Intravascular Radiation Delays Recurrence of Recalcitrant Coronary In-Stent Restenosis Authors: Kent Wallner, Kathleen E. Kearney, Joseph M. DeCunha, Edward Y. Kim, Lorenzo Azzalini, Michael R. Folkert, George Sandison, William L. Lombardi, Sant Kumar, Bishwambhar Sengupta, Creighton Don, Minsun Kim Published: 2026 ABSTRACT PurposeIntravascular brachytherapy (IVBT) has a well-proven activity against coronary restenosis of a native vessel or a single bare metal stent. However, its effectiveness for multiple recurrent same-vessel in-stent restenoses (ISRs) is unproven. The objective of this study was to determine if IVBT prolongs the ISR-free interval in recalcitrant lesions.Methods and MaterialsISR-free intervals before and after IVBT for individual patients were compared. T1 was the ISR-free interval prior to IVBT, and T2 was the ISR-free interval after IVBT. Accrual was sought from patients with a prior ISR-free interval of less than 2 years. Standard radiation techniques were used, with doses of 18, 23, or 34 Gy prescribed at 2 mm from the radiation source. ISR-free intervals before (T1) and after IVBT (T2) were compared.ResultsTwenty-six vessels in 19 patients were studied. ISR-free intervals prior to IVBT ranged from 25 to 317 days (median, 96 days) compared with ISR-free intervals after IVBT that ranged from 30 to 612 days (median, 350 days; P < .001; log-rank test).ConclusionThe results support the use of IVBT to prolong the ISR-free interval in recalcitrant ISRs. Introduction Catheter-based interventions to correct coronary stenosis date back to the introduction of balloon angioplasty in 1974.1 Since that time, there have been a variety of percutaneous coronary intervention (PCI) technical innovations.2 Although most patients remain stenosis-free after PCI, a substantial minority develop recurrent same-site in-stent restenosis (ISR) (Fig. 1). PCI for ISR accounts for 11% of the National Cardiovascular Data Registry and represents a higher share of overall procedures (19%) at centers with a dedicated complex ISR program.3,4 Multiple same-site ISRs are considered “recalcitrant.”Figure 1This 69-year-old man had 2 prior ISRs of his right coronary artery, 13 and 8 months before PCI + IVBT. Yellow arrows show multiple recurrent ISR sites over time. The patient's artery was cleared of obstruction at each PCI (bottom row). He has remained free of ISR at 10 months following PCI + IVBT.Abbreviations: ISR = in-stent restenosis; IVBT = intravascular brachytherapy; PCI = percutaneous coronary intervention.Figure 1 dummy alt text This 69-year-old man had 2 prior ISRs of his right coronary artery, 13 and 8 months before PCI + IVBT. Yellow arrows show multiple recurrent ISR sites over time. The patient's artery was cleared of obstruction at each PCI (bottom row). He has remained free of ISR at 10 months following PCI + IVBT. Abbreviations: ISR = in-stent restenosis; IVBT = intravascular brachytherapy; PCI = percutaneous coronary intervention. Intravascular brachytherapy (IVBT) has a well-proven activity against coronary restenosis of native vessels or a single bare metal stent (Fig. 2).5,6 However, its effectiveness was rigorously tested only in patients with native vessels or first-time ISRs. With advances in drug-eluting stents, such patients are typically treated with a first or second drug-eluting stent, rather than IVBT.Figure 2This panel of images shows a typical IVBT procedure. The patient presented with symptomatic stenosis of the right coronary artery (RCA). It was opened with a 30-mm length catheter-based balloon, followed by a 6-minute dwell positioning of a 40-mm Sr-90 source.Abbreviation: IVBT = intravascular brachytherapy; Sr-90 = strontium-90.Figure 2 dummy alt text This panel of images shows a typical IVBT procedure. The patient presented with symptomatic stenosis of the right coronary artery (RCA). It was opened with a 30-mm length catheter-based balloon, followed by a 6-minute dwell positioning of a 40-mm Sr-90 source. Abbreviation: IVBT = intravascular brachytherapy; Sr-90 = strontium-90. The optimal therapy for multiple recurrent ISRs is unclear.7 Placing additional stents at an affected site may increase the risk of subsequent stenosis.8 Accordingly, IVBT is commonly used for multiple same-site restenoses.3 Several investigators have published retrospective series of IVBT as second- or third-line treatment of ISR.9, 10, 11, 12, 13, 14, 15 The restenosis rate among such series has been fairly consistent, ranging from 33% to 46% at 3 years.9,11 However, no prospective, randomized trial of IVBT compared with other treatments has been conducted in complex patients, due in part to their highly heterogeneous nature. A randomized controlled trial (RCT) would require a large number of subjects to account for multiple clinical and angiographic features. Considering that IVBT is an invasive, expensive procedure, it would be valuable to determine its effectiveness in today’s complex patients. Although RCTs remain (rightly) the gold standard for testing therapies, there has been interest in using within-patient controls when RCTs are impractical.16, 17, 18, 19 Such trials have been referred to as “N-of-1” or “single-case experimental design” trials. Prior IVBT investigators have used N-of-1 studies to evaluate IVBT outcomes. Kearney et al15 compared the ISR-free interval before and after combined IVBT + drug-coated balloon treatment, showing that the combination increased the ISR-free interval. However, their study was limited by being retrospective, open to selection bias, and with incomplete follow-up. Dilcher et al20 used 14 patients as their own controls in comparing ISR-free intervals after the first versus second IVBT. They reported longer ISR-free intervals after patients’ second IVBT, suggesting an enhanced anti-ISR effect from cumulative radiation effect with retreatment. In an effort to determine the benefit of IVBT added to PCI in patients with multiple ISRs, the current study compares patients’ stenosis-free interval before versus after IVBT. Methods Patients with an ISR-free interval of less than 2 years prior to IVBT were enrolled on an institutional review board–approved prospective registry. All patients signed informed consent prior to their IVBT procedure. “PCI-1” refers to the PCI procedure performed preceding the IVBT-treated ISR (Fig. 3). IVBT was always performed in conjunction with PCI, referred to as “PCI-0.” For the current analysis, the first 19 patients with a pre-IVBT ISR-free interval of less than 12 months were studied, to allow for follow-up time exceeding their pre-IVBT ISR-free interval. Seven enrolled patients with prior IVBT or pre-IVBT ISR-free intervals of longer than 12 months were excluded from this analysis.Figure 3Schema for time points analyzed. PCI-1 is the procedure to clear the ISR preceding the IVBT-treated ISR. T1 is the time from PCI-1 until the next ISR, treated with PCI + IVBT.Abbreviations: ISR = in-stent restenosis; IVBT = intravascular brachytherapy; PCI = percutaneous coronary intervention.Figure 3 dummy alt text Schema for time points analyzed. PCI-1 is the procedure to clear the ISR preceding the IVBT-treated ISR. T1 is the time from PCI-1 until the next ISR, treated with PCI + IVBT. Abbreviations: ISR = in-stent restenosis; IVBT = intravascular brachytherapy; PCI = percutaneous coronary intervention. Beta-radiation brachytherapy was performed with a Novoste Beta-Cath System using a strontium-90 (beta) source (Best Vascular). Source lengths of 40 or 60 mm were used. A dose of 18, 23, or 34 Gy was prescribed at 2 mm from the source center, with higher doses used for larger diameter vessels.21 Dwell time was 5 to 10 minutes. Two time intervals were compared. T1 was the ISR-free interval prior to IVBT-treated ISR. T2 was the time from IVBT to subsequent ISR (Fig. 3). For most patients, the ISR verification and IVBT dates were the same. However, 5 patients had ISR documented preceding IVBT referral. They were treated with IVBT as a staged procedure, in which case the T1 interval was calculated from PCI-1 to ISR verification prior to IVBT. The time from IVBT to subsequent target vessel restenosis was the primary endpoint. Post-IVBT ISR documentation was always preceded by symptoms suggestive of ISR. Post-IVBT angiograms were not done in the absence of ischemic symptoms. The T1 and T2 intervals were compared by Kaplan-Meier log-rank. Results Twenty-six vessels in 19 patients were studied (left anterior descending, 6; circumflex, 10; right coronary artery, 9; and left main, 1; Table 1). Each vessel was analyzed separately. All patients had 2 or more previous ISRs of the IVBT-treated site (Table 1). Patient characteristics and immediate procedural outcomes were typical of other complex PCI series (Tables E1-E3).Table 1Patient characteristics and interventional tools used for PCI-1 and PCI-0 + IVBTTable 1 dummy alt textPt #AgeSexVesselPrior ISRsToolsPCI-1ToolsPCI-0 + IVBTXRT dose (Gy)LengthXRT (mm)ISR afterIVBT?165MaleoRCA4I, R, La, CI, L, C3460277MaleCxOM13I, N, SL, I, C, N2340277MaleRCA3Co, I, N, SL, I, C, N2340361MaleCxOM12C, NI, C, N2340Yes480FemaleOM15I, L, NI, C2360542MaleLAD3B, II, U2360676FemaleRCA2BB, I, N, C2340Yes676FemaleLAD1B, DES, NI, N23/2340/40757MaleOM12I, NI, C2360757MaleOM24I, NI, C1860865MaleOM3I, CO, C1860Yes980MaleRCA2L, UN, O, L, C23/2360/60Yes1065MaleLAD3I, L, CI, C, S, N23/2360/601065MaleRCA3I, La, DCB, DESI, C, N3460Yes1161FemaleRCA3I, B, DESx2, NB3460Yes1289FemaleRCA2B, NN, I, C, U23/2340/401385MaleLM LAD2I, NC, I, N23601471FemaleoCx2NI, C, N23601471FemaleoLAD3NI, C, N23601471FemaleLM2NI, C, N23601564MaleRCA2I, N, CoI, C, U34601669MaleD12I, C, BI, N, C, S23401769MaleRCA3I, L, S, DESx2O, N, C23601879MaleCx/OM13I, N, LI, N, S, C23601964MaleCx2I, NI, S, L1860Yes1964MaleoOM12I, CI, C, N1860YesAbbreviations: B = plain old balloon angioplasty; C = cutting; I = IVUS; L = lithotripsy/shock; N = noncompliant balloon; OCT = ocular coherence tomography; o = ostial; R = rotational; S = scoring; U = ultra-high-pressure balloon; La = laser; Cx = circumflex; DCB = drug-coated balloon; DES = drug-eluting stent; ISR = in-stent restenosis; IVBT = intravascular brachytherapy; LAD = left anterior descending; LM = left main; OM = obtuse marginal; RCA = right coronary artery; XRT = radiation. Table 1Patient characteristics and interventional tools used for PCI-1 and PCI-0 + IVBTTable 1 dummy alt textPt #AgeSexVesselPrior ISRsToolsPCI-1ToolsPCI-0 + IVBTXRT dose (Gy)LengthXRT (mm)ISR afterIVBT?165MaleoRCA4I, R, La, CI, L, C3460277MaleCxOM13I, N, SL, I, C, N2340277MaleRCA3Co, I, N, SL, I, C, N2340361MaleCxOM12C, NI, C, N2340Yes480FemaleOM15I, L, NI, C2360542MaleLAD3B, II, U2360676FemaleRCA2BB, I, N, C2340Yes676FemaleLAD1B, DES, NI, N23/2340/40757MaleOM12I, NI, C2360757MaleOM24I, NI, C1860865MaleOM3I, CO, C1860Yes980MaleRCA2L, UN, O, L, C23/2360/60Yes1065MaleLAD3I, L, CI, C, S, N23/2360/601065MaleRCA3I, La, DCB, DESI, C, N3460Yes1161FemaleRCA3I, B, DESx2, NB3460Yes1289FemaleRCA2B, NN, I, C, U23/2340/401385MaleLM LAD2I, NC, I, N23601471FemaleoCx2NI, C, N23601471FemaleoLAD3NI, C, N23601471FemaleLM2NI, C, N23601564MaleRCA2I, N, CoI, C, U34601669MaleD12I, C, BI, N, C, S23401769MaleRCA3I, L, S, DESx2O, N, C23601879MaleCx/OM13I, N, LI, N, S, C23601964MaleCx2I, NI, S, L1860Yes1964MaleoOM12I, CI, C, N1860YesAbbreviations: B = plain old balloon angioplasty; C = cutting; I = IVUS; L = lithotripsy/shock; N = noncompliant balloon; OCT = ocular coherence tomography; o = ostial; R = rotational; S = scoring; U = ultra-high-pressure balloon; La = laser; Cx = circumflex; DCB = drug-coated balloon; DES = drug-eluting stent; ISR = in-stent restenosis; IVBT = intravascular brachytherapy; LAD = left anterior descending; LM = left main; OM = obtuse marginal; RCA = right coronary artery; XRT = radiation. Patient characteristics and interventional tools used for PCI-1 and PCI-0 + IVBT Abbreviations: B = plain old balloon angioplasty; C = cutting; I = IVUS; L = lithotripsy/shock; N = noncompliant balloon; OCT = ocular coherence tomography; o = ostial; R = rotational; S = scoring; U = ultra-high-pressure balloon; La = laser; Cx = circumflex; DCB = drug-coated balloon; DES = drug-eluting stent; ISR = in-stent restenosis; IVBT = intravascular brachytherapy; LAD = left anterior descending; LM = left main; OM = obtuse marginal; RCA = right coronary artery; XRT = radiation. A wide variety of interventional devices were used for the PCI-1 and PCI/IVBT interventions (Table 1). In general, more devices were used for the IVBT PCIs, although the type and number of instruments was highly variable and not amenable to meaningful statistical analysis. The PCI-1 procedures used an average of 2.4 devices (median, 2), compared with an average of 3.2 devices for the IVBT-related PCIs (median, 3). ISR-free intervals prior to IVBT ranged from 25 to 317 days (median, 96 days) compared with ISR-free intervals after IVBT that ranged from 30 to 612 days (median, 350 days; P < .001; log-rank test) (Fig. 4). As of this analysis, 5 vessels have restenosed after IVBT, at from 98 to 504 days. Median time to ISR preceding IVBT was 107 days, whereas the median time to ISR after IVBT has not been reached. All ISRs following IVBT occurred at the irradiated site.Figure 4Kaplan-Meier comparison of ISR-free intervals preceding versus following PCI/IVBT. The PCI-1 line represents freedom from ISR after the PCI for stenosis prior to the IVBT-treated ISR. PCI + IVBT shows the freedom from ISR for the same patients, following treatment with PCI + IVBT.Abbreviations: ISR = in-stent restenosis; IVBT = intravascular brachytherapy; PCI = percutaneous coronary intervention; RCA = right coronary artery.Figure 4 dummy alt text Kaplan-Meier comparison of ISR-free intervals preceding versus following PCI/IVBT. The PCI-1 line represents freedom from ISR after the PCI for stenosis prior to the IVBT-treated ISR. PCI + IVBT shows the freedom from ISR for the same patients, following treatment with PCI + IVBT. Abbreviations: ISR = in-stent restenosis; IVBT = intravascular brachytherapy; PCI = percutaneous coronary intervention; RCA = right coronary artery. Discussion The choice of intervention technique for patients with 2 or more same-site ISR has long been a subject of controversy because of the lack of RCTs comparing modalities. Prospective randomized trials testing IVBT would require a large number of patients, given the many potentially confounding factors in multi-ISR patients, prompting this N-of-1 study. When feasible, it could be argued that N-of-1 trials are preferable to conventional randomized studies and that the use of patients as their own controls addresses the challenge of multiple confounding factors. The results reported here support the use of IVBT to prolong the ISR-free interval in patients with multiple prior ISRs. That conclusion is, admittedly, confounded by the use of different instruments in the PCI-1 versus the PCI/IVBT procedures. It is possible that the use of more instruments for some patients receiving IVBT could explain their longer ISR intervals. However, the lack of clear benefit to any instrumentation over balloon angioplasty alone makes it unlikely that the instruments employed account for the favorable IVBT results.22,23 The results reported here call into question prior studies evaluating the effectiveness of IVBT compared with other modalities. The authors are unaware of any prospective trials comparing IVBT against other modalities in the setting of drug-eluting stents. The authors are aware of only 2 retrospective comparisons of outcomes of IVBT versus other interventions, both of which used matched-pair control groups. A 2018 publication from Mt. Sinai Medical Center reported that IVBT was more effective than contemporary, routine PCI.10 By contrast, a 2019 publication from Erasmus Medical Center reported that IVBT was no more effective than routine PCI.24 Prior to the results reported here, the effectiveness was open to question. The favorable IVBT results documented here may engender a debate regarding its effectiveness compared with recently introduced drug-coated balloons. There is, to date, no such comparison study. Regardless of relative effectiveness, an advantage of IVBT is the ability to treat ISR segments greater than 30-40 mm, which would require multiple drug-coated balloons. In fact, the authors view the 2 interventions as complementary rather than competitive. The study has several limitations. First, it includes heterogeneous patients, such that it may not be applicable to all complex ISR patients. Second, only the first 26 lesions in 19 patients were analyzed. Despite the limited patient numbers, the prolongation of ISR-free intervals was substantial enough that IVBT is quite likely to be at least somewhat effective—analyzing more patients is unlikely to change that conclusion. Third, the conclusion could change with longer follow-up. However, longer follow-up can only impact the IVBT results favorably, as the pre-IVBT ISR-free interval is fixed. Conclusion The results support the use of IVBT to prolong the ISR-free interval in patients with recalcitrant ISRs. Disclosures The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.