Small vessel disease and diffuse coronary artery disease represent particularly challenging subsets for treatment with transcatheter coronary interventional therapies. This pathology is associated with higher risk comorbidities, such as diabetes, and is more frequently associated with female gender and diffuse coronary involvement.¹ Small and diffusely diseased vessels are frequently noncompliant, calcified, tortuous, and distal in location, making these targets more technically challenging for intervention. Consequently, a higher incidence of acute complications, including significant vessel dissection, acute vessel closure, myocardial infarction, and emergent coronary bypass grafting, has historically complicated intervention in small and diffusely diseased vessels.^2,3 Along with poorer acute outcomes, these subsets are plagued by high restenosis and thrombosis rates, often necessitating repeat intervention or bypass surgery.^4,5 Despite the challenges and complexities inherent to small vessel intervention, the problem is common. Between 30% and 67% of all percutaneous coronary interventions involve small vessels, depending on the definition of a small vessel.^6-9

This chapter will review the definition and diagnostic approach to small and diffuse coronary artery disease. Acute and long-term outcomes associated with percutaneous interventional therapies, particularly the favorable impact of drug-eluting stents, will be reviewed. Finally, the future role in the treatment of these subsets with drug-eluting balloons and bioabsorbable stents will be examined.

The definition of what constitutes a small vessel in the context of coronary interventional therapy has been quite variable. A number of studies examining small vessel intervention have defined vessels less than 3.0 mm in reference size as being small, and this criterion is in part derived from early stent trials in which patients with vessels less than 3.0 mm were excluded from enrollment.^10,11 Other studies have defined small vessels as those less than 2.5 to 2.8 mm in diameter.^9,12-14 Certainly, less than 3.0 mm is the most sensitive descriptor, although necessarily less specific. Despite whatever arbitrary cutoff is used, the relationship between vessel size and outcome is not a step function; instead, a continuous inverse relationship exists between outcomes and size.^8,15 Consequently, differences in the definition and enrollment criterion frequently help explain apparently conflicting results from various studies of small vessel intervention. Like the small vessel, diffuse disease has had various definitions; however, the most consistent definition is lesion length greater than 20 mm.¹⁶

The definition of small vessel has been based on angiography; nevertheless, results of intravascular ultrasound (IVUS) have demonstrated that many angiographically small vessels are in fact “pseudosmall” as a result of angiographically undetected disease in the reference segment and positive remodeling at the lesion site.¹⁷ Briguori and colleagues¹⁸ compared 344 consecutive patients having 419 lesions in small vessels (angiographic reference vessel ≤2.75 mm) with 953 patients having 1161 lesions in large vessels (>2.75 mm); all underwent concomitant IVUS. The difference in angiographic and IVUS reference vessel dimension delta (IVUS-angiography) was calculated for all lesions.¹⁸ The tendency to underestimate vessel size was significantly greater for small vessels (difference between IVUS and angiography, 1.3 ± 0.5 mm vs 1.0 ± 0.6 mm; P < .001). Angiography underestimated vessel size by more than 1.0 mm in 71% of small vessels and in only 49% of large vessels (P < .001). There was a stronger correlation between plaque burden and the delta (IVUS-angiography) in small vessels (r = 0.80; P < .001) than in large vessels (r = 0.59; P < .001). Predictors of delta (IVUS-angiography) were proximal or middle lesion location and female gender.¹⁹ Extending these observations, Moussa et al¹⁹ examined the predictors of large discrepancies between IVUS and angiography. Independent predictors of a delta (IVUS-angiography) greater than 1.0 mm were small vessel size (<3.0 mm), location of the lesion in the proximal vessel, and diabetes (Fig. 39-1). The authors suggested that IVUS might be particularly helpful in small vessel intervention when the lesion was proximally located or the patient had diabetes.

Early studies of percutaneous transluminal coronary angioplasty (PTCA) demonstrated higher risks associated with small vessel intervention.²⁰ Overall procedural success with plain old balloon angioplasty (POBA) was significantly less for small vessels compared to interventions on larger vessels.¹ Additionally, major adverse cardiac events occurred more frequently in small vessel cohorts. Hypotheses explaining the higher event rates include the association of small vessels with other high-risk comorbidities, greater lesion complexity, and the higher incidence of diffuse plaquing. The combination of small vessel size and diffuse disease is particularly difficult, with a much higher incidence of abrupt closure compared to focal lesions. Strategies that have been shown to be beneficial in reducing these complications and improving restenosis rates include utilization of longer balloons, gradual and prolonged balloon angioplasty, and cutting balloon (CB) angioplasty. Data from the REDUCE (Restenosis Reduction by Cutting Balloon Evaluation) and CAPAS (Cutting Balloon Angioplasty Versus Plain Old Balloon Angioplasty Study) trials collectively suggested that CB angioplasty may be a superior technique to POBA, with fewer coronary dissections, improved binary restenosis rates, and reduced target lesion revascularization.^21,22

Because the results of POBA had been demonstrated to be suboptimal, interventional therapies using ablative techniques, particularly rotational atherectomy and laser angioplasty, have also been evaluated. Several studies done in the early 2000s, including a meta-analysis of trials examining PTCA versus atherectomy, CB angioplasty, or laser angioplasty, demonstrated higher acute complication rates without restenosis benefits in patients who underwent these procedures.^23-25 Despite limited data to suggest plaque ablative techniques should be used routinely to treat small, diffuse disease, it continues to have a niche role in heavily calcified vessels, facilitating more complete dilation and delivery of stents.²⁵

Early reports of stenting in small vessels suggested higher rates of complications and no improvement in restenosis.^26-28 Diffuse disease treated with multiple or long stents was also associated with a higher risk of stent thrombosis as well. A meta-analysis of the Belgian-Netherlands Stent (BENESTENT) study and the Stent Restenosis Study (STRESS) suggested no improvement in restenosis in small vessels when elective stenting was compared with balloon angioplasty.²⁹ These reports led to initial American College of Cardiology/American Heart Association recommendations that small vessels should not be routinely stented.³⁰

Despite these mediocre results, subsequent reports of bailout stenting in the setting of abrupt or threatened closure in small vessels were encouraging.^31,32 An analysis of small vessel lesions enrolled in the STRESS I and II trials suggested that in vessels less than 3.0 mm, elective stenting was beneficial.¹⁰ Angiographic restenosis was significantly less in those receiving elective stenting (34% vs 55% restenosis for stent vs balloon; P < .0001), and there was no difference in abrupt closure between the POBA and stent group (3.6% in both groups). Many other elective registries also suggested favorable outcomes with stenting when compared to historical controls.^33-42

Following these early reports, there have subsequently been a large number of randomized trials comparing bare metal stenting to PTCA.^9,13,42-50 There has been considerable variability in the results of these studies. In 2004, a meta-analysis of these randomized trials was performed and included 3541 patients; 1672 patients were allocated to balloon angioplasty, and 1869 were allocated to bare metal stenting.⁵¹ Of these, 84% had angiographic follow-up. The pooled restenosis rate was 25.8% in patients assigned to stenting and 34.2% in patients allocated to balloon angioplasty (relative risk, 0.75; 95% confidence interval [CI], 0.67-0.84; P < .001). A smaller reference vessel diameter at baseline was associated with higher risk reduction of restenosis in the stent group (P = .012). In summary, these studies suggested that stenting improves restenosis in small vessels, particularly when PTCA outcomes are not optimal and possibly when reference vessel size is smaller.

Stent Strategies

Initially, the treatment of diffuse disease, particularly in small vessels, was discouraging because of higher rates of stent thrombosis and restenosis. Subsequent studies suggested more favorable results. In long lesions, Serruys et al⁵² showed the superiority of IVUS-guided stenting compared to optimal balloon angioplasty in the Additional Value of NIR Stents for Treatment of Long Coronary Lesions (ADVANCE) trial.⁵² In this study, 437 patients with lesions 20 to 50 mm in length were angioplastied with long balloons; 149 of the patients (34%) required bailout stenting for flow-limiting dissections. The remaining 288 patients who achieved an optimal PTCA result were randomized to IVUS-directed stenting or no further therapy. The restenosis rate in the stent group was significantly lower than in the PTCA-only group (27% vs 42%; P = .022), with no difference in early complications. Using IVUS guidance, Colombo et al⁵³ described a technique whereby the vessel was aggressively dilated and segments within the lesion with minimal lumen areas less than 5.5 mm received spot stenting. Compared to usual stenting, spot stenting was associated with no difference in 30-day major adverse cardiac events (MACE) but with significantly less restenosis (25% vs 39%; P < .05). Although single-center reports of spot stenting have been encouraging, the time-consuming and operator-dependent technique has not been embraced as a mainstream strategy.

Another strategy examined in the setting of small vessel intervention has been direct stenting.⁵⁴ Garcia and colleagues⁵⁵ randomized 350 patients with reference vessel sizes between 2.2 and 2.7 mm to either direct stenting or predilation followed by stenting using the Pixel stent (Abbott Vascular, Santa Clara, CA). Of those assigned to direct stenting, 83% were successfully stented without predilation. Binary angiographic restenosis at 180-day follow-up was 16% for direct stenting versus 25% for the control (P = not significant [NS]). Target lesion revascularization was low in both groups (3.4% vs 4.3% for direct stent vs control; P = NS).

Finally, the use of glycoprotein IIb/IIIa inhibitors as a routine strategy in small vessel intervention has not been shown to reduce restenosis. Hausleiter and colleagues⁴⁵ randomized 502 patients in a 2 × 2 factorial design to stent (phosphochlorine-coated) versus PTCA and abciximab versus control. There was no difference in 30-day MACE, restenosis, or target vessel revascularization (TVR) between the abciximab and control groups.

First-Generation Drug-Eluting Stents

The fundamental problem with stenting in small arteries is that the amount of neointimal hyperplasia is independent of vessel size, with small arteries suffering on average the same amount of late loss (0.8 mm) as larger vessels.⁸ Because of lower postprocedure stent area, small vessels are less able to accommodate neointimal accumulation than larger arteries. Lesion length independently exacerbates the negative impact of smaller vessel size.⁸ Even when in-stent minimal vessel areas are large, small vessels have higher rates of TVR than larger vessels with similar postprocedural minimal in-stent lumen areas (Fig. 39-2).⁵⁶

The solution to improving long-term outcomes in small and diffuse coronary artery stenting is to reduce neointimal proliferation, thereby attenuating late lumen loss. The site-specific application of antiproliferative agents such as sirolimus, paclitaxel, everolimus, and zotarolimus dramatically and unequivocally reduces restenosis in larger vessels.^57-64 Subset analysis from randomized trials and subsequent small vessel–specific studies have demonstrated remarkable effectiveness in the setting of small vessel disease as well. Their impact on diffuse disease has been equally favorable.

An early analysis compared the use of sirolimus-eluting stent (SES) versus balloon angioplasty.⁶⁵ In vessels with a reference size less than 3.0 mm, the effectiveness of the SES in the SIRIUS (Sirolimus) trial was compared to the results of balloon angioplasty in patients treated in BENESTENT I and II and STRESS. Subacute occlusion was 0.4% versus 1.4% (SES vs POBA), and the SES reduced target lesion revascularization (TLR) from 24% to 5.1% (difference, –18.9%; 95% CI, –23.5% to –14.3%), suggesting that the drug-eluting stent was far more effective than balloon angioplasty.

In the evaluation of first-generation drug-eluting stents (DESs) on small vessel disease, the SIRIUS trial demonstrated that the Bx-VELOCITY SES (Cordis, Hialeah, FL) dramatically reduced in-stent late loss in smaller vessels with some attenuation of the effect over the analysis segment due to proximal edge restenosis (Fig. 39-3).⁵⁸ TVR was reduced in a similar fashion, although the relative reduction was somewhat less for smaller vessels. Likewise, the results from the e-SIRIUS and c-SIRIUS trials, looking at the benefits of CYPHER stents (Cordis) in small vessels compared to the Bx-VELOCITY bare metal stents,⁶¹ demonstrated decreased late lumen loss across all vessel sizes, particularly small vessels (~2.2 mm), of 0.12 mm. The lack of edge restenosis shown in these studies suggests that avoiding geographic miss and stenting “normal to normal” is particularly important in small vessels.

Like the SES, the paclitaxel-eluting stent (PES) has been shown to be effective in small vessel disease.⁶⁰ Results from the TAXUS IV (Treatment of De Novo Coronary Disease Using a Single Paclitaxel-Eluting Stent) trial demonstrated substantial reductions in in-segment restenosis, with no edge restenosis and with concomitant reductions in TVR and TLR (Fig. 39-4). However, older trials comparing PES with SES in small vessel and diffuse coronary disease were inconsistent. The REALITY trial randomized patients having a mean vessel diameter of 2.4 mm and a lesion length of 17 mm to either the TAXUS (Boston Scientific, Marlborough, MA) or CYPHER SES stent.⁶⁶ Although the in-stent late loss was significantly less in the CYPHER group, there was no overall significant difference in binary angiographic restenosis and TLR; in particular, neither stent was superior in the subgroups of “long lesions” and “small vessel.” In contradistinction, the SIRTAX (Sirolimus-Eluting Versus Paclitaxel-Eluting Stents for Coronary Revascularization) trial exhibited that the angiographic restenosis and TLR associated with CYPHER was significantly lower overall in small vessels and long lesions compared to the PES.⁶⁷

Newer Generation Drug-Eluting Stents

The benefits of newer generation DESs in treating coronary stenosis by percutaneous transcatheter intervention compared to older therapies extend to the treatment of small vessel coronary disease, as well. The newer generation everolimus-eluting stents (EESs) have particularly been demonstrated to be effective in reducing clinical and angiographic events in diffuse, small vessel disease. The XIENCE V USA study was a “condition of approval” postmarket study involving 5054 patients to evaluate the safety of XIENCE V (Abbott Vascular), everolimus-eluting platform in a real-world setting.⁶⁸ Two cohorts were identified; a small vessel cohort comprising 1869 patients who received at least one 2.5-mm XIENCE V stent and a non–small vessel cohort comprising 2905 patients who received XIENCE V stents larger than 2.5 mm. Baseline characteristics were similar between both patient populations, with the exception that small vessel patients were more likely to be women; had a higher prevalence of diabetes, hypertension, and anemia; and were more likely to have had a prior myocardial infarction, multivessel disease, and prior coronary artery bypass graft. Despite these differences in baseline characteristics and clinical risk factors, 1-year clinical outcomes were similar in both the small vessel and non–small vessel groups, which included stent thrombosis rates, and composite rates of cardiac death or Academic Research Consortium–defined myocardial infarction. TLR was higher in the small vessel group (4.9% vs 3.8%).

More consistent differences between drug-eluting platforms were demonstrated between EES and PES in a subset in a post hoc analysis of the SPIRIT (Clinical Evaluation of the XIENCE V Everolimus Eluting Coronary Stent System in the Treatment of Patients with de novo Native Coronary Artery Lesions) III trial, which sought to compare clinical and angiographic outcomes in patients treated with 2.5-mm XIENCE V (EES) or TAXUS stents; lower rates of late loss were observed in vessels treated with EES.⁶⁹ The subgroup of patients in the original trial who received at least one 2.5-mm stent consisted of 160 patients (total of 190 stents implanted) in the XIENCE V arm and 59 (total of 67 stents) in the TAXUS arm, with a mean vessel diameter of 2.36 ± 0.3 mm and 2.34 ± 0.33 mm in the XIENCE V and TAXUS arms, respectively. In-hospital and 30-day event rates for MACE and target vessel failure (TVF) were low and similar in both groups. Through 9 months of follow-up, however, TVF and MACE were significantly lower in the group receiving EES compared to PES. The difference in MACE was driven mainly by a reduction in TLR (1.3% vs 12.5%; P = .0016). Follow-up angiography also demonstrated a significant difference in in-stent and in-segment late loss between the stents, and binary in-segment angiographic restenosis was significantly lower for the EES group (4.1% vs 20.8%; P = .02; Fig. 39-5).⁶⁹