The effect of percutaneous coronary intervention (PCI) on coronary microvascular function and the prognostic implication of pre and post-procedural index of microvascular resistance (IMR) has been shown in previous studies [1, 2, 3]. Coronary rotational atherectomy (RA) is an efficient way to facilitate balloon or stent delivery and optimize stent expansion by physical removal of hard plaque via lumen enlargement [4, 5]. Current PCI guidelines state that RA is a reasonable approach for the treatment of heavily calcified plaques that cannot be crossed by a balloon catheter or adequately dilated before stent implantation [6]. However, the RA procedure has previously been reported to be associated with microvascular disorder resulting from microcirculatory obstruction [5].

The pressure-temperature wire-derived coronary flow reserve (CFR) and IMR have constituted the reference standard to assess the status of coronary microcirculation thus far [7, 8]. Prior studies have indicated that there are major limitations to the pressure wire-derived CFR calculation; the maximal hyperemic coronary blood flow is strongly pressure-dependent, and the pressure wire-derived method appears to systematically underestimate the CFR values [9]. However, the pressure-temperature wire-derived IMR shows good specificity and reproducibility compared with CFR [10, 11], whereas the invasive measurement increases additional intracoronary performance and prolongs the operation time. Thus, to a certain degree, the invasive measurement raises unpredictable procedural risks, especially when faced with treating complicated lesions or in urgent situations. Alternatively, multiple pressure-wire-free tools, such as angiography-derived index of microcirculatory resistance, to assess coronary microvascular dysfunction have been developed [12]. The pressure-wire-free method was revealed to be well correlated with wire-derived IMR for estimation of microcirculatory function [13]. A novel coronary angiography-derived index of microvascular resistance (caIMR) shows good agreement with pressure-temperature wire-based IMR and has similar accuracy; thus, it has been proposed as a well-adopted non-invasive physiological assessment of coronary microcirculation function [14, 15, 16]. The effect of RA on coronary microcirculation function remains unclear, and this study aimed to investigate coronary microcirculatory function indicated by caIMR in patients undergoing RA.

The aim of this study was to evaluate coronary microcirculation function indicated by caIMR in patients undergoing RA. Our main findings are as follows: (1) Post-RA caIMR, which indicates coronary microcirculation function, was greater than pre-RA caIMR in the treated vessels. Patients with $\ge$25 post-RA caIMR accounted for nearly 12% of those with pre-RA caIMR 25; (2) among patients without increased pre-RA caIMR, those with post-RA caIMR $\ge$25 were associated with a significantly increased incidence of myocardial injury compared to those with post-RA caIMR 25; (3) patients with post-RA TIMI flow grade 3 showed significant differences in both pre- and post-RA caIMR compared with those of patients with normal TIMI flow; (4) among patients who underwent RA, those receiving longer RA time of each pass and with pre-RA caIMR $\ge$25 had worse outcomes.

It has been demonstrated that RA facilitates procedural success in treating calcified plaques, especially in complex ostial lesions and bifurcation lesions, which feature bulky plaque and unfavorable geometry for stent deployment [24, 25]. While there have always been concerns regarding microcirculatory dysfunction associated with RA, analyzing the CFR and coronary microvascular resistance using intracoronary Doppler guidewire has been considered to be the most reliable method for coronary microcirculation assessment [26, 27]. However, intracoronary Doppler guidewire is unavailable in current practice. The pressure-temperature wire-derived CFR is associated with variations in measurement and has unsatisfied reproducibility [9, 28]. The pressure-temperature wire-derived measurements of coronary microcirculation, indicated by IMR, seem impracticable with regard to real-world applicability, particularly when applied in urgent situations or complex PCI. Previous studies have demonstrated that caIMR is a feasible alternative for the evaluation of coronary microcirculatory function [14, 15, 16].

In the present study, post-RA caIMRs were significantly higher than pre-RA caIMRs in the treated vessels. Nearly 1/8th of patients without demonstrated microcirculatory dysfunction indicated by pre-RA caIMR 25 had an increased post-RA caIMR ($>$25). To our knowledge, this is the first report on the evaluation of microcirculation function indicated by pre-RA caIMR in patients undergoing RA. It has been revealed that the RA debris containing atheromatous particles and platelet-rich tissue might be apt to induce embolic formation, subsequently resulting in clogging within the distal coronary microcirculation [29, 30]. We presumed that the resultant elevation of post-RA caIMR was mainly attributable to the microvascular embolization of atherosclerotic debris and associated thrombi [5]. Further, the possibility of microvascular spasm induced by RA cannot be excluded, even though nitroglycerin was continuously administered during the procedures.

In our study, we observed that among patients without microcirculation dysfunction reflected by pre-RA caIMR 25, the incidence of myocardial injury was significantly higher in those with post-RA caIMR $\ge$25 than in those with post-RA caIMR 25, while the rates of MI were comparable. This implies increased microvascular resistance resulting from micro-embolization of the debris generated during debulking of the lesion might cause detrimental effects such as myocardial injury. Most particles created by the RA procedure are 10 $\mu$m and have a mean diameter of 5 $\mu$m. They are smaller than normal, mature erythrocytes and can traverse coronary microvasculature cleared by the reticuloendothelial system [31]. However, in certain circumstances, the microdebris might be either too large to penetrate through the distal microcirculation or too abundant to be readily absorbed, and is consequently followed by increased caIMR, which indicates distal microvascular dysfunction [32]. However, procedure-related MI alone cannot adequately explain the difference observed between those with and without increased post-RA caIMR in our study, and more factors might be involved during the RA procedure. For instance, the release of adenosine from the ischemic myocardium due to the aggregation of the ablated microdebris could have made the RA procedure even more complicated.

Slow flow or no flow is not a rare phenomenon during RA, with reported rates varying from 7–10% [32, 33]. The incidence of instant TIMI flow grade 3 in the present study was 8.4%, and no flow occurred only in one case. Slow flow may lead to hemodynamic instability due to serious hypoperfusion, which is thought to be related to reduced coronary artery conductance [33, 34], which in turn is associated with increased occurrence of MI, rather than myocardial injury, as shown in our study. In our study, a reduction in coronary conductance was indicated by the finding that more patients with TIMI flow grade 3 had higher post-RA caIMR. RA did not definitely have a detrimental effect on the microcirculatory status in patients with post-RA TIMI flow grade 3, however, the observation that more patients with post-RA TIMI flow grade 3 had higher pre-RA caIMR indicates that the embolization of ablated microdebris and associated microthrombi subsequently followed by slow flow more likely occurred in patients with baseline coronary microcirculatory dysfunction.

In our adjusted analysis for various related variables, patients with maximum RA time of each pass had a significantly increased risk of the primary outcome. Prolonged rotational duration does not necessarily confer beneficial effects, which is consistent with a prior report which found that adopting an aggressive RA strategy did not offer advantage, and was sometimes even detrimental [35]. Notably, our study revealed that pre-RA caIMR $\ge$25 in the treated vessels, but not post-RA caIMR $\ge$25, was identified to be an independent predictor of the primary outcome with approximately a $>$3-fold increase in risk of the primary outcome among patients undergoing RA. The underlying coronary microcirculatory dysfunction indicated by an elevation of pre-RA caIMR was associated with a benign outcome in patients who underwent RA, and increased post-RA caIMR might only suggested to be a resultant slow flow.

This study has several limitations. First, this was a single-center retrospective observational study, and the lack of a control group weakened the strength of the study’s implications. However, we performed a self-control analysis and measured caIMR in the reference vessels. Second, the decision to perform RA was at the discretion of the interventionist, and there was sustained improvement in the RA techniques across the cases. Thus, the results of our study should be interpreted with caution. Third, as the caIMR measurement was related to Pd${}_{\text{hyp}}$ and V${}_{\text{diastole}}$, as mentioned above in the methods section, severe stenosis in the target vessels might influence the pressure and velocity in the distal vessel to some degree. Nevertheless, it has been reported that minimal microvascular resistance does not change with epicardial stenosis severity, and IMR is a specific index of microvascular resistance when coronary wedge pressure was taken into account [10, 36]. Despite the value of caIMR in our study was a corrected IMR following Yong formula, whether wire-derived IMR could be translated to caIMR deserve further study. Fourth, in this study, we only observed the instant and short-term impact of RA on coronary microcirculation, and long-term angiography follow-up data were not obtained. Analyzing the peri-procedural and in-hospital outcomes combined with long-term outcomes of RA affecting coronary microcirculation should be considered in future studies.

There were significant changes in the coronary microcirculation function of the target vessel after receiving RA, as indicated by a increase in post-RA caIMR compared with pre-RA caIMR. Post-RA TIMI flow grade 3 was more likely to be observed in patients with baseline coronary microcirculatory dysfunction. Patients receiving longer RA time of each pass and with pre-RA caIMR $\ge$25 had worse outcomes. In the future, further investigation of the impact of RA on long-term coronary microcirculatory function is warranted. To justify the clinical performance of caIMR, prospective and mode in-depth analysis should be designed also for short-term outcomes.

RA, rotational atherectomy; caIMR, coronary angiography-derived index of microvascular resistance; PCI, percutaneous coronary intervention; CFR, coronary flow reserve; CMR, coronary microvascular resistance; IMR, index of microvascular resistance; ACS, acute coronary syndrome; PPAE, peri-procedural adverse event; MI, myocardial infarction; rpm, revolutions per minute; TIMI, thrombolysis in myocardial infarction; cTn, cardiac troponin; CFD, computational fluid dynamics; MAP, mean aortic pressure.

Study conception and design—H-PZ, HA. Acquisition of data—H-PZ, XP, LL, G-DT, YZ, G-JY, N-XZ, F-CS. Analysis and interpretation of data (e.g., statistical analysis, computational analysis)—HL, XP, H-PZ, HA. Writing, review, and/or revision of the manuscript—HL, XP, Y-DF, H-PZ. Study supervision—H-PZ. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

The study was approved by the institutional Ethics Committee (Approval No. 2019BJYYEC-021-02), all patients signed informed consent to undergo coronary angiography and the intervention procedure. Because data on caIMR were collected retrospectively, informed consent on the use of caIMR was waived given the institutional ethics regulations with regard to observational study nature.

Not applicable.

The present study was supported by National High Level Hospital Clinical Research Funding (BJ-2018-201), Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (Grant Number 2021-I2M-C&T-A-019), and Beijing Municipal Science & Technology Commission, Administrative Commission of Zhongguancun Science Park (Grant Number Z211100002921008).

The authors declare no conflict of interest.