Coronary angioplasty was not the first interventional procedure in cardiology (Table 1), but it clearly launched the discipline called interventional cardiology as we know it today.
Percutaneous coronary intervention (PCI), as it is customarily called today, had a slow start. Three years into its existence, the world experience still totaled less than 1,000 interventions (Table 2). Andreas Grüntzig showed impressively and unmistakably that PCI could be applied in a successful and safe manner even before the development of stents. The first case showing a good result up to the current day (Figure 1) may be considered anecdotal. However, Andreas Grüntzig’s lifetime series, cut short by his demise 5 months before the first coronary stent implantation, 1 confirms it unequivocally. He and his close collaborators performed more than 2,000 balloon-only procedures without any in-hospital mortality (Table 3).

PCI has seen lots of adjunctive therapies and even so-called balloon replacement therapies over the past 30 years. Few have prevailed, and even today, what is really important boils down to 5 components: state-of-the-art balloons, guidewires, guiding catheters, X-ray equipment, and stents. Add to these novel antiplatelet agents and perhaps percutaneous left ventricular assist devices that save an occasional life in large-volume centers (Table 4). Techniques to close the femoral puncture site are important comfort items for both physician and patient. All of the other presumable assets were either useless (some even dangerous) or remain completely optional. Nonetheless, many of them have made it into routine use in most catheterization laboratories.
Even the coronary stent is clearly overused these days. First it went through a difficult initial phase where its 20% thrombosis rate was not sufficiently corrected for the situations in which the stents were used (bailout for impending or established abrupt closure), but rather compared with the 7% thrombosis rate of plain balloon angioplasty. This bane vanished when stents were implanted mostly in elective cases. Erroneously, the drastically decreasing stent thrombosis rate was not initially attributed (as it should have been) to the transition from bailout to elective stenting, but rather to changes in implantation techniques (higher pressures) and the addition of a second antiplatelet agent (thienopyridine). The latter contributed to improvement in stenting results, but only to a minor degree.

The community of angioplasty operators then went to the other extreme and made stenting the default procedure, turning a completely blind eye on the fact that without stents, at least 70% of lesions never had a problem, neither acutely nor during follow up. In these 70% of patients (who of course never ceased to exist), a stent cannot be of any benefit, but it can harm during the “unnecessary” implantation, or it can harm during follow up by creating in-stent restenosis that is much more difficult to treat than restenosis after plain balloon angioplasty or by late thrombosis, unseen after balloon angioplasty. As soon as default stenting was ubiquitously adopted, the initial advantage of stenting in terms of survival and avoidance of myocardial infarctions due to abrupt vessel closure was lost again during mid-term follow up. The advent of drug-eluting stents basically eliminated the problem of in-stent restenosis. Strangely, when for a while their safety was questioned, people considered a fallback to bare-metal stents, but not to plain balloon angioplasty, though only the latter completely precludes late thrombosis.

A temporary setback of PCI is currently witnessed in light of the falsely-accused drug-eluting stent and the falsely-interpreted comparisons between PCI and conservative treatment in the Mass II,2 PET and COURAGE trials.4 These trials found no difference in survival or instances of myocardial infarctions in stable coronary patients treated with either PCI (i.e., bare-metal stenting) or medical therapy plus PCI only when symptoms did not abate. The interpretation that this meant PCI was not warranted cannot be condoned. In contrast, the trials should be interpreted such that in spite of having repaired the problem in the group that underwent PCI, the mid-term outcomes were not worse, although the PCI-arm patients already had all the possible problems from the intervention behind them, whereas the conservative group still had them ahead of them. In other words, if you know your brakes need fixing, you may well drive on for a certain time without fixing them. The fact that you did not have an accident does not mean that your brakes never needed fixing.
These storms in a teacup could, however, have called us back to reason that, yes, we should do the interventions but, no, we should not fall victim to the temptation of overdoing things. The case in Figure 2 shows that interventional cardiology can do complex things with rather frugal techniques and materials.5 Table 5 is a summary of procedures conducted in this patient and the material used. The patient was released 3 days postprocedure.