To the Editor:

Pulmonary arterial hypertension (PAH) is an incurable disease characterized by vascular obliteration and luminal narrowing (1). Current vasodilator therapies have no proven effect on the underlying molecular and cellular changes in the PAH lung, which include the emergence of apoptosis-resistant cell populations (2, 3) and proliferation of endothelial cells (4).

We recently showed that 6-mercaptopurine (MP, Purinethol) augmented BMP (bone morphogenetic protein) signaling and reversed abnormal vascular remodeling and right ventricle (RV) hypertrophy in the Sugen-hypoxia rat model of PAH (5). MP has been used in leukemia and inflammatory bowel disease for decades with an acceptable and manageable toxicity profile (6, 7). MP is an immunosuppressant that reduces inflammation and proliferation of vascular cells, which is one of the pathophysiological hallmarks of PAH (4). Here, we report the results of an open-label, proof-of-concept, single-center study of MP in patients with PAH (EudraCT 2017-000137-31, Dutch Medical Ethical Committee 2017.059). The primary aim was to evaluate the efficacy and safety of MP treatment in patients with PAH, using a dose of 1.2–1.5 mg/kg (80–100%) once daily for 16 weeks, rounded in tablets of 25 mg. The patient population consisted of idiopathic, hereditary, or drug-induced PAH; New York Heart Association (NYHA) functional class II, III, or IV; and stable on current PAH medication (no dose adjustments in PAH-specific therapy for ≥3 mo or changes in diuretics for 30 d). Eligible patients were required to show precapillary pulmonary hypertension with mean pulmonary artery pressures (mPAPs) ≥25 mm Hg, pulmonary artery wedge pressure 915 mm Hg, and, to enrich the population under study, a pulmonary vascular resistance (PVR) of ≥480 dyn · s⁻¹ · cm⁻⁵. Thiopurine methyltransferase activity was tested before enrollment to prevent MP toxicity. Safety assessments included recording of all adverse events and blood sampling at Weeks 1, 2, 4, 6, 8, 12, and 16. The primary study endpoint was a change in PVR, and secondary study endpoints included changes in mPAP and cardiac index, changes in RV volumes measured by cardiac magnetic resonance imaging, NT-proBNP (N-terminal brain natriuretic peptide) levels, 6-minute-walk distance, NYHA functional class, and quality-of-life scores. With a statistical power of 80% (α = 0.05), we expected to detect a clinically relevant decrease in PVR of >240 dyn · s⁻¹ · cm⁻⁵ (SD, 400 dyn · s⁻¹ · cm⁻⁵). With an expected responder rate of 50% (8) and a dropout rate of 15%, we estimated the sample size at 50 patients.

Out of 77 patients who were eligible for inclusion in our center, 60 patients (78%) declined to participate, mainly because of high study burden (time, need to travel and undergo procedures, and fear of side effects). A total of 15 patients were enrolled into the study from 2017 to 2019. The mean PVR was 878 dyn · s⁻¹ · cm⁻⁵ and was significantly decreased after treatment (P = 0.0273 intention-to-treat analysis and P = 0.0044 per-protocol analysis in Figure 1). This decrease in PVR remained significant after correcting for confounding effects of Hb (P = 0.0071). mPAP was significantly decreased in the per-protocol analysis (−5 mm Hg, P = 0.04) but not in intention-to-treat analysis. Although this decrease in PVR was significant, the magnitude of reduction was relatively small, leading to no improvement in cardiac index, RV function, NYHA functional class, 6-minute-walk distance, or NT-proBNP concentrations.

Figure 1. Intention-to-treat analysis on all patients who started therapy in white, and per-protocol analysis on patients who finished 16-week treatment in blue. Data are presented as mean (SD). The P values were determined using a paired t test. Significant P values are depicted in red. *Nonparametric data are depicted as median [interquartile range]; P values were determined using Wilcoxon nonparametric test. ^†Generalized estimation equation analysis of PVR corrected for decrease in Hb. BMPR2 = bone morphogenetic protein receptor type II; DIPAH = drug-induced PAH; HPAH = hereditary PAH; IPAH = idiopathic PAH; NT-proBNP = N-terminal brain natriuretic peptide; NYHA = New York Heart Association; PAH = pulmonary arterial hypertension; TBX4 = T-box transcription factor.

Download Figure | Download Powerpoint

[More] [Minimize]

Although the dosage of MP used in this study is generally well tolerated by patients with inflammatory bowel disease (6), it appeared too burdensome for this PAH population. Two patients stopped the study after 2 (data not shown) and 13 weeks because of nausea, vomiting, and fatigue. Two other patients were excluded after 6 and 8 weeks because of migraine and myelosuppression. Hb and leukocyte counts were decreased in almost all patients (Hb, −0.81 g/dl, P = 0.0399; leukocytes, −2.7 × 10⁹/L, P = 0.0001). Leucopenia was higher than expected (5–25% reported [6] vs. 40% in our study), and quality-of-life scores were significantly worse. Eight out of 15 patients who finished the protocol (53%) received dose reductions, indicating poor tolerability. Interestingly, none of the patients showed liver toxicity. All side effects were dose dependent and resolved after dose adjustments or treatment discontinuation. Low thiopurine methyltransferase enzymatic activity, the best-known and most frequently studied risk factor for thiopurine-induced leukopenia (6), was not present in our patient cohort.

In accordance with our preclinical study (5), this proof-of-concept study provides first evidence that MP decreases PVR coinciding with an increased BMPR2 (bone morphogenetic protein receptor type II) mRNA expression in peripheral blood mononuclear cells of 10/11 patients after MP treatment (P < 0.006 in Figure 1 and BMPR2 expression per patient in Figure 2). Impaired BMP signaling is observed in both hereditary PAH and idiopathic PAH lungs, and pulmonary vascular remodeling can be attenuated by enhancing BMPR2 activity (9). Therefore, targeting the BMPR2 pathway has repeatedly emerged as a novel treatment strategy for PAH (10), and improvement of BMPR2 expression in circulating peripheral blood mononuclear cells could have contributed to the decrease in PVR.

Figure 2. Percentage difference of pulmonary vascular resistance and mean pulmonary artery pressure depicted per patient. Patients with TBX4 (T-box transcription factor) mutations are depicted in green. Patients with a BMPR2 (bone morphogenetic protein receptor type II) mutation are depicted in blue. Patients with idiopathic and drug-induced pulmonary arterial hypertension are depicted in red. Diamonds indicate patients that prematurely ended the study. The right panel depicts the used dosage per patient during the time course of the study. The starting dose was 80–100% (1.2–1.5 mg/kg rounded to tablets of 25 mg), and light color means dose reduction. PAH = pulmonary arterial hypertension.

Download Figure | Download Powerpoint

[More] [Minimize]

In addition to the small inclusion capacity of a single center, this study suffered from patient reluctance to participate after sharing of negative experiences by study participants in the PAH patient community (e.g., social media and patient supportive groups). High toxicity rates together with declining inclusion led us to prematurely end the study. We did not measure thiopurine end‐products 6-thioguanine nucleotides or 6-methylmercaptopurine, which are known to generate the cytotoxic and apoptotic effects of thiopurines and give rise to an increased risk for thiopurine-induced leucopenia. Literature shows that alternative thiopurine analogs, such as azathiopurine, thioguanine (Thiosix), or a (metabolite-sensed) step-up dosing, may result in less toxicity (6, 7). This is also shown by the fact that patients who started with 1.2 mg/kg (80%) all completed the study because of lower side effects (Figure 2). As myelosuppression is dose dependent, we hypothesize that a lower starting dosage in a step-up scheme increases the tolerability of MP treatment (6) but at the same time still effectively reduces pulmonary pressures.

The observed decrease in PVR confirms that targeting excessive vascular remodeling by inhibiting cell proliferation and inducing apoptosis with MP is of therapeutic interest. As frequency and severity of side effects were higher than reported and expected in the current design, we conclude that antiproliferative therapy with MP as add-on treatment for hereditary and idiopathic PAH has an unfavorable risk/benefit ratio. However, improvements in dosing schemes and/or use of other thiopurine analogs deserve further study.

References

Section:

ChooseTop of pageReferences <<CITING ARTICLES