Riociguat for the Treatment of Raynaud’s Phenomenon: A Single-Dose, Double-Blind, Randomized, Placebo-Controlled Cross-Over Pilot Study (DIGIT)

DIGIT was a single-dose, double-blind, randomized, placebo-controlled, cross-over, exploratory study (Supplementary Fig. 1). Patients aged 18–70 years were eligible for inclusion if they had primary RP or secondary RP associated with limited cutaneous SSc (lcSSc), diffuse cutaneous SSc (dcSSc), undifferentiated SSc, or SSc overlap syndrome for ≥ 1 year prior to screening. Exclusion criteria consisted of co-medication with vasodilators including PDE5i, endothelin receptor antagonists, CCBs, prostacyclin analogs, or nitrates. All vasodilator medications had to be discontinued at least 1 week prior to enrollment until 24 h after last study drug intake. Other exclusion criteria were: systolic blood pressure < 105 or > 160 mmHg at rest; diastolic blood pressure < 50 or > 95 mmHg at rest; any condition that may have interfered with normal pharmacokinetics of riociguat; smoking; abnormal electrocardiographic or heart rate findings; or clinically relevant findings on physical examination.

Eligible patients were randomized to receive a single oral dose of either riociguat (Adempas^®, Bayer AG, Berlin, Germany) 2 mg or matching placebo, with cross-over separated by 7 ± 3 days before receiving the other respective study drug. Some studies of PDE5i have reported prolonged effects on vascular function, which exceed their plasma half-life. The separation period ensured adequate washout and allowed time for any long-lasting vascular effects of riociguat to fade between treatments [15, 17, 21].

Perfusion of the right index finger was measured by laser-speckle contrast analysis (PeriCam PSI Systems, Perimed, Järfälla, Sweden) at baseline and 2 h post-dose at room temperature (RT) (20–24 °C for ≥ 20 min). Perfusion was also measured immediately after cold water exposure (CE) (both hands at 12 °C for 5 min) at baseline and at 2 h post-dose. After CE, the average of the 5 different time points (mean of 2 measurements each at 0, 3, 5, 7, and 15 min) of the distal phalanx of the right index finger was calculated. In the case of a contraindication to using the right index finger (2 patients), measurements were performed on the left index finger. At 2 h post-dose, venous blood samples were collected, preferably from the cubital fossa of the left arm, for the measurement of riociguat plasma concentration at the time of digital blood flow measurement.

The efficacy of riociguat was assessed using placebo-corrected change from baseline in digital blood flow 2 h post-dose at RT and following CE. The pre-specified responder criterion was a placebo-corrected increase in digital blood flow by ≥ 10% versus baseline at 2 h post-dose at RT or after CE. The responder cut-off was based on a previous study in patients with RP, which showed a digital blood flow increase of approximately 10% 1 h after a single dose of the PDE5i vardenafil at RT or after CE [17]. Adverse events (AEs) were recorded for at least 7 days, with active monitoring for serious AEs (SAEs) for a further 23 days. All AEs were assessed and documented by the investigator in terms of seriousness, intensity (mild, moderate, or severe), and causal relationship to administration of treatment and protocol-required procedure(s). SAEs were defined as any AEs that resulted in death or were life threatening, or that required inpatient hospitalization or prolongation of hospitalization. An AE was also considered serious if it resulted in persistent or significant disability/incapacity (disability meaning a substantial disruption of the patient’s ability to conduct normal life functions); if it was a congenital abnormality or birth defect; or if it was another medically important serious event as judged by the investigator. Blood and urine samples were collected at screening and follow-up for determination of hematologic parameters (leukocytes, erythrocytes, hemoglobin, hematocrit, and platelets) and clinical chemistry (aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, creatine kinase, urea, sodium, and potassium).

Riociguat concentration was determined in plasma after protein precipitation with methanol containing an internal standard followed by separation using high-pressure liquid chromatography and tandem mass spectrometric detection. The calibration range was 2.00 [lower limit of quantification (LLOQ)] to 500 μg/L (upper limit of quantification). Mean inter-assay accuracy of back-calculated concentrations (except LLOQ) in calibrators was 96.8–104% and precision ≤ 8.63%. Accuracy and precision at the LLOQ were 99.8 and 4.63%, respectively. Quality control samples in the concentration range 6.00–400 μg/L were determined with an accuracy of 99–102% and a precision of 2.74–5.59%. All samples were stored at or below − 15 °C and analyzed within 9 months after sampling.

Based on previous experience [17], the chosen sample size of 20 valid patients measured in a cross-over design was considered sufficient to fulfill the objectives of the study. No formal statistical sample size estimation was performed in this exploratory study. The number of digital blood flow responders at RT and after CE, including subgroup analyses according to medical grouping, was calculated. Summary statistics for digital blood flow and percentage change from baseline were calculated by response. The percentage change from baseline of digital blood flow was analyzed by an explorative analysis of variance. Point estimates and 2-sided 90% confidence intervals (CIs) for the difference “riociguat–placebo” were calculated. Individual subject profile plots for placebo-corrected change in digital blood flow were used for visualization.

Of the 23 randomized patients, 20 completed the study without major protocol deviations and entered the pharmacodynamic analysis group (Supplementary Fig. 2). All 23 patients who received study drug were included in the safety analysis group, and 22 were included in the pharmacokinetic analysis group. The baseline characteristics of the 20 patients who completed the study without major protocol deviations are presented in Table 1. Most patients (65%) were treatment naïve, and of the previously treated patients (n = 7), 5 had received CCBs, which were discontinued at least 1 week prior to enrollment.

In the 20 patients who completed the study, the placebo-corrected change in digital blood flow at RT 2 h post-dose was + 46% (90% CI − 6 to + 98) (Fig. 1a; Table 2), as riociguat increased mean [standard deviation (SD)] digital blood flow by + 41% (109) compared with −5% (59) after placebo. Following CE, the placebo-corrected change from baseline in digital blood flow at 2 h post-dose was − 9% (90% CI − 63 to + 44) (i.e. digital perfusion was higher with placebo) (Fig. 1b; Table 2), as mean (SD) digital blood flow increased by + 15% (51) with riociguat and by + 25% (114) with placebo. Variability in digital blood flow between subjects was high following both single doses of riociguat and placebo; individual placebo-corrected changes from baseline in digital blood flow following treatment with riociguat at RT and after CE are shown in Fig. 2. Figure 3 shows laser speckle contrast analysis images after dosing with riociguat (a) or placebo (b).

A total of 8 patients (40%) were classified as responders to riociguat at RT, and 12 patients (60%) as responders after CE, according to the predefined criterion. In responders, riociguat led to a mean (SD) increase from baseline in digital blood flow of + 136% (114) at RT (Table 2) and + 39% (53) after CE (Table 2). The highest response rates following CE were seen in patients with lcSSc (80%; n = 4) and primary RP (67%; n = 2) (Fig. 4).

The geometric mean (SD) concentration of riociguat in plasma 2 h post-dose was 76 (1.5) μg/L (range 35–121 μg/L). Plasma concentration of riociguat 2 h post-dose was comparable in responders [geometric mean (SD) 81 (1.5) μg/L; range 35–121] and non-responders [geometric mean (SD) 71 (1.6) μg/L; range 38–117].

All patients who received a dose of riociguat were included in the safety analysis set (n = 23). AEs were reported in 6 patients while receiving riociguat (headache, n = 5; and dyspepsia, n = 1) and in 2 patients while receiving placebo (malaise, n = 1; and pain in extremity, n = 1). One patient discontinued study drug and withdrew consent after the first treatment period due to side effects; after unblinding, this patient’s treatment was shown to be placebo.

No SAEs were reported, and no clinically relevant hematologic or biochemical findings were observed. Changes in heart rate and blood pressure were minimal throughout the treatment period, with no patient experiencing symptomatic hypotension (Supplementary Table 1).

Blood collection was performed without difficulties. No patient reported anxiety or pain during collection, and no clinical sign of a vasovagal reaction was observed.