Patients with obstructive coronary artery disease (CAD) and ischemic symptoms are helped from coronary revascularization. Most would not argue that in the non-CTO population, angina status is improved post PCI. This reality forms the basis for the majority of elective ad-hoc PCI. Although challenged by some for not being hard outcomes, angina relief and improvement in quality of life remain valuable and safely achievable goals. This reasoning is rarely put into question for patients with severely narrowed vessels, but yet is often challenged for occluded vessels. A sense of security is present when collaterals are well formed, ignoring the fact that myocardium remains frequently and severely ischemic. Physiologic interrogation of the distal bed of a collateralized CTO with a pressure wire reveals significant pressure gradients before and after vasodilator stimuli [9]. Such occluded vessels derive the same benefits in term of normalization of the fractional flow reserve after PCI than severely narrowed but non-occluded vessels [10]. Thus patients often remain chronically ischemic and symptomatic despite collateralization of the occluded vessel(s), and must be maintained on an intensive medical regimen.

The majority of the observational studies where angina status was assessed reported improvement in Canadian Cardiovascular Society (CCS) classification. In an analysis of six studies that reported residual/recurrent angina, successful recanalization was associated with a significant reduction in residual/recurrent angina (OR 0.45, 95 % CI 0.30–0.67) when measured in a dichotomous fashion i.e. presence or absence of angina at follow-up [11]. In one of the observational studies, patients with successful recanalization were more likely to have a negative exercise test result at follow up (73.0 % vs. 46.7 %, P = 0.0001) [12]. In the prospective, randomized PRISON II trial comparing DES and BMS in CTO, the overall proportion of patients with CCS angina class ≥3 was reduced from 62 % at baseline to 25 % at 6 months after successful CTO PCI [13]. Quality of life was assessed in the FACTOR trial [14], which examined the health status benefits of successful vs. unsuccessful CTO PCI on symptoms, function, and quality of life. By using the Seattle Angina Questionnaire (SAQ), procedural success was independently associated with angina relief, improved physical function and enhanced quality of life 1 month after successful PCI. These findings were found mainly in the symptomatic patients, who derived large and clinically important improvement in reported health status. The magnitude of benefits was found to be similar to improvement of SAQ scores after surgical revascularization and PCI of non-CTO lesions. In a long-term registry of attempted CTO recanalizations, successful procedures were also associated with significant improvement in angina-related quality of life at a median follow up of 4 years [15]. Recently, a 10-center prospective registry confirmed the observation that symptoms and quality of life improve to the same degree in CTO PCI compared to non-CTO PCI [16].

The evidence thus far suggests angina reduction or disappearance with successful CTO PCI, as is frequently observed with non-CTO PCI. Quality of life, functional status, exercise capacity and medication reduction are all noble goals to consider in the care of patients with CTO (Fig. 1.1).

Historically, the presence of a CTO was a strong predictor of referral to coronary artery bypass grafting (CABG) compared to CTO PCI [5]; the purpose being a more probable achievement of complete revascularization (CR) with CABG [17]. This concept is not new as CR has been a long-standing objective in coronary revascularization. CR has been associated with better long-term outcomes with both PCI [18, 19] and CABG [20]. Most of the data however comes from the surgical literature, as surgical techniques have allowed for CR since almost their inception. Whether achieved by CABG or PCI, the goal of CR is rarely challenged unless significant co-morbidities are present to accept incomplete revascularization (IR).

The outcomes of CR have been studied in the different eras of coronary revascularization with mostly observational studies or subgroup analysis of RCT. A recent meta-analysis by Garcia et al. [21] included 35 studies and near 90,000 patients. CR was more often achieved with CABG than with PCI (75 % vs. 44 %), due to the historical difficulty in achieving CR with PCI. Overall, CR was associated with hard outcomes reductions, namely a 30 % reduction in long-term mortality and a 22 % reduction in MI. Importantly, similar mortality reductions were observed in both PCI and CABG-treated patients and were independent of the study design and definition of CR.

Designed a decade ago, the SYNTAX trial [17] compared the long-term outcomes of the two accepted and favored approaches for patients with complex multi-vessel disease i.e. CR with PCI or CR with CABG. A criterion for randomization was that equivalent anatomical revascularization could be achieved with either treatment, avoiding goal-oriented IR. CR was achieved more frequently in the CABG group (63 % vs. 57 % in PCI group) [22]. Also, the presence of a CTO was a strong predictor of IR in the PCI group (OR = 2.46, 95 % CI 1.66–3.64, P < 0.001) [22] reflecting the era before specialized technique and dedicated material for CTO recanalization. The residual SYNTAX score (rSS) defined as the delta between the baseline and post-revascularization SYNTAX score (SS) correlates with adversed long-term outcomes [23]. A rSS > 8 was associated with 35.3 % all-cause mortality at 5-years in the SYNTAX PCI cohort [24]. Not surprisingly, the presence of at least 1 CTO was observed in half (50.7 %) of patients with a rSS > 8 [24], suggesting a strong impact of failed CTO recanalization on the rSS. In a registry of patients with multi-vessel PCI, the rSS was an independent predictor of mortality, whereas the SS was not [25]. In a large all-comers DES registry, the rSS was associated with adverse cardiac events and larger rSS values were found in patients with multiple comorbidities such as diabetes, hypertension, previous PCI, and MI histories [26]; characteristics frequently found in patients with CTO. These findings suggest that the rSS is a reflection of residual ischemia burden. Significant ischemia is linked to adverse outcomes, and revascularization of patients with moderate-severe ischemia is linked to better prognosis [27, 28].

When assessing patients with multi-vessel disease and the presence of a CTO, one must assess the probability of achieving CR, which is most often derived from the probability of revascularization of the CTO. CR, as a pre-determined goal, is then a key determinant as to which revascularization approach is chosen. When multi-vessel PCI is decided, for example in low to intermediate SS, careful anatomical evaluation is done to decide the sequence of PCI procedures. Several factors come into play, such as the jeopardy score, complexity of lesions, collateral circulation etc. On many occasions the non-CTO vessels may be intervened upon first as to facilitate and improve the safety of CTO recanalization. The retrograde technique, for example, requires a reasonably healthy donor vessel which is often treated before the index contralateral CTO PCI. Hence, the subsequent CTO recanalization is then performed to achieve the pre-determined goal of CR, for the benefits described above. Failing to do so prevents the achievement of CR in patients with multi-vessel disease, in whom CABG might otherwise have been a more reasonable option.

As said, CABG is often selected over PCI for multivessel disease, especially when some arteries have total occlusions. However, the effectiveness of CABG for the revascularisation of CTOs is questioned. Although it is common for surgeons to perform CABG on totally occluded vessels, those arteries present some additional challenges compared to non-occluded vessels. As the surgeon needs to perform an anastomosis on the vessel distal to the occlusion, a CTO, which is usually associated with limited contra-lateral flow to the distal bed and significant negative remodeling (or vessel shrinkage) [29] will present additional difficulties to the surgical manoeuvres. In the large PRAGUE-4 trial, which compared off-pump vs. on-pump CABG on long-term graft patency, although all bypass grafts placed distal to a collateralized LAD CTO remained patent at 1-year, only 23 % of those grafts remained patent when placed on the LCX or the RCA system [30]. More recently, in the SYNTAX trial, 543 patients were randomized despite the presence of a totally occluded vessel. Of the 266 patients randomized to surgery, 32 % of the totally occluded vessels finally never received a graft, leading to an incomplete revascularisation. Reasons for not bypassing the totally occluded vessel were not specified in close to half of patients. Otherwise, reasons quoted by surgeons for not grafting the vessels were multiple, including a too small vessel, a too diseased vessel, having no intention to graft, etc. In the end, an incomplete revascularisation procedure, either with PCI or CABG, in the presence of a totally occluded vessel, was associated with increased mortality [31]. In summary, many CTOs will not receive a graft when referred for surgery. And in the cases where a graft was placed, there is evidence that long-term patency is poor. Therefore, effectiveness of CABG specifically for CTOs is questioned, and should be better studied.

Thus far, the data of CTO PCI on hard outcomes is unfortunately only supported by observational studies comparing successful and unsuccessful CTO PCI. The majority of these studies have found improved survival at long-term follow-up after successful CTO PCI. Four meta-analyses, designed alike, have all yielded similar findings in terms of reduction of mortality [11, 32–34]. A summary of their major findings is depicted in Table 1.1. Additional studies have since been reported, adding to the pool of studies, but not to the quality of the data. The latest study includes data from the U.K. Central Cardiac Audit Database [35], which analyzed outcomes in over 14,000 CTO PCI procedures. Successful CTO PCI was again associated with improved survival (hazard ratio [HR]: 0.72; 95 % CI: 0.62–0.83; p < 0.001), with the biggest survival advantage found in patients with complete, compared to those with partial or failed revascularization.

The major limitations of these observational studies are the presence of unmeasured confounders, with the unsuccessful CTO PCI patients likely representing a higher risk group. There will always remain a certain degree of bias that cannot be fully negated even by the statistical adjustment for the most common confounders. Patients with complex CTOs often have complex non-CTO lesions, with high atherosclerotic burden, inferring worse prognosis irrespective of the CTO itself.

However, the current body of evidence does suggest a survival benefit with CTO PCI. Some have found the survival benefit to be confined to LAD CTO PCI [36], while others have found it to be related to both LAD and CX CTO PCI [37]. The LAD supplies the largest area of myocardium and its patency has the largest effect on ventricular function and electrical stability. Proximal LAD occlusion are frequently associated with greater than 10 % of ischemic myocardium, and this threshold has been demonstrated to confer worsen prognosis [27]. It is likely and intuitive that the greatest survival benefit be associated with successful LAD (ideally proximal) CTO PCI. Until conclusive data is available, a recanalized proximal LAD supplying a large area of ischemic myocardium must be perceived as providing a better prognosis.

The presence of a CTO in a non-infarct related artery is associated with increased mortality in both patients with STEMI [38] and NSTEMI [39]. In patients with MVD and myocardial infarction, the presence of a CTO in a non-infarct related artery (IRA) is the main driver of increased mortality, while MVD without a CTO is a weaker predictor [40]. In patients undergoing primary PCI, markers of reperfusion such as ST segment resolution, TIMI −3 flow and myocardial blush are affected by the presence of a CTO [41]. The presence of a CTO in a non-infarct related artery is also predictor of hemodynamic instability and cardiogenic shock. In patients with cardiogenic shock, both MVD and the presence of a CTO affect short-term mortality, but long-term mortality is mostly affected by the presence of a CTO [38]. These findings underscore the relative fragility of the safety net provided by chronic collateral supply to a major occluded epicardial vessel, leading to coronary inter-dependence. Acute loss of a donor artery often leads to myocardial infarction in multiple interdependent territories.

Several trials are underway to determine the long-term outcomes of CTO PCI compared to optimal medical management. The EXPLORE trial [42] is evaluating the value of recanalization of a non-IRA CTO after primary PCI. The primary endpoints are left ventricular function and left ventricular end diastolic volumes at 4 months, with clinical follow up at 5 years, which should provide insight on clinical outcomes. Two trials, in different geographical regions, are assessing hard outcomes on CTO PCI vs. optimal medical therapy (OMT). The EuroCTO trial is randomly assigning 1200 European patients, with a primary endpoint of QOL at 1 year and a cumulative composite end point of all-cause death, non-fatal MI at 3 years. The estimated completion date is in 2017 but may be delayed based on slow recuitment. The DECISION-CTO is randomly assigning 1284 Asian patients, with a primary endpoint of composite outcomes of all cause death, myocardial infarction, stroke, and any revascularization at 3 years. The estimated completion date is in 2018.

Until RCT data are available to settle the controversy on survival from CTO PCI, sound clinical decision-making must be made in each patient with a CTO when considering the potential impact on survival. The extent of ischemia, the location of the CTO and the plaque burden and risk of plaque rupture in the donor vessel must be taken into account.

Chronic ischemia related to a CTO can cause LV dysfunction, and may lead to exercise intolerance and finally heart failure. It therefore seems logical that the opening of an occluded artery, which irrigates dysfunctional myocardium, could reverse this dysfunction. Several studies have assessed the effects of CTO PCI on LV function and remodelling. In these studies, statistically significant improvement in regional wall motion, global LV ejection fraction (LVEF) and/or decrease in LV volumes have been demonstrated at 5–6 months after the procedure [43–51]. However, the degree of LVEF improvement was generally small, typically <5 % [44–49]. The greatest LVEF gains have been limited to patients with a patent target vessel at follow-up, those without prior MI in the distribution territory of the CTO, and those with baseline regional or global LV dysfunction [44–46]. The improvement does not appear to depend on the presence of pre-existing collaterals, but probably on preserved microvascular integrity [46]. Improvement in regional and global LV function after CTO PCI has also been related to the extent of baseline transmural necrosis, as shown in studies using contrast-enhanced magnetic resonance imaging (MRI) [48–52].

The combination of multiple MRI derived viability parameters including dobutamine contractile reserve assessment, transmural extent of infarction (TEI), and segmental wall thickening (SWT) of normal residual myocardium was shown to reduce the proportion of false-positive patients, that is, patients with viable myocardium but without improvement in LV function after successful CTO PCI [49]. Indeed, this combination of parameters was a better predictor of improvement of dysfunctional segments than the single widely used parameter of TEI [48]. The expected beneficial prognostic effect of CTO PCI is thought to be associated with the amount of ischaemic myocardium, as has been observed in patients with CAD in general [27, 53, 54]. However, CTO–PCI might be beneficial in some cases despite the absence of ischemia. In one study, patients with successful CTO–PCI of the LAD coronary artery (n = 99) were stratified according to the presence of perfusion defects on nuclear imaging before the procedure [55]. Both those with reversible (n = 40) and those with fixed (n = 50) perfusion defects had significant improvement at 1 year in perfusion abnormalities (−20 %, P = 0.001 and −15 %, P = 0.041, respectively), LVEF (6 %, P = 0.002 and 4.1 %, P = 0.006), quality of life measured as improved 6 min walking distance (~50 m, P < 0.05 and ~25 m, P < 0.05), and frequency of angina measured with the SAQ (mean score 18, P <0.05 and mean score 15, P <0.05). No benefit of CTO–PCI was observed in patients who had no perfusion defects (n = 9) [55].

Currently, limited evidence is available to show that myocardial electrical stability is improved after successful CTO–PCI. However, in patients with an implantable cardioverter–defibrillator (ICD) for ischaemic cardiomyopathy (n = 162), a CTO was significantly associated with ventricular arrhythmias requiring ICD therapy (HR 3.5, 95 % CI 1.5–8.3, P = 0.003) [56]. Two previously established arrhythmogenic factors might be responsible for the ventricular tachycardia: ischemia owing to inadequate perfusion of the myocardium can lead to abnormal automaticity of the ventricular myocardial cells, and re-entry circuits in patients with a previous myocardial infarction and fibrous tissue interspersed with islands of viable tissue [57]. Restoring antegrade flow after successful CTO–PCI could resolve the ischemia and might, therefore, enhance electrical stability in patients with ventricular arrhythmia, regardless of the presence of an ICD, which only treats the arrhythmic defect and not the cause of ischemia.

Despite procedural complexity, increased operator volumes have led to an improved success rate of CTO-PCI from approximately 68 to 85 % [58–62], and even higher success rates with highly trained operators [59, 63–67]. This improvement is accompanied by a low risk of procedural complications, regardless of procedural success [68]. These figures are similar to non-CTO procedures, with the exception of a significantly increased use of contrast agent and fluoroscopy time [60].

Most of the potential procedural complications are clinically uneventful; the in-hospital major adverse cardiac events (MACEs) after elective CTO PCI range from 0.9 to 6.5 % [43, 60, 68]. However, recent reports from registries performed by experienced operators showed a similar in-hospital MACE rate for CTO PCI as compared to non-CTO interventions. Of note, a higher reported in-hospital mortality rate among patients in whom the PCI of the CTO failed, ranging from 1.0 to 2.6 %. The 30-day procedure-related outcomes of CTO PCI are shown (Table 1.2) and are thus far reported in only two studies [60, 69]. Thirty-day mortality in the overall CTO cohort regardless of procedure success was 1.1 %.

Hence, under experienced hands, CTO PCI have high success rate nowadays with an acceptable rate of meaningful complications. Dedicated material, improved techniques and dissemination of knowledge within the CTO community have led to safer and more efficient procedures.

Contrary to popular belief, current American and European guidelines generally support CTO PCI for ischemic patients. Although worded differently, both the ACCF/ACC/SCAI and ESC guidelines give a Class IIa indication for CTO PCI [70, 71]. The ACCF/ACC/SCAI guidelines require “appropriate clinical indication”, “suitable anatomy” and “appropriate expertise”. The ESC guidelines are more liberal by mainly mentioning an “expected ischemia reduction”. The ACCF/SCAI/STS/AATS/AHA/ASNC appropriateness criteria provide several scenarios to help guide in the decision making to recanalyse a CTO [70]. The 18 scenarios are based on degree of symptoms, ischemia quantification and intensity of medical regimen. However, the criteria mainly pertain to isolated CTOs, as they fail to address the issue of completeness of revascularization in patients with MVD and a CTO. Regardless, in more than 2/3 of the clinical scenarios CTO PCI is rated as either “appropriate” or “uncertain”, leaving a large role to clinical judgement and discussion with patients regarding the expected benefits. A summary of the clinical guidelines and appropriateness criteria are provided in Tables 1.3 and 1.4 respectively.