Chronic thromboembolic pulmonary hypertension (CTEPH) is a rare entity, found in just 0.5%–9% of patients with pulmonary embolism (PE).1,2 Various mechanisms behind the development of CTEPH have been suggested: recurrent embolism,3 growth of the thrombus in the bronchial tree4 and vascular disease secondary to endothelial injury.5 Valvular disease, involving microvascular changes that are indistinguishable from idiopathic pulmonary hypertension (PH), can occur in areas of the vascular tree unaffected by PE. The treatment of choice for this disease, wherever possible, consists of pulmonary thromboendarterectomy (PTE) to remove obstructive material from the pulmonary arteries.6–8

The intervention pursues three main goals: hemodynamic stability, by reducing the effect of PH on the right ventricle; respiratory stability, improving ventilatory efficiency by relieving alveolar dead space; and prophylaxis, by preventing right ventricle failure, retrograde propagation of thrombi in the bronchial tree, and secondary valvular disease in permeable vessels.

We present the results of the first 106 PTEs performed in the Hospital Universitario 12 de Octubre. We report the immediate postoperative results in terms of mortality, morbidity and hemodynamic parameters. We also analyze results of long-term follow-up in terms of survival, functional capacity, hemodynamic parameters and right ventricle remodeling.

We performed 106 consecutive PTEs between February 1996 and June 2014 in our hospital. The mean age of the population was 53±14 years (range 23–77); 57% were men. The clinical and hemodynamic characteristics of the patients are shown in Table 1. In 29 patients, PVR was higher than 1000dynasscm−5. Seven patients (6.6%) had undergone previous heart surgery; 3 of these interventions were PTE (one of which was performed in our hospital). Mean time between diagnosis of CTEPH and surgery was 5 months (IQR 8).

Ninety-six (91%) of the PTEs were bilateral. Mean time with cardiopulmonary bypass, ischemia and circulatory arrest was 205±38, 116±28 and 40±14min, respectively. In 20 patients, PTE was combined with another procedure: 7 tricuspid repairs, 7 myocardial revascularizations, 5 permeable oval foramen closures, and 1 right ventricle thrombectomy. Of the biological material extracted, 49% was classified (under the San Diego system) as type 1, 42% as type 2, and 9% as type 3.

Median time on ventilation in the recovery room was 27h (IQR 92). The most relevant morbidity was lung reperfusion injury in 21 patients (20%), pulmonary hemorrhage in 4 patients (4%), reintervention due to bleeding in 6 patients (6%), heart failure in 3 patients (3%), transient ischemic attack in 1 patient (1%), and need for ECMO or ventricular assist device in 5 patients (5%). Indication for ECMO was heart failure in 2 patients, and respiratory failure due to pulmonary hemorrhage in 3 patients. Mean ICU stay was 5 days (IQR 7), and mean hospital stay was 13 days (IQR 9). In the recovery room, mean PAP was 28±7mmHg and mean PVR was 311±130dynasscm−5, significantly (P<.05) lower than preoperative levels.

Overall in-hospital mortality for the series was 6.6% (7/106, 95% CI: 2.8%–13.5%). Causes of mortality were respiratory failure in 4 patients, pulmonary hemorrhage in 2 patients, and cardiogenic shock in 1 patient.

In the univariate analysis, being among the first 30 cases (P=.019), PVR≥1000dynasscm−5 (P=.019), functional class IV (P=.006) and postoperative lung reperfusion injury (P=.003) were associated with in-hospital mortality. Following multivariate analysis, only reperfusion lung injury remained as an independent risk factor for in-hospital mortality (P=.015), (Table 2).

Mean follow-up was 31 months (IQR 50). Survival at 3 and 5 years was 90%±3% and 84%±5% respectively, including in-hospital mortality (Fig. 2). Changes in clinical, echocardiographic and hemodynamic parameters at 1-year follow-up are shown in Table 3.

Fig. 2.

Overall survival curve, including in-hospital mortality.

(0.07MB).

In total, 6 patients died during follow-up. Causes of death were heart failure (3 patients), infection (1 patient), adenocarcinoma of the bladder (1 patient) and pulmonary thromboembolism (1 patient).

Fourteen patients were diagnosed with persistent PH following PTE. At 1 year, 64% of these patients were classified as functional class I–II. Nine (64%) patients required specific treatment for PH: endothelin receptor antagonists, phosphodiesterase type 5 inhibitors, prostacyclin analogs or soluble guanylate cyclase stimulators. Two patients died due to heart failure. Actuarial survival in the persistent PH group is 91±9 at 3 years, and 73±11 at 5 years. Fig. 3 shows the survival curve for this patient group and that of patients with no residual PH, after adjusting for in-hospital mortality. There is no significant difference between groups. It is interesting to note that despite persistence of PH, PVR was significantly reduced compared to preoperative levels (924±245 vs, 641±269dynasscm−5; P=.033).

Fig. 3.

Survival curves in patients with and without residual pulmonary hypertension, excluding in-hospital mortality.

(0.09MB).

CTEPH is a rare disease, with only 8.9 cases per million inhabitants in our region.20 Studies have shown that CTEPH is a long-term complication of symptomatic PTE, with an accumulative incidence of 0.1%–9.1% at 2 years.1,2 However, in a significant number of cases, PTE was asymptomatic.5 This is why CTEPH is thought to be under-diagnosed, and patients are all too often referred late for treatment. Clinical guidelines10 and consensus documents15 establish the need to form multidisciplinary teams of expert surgeons to treat CTEPH. Surgery should not be definitively ruled out in any patient until they have been assessed by this team. In the international register,21 up to 43% of the patients evaluated were not considered candidates for surgery, and the same was true of 70% of patients in the Spanish register (REHAP).20 Our team has performed surgery on 53%22 of cases submitted to the multidisciplinary team meeting. In many patients, however, there was a lengthy delay between diagnosis and surgery due to late referral, and 13% of our patients were classified as functional class IV at the time of surgery.

The results of our series are excellent, and are comparable to the largest series reported in the literature.21,23 Our success is due to the expertise of our surgeons and the procedural protocol developed by the Multidisciplinary Pulmonary Hypertension Team of the Hospital 12 de Octubre.

The PTE series presented in this study is the largest and most consolidated cohort studied in Spain. In 2009, we published the initial24 results of a series of 30 PTEs. These findings were similar to those published by the Hospital Clìnic de Barcelona25 in a series of 32 patients; in-hospital mortality was 17% and 18.8%, respectively. Both studies found that surgical outcomes improved as the multidisciplinary team acquired more experience. Greater experience is also a factor in improved survival and a gradual increase in the number of patients considered candidates for surgery. In addition, surgical outcomes are better than those of medical treatment. For all these reasons, PTE is the treatment of choice in this disease, and is associated with good outcomes in our hospital.

The accessibility of lesions varies according to the level of experience of the surgical team. In the most experienced teams, the percentage of patients with segmental branch involvement (San Diego group 3) increases over time, although this is not detrimental to outcomes.26–28 Advances in surgical techniques now enable surgeons to dissect and extract material at both the segmental and subsegmental level.

Although the exact PVR or right ventricular dysfunction level that would rule out surgery is unknown, surgical risk and prognosis in the long-term worsen with higher PVR (above 1000dynasscm−5).29,30 Although the risk of surgery is increased in these patients, they can still benefit from the intervention.29 We did not rule out surgery in any patient on the basis of high PVR or severe ventricular dysfunction. In this study, PVR greater than 1000dynasscm−5 was a risk factor for in-hospital mortality in the univariate analysis, but not in the multivariate analysis, even though it is a major risk factor for perfusion injury.

The standard surgical approach to PTE is based on the protocol created by the San Diego group,4 namely, endarterectomy in circulatory arrest with deep hypothermia. This technique has proved both safe and effective. In their most recent series,27 with more than 2700 cases, the group reported a mortality rate of 2.2% in the last 500 patients operated. Although a number of studies have described PTE without circulatory arrest, or at least, with continuous cerebral perfusion,31–34 these techniques did not improve outcomes or diminish neurological complications. All studies in this field have mentioned the difficulty involved in reaching distal branches in the absence of circulatory arrest. In our series, only 1 transient neurological complication occurred, and this was not solely due to circulatory arrest.

Reperfusion injury is inherent to PTE. Incidence ranges from 5% and 20% of all interventions,23 according to different series, and it is a risk factor for in-hospital mortality. This was also found to be true in our series. Manifestation of reperfusion injury varies from mild hypoxia with insignificant radiological findings to severe respiratory failure, similar to respiratory distress. For this reason, it is important to establish a definition of reperfusion injury that limits cases to a clinical entity with an impact on the patient's evolution.

Other series have described the need for extracorporeal membrane oxygenation (ECMO) in heart failure secondary to PH or postcardiotomy syndrome, and severe respiratory failure due to pulmonary hemorrhage or edema.35,36 Berman et al.36 reported a survival rate of 57% in these patients; in our series, it was 40%. These patients would have died without this treatment, and for this reason ECMO is a therapeutic step that should be available in all PTE units.15

Post-PTE survival in our series was excellent, and comparable to that obtained in the most major studies.23 Moreover, in most of our patients, functional class improved, pulmonary pressure was normalized, and as a result, right ventricular remodeling occurred.

Diagnosis of persistent PH is widely debated, both in terms of its definition (mPAP>25, 30 or 35mmHg; PVR>400, 450, 500 or 550dynasscm−5),21,30,37 and the right moment to establish a diagnosis. Incidence varies from 6% to 31%, according to the definition used. Our team decided to define persistent PH as PVR>400dynasscm−5, because unlike mPAP, PVR has both a pre- and postoperative prognostic value. As far as the timing of the diagnosis is concerned, in our series it was made during the 6-month follow-up for two reasons: hemodynamic changes immediately following surgery will affect estimates of PVR, and the prevalence of residual PH increases over follow-up37 when the cause is distal vascular disease.

Despite persistence of PH, this subgroup showed good survival rates and improved functional class, possibly because PVR decreased significantly even in patients with PH. The findings of other groups, such as Freed et al. in Papworth Hospital30 (Cambridge) are largely consistent in terms of patients with residual PH.

Our study contributes further evidence that PTE is the treatment of choice in CTEPH,10 and can be performed successfully in Spain. This surgical intervention has a positive impact on disease prognosis and long-term survival. It also resolves pulmonary hypertension and improves the functional status of most patients.

This is a retrospective, observational study in the initial series of a single hospital.

No information was available on the evolution of NO-pro-BNP markers in patients operated prior to 2009.

One-year follow-up results were analyzed in the 85 patients that reached this stage of the study. Of these, we obtained 76 catheterizations; the remaining 9 were patients that had been referred to their local hospitals, and this information was not forwarded to our group.

López Gude MJ and Cortina Romero JM collected data, performed the statistical analysis, interpreted the results and drafted the manuscript.

Pérez de la Sota E, Forteza Gil A and Centeno Rodríguez J performed the surgical procedures.

Eixeres A collected data.

Ruiz Cano MJ, Velázquez MT, Gómez Sánchez MA, Pérez Vela JL, and Sánchez Nistal MA established diagnosis and treated the patients in the clinical setting.

Escribano Subías P established diagnosis and treated the patients in the clinical setting, interpreted the results and drafted the manuscript.

All the authors have reviewed the manuscript.

The authors declare they have no conflict of interest.