: S. Breuils-Bonnet has nothing to disclose.
Conflict of interest: E. Tremblay has nothing to disclose.
Conflict of interest: S. Martineau has nothing to disclose.
Conflict of interest: R. Paulin has nothing to disclose.
Conflict of interest: S. Bonnet has nothing to disclose.
Conflict of interest: S. Provencher has nothing to disclose.
Conflict of interest: F. Potus has nothing to disclose.
Conflict of interest: O. Boucherat has nothing to disclose.
- Received August 25, 2020.
- Accepted January 2, 2021.
- Copyright ©The authors 2021. For reproduction rights and permissions contact permissions{at}ersnet.org