The use of new-generation drug eluting stents (DES) has decreased the risk of in-stent restenosis (ISR), but it remains a fairly common and challenging problem with an incidence up to 10% [1]. In the United States, 1 in 10 percutaneous coronary interventions (PCI) are done to treat ISR [1]. In contemporary practice in Europe, and soon in the United States, common treatment options include additional DES or drug-coated balloons (DCB), with both of these proven superior to balloon angioplasty or bare metal stenting. Controversy remains regarding the superiority of DES over DCB. In a meta-analysis of 10 clinical trials including 1976 patients, DCB was associated with an increased risk of target lesion revascularization (TLR) at 3 years [hazard ratio (HR) 1.32, 95% CI 1.02–1.70, P = 0.035], but the primary safety endpoint of all-cause death, myocardial infarction, or target lesion thrombosis showed no difference between the treatments (HR 0.80, 95% confidence interval (CI) 0.58–1.09, P = 0.152) and long-term all-cause mortality was also not different (HR 0.81, 95% CI 0.53–1.22, P = 0.310) [2]. A more recent study following 402 patients with a median follow-up over 10 years showed no difference in TLR between patients who underwent DES vs. DCB (HR = 1.54 [0.89–2.69]) at 1 year and (HR = 1.26, CI 0.82–1.92) after 1 year [3]. While in general DES may have an edge over DCB, patients with multiple layers of stents are an exception, since ≥ 3 layers of stents correlate with major adverse cardiovascular events (MACE) [4]. Current recommendations are to avoid another layer of DES in patients who already have more than 1 layer of stents [5].

In this issue of Cardiovascular Drugs and Therapy [6], Zhang et al. report the results of a retrospective cohort study comparing patients with ISR and chronic total occlusion (CTO) who were treated with DES vs. DCB, specifically paclitaxel-coated balloons. They retrospectively analyzed data for 214 patients who underwent ISR-CTO treatment at their center, (65% DES and 36% DCB). Most patients were men (84%). Baseline characteristics were comparable between the two groups except for more stent layers in the DCB group. At a median follow up of 39 months, no significant difference in the incidence of MACE was found between DES and DCB (26.5% vs. 28.2%, p = 0.78), nor was DCB associated with MACE in multivariate analysis (HR 1.25, CI 0.64–2.46). The authors conclude that DCB is an effective alternative treatment to repeat stenting with DES for ISR-CTO, offering similar long-term outcomes.

Strengths of this study include propensity analysis to minimize confounding, although the DES and DCB groups were similar at baseline except for preferential use of DCB in left anterior descending lesions and in patients with more previous stent layers. Limitations of the study include those common to all cohort studies. Use of intravascular imaging (ultrasound or optical coherence tomography) was low (15%) and less than recommended for treatment of re-stenotic lesions. Preferential use of DCB in patients with multiple prior stent layers may have biased results against the DCB group. Bias against the DES group may have been caused by preferential stenting in patients where angioplasty of the ISR-CTO lesion caused dissection or inadequate stent expansion, risk factors for repeat restenosis.

The authors also noted that patients treated with a ‘drug-switching” strategy (i.e., ISR treated with a different drug [on stent or balloon] than the initial stent drug) had lower MACE (HR 0.45, CI 0.24–0.85, P = 0.014) compared to a non-switching strategy. When the DES arm was considered alone, drug-switching still showed an advantage (HR 0.35, CI 0.15–0.79, P = 0.012). However, this finding should not be considered definitive because it was not a pre-specified analysis and prior studies have produced conflicting results regarding the benefit of a drug-switching strategy [7,8,9].

An important principle in business is also true for coronary intervention: the final choice of strategy may matter less than the quality of performance of that strategy. The authors remind us of several important aspects of high-quality treatment of in-stent restenosis:

  1. 1.

    Use intra-vascular imaging to assess the lesion and the stent in the lesion.

  2. 2.

    Optimize expansion of the previous stent, where necessary, using intravascular lithotripsy, rotational or laser atherectomy, or high-pressure balloon dilatation.

  3. 3.

    Prepare the lesion before stenting or DCB (or even brachytherapy). Use of a scoring balloon for this has been associated with lower re-restenosis rates [10].

  4. 4.

    After stenting or DCB, use intravascular imaging again and optimize the final result as needed.

In summary, this study confirms for ISR CTO what has become clear for ISR non-CTO lesions: DCB is a reasonable alternative to adding another layer of stent, and the attractiveness of the DCB option corresponds directly to the number of previous layers of stent. While DCBs are a valuable tool for interventionalists in the battle against ISR the war against ISR continues. ISR-CTO patients continue to have poor long-term outcomes even with advances in balloon and stent technology. New innovative strategies are needed to improve long-term results, with prospective randomized trials assessing the nuances of different treatment strategies.