Room with a View: Towards Molecular Imaging of Pulmonary Vascular Remodeling

The inaccessibility of the pulmonary vasculature to non-invasive assessment or routine tissue sampling has limited our ability to diagnose pulmonary arterial hypertension (PAH) or assess histopathologic changes with or without therapy. A modality that can measure cellular events contributing to pulmonary vascular remodeling such as vascular cell proliferation, apoptosis, or inflammation could guide its diagnosis and treatment. In PAH, resistance-determining vessels of the pulmonary vasculature undergo progressive obstruction and obliteration, with aberrant proliferation of endothelial, smooth muscle, and myofibroblast lineages, marked by neo-intimal lesions, medial hypertrophy, perivascular infiltrates, and complex, multi-channeled plexiform lesions.¹ Despite currently approved therapies that act principally as vasodilators, transplant-free survival following the diagnosis of PAH remains slightly better than 50% at 5 years.² Several novel anti-proliferative have recently been investigated with the potential to modify PAH by targeting remodeling directly.^3–5 It is thought that outcomes in PAH might be improved with earlier diagnosis, by either facilitating the deployment of conventional vasodilator therapies before the loss of vasoreactivity and presumed efficacy, or by targeting remodeling before irreversible loss of vessels has occurred. A non-invasive imaging modality that reflects underlying biological events could reveal important endophenotypes within the spectrum of PAH with distinct prognoses and responses to specific agents. In PAH, Positron Emission Tomography (PET) of the chest using ¹⁸Fluorodeoxyglucose (¹⁸FDG) has been employed experimentally to reveal enhanced metabolic activity of the right ventricle (RV)⁶ and in the lungs,⁷ reflecting the hypertrophic response to pressure overload as well as a shift to a glycolytic metabolism due to relative hypoxemia and metabolic adaption, akin to the Warburg effect seen in cancer biology.⁸ The RV ¹⁸FDG PET signal has prognostic significance in PAH,^{9, 10} tracks therapeutic responses,^{6, 11} and has prompted investigation of anti-proliferative or metabolic therapies to promote more efficient oxidative phosphorylation.^{11, 12} While ¹⁸FDG PET detects metabolic adaptations and increased cellular metabolism that may correspond with hyperproliferative phenotypes, neither this nor other currently available imaging modalities can detect vascular remodeling in PAH specifically.

In the current issue, Ashek and Spruikt et al.¹³ investigate ¹⁸F-3’-fluoro-3’-deoxythymidine (¹⁸FLT) PET as a method for imaging proliferative activity in the pulmonary vasculature of experimental PH, and in patients with idiopathic PAH. ¹⁸FLT is fluorine-modified thymidine analogue that assesses cellular uptake of thymidine as a surrogate for DNA synthesis, and has been used successfully for imaging tumor growth and predicting the outcomes of anti-proliferative therapies.¹⁴ Capturing a different biological activity than ¹⁸FDG PET, in tumor imaging ¹⁸FLT appears to provide a better signal to noise ratio, with low background activity in the thorax particularly for the imaging of lung cancer.^15,16 Consistent with the previously known hyperproliferative phenotypes evident in the histology of PAH, as well as in isolated vascular endothelial cells, smooth muscle cells and fibroblasts from individuals with PAH,^17,18 enhanced uptake of ¹⁸FLT was observed within the lung fields in experimental and human PAH. The quantitation of ¹⁸FLT uptake and retention by dynamic scanning was most optimal using kinetic modeling of the rate limiting phosphorylation of thymidine (k3), and a reversible two-tissue model in which lung uptake was calculated in relationship to spillover from the pulmonary circulation blood pool. Taking advantage of two well-characterized experimental models of PH, the authors found that dynamic ¹⁸FLT scanning of rats with monocrotaline- or SU5416/hypoxia-induced PH correlated closely with the hemodynamic severity of the phenotype, as well as the degree of Ki67 and TK1 staining in lesions of diseased lungs, all of which were reduced by treatment with dichloroacetate or imatinib, two experimental therapies with known anti-proliferative effects on pulmonary vascular remodeling. Importantly, there was evidence of an elevated signal in rats as early as 1 week after treatment with monocrotaline, preceding the development of increased right ventricular systolic and pulmonary artery pressures by at least 1 week.^19,20 These exciting findings suggest that ¹⁸FLT PET imaging might provide a window for the early recognition of pulmonary vascular disease.

Interestingly, among the human subjects the specific lung uptake of ¹⁸FLT was most proportional to measures of RV performance and dilatation via cardiac magnetic resonance (MR) imaging, but not with invasive hemodynamic parameters such as mean pulmonary arterial pressure or pulmonary vascular resistance. Consistent with a hyperproliferative phenotype, enhanced thymidine kinase 1 (TK1) expression was found in the lungs of patients with idiopathic PAH, coinciding with cells staining for smooth muscle alpha actin, as well as adventitial fibroblast cells. There was considerable heterogeneity in the ¹⁸FLT uptake (k3) between different patients with IPAH, and evidence of patchy heterogeneity within the lung fields, and most evident in the subject with the highest pulmonary vascular resistance. These findings are consistent with the known heterogeneity in proliferative remodeling seen between patients with PAH, and within the lung tissues of individual patients. These results are reminiscent of patterns observed in a recent clinical study investigating single-photon emission computerized tomography (SPECT) lung imaging with ^99mTc-PulmoBind, an adrenomedullin receptor ligand probe, in which subjects with PAH exhibited increased spatial heterogeneity in lung activity versus controls,²¹ also consistent with regionalized disease activity.

The pattern of heterogeneity in these imaging approaches suggested that focal proliferative activity may be detected, potentially as a surrogate of disease activity. However, it is uncertain to what extent the ¹⁸FLT signal corresponds to the activity of vascular cells, versus proliferation of infiltrating or locally-derived inflammatory cells, all of which are associated with disease activity in PAH. The heterogeneous but significantly elevated ¹⁸FLT uptake in a small sample of IPAH patients in this report prompts several questions: How would ¹⁸FLT uptake (k3) perform across a larger group of patients with a wide range of known disease burdens, and could the intensity of ¹⁸FLT uptake be used to “stage” disease activity in patients, as might be suggested from the analysis of disease progression in the monocrotaline rat model? Conversely, would its discriminating power be diminished in severe, end-stage disease in which the pulmonary vascular tree is already severely obliterated or “pruned”? How might the anatomic pattern and intensity of ¹⁸FLT uptake differ in patients with other etiologies of World Health Organization Group 1 PAH, or Group 2 or 3 PH due to cardiac or pulmonary disease? Would comparative scans in patients with PAH associated with connective tissue disease (i.e., scleroderma, rheumatoid arthritis or lupus) have potentially higher activity due to the contribution of inflammation from their underlying rheumatologic disorder? And perhaps most critically, how might this test change in response to taking standard vasodilator versus investigational anti-proliferative, anti-inflammatory, or metabolic PAH therapies in humans? While the authors show provocative data that DCA and imatinib both diminish ¹⁸FLT uptake in rat PH models, it would be important to know how these changes compare and contrast with vasodilator therapy, or up-front combination vasodilator therapy that has shown exciting potential for impacting the natural history of disease.²² These questions are the subject of follow-up investigations, and given the widespread availability of ¹⁸FLT and PET scanning in tertiary care nuclear medicine facilities, could be rapidly incorporated into innovative clinical trial designs for novel mechanism-targeted therapies for PAH. While several functional, hemodynamic, genetic and demographic factors are known contributors to transplant-free survival in PAH, perhaps none correlate more strongly with survival than measures of RV performance. Given the strong correlation of ¹⁸FLT dynamic PET scanning with measures of RV performance, and its potential consideration as a surrogate endpoint of efficacy, it would be important to define whether or not ¹⁸FLT dynamic PET reflects survival, and if so, if its predictive value is dependent or independent of other measures of RV performance.

Given that there are currently no reliable imaging biomarkers of pulmonary vascular disease activity, application of ¹⁸FLT-PET as a pulmonary vascular imaging modality could have immediate impact, with potential roles in i) tailoring regimens of approved therapies; ii) screening for early pathogenetic changes in individuals at high risk for PAH, i.e., patients with severe liver dysfunction being considered for liver transplantation, patients with scleroderma or other high-risk associated populations, and asymptomatic individuals that are known mutation carriers or relatives of individuals with heritable PAH. Application of ¹⁸FLT-PET in the entities of borderline PAH (PA mean of 19–24 mmHg, in contrast to the current WHO PAH definition of ≥ 25 mm Hg) and exercise-associated PAH (PA mean <25 mmHg at rest, but reaching 30 mmHg only with exercise), both of which may represent early forms of PAH,^23,24 could demonstrate the potential of ¹⁸FLT-PET as a screening tool for pulmonary vascular disease even in individuals without overt PAH, while potentially discriminating against other confounding conditions. Further validation and characterization of ¹⁸FLT-PET has potential for expediting diagnosis, identifying more effective treatments, and advancing the deployment of precision medicine in PAH.