Introduction

Physiological assessment of borderline stenosis in the epicardial coronary arteries has emerged as a fundamental method for determining ischemia‑causing lesions. Consequently, the most recent guidelines strongly recommend assessing the hemodynamic significance of intermediate coronary stenoses using invasive physiological methods before performing revascularization procedures (class IA recommendation).^1,2 Among the current diagnostic techniques, pressure wire–derived fractional flow reserve (FFR) is regarded as the gold standard for assessing the ischemic significance of coronary artery stenoses and guiding decisions regarding revascularization.^2-4 Invasive physiological assessment of myocardial ischemia encompasses not only hyperemic indices but also several nonhyperemic, pressure‑derived parameters, including the instantaneous wave‑free ratio (iFR), resting full‑cycle ratio (RFR), and others, whose advantages are strongly related to their hyperemia‑independent character resulting in better patient tolerance of the procedure. The clinical reliability of both hyperemic and nonhyperemic parameters has already been proven, with respective FFR and iFR/RFR cutoff values of up to 0.8 and lower than 0.89 indicating myocardial ischemia.^1-3

FFR is defined as the ratio of the pressure distally to the assessed coronary lesion to aortic pressure during maximal hyperemia that is achieved through the administration of vasoactive substances, most commonly adenosine.⁵ Intravenous adenosine provides more consistent and sustained vasodilation and is particularly useful for assessing tandem lesions or diffuse coronary artery disease. In contrast, intracoronary bolus administration is more cost‑effective and easier to perform, with a rapid onset of action and short half‑life, making it ideal for repeated measurements.⁶ However, despite its critical role in FFR assessment, the protocol for hyperemia induction remains unstandardized, and submaximal hyperemia during the procedure may contribute to underestimation of the lesion’s hemodynamic significance, leading to false‑negative results.

Furthermore, some clinical and anatomical factors have been suggested to influence the results of FFR assessment, including coronary lesion morphology,^7,8 coronary disease pattern,⁹ myocardial area supply,¹⁰ and noncoronary cardiac conditions, such as aortic stenosis (AS)^11,12 and atrial fibrillation.^13,14 Thus, the aim of this study was to evaluate the potential factors influencing variations in FFR values following intracoronary administration of escalating doses of adenosine.

Patients and methods

This is a retrospective, observational, single‑center study gathering data of patients hospitalized at the Clinical Department of Cardiology and Cardiovascular Interventions of the University Hospital in Kraków between January 2020 and January 2025. All patients underwent invasive physiological assessment as part of routine diagnostic coronary angiography procedures for intermediate coronary lesions, defined as visual vessel luminal narrowing of 40% to 80%. Only the patients who had at least 2 measurements performed in the same lesion with the administration of a progressive dose of adenosine were included in the study, regardless of the number of assessed vessels. Routine blood tests included standard biochemical and hematological indices. Patient demographic and clinical data were collected from medical records. Hypertension was defined as blood pressure equal to or greater than 140/90 mm Hg recorded on 2 separate occasions or a documented history of hypertension and use of antihypertensive drugs. Dyslipidemia was diagnosed if total cholesterol level exceeded 5 mmol/l, low‑density lipoprotein cholesterol level exceeded 3 mmol/l, or if lipid‑lowering therapy was being used. Diabetes was diagnosed based on a prior medical history or if fasting blood glucose levels were greater than 7 mmol/l (126 mg/dl) on 2 separate occasions. Chronic obstructive pulmonary disease was diagnosed in accordance with the Global Initiative for Chronic Obstructive Lung Disease guidelines. Renal disease was identified as estimated glomerular filtration rate below 60 ml/min/1.73 m². Peripheral artery disease was defined as a confirmed stenosis of 50% or greater in at least 1 peripheral artery and / or a prior intervention within any of the peripheral arteries. Severe AS was defined by echocardiographic criteria as an aortic valve area smaller than or equal to 1 cm², a mean transvalvular pressure gradient equal to or greater than 40 mm Hg, or a peak aortic jet velocity of 4 m/s or greater.

Results

The study included 173 patients (191 vessels) out of 892 FFR procedures performed in our center during the study period, and all analyses were performed per vessel. Mean (SD) age of the study population was 67.9 (10.1) years, and 71.2% were men. A majority of the patients presented with common cardiovascular risk factors: hypertension was present in 81.7%, diabetes mellitus in 41.9%, dyslipidemia in 70.2%, and nearly half of the study population (46.6%) had a history of myocardial infarction. Median (IQR) left ventricular ejection fraction was 55% (45%–60%), and most lesions were located in the left anterior descending artery (62.8%). Median (IQR) visual diameter stenosis was 55% (50%–70%), and median (IQR) baseline FFR value at the standard adenosine dose was 0.84 (0.81–0.88).

Detailed demographic and clinical characteristics of the study population are presented in Table 1.

Among the vessels studied, the FFR value decreased after adenosine dose escalation in 89 cases (46.6%), while in 102 vessels (53.4%) FFR value remained stable. FFR value changes are presented casewise in Supplementary material, Figure S1. There were no differences between these groups in terms of baseline clinical characteristics; laboratory parameters were also comparable. The group with FFR value reduction showed larger angiographic stenosis than the group with stable FFR (median [IQR], 60% [50%–70%] vs 50% [50%–60%], respectively; P = 0.045). Baseline nonhyperemic indices and initial FFR values at the standard adenosine dose did not differ significantly. Detailed data are presented in Table 1.

Among the individuals with FFR values negative for ischemia (FFR >0.8) following administration of the standard adenosine dose (200 μg), the patients with FFR values persistently exceeding the cutoff (n = 134) and those whose FFR became indicative of ischemia after dose escalation (FFR ≤0.8; n = 17) did not differ significantly in terms of baseline clinical characteristics or laboratory parameters (reclassification rate, 11%; Wilson CI, 0.067–0.174). The patients whose FFR fell below the ischemic threshold after dose escalation had larger diameter stenosis on angiography than the group with FFR values consistently above 0.8, but the difference was not significant (median [IQR], 60% [60%–70%] vs 50% [50%–60%], respectively; P = 0.06). Additionally, nonhyperemic indices (median [IQR], 0.9 [0.87–0.91] vs 0.93 [0.91–0.96]; P = 0.001) and initial FFR values (median [IQR], 0.81 [0.81–0.83] vs 0.86 [0.84–0.89]; P <⁠0.001) were lower in this group, with a greater magnitude of FFR change after dose escalation (median [IQR], –0.02 [–0.04 to –0.01] vs 0 [–0.01 to 0]; P <⁠0.001). Severe AS was numerically more frequent in the group with FFR values indicative of ischemia following dose escalation, as compared with the ischemia‑negative group (17.6% vs 5.3%; P = 0.09), though the difference was not significant. All details are presented in Table 1.

Discussion

The study evaluated the influence of various clinical factors on changes in FFR values following escalation of intracoronary adenosine bolus doses. Our findings demonstrated that escalating the intracoronary adenosine dose resulted in a decrease in FFR values in nearly half of the examined vessels, with 11% of the lesions having been reclassified as ischemia‑positive. Notably, dose escalation was safe and did not precipitate any major complications. Minor side effects, such as transient bradycardia or a need for atropine administration, were occasionally observed but did not require additional intervention. Moreover, the lesions that showed FFR value reduction following adenosine dose escalation had both lower nonhyperemic indices and lower FFR values after the conventional dose, indicating that the vessels with FFR values closer to the ischemic threshold are more susceptible to further FFR value decline with dose escalation. Nevertheless, most clinical characteristics, comorbidities, and laboratory parameters showed no significant correlation with FFR changes following progressive adenosine dose administration.

Achieving maximal hyperemia is essential for accurate FFR measurement. Large randomized trials have established intravenous adenosine infusion at 140 μg/kg/min as the standard approach, as it produces a stable and sustained hyperemic response.¹⁵ It also allows for measuring pressure pullback gradient (PPG)—an index differentiating focal atherosclerosis from the diffuse type. This index was proved to predict symptom relief after percutaneous coronary intervention (PCI); however, its predictive ability for hard clinical outcomes is yet to be determined.¹⁶ A confirmation of PPG as an outcome predictor after a potential PCI would increase the importance of continuous intravenous infusion, and decrease the importance of intracoronary boluses.

While intracoronary adenosine offers a practical alternative, the dosing remains under discussion. Early studies used much lower intracoronary doses, typically 8–12 μg for the right coronary artery (RCA) and 15–18 μg for the left coronary artery (LCA), but maximal hyperemia was not achieved in approximately 10%–15% of cases.¹⁷ More recently, dose–response studies conducted in angiographically normal arteries have generated recommendations for higher intracoronary doses—up to 100 μg for the RCA and 200 μg for the LCA—to ensure sufficient hyperemia.⁵ Interestingly, De Luca et al¹⁸ demonstrated a clear dose–response relationship with even higher adenosine doses (up to 720 μg), showing a greater ischemia detection rate at the highest dose (51.2% at 720 μg vs 38% at 120 μg). Interestingly, Wilson et al¹⁹ showed that coronary blood flow fluctuates considerably under submaximal hyperemic conditions, but this variability can be mitigated by escalating the adenosine dose. Therefore, increasing the hyperemic stimulus, particularly in the cases with borderline FFR values, appears to be a rational strategy to overcome biological variability and improve diagnostic accuracy. In our study, the rate of reclassification after dose escalation to 400 μg reached 11%, and was consistent with previous reports indicating that around 10% of cases fail to achieve maximal hyperemia at conventional adenosine doses.^17,20,21

Similarly to previous studies, our analysis showed that common atherosclerotic risk factors or prior myocardial infarction did not significantly influence FFR values following high‑dose adenosine escalation.^19-21 However, when considering physiological and angiographic factors, we observed that a lower baseline nonhyperemic index and lower FFR values after the conventional adenosine dose were predictive of FFR conversion from ischemia‑negative to ischemia‑positive, aligning with earlier reports.²¹ Additionally, greater diameter stenosis was associated with a further decrease in FFR values following adenosine dose escalation, which is in line with other studies.^18,21

Interestingly, in the subgroup of lesions where FFR changed to ischemia‑positive after dose escalation, we observed a higher prevalence of severe AS (17.6% vs 5.3%), as compared with the lesions with FFR values consistently negative for ischemia. This finding implies that severe AS might affect coronary hemodynamics, warranting further investigation into this relationship. Importantly, a lack of significance in this setting should not be taken as evidence of no effect. Given the small sample size, the analysis was likely underpowered and therefore susceptible to a type II error (false negative). We therefore interpret this result as an inconclusive but potentially meaningful signal that is consistent with the pathophysiological mechanisms proposed below. The chronic pressure overload associated with AS leads to left ventricular hypertrophy, which in turn increases myocardial oxygen demand. Therefore, AS is associated with reduced coronary microvascular density, enhanced autoregulatory vasodilation, and lower basal microvascular resistance—all contributing to an elevation in resting coronary blood flow. Consequently, patients with AS present increased baseline flow but blunted hyperemic response, resulting in an attenuated coronary flow reserve and reduced microvascular resistance reserve. These hemodynamic nuances may explain the heightened sensitivity of FFR to adenosine dose escalation in AS patients, as the limited hyperemic capacity may only be fully unmasked under maximal vasodilatory stimulus.²²

Coronary microvascular dysfunction, previously shown to cause FFR/RFR discordance, may be the key mechanism underlying attenuated vasodilatory response to standard adenosine doses. With a damaged endothelium, the vasodilation ability is impaired and response to the hyperemia‑inducing medication is restricted. This causes a discrepancy between the values of nonhyperemic indices and FFR, with the FFR value usually being negative for ischemia despite ischemia‑positive iFR/RFR.²³ However, among the factors associated with this discrepancy, including insulin‑treated diabetes, chronic obstructive pulmonary disease, and atrial fibrillation (causing significant beat‑to‑beat variability),^24,25 none has been found to influence the dose response in our study. Therefore, further research on this phenomenon and its implications on clinical decision‑making is needed.

Additionally, as FFR values were found to be dose‑dependent in some patients in our study, optimal thresholds for ischemia‑positive results should be verified for different adenosine doses when applying intracoronary boluses.