Target and non-target vessel related events at 10 years post percutaneous coronary intervention

Aims To define the incidence of events related to the stented vessel (target vessel related events: TVRE) and events related to non-stented vessels (non-target vessel related events: NTVRE) through to 10-year follow-up in patients post-PCI with newer generation drug eluting stents (DES). Methods and results The current study is a post-hoc analysis of patient level data from two randomised controlled trials in Germany. Patients older than 18 years with ischemic symptoms or evidence of myocardial ischemia in the presence of ≥ 50% de novo stenosis located in the native coronary vessels were considered eligible. The endpoints of interest were TVRE (a composite of first target vessel myocardial infarction or target vessel revascularization) and NTVRE (a composite of first non-target vessel MI or non-target vessel revascularization) through to 10 years post PCI. We included 4953 patients in this analysis. Through to 10-years post-PCI, TVRE occurred in 1238 of 4953 patients (cumulative incidence: 25.8%) and NTVRE occurred in 1442 of 4953 patients (cumulative incidence: 30.3%). The majority of TVRE and NTVRE were revascularization events. From 0 to 1 years, the cumulative incidence of TVRE was 15.9% and of NTVRE was 12.3%. From 1 to 10 years, the cumulative incidences of TVRE and NTVRE were 11.2% and 22.4%, respectively. Conclusion At 10-year post-PCI with new generation drug eluting stents, events related to remote vessel disease progression account for a higher proportion of events than events related to the stented vessel. Trial registration ISAR TEST 4 ClinicalTrials.gov Identifier: NCT00598676. ISAR TEST 5 ClinicalTrials.gov Identifier: NCT00598533. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s00392-022-01986-4.


Introduction
Patients who undergo percutaneous coronary intervention (PCI) are at risk of subsequent cardiac events, including myocardial infarction (MI) and repeat revascularisation [1,2]. Coronary stent trials traditionally focus on target vessel/lesion related clinical outcomes in the shorter term with a particular emphasis on outcomes in the first 12-months post-PCI [3].
It is important to consider that patients who undergo PCI are at risk of not only events related to the stented vessel (target vessel related events: TVRE) but also events related to non-stented vessels (non-target vessel related events: NTVRE) [1,2]. While data have been published on the incidence of these event types up to 5-years post-PCI [1,2,[4][5][6][7], their incidence at longer term follow-up has not been defined.
An analysis of TVRE and NTVRE through to 10 years after PCI in a large number of patients treated with new generation DES may be useful and help guide strategies to improve cardiovascular outcomes after PCI. Therefore, we conducted the current analysis, with the primary aim of defining the cumulative incidence of TVRE and NTVRE through to 10 years post PCI in patients enrolled in the ISAR-TEST 4 and ISAR-TEST 5 trials.
The current analysis includes all patients from the ISAR-TEST 4 and ISAR-TEST 5 trials, with the exception of the early generation permanent polymer sirolimus eluting stent group from ISAR-TEST 4. This group of patients were excluded from this analysis as early generation DES have been associated with a higher risk of adverse events compared to newer generation DES and their use was not felt to reflect current practice or be relevant to the study question, which is focused on new generation DES [12]. This study conforms to the declaration of Helsinki and the study protocol was approved by the ethics committee of the two participating centers in Munich, Germany (Deutsches Herzzentrum München and 1.Medizinische Klinik, Klinikum rechts der ISAR).
Enrolment criteria for both studies have been previously reported [8,9]. Patients older than 18 years with ischemic symptoms or evidence of myocardial ischemia in the presence of ≥ 50% de novo stenosis located in the native coronary vessels were considered eligible. Of note, patients with a target lesion in the left main stem or in cardiogenic shock were considered ineligible for both studies. Relook angiography was recommended (but not mandated) by the trial protocols at 6-8-months after the index PCI.

Endpoints and definitions
Detailed descriptions of study endpoints for both trials have been published previously [10,11]. The primary endpoints assessed for the current analysis were TVRE (a composite of first target vessel MI (TVMI) or target vessel revascularization (TVR)) and NTVRE (a composite of first non-target vessel MI (NTVMI) or non-target vessel revascularization (NTVR)). The target vessel was defined in this analysis as the vessel/vessels treated during the index PCI. All other vessels were classified as 'non-target vessels'. Therefore, patients could experience no events, TVRE, NTVRE or Both Events (TVRE and NTVRE) during follow-up.
Secondary endpoints included the individual components of the primary endpoints. We also analysed the median time to event for patients experiencing an endpoint through to 10 years post PCI and the interval between events for patients who experienced both a TVRE and NTVRE through to 10 years.

Statistical analysis
Continuous data are presented as means and standard deviation (± SD) or medians and interquartile ranges. Categorical data are presented as counts and proportions. Differences between groups were checked for significance using an analysis of variance test (ANOVA) for continuous data. Chisquared test (χ2) was used to check for differences between categorical variables. Survival was analyzed by the Kaplan Meier method to estimate the time to first event for each endpoint of interest. Hazard ratios were calculated using a Cox proportional hazards model after checking for fulfilment of the proportional hazards assumption as per the method of Grambsch and Therneau. The analysis of all endpoints accounted for the competing risk of death. In addition to the primary analysis, the cumulative incidences of TVRE and NTVRE were analysed as per stent polymer type, dividing patients into three groups; BP-DES, PP-DES and PF-DES. Statistical analysis was performed using the R 3.6.0 Statistical Package (The R Foundation for Statistical Computing, Vienna, Austria). A two tailed p value of < 0.05 was taken to confer statistical significance.

Study population
We included 4,953 patients in this analysis. This represents 88.4% of the total cohort enrolled in the ISAR-TEST 4 and ISAR-TEST 5 trials. Of these patients

Baseline characteristics
Baseline characteristics for the patients included in this analysis as per the event type experienced are demonstrated in Table 1. Female sex was more common in patients who experienced no events. Patients who experienced no events also tended to be older at index PCI, less commonly diabetic and had a lower frequency of triple vessel coronary artery disease. Trial protocol recommended angiography (recommended to be performed at 6-8 months after PCI) was performed less frequently in the no-events group.

Angiographic and procedural characteristics
The angiographic and procedural characteristics for the patients included in this analysis as per the event type experienced are summarized in Table 2. Patients who experienced no events had a lower proportion of complex lesions treated during the index procedure, as well as a shorter lesion and total stented length. The left anterior descending artery was more commonly treated in the group of patients who experienced no events.
Primary endpoint: cumulative incidence of TVRE and NTVRE through to 10 years post PCI Through to 10-years post-PCI, a TVRE occurred in 1,238 of 4,953 patients (cumulative incidence: 25.8%) and a NTVRE occurred in 1,442 of 4,953 patients (30.3%). These data are demonstrated in Fig. 1 and Table 3. The cumulative incidence of TVRE, NTVRE and 'Both Events' are shown in Supplemental Fig. 1.

Landmark analysis
A landmark analysis was then performed to assess the cumulative incidence of TVRE and NTVRE from 0 to 1 years and 1-10 years post PCI. From 0 to 1 years, the cumulative incidence of TVRE was 15.9% and of NTVRE was 12.3%. From 1 to 10 years, the cumulative incidences of TVRE and NTVRE were 11.2% and 22.4%, respectively. These data are demonstrated in Fig. 2.

Primary endpoint: As per stent polymer technology
The frequency of both TVRE and NTVRE through 10 years was comparable for patients treated with PP-DES, BP-DES and PF-DES. This is shown in Supplemental Figs. 2, 3.

Secondary endpoints: TV/NTV MI and TV/NTV revascularisation through to 10 years post PCI
TVMI occurred in 210 of 4,953 patients through to 10 years post PCI (4.4%) and NTVMI occurred in     Fig. 4. For the majority of patients, the time interval between the two event types was less than 1 year.

Discussion
In this analysis of TVRE and NTVRE through to 10-years post-PCI with new generation DES, the main findings are as follows: 1. NTVRE account for a higher proportion of total events than TVRE at 10-year follow-up. 2. TVRE and NTVRE demonstrate distinct temporal patterns, with a higher incidence of TVRE in the first year post-PCI, followed by a higher incidence of NTVRE from year 1 to year 10. 3. Analysis of the median time to first event echoed these results, with a shorter time to first TVRE compared to first NTVRE. 4. For the majority of patients who experience both a TVRE and a NTVRE, the time interval between these events is less than 1 year.
The current report presents the only analysis of TVRE and NTVRE through to 10-years post-PCI. It also defines for the first time the incidence of TVRE and NTVRE in patients undergoing PCI with all three newer generation DES polymer technologies (BP, PP, PF) used in current practice.
The findings of this analysis build on the sometimes divergent results of several other studies reporting on events related to stented and remote vessels/lesions post-PCI [1,2,4,5,13]. Cutlip et al. demonstrated that in patients receiving PCI with second generation, bare metal coronary stents, outcomes at 5 years were determined by a high rate of events related to disease progression in segments other than the stented lesion [4]. In contrast to our study, this analysis divided events into target lesion events (TLE) and non-target lesion events (NTLE). Subsequently, Stone et al. reported that in patients with ACS undergoing PCI, major adverse cardiovascular events were equally attributable to TLE and NTLE at a median follow-up of 3.4 years [1]. This study also reported that although lesions responsible for TLE were frequently angiographically mild, intravascular ultrasonography (IVUS) analysis demonstrated a combination of characteristics in these lesions, including thin-cap fibroatheroma, large plaque burden and small luminal area. In contrast to our current analysis, where all patients were treated with newer generation DES, early generation DES were implanted in only 66% of lesions in this study. Zellweger et al. subsequently reported that in an analysis of patients from 7-to 60-months post-PCI, remote events accounted for 37.1% of all non-fatal events in a patient population treated with both bare metal stents (BMS) and early generation DES.
With regard to newer generation DES, Abdel-Wahab et al. described the incidence of non-target lesion revascularisation (NTLR) at 3-year follow-up in a large cohort of  [7]. They reported that the cumulative incidence of NTLR was almost three times more frequent than TLR at 3-years post-PCI. However, this differed to our analysis in that it focused only on TLR/NTLR and excluded patients who had both events (TLR and NTLR). In addition, the patient cohort analysed was derived from a combination of randomised controlled trial and registry data rather than exclusively from randomised controlled trial data [7].
The timing of events in our study merits further discussion. As demonstrated in the median time to first event analysis, TVRE tend to occur earlier post-PCI than NTVRE. Accordingly, TVRE accounted for a greater proportion of events than NTVRE from in the first year post-PCI. However, after the first year NTVRE began to predominate and at 10-year follow-up NTVRE account for a higher proportion of total events than TVRE. This is similar to 5-year results previously reported by Cutlip et al. [4]. It is also notable in our analysis that, after undergoing PCI to a target vessel, over one quarter of patients will have a further event related to this vessel at long term follow-up. Given the classification of events in this analysis as target vessel rather than target lesion related, these events could have been related to either the previously implanted stent or to disease progression in remote regions of the vessel that were not treated at the index PCI. Notwithstanding this limitation, it is evident from our data that both TVRE and NTVRE contribute importantly to overall event rates at long term follow-up.
Whether novel stent polymer technologies (BP, PF) influence the proportion of TVRE and NTVRE at follow-up has not been previously studied. This is important to define, because concern has been raised regarding chronic vascular inflammation with PP-DES secondary to the presence of a permanent polymer in the vessel wall [14][15][16][17]. In addition, it has been shown in animal models that focal vascular inflammation may accelerate atherosclerotic disease in remote arterial segments [18]. Therefore, it may be logical to consider that different polymer technologies could theoretically demonstrate different frequencies of non-target vessel disease progression at long term follow-up. However, in our study, the frequency of both TVRE and NTVRE at 10 years were comparable between the three studied stent polymer technologies (PP, BP, PF).
The current analysis may also serve to highlight the importance of instituting long term secondary prevention measures after PCI to improve outcomes and challenges a solely stenting/stenosis focused interventional paradigm [19,20]. Given that the incidence of NTVRE is unlikely to improve with iterative improvements in stent technology, novel strategies may be required, to improve long term outcomes for patients in this regard.

Limitations
The major limitation of this analysis is that it is a post-hoc analysis of pooled data from two randomised controlled trials. This is subject to the usual limitations associated with posthoc analysis and the findings should be regarded as hypothesis generating. Due to the inability to accurately ascribe mortality to TVRE or NTVRE, we were not able to include mortality or cardiac mortality in our primary endpoint. Planned angiographic follow-up at 6-8-months after PCI was recommended as part of the study protocols and this may have artificially altered the incidence of TVRE and NTVRE in comparison to standard clinical practice. In detail, 6-8-month follow-up angiography was performed in 68% of the patients in the no-events group, 84.1% of patients in the TVRE group, 89.7% of patients in the NTVRE group and 93.3% of patients in the both events group (Table 1). Intracoronary imaging or coronary physiological testing was not performed as part of the study protocols to the target or non-target vessels, and therefore, we have no information on intracoronary imaging or physiology predictors of TVRE and NTVRE. This analysis considered only the first event for both TVRE and NTVRE, and therefore, recurrent events were not accounted for. Information as to whether the procedure was staged or not was also not collected in the trial database. A final limitation of this analysis is the absence of data on rates of dual anti-platelet therapy and cardiovascular secondary prevention measures up to 10 years, both of which may have impacted on the incidence of long-term events.

Conclusions
At 10-year follow-up post-PCI, events related to non-target vessels account for a higher proportion of total events than events related to the target vessel.
Author contributors JJC was primarily responsible for the drafting of this manuscript and was also involved in the conception and design of the project, analysis and interpretation of the data. AK was primarily responsible for conception and design of the project, performed statistical analysis of the data, provided revision of the manuscript for important intellectual content and is responsible for the overall content as guarantor. AA assisted with preparation of the manuscript, interpretation and analysis of the data and revision of the initial draft. SK was involved in the design and conception of the project, interpretation of the data and revised the manuscript for important intellectual content. EX, SC, MJ, TK, JW, TL, TR, CP, SG, TI, KLL and HS were involved in the conception of this project and provided their expert review and revision on the initial manuscript draft for important intellectual content. All authors reviewed and approved the final version of the manuscript prior to submission. AK is responsible for the overall content of this manuscript as guarantor.