Placing a new technique into clinical practice

Diseases of the heart valves are common, affecting an increasing number of patients as the population ages. Until recently, definitive treatment of heart valve disease has required surgical repair or replacement of the affected valve, usually at cardiopulmonary bypass. Advances in catheter techniques have introduced nonsurgical options for treatment, which will probably become feasible for all heart valves in the coming years. Major clinical and economic factors will ultimately influence whether such novel techniques will be accepted into clinical practice. However, the determinants of the success of the procedure is difficult to measure. It depends on the interplay of multiple parameters, such as the risks and the proven and potential benefits compared with established surgical therapy. The parameters emerging for aortic [1], pulmonary [2] and mitral [3] valve and related clinical situations vary and lead to complex problems in the strategy to ultimately bring new technology into accepted clinical practice. We would like to outline, in this editorial, our own experience of percutaneous pulmonary valve implantation (PPVI), which represents the longest clinical experience and the largest coherent patient population in the field of transcatheter valve implantation [2].

When a new medical device becomes available, the most efficient match between the new technology and the patients’ need must be determined. Therefore, two fundamental steps are necessary. First, what the device is expected to achieve must be defined. Second, the patient population has to be selected carefully, so that it can be proven whether the device performs according to the intended benefits.

In our experience, we intended to prolong the surgical result (conduit life) of patients who undergo right ventricle (RV)-to-pulmonary artery (PA) conduit placement and, over time, developed conduit stenosis or regurgitation. Therefore, this definition of intended benefit did not allow us to make a direct comparison between our results and the results of surgery, despite surgery being indicated for our patients.

Patient selection, for the percutaneous aortic programs, included patients with severe aortic valve disease, who were unable to undergo alternative surgical treatment through surgery owing to formal surgical contraindications [1]. By contrast, we had conversely selected patients who could undergo a surgical alternative to the treatment we proposed. This allowed us to work with surgical backup, which could add further safety to our procedures should complications occur, therefore minimizing the risks and increasing the benefits. As a rationale in decision making, we first judged technical feasibility and accepted that clinical indications might be infrequent. Most of our patients underwent RV-to-PA homograft insertion, which represents the preferable technical environment to deploy the valved stent safely. The implantation site did not exceed 22 mm in diameter since bovine jugular venous valves are not available in larger sizes and this allowed for good device anchorage, hence reducing the risks of device instability. Patients with transannular patches were, therefore, usually unsuitable for this approach [4].

Advanced imaging techniques, including magnetic resonance imaging (MRI) and tissue-Doppler echocardiography, helped us to determine anatomy and function of the myocardial and vascular substrate [5]. Clinically, indications for PPVI were symptoms of RV outflow tract (RVOT) dysfunction of a sufficient degree to warrant surgical intervention on the basis of conventional criteria [4]. This included RV hypertension (two thirds of systemic blood pressure or greater), with outflow tract obstruction, significant pulmonary insufficiency, RV dilatation or RV failure [6].

With increasing confidence, we performed PPVI as a palliative approach in four patients who had formal contraindications to surgery. Two had severe pulmonary hypertension and two presented with severe heart failure.

As expected, there were procedural and follow-up complications after PPVI. Since we mostly selected patients who could potentially undergo cardiac surgery as an alternative [7], we hoped the procedural complication profile of PPVI would be equal or less than the established therapy. Although PPVI is not strictly an alternative to surgery, we did make an assessment to compare the rate of complications between the percutaneous approach and surgery. The procedural complications are different in nature between the surgical treatment and PPVI and are therefore difficult to scientifically weigh against each other. Despite this, procedural complications of PPVI were comparable or less prevalent than in surgical intervention [6]. In our experience, six patients had major procedural complications requiring periprocedural surgery (two homograft rupture, one obstruction of the right PA, one coronary compression and two valve dislodgements), out of which all but the homograft rupture could have been avoided in retrospect. Therefore, most complications are now interpreted as learning-curve related. During patient selection, we now perform an aortogram or even selective coronary angiogram during catheterization to further clarify the anatomical interaction between the implantation site and the coronary arteries. In another four patients, procedural complications were minor in nature.

While the size of the implantation site and the course of the coronary arteries are known predictors for procedural risks, follow-up complications are difficult to predict during patient selection. The known follow-up complications, which represent early device failure, include stent fracture, residual stenosis, ‘hammock effect’ (i.e., prolapse of the venous wall into the lumen associated with a design imperfection in the first PPVI generation) and outgrowth. Most of these complications could be treated percutaneously with either balloon valvuloplasty or a second PPVI within the first percutaneous valve. The remaining patients required surgical intervention, which was then performed electively.

Since our long-term experience is limited, we do not know the longevity of the valve. Preliminary data suggested, however, that the valve performance measured with MRI is better in the PPVI group compared with surgically-placed homografts 1 year following intervention. While not encountered in our follow-up, if the PPVI proves to be significantly regurgitant, the implantation of a second device within the first could be performed with ease. In analogy to our experience with repeat PPVI as treatment for early device failure, which demonstrated feasibility and safety, the application of repeat PPVI could further prolong conduit longevity.

With a safe strategy for patient selection and a favorable risk profile, the success of our technique is then primarily judged on its benefits. Proven benefits of PPVI compared with surgery include: the avoidance of cardiopulmonary bypass and less trauma since a sternotomy is not required, which leads to a significantly shorter hospital stay [6].

In terms of physiological benefits, we could demonstrate improvement in RV and left ventricle (LV) performance assessed with objective parameters, immediately post intervention. RV and LV volumes measured with MRI and cardiopulmonary exercise tolerance improved significantly after PPVI [2,5]. This was paralleled by marked subjective improvement in symptoms and exercise tolerance.

The judgement of potential benefits is important since it has implications for treatment strategies. As treatment with PPVI is less invasive compared with surgery, a potential benefit could be earlier intervention. This parameter is more difficult to interpret, because success would be judged mainly by prevention from deterioration, rather than by improvement in hemodynamic parameters. To date, the optimal timing for intervention/reintervention has not been determined. Onset of RV dysfunction, mirrored by increasing symptoms, may be too late. Ongoing research in this field will help identify the correct time for intervention, before ventricular dysfunction becomes irreversible.

Besides the obvious economic impact of reducing intensive care and hospital stay, patient acceptance of less-invasive techniques produces an unmeasurable advantage. Our patient population of young adults, often with a surgical history of multiple operations, are very reluctant to accept further surgery. The indirect clinical and economic effect of patients who go towards hemodynamic deterioration and irreversible cardiac damage could be reduced with the application of less-invasive techniques.

The difficulty in placing new technology into a clinical context is enhanced by the different viewpoints of three major players: the clinicians, the regulators and the industry, which drives necessary investment for a successful technical approach. For all percutaneous valve techniques, there are differences in parameters that must be balanced to introduce this fundamentally different approach to valve disease into routine clinical practice. The results of surgery, as an established therapy help in the decision as to whether a new approach might find clinical justification, but surgery does not serve as ‘gold standard’ to which new techniques can be compared. From a regulatory perspective, this presents a very difficult problem. This, in turn, might lead to the fact that clinically obvious benefits cannot be introduced into medical practice, since the objective comparison between the ‘old’ and the ‘new’ is difficult in the field of new percutaneous techniques.