Jamil F. Borgi and Michael S. Mulligan
Introduction
Approximately 600,000 acute pulmonary emboli (PEs) occur yearly in the United States, resulting in 200,000 to 300,000 deaths annually. In the vast majority of cases the clots resolve completely. However, in 0.1% to 0.5% of these patients, the resolution of the thromboembolism is incomplete and ultimately these patients progress to the development of chronic thromboembolic pulmonary hypertension (CTEPH). It is estimated that 500 to 2,500 new cases of CTEPH are generated annually but a significantly lower number of surgeries is performed.
INDICATIONS
Patients with CTEPH usually present with dyspnea on exertion. A history of prior symptomatic pulmonary embolus is not readily documented in many of these patients. Without that history, the diagnosis can be particularly difficult to make and is often quite delayed. Not uncommonly, however, a patient will provide a history of having had a large pulmonary embolus. Typically the therapy for that event will have been entirely appropriate. Other patients may report that the diagnosis was not made until few months after the acute event or perhaps that therapy was suboptimal.
Patients then progress to hypoxemia and will ultimately manifest signs and symptoms of right heart failure. They may then present with ascites or lower extremity edema. This is obviously a sign of progression of the disease. It was formally believed that the reason for this progression was recurrent pulmonary emboli. Though this certainly may be true, a number of these patients will progress with vena caval filters in place and on appropriate anticoagulation. There appears to be a secondary fibroproliferative response within the pulmonary arteries, which develops producing progressive more distal obstructions.
Younger patients may be relatively asymptomatic at rest, but become profoundly short of breath with minimal exercise. This is explained by the presence of a fixed obstruction and the lack of a normal response to exercise. Patients with coronary disease may present with angina. Alternatively, excessive right heart strain can also produce angina even in the absence of significant coronary obstructions. Most patients will relate a history of lightheadedness or possible syncope if questioned carefully. This is more common following a cough or other maneuvers, which elicit a Valsalva and produce a transient increase in pulmonary vascular resistance.
Once the diagnosis has been established, proper selection of patients for surgical intervention can occur. Candidates for operation must have pulmonary vascular obstructive disease with significant hemodymanic and cardiopulmonary impairment. A pulmonary vascular resistance of at least 300 dynes/sec/cm-5 at rest or after exercise should be seen. The thrombi should also be surgically accessible. The development of a classification scheme of the anatomic distribution has been helpful in identifying appropriate patients. Type I disease is associated with obvious central thrombus. Type II disease demonstrates no major vessel thrombus but rather intimal thickening and webs at the main lobar and segmental level. Type III describes disease that is restricted to the segmental and subsegmental levels. Finally type IV disease affects only very peripheral resistance vessels and is nonoperative. Type I and II diseases are quite amenable to endarterectomy but type III disease should only be approached by the surgeon experienced in pulmonary endarterectomy.
Once mean pulmonary arterial (PA) pressures exceed 30 mm Hg, the 5-year survival is only 30%. Once that pressure exceeds 50 mm Hg, the 5-year survival is only 10%. The majority of the patients who we have cared for have had mean PA pressures averaging nearly 50 mm Hg. That would imply a 90% risk of death over the next 5 years. With our current perioperative survival rate of >95%, the risks and extensive nature of this surgical operation are justified and seen as quite acceptable by our patients.
CONTRAINDICATIONS
As stated above, patients with type IV disease should not undergo endarterectomy. Patients selected for endarterectomy should also have limited comorbidities. Significant parenchymal lung disease is a risk factor for prolonged postoperative mechanical ventilation, limited improvement in dyspnea, and increased perioperative mortality. Peripheral vascular disease is also a strong relative contraindication. Chronic renal insufficiency makes postoperative fluid management and establishment of a brisk and timely diuresis troublesome. Those with compromised renal function tend not to fare as well. When considering patients over the age of 80 one should be very selective.
Finally patients must understand and accept the risks of surgery. In addition to complications that are relatively common to open heart surgery, there are added risks associated with circulatory arrest and operating on patients in advanced stages of right heart failure. A generous amount of time is, therefore, spent with patients and their families counseling them on the specifics of the operation, the anticipated postoperative course and the potential complications. In general patients are very enthusiastic about surgical intervention after the natural history of their disease has been explained.
PREOPERATIVE PLANNING
Without attention, the evaluation can quickly become redundant or misdirected. Therefore, it is important to organize existing information and order further studies with three specific goals in mind.
The goals are:
1. To determine whether the patient has pulmonary hypertension and assess its severity.
2. To determine the etiology of that pulmonary hypertension.
3. To determine if the patient’s disease is surgically accessible.
The diagnostic evaluation has often begun long before the patient presents to a surgeon and the first two goals are usually established. The surgeon role is to determine operability and surgical candidacy.
These are the diagnostic modalities that have become our standard routine:
Pulmonary Function Testing (PFT)
PFT is critical. More important than documenting any specific numbers or threshold values, one must be able to recognize the pattern of abnormalities. Specifically, the symptoms and gas exchange abnormalities are far greater than any spirometric defects. There is often a significant reduction in the diffusion capacity that goes with the patients’ experience of pronounced dyspnea. However, there are typically lesser restrictive or obstructive defects. Alternatively, patients with CTEPH may have other parenchymal lung disease that increases the risk of surgery or contraindicates it altogether.
ABGs
The PO2 on blood gas analysis is often low. If this is not the case at rest, there may be a precipitous decrease with exercise. This is due at least in part to patients’ inability to increase their cardiac output adequately with exercise because of fixed obstructions to pulmonary blood flow. As they consume oxygen with exercise, their mixed venous oxygen content drops. Since reoxygenation is limited by inadequate cardiac output traversing functioning lung, progressive hypoxemia ensues. Furthermore, up to 25% of these patients have a patent foramen ovale. With exercise, systemic vascular resistance drops and there is a rapid rise in PA and right heart pressures. The resultant right to left shunting, at the atrial level, will further exacerbate any hypoxemia.
Echocardiography
It provides an estimate of pulmonary pressures and also screens for any coexistent cardiac pathology. Right atrial and ventricular enlargement are typically obvious and both the atrial and ventricular septae may be displaced from right to left (Figs. 46.1 and 46.2). A patent foramen ovale in 25% of these patients however its detection may require an agitated saline contrast study. The tricuspid regurgitation is typically moderate to severe. Once the echocardiogram has indicated the presence of pulmonary hypertension the first goal of the diagnostic evaluation is essentially achieved. Next one must try and discern the etiology.
Ventilation–perfusion (VQ) Scan
VQ scanning typically demonstrates segmental or larger mismatched defects (Fig. 46.3). With primary pulmonary hypertension, the defects may be patchy or subsegmental or the scan may be entirely normal. This, therefore, allows a better distinction between small resistance vessel disease and chronic thromboembolic pulmonary hypertension. Unfortunately, VQ scanning tends to underestimate the degree of central obstruction. Recanalized vessels may offer hemodynamical resistance to flow but still allow isotope to reach the periphery. As a result scans may look deceptively normal. Furthermore, VQ scanning does not adequately demonstrate the magnitude, exact location or proximal extent of the thickening of the diseased arterial wall. It, therefore, is inappropriate to use VQ scanning alone to select patient for operation. VQ scanning is used however, as a screening tool to determine which patients should be appropriately referred for more invasive testing.
Figure 46.1 TEE showing enlarged right atrium.
Figure 46.2 Echocardiogram showing bulging of the interventricular septum to the left (D-shaped left ventricle).
Right Heart Catheterization (RHC) and Pulmonary Angiography
Once pulmonary hypertension has been documented by echocardiography and an etiology has been suggested by VQ scanning, patients proceed to RHC and pulmonary angiography.
Venous access is preferably obtained via the neck to avoid disturbing possible residual iliofemoral thrombus. Patients require careful monitoring in the radiology suite and it is preferable to use single power injections per side with very limited volumes. Selective injections should only be used when absolutely necessary to provide details important to operative decision making. The nature of their pathophysiology implies that these patients typically have significantly increased intravascular volume and excessive use of contrasts can lead to cardiopulmonary decompensation. To optimize the efficiency of the studies and limit the number of injections, it is preferable for the surgeon to be present in the angiography suite to assist in selection of quality views. Typical findings of CTEPH include pouch defects (produced by spherical filling defects), webs or bands, intimal irregularities, abrupt narrowing of pulmonary branch vessels. Normal branching patterns should produce gradual tapering of arterial caliber. In CTEPH the caliber may narrow abruptly with loss of peripheral arborization or delayed filling of segmental and subsegmental vessels (Figs. 46.4 and 46.5).
Figure 46.3 Significant perfusion defect particularly in the left lower lung field.
Figure 46.4 Pulmonary angiography showing abrupt narrowing and loss of peripheral arborization in the left lower lobe.
Assessment of pulmonary pressures and pulmonary vascular resistance may require provocative exercise testing to demonstrate critical elevations. This is particularly true in younger patients who are only symptomatic with exertion. While there may be mild elevations in pulmonary pressures with exertion in many patients, patients with CTEPH typically demonstrate a steep rise in pressures with minimal exertion. This may be the only way to demonstrate that the patient with relatively normal pressures at rest has a valid physiologic explanation for their symptoms.
Figure 46.5 Pulmonary angiography showing intimal irregularities in and spherical filling defect.
Figure 46.6 CT scan with angiography showing thrombus in the right and left PA branches.
CT Scan
Defining the proximal extent of disease is essential in determining surgical accessibility. As such, adjunctive studies are often used to confirm the site where thickening of the vessel wall or luminal irregularities begins. We currently employ spiral CT scanning with a PA-specific protocol to examine for thickening of the arterial walls and residual thrombus in the main and proximal lobar pulmonary arteries (Fig. 46.6).
Finally, any patient who presents with angina, men over the age of 40 and women over the age of 50 typically undergo left heart catheterization to rule out coexisting coronary artery disease. If significant coronary disease is identified, revascularization is planned concurrently with pulmonary endarterectomy.
SURGICAL TECHNIQUE
The PA and bronchial arterial systems provide a unique dual blood supply to the lung tissue. When the PA system becomes obstructed, the nutrient bronchial flow maintains tissue viability. Accordingly, when the obstructions to PA flow are relieved, the lung can once again participate in normal gas exchange. Unfortunately however, bronchial collateral flow can also significantly obscure operative visualization (even on full cardiopulmonary bypass). Certain technical maneuvers must, therefore, be undertaken to compensate for this.
This operation is not an embolectomy or so called Trendelenburg procedure done for acute PE. The simple removal of residual central thrombus will not result in an effective reduction in pulmonary vascular resistance. Rather the operation requires an actual endarterectomy with the dissection plane that is in the middle of the media. A dissection plane that is too superficial will not achieve the appropriate hemodynamic result and one that is too deep runs the risk of perforation. A plane that is too deep will yield a pink raw appearance to the remaining vascular wall. This is the adventitia and the dissection should be redirected to more superficial plane.
Figure 46.7 Operative layout showing ascending aortic cannulation and bicaval venous cannulation. The right superior pulmonary vein is used for LV venting.
The operation must be considered a bilateral procedure. This is true even if angiography suggests a predominance of disease on one side. The obstruction must be relieved on both sides. This will allow for an optimal redistribution of pulmonary blood flow, improved oxygenation, and ultimately maximal perioperative and long-term survival. Therefore, the operation is conducted through a median sternotomy and not through a thoracotomy.
To optimize visualization and counter the impaired visualization imposed by excessive bronchial collateral flow, the operation is performed using cardiopulmonary bypass and deep hypothermic circulatory arrest.
Cardiopulmonary bypass is commenced with ascending aortic canulation. Venous drainage is via the right atrium or bicaval access depending on whether a patent foramen ovale has been detected on agitated saline contrast echocardiogram. The left ventricle is vented via the right superior pulmonary vein. An ascending aortic cannula is placed for administration of antegrade cardioplegia and later on for daring the heart (Fig. 46.7).
Cardiac protection is rarely modified based on the presence of concomitant cardiac valvular or coronary pathology. But generally, antegrade cardioplegia is preferred and uniformly used whenever possible because of better right ventricular protection that is usually hypertrophied and at risk with this operation.
To expose the right pulmonary artery, extensive mobilization of the SVC is required. Use of the cautery will limit bleeding at the end of the case but great care must be taken to avoid injury to the phrenic nerve. The SVC is surrounded with a vessel loop. The ascending aorta is also dissected circumferentially and surrounded with an umbilical tape, for later retraction (Fig. 46.8).
This dissection is performed while the patient is being cooled. Once a temperature of 18°C is reached, an aortic cross clamp is applied and cardioplegia is administered. The SVC is retracted anteriorly and laterally and the aorta retracted medially to expose the main right PA. An incision is made in the right PA with an 11 blade starting from underneath the aorta. This is carried out with Venous Potts scissors out toward the lower lobe division. This dissection is contained within the pericardium (Fig. 46.9). In general any central thrombus can be removed prior to circulatory arrest. The endarterectomy plane is developed with a 15 blade along the transverse axis of the pulmonary artery (Fig. 46.10). The depth of the plane is of crucial importance and should be in the middle of the media. After raising the endarterectomy plane in the main PA, it is carried out to the segmental and then on to the subsegmental vessels. This step is best performed using the special suction tip dissector (Fig. 46.11).
Figure 46.8 Both the SVC and the ascending aorta are encircled with vessel loops.
Figure 46.9 Exposure of the right PA is sometimes facilitated by the placement of a self-retaining retractor.
Figure 46.10 The dissection plane is developed transversely with a no. 15 blade. Separation of the thrombus is completed using a suction dissector (insert).
Once at the level of segmental and subsegmental branches, visualization will be obscured with back bleeding. Intermittent circulatory arrest here is used with arrest periods lasting no more than 20 minutes. In between arrest periods cardiopulmonary bypass is recommenced for 10 minutes or until the mixed venous saturation returns to 90%. One period of circulatory arrest is usually sufficient to complete one side. Occasionally another period is needed.
Care should be taken to avoid fracturing the endarterectomy specimen at the subsegmental level, as the distal portion will retract distally beyond reach. After completion of the right side the vessel is closed with running 5-0 polypropylene suture and the retraction on the aorta and SVC is released (Fig. 46.12). Additional cardioplegia may be administered at this time or regularly every 20 to 25 minutes.
Attention is then given to the left side. The heart is retracted anteriorly and to the right. We routinely use the Janke-Barron Heart Support (Edwards Lifesciences LC Irvine, Ca, USA) to keep the heart in the retracted position. The left pulmonary artery is opened from the main PA into the left PA and extends to the pericardial reflection in the same technique described for the right PA (Fig. 46.13). An endarterectomy is then performed on that side, the patient is reperfused and the left PA closed with 5-0 polypropylene in a running fashion. The specimen is reconstructed on the back table and typically demonstrates some degree of acute or subacute thrombus with the more mature fibrotic disease originating at the segmental level and beyond (Fig. 46.14A,B).
Figure 46.11 Suction tip dissector (Jamieson).
Figure 46.12 Closure of the arteriotomy with a running 5-0 polypropylene suture.
Figure 46.13 Left PA arteriotomy and endarterectomy. Separation of the thrombus from the arterial wall (insert).
If the echocardiogram demonstrated a patent foramen ovale or if significant suspicion exists, the atrial septum is then inspected after rewarming has begun. Likewise any revascularization or adjunctive procedures may be undertaken at this point if they were not completed during cooling. It is critical to use gradual cooling and warming so as to accomplish more uniform tissue temperatures and optimize metabolic protection. Maintaining an 8°C to 10°C gradient while rewarming may also help limit reperfusion injury that would otherwise be exacerbated by a hyperthermic perfusate. EEG monitoring and cerebral oximetry are uniformly used in our cases of hypothermic arrest.
Specific tricuspid valve repair is generally not required. There is marked resolution of the tricuspid regurgitation that occurs to a great degree in the operating room and continues postoperatively. This relates to right ventricular remodeling that occurs acutely after surgery. Valvular regurgitation is typically absent by 4 to 5 days postoperatively.
Figure 46.14 A,B: Back table reconstruction of the endarterectomy specimen.
Typically the heart is defibrillated once a temperature of 28°C is reached. Cardiopulmonary bypass is weaned at a temperature of 36°C and the patient is decannulated. Closure then follows after placement of drainage tubes. A closed-type suction tube is also placed in the pericardium to prevent any delayed pericardial fluid formation.
POSTOPERATIVE MANAGEMENT
Every attempt should be made to minimize pulmonary vascular resistance. A target pCO2 on blood gas analysis of 30 to 35 mm Hg is desirable but permissive hypercapnea is acceptable when oxygenation is problematic. The FIO2 is minimized so long as oxygen saturations of 92% are maintained. Inhaled nitric oxide may be helpful not only for potentiating the reductions in pulmonary vascular resistance that will develop over time but also to help mitigate reperfusion injury and to avoid ventilation/perfusion mismatch. Inverse ratio and pressure control ventilation have at times been helpful at minimizing plateau inspiratory pressures and should be strongly considered. Subcutaneous heparin and the use of sequential compression devices are begun immediately postoperatively and coumadin or heparin therapy are initiated 48 hours after surgery and maintained for life. As mentioned these patients generally have increased intravascular and interstitial volumes. With the abrupt reduction in pulmonary vascular resistance produced by surgery, they will typically require assisted diuresis 24 hours following surgery. This diuresis can be brisk and sustained.
POSTOPERATIVE COMPLICATIONS
Our 30-day mortality at present is less than 5%. Considerably higher mortality rates have been reported in the literature and anecdotal institutional reports have revealed prohibitive perioperative mortality. Certainly as institutional experience and in particular, a single surgeon’s experience is gained, mortality rates decline. At the University of Washington our results are now on par with those of the most experienced centers.
In previous reports, the most common cause of 30-day mortality was unrelieved pulmonary hypertension. Less commonly, mediastinal hemorrhage, intraoperative cardiac arrest, and severe reperfusion pulmonary edema were cited. Cerebral vascular accidents and cannulation site dissections accounted for occasional deaths.
Reperfusion pulmonary edema develops in 10% to 25% of patients. It is often mild but may be hemorrhagic and fatal. No preoperative factors reliably predict its development. It typically manifests within 8 to 12 hours of operation but may not develop for up to 72 hours. It only develops in previously obstructed segments and, therefore, may have a patchy appearance on x-ray as compared to reperfusion injury, which may develop after lung transplantation. Therapy is generally supportive although a brief pulse of steroids may be helpful. PA steal can also develop in 10% to 15% of patients and is associated with significant hypoxemia. The diseased segments that underwent endarterectomy will have a lower resistance than relatively nondiseased segments that were not manipulated. Unfortunately, since the manipulated (lower resistance) segments are the only ones vulnerable to reperfusion injury, the preponderance of pulmonary blood flow may be directed to the most edematous. Much of the resultant hypoxemia will resolve over the first several days postoperatively but 7 to 10 days of mechanical ventilation may be required. Ultimately the pulmonary circulation patterns will normalize completely over 6 to 8 weeks and patients should no longer require any supplemental oxygen.
Complications that are not specific to pulmonary endarterectomy but associated with cardiac surgery in general include arrhythmias, atelectasis, wound infections, and phrenic nerve injury (particularly the right). Transient delirium or mental status changes are present in 10% of patients, but virtually all of these problems are transient. Pericardial effusions may develop late in the postoperative course. To prevent tamponade the posterior pericardium can be fenestrated or a closed suction drain can be left in place as described above.
RESULTS
Significant improvements in cardiopulmonary hemodynamics are seen immediately after operation. In addition the pulmonary vascular resistance may continue to fall over the first few days or weeks after surgery. In our initial series the mean PA pressures fell from 48 mm Hg preoperatively to 22 mm Hg postoperatively. Pulmonary vascular resistance fell dramatically to normal values and the cardiac output nearly tripled. Tricuspid regurgitation resolved in virtually all patients and NYHA class improved from a mean of 3.7 to 1.3 postoperatively. These results were consistent with those of other major centers.
In addition to profound improvements in hemodynamics, lower extremity edema resolved or significantly improved in up to 90% of patients. Some lower extremity edema likely persisted since some of those patients had chronic postphlebitic syndrome from previous deep venous thrombosis. Ascites likewise resolves in virtually all patients who achieve the desired hemodynamic result. Although up to 75% of patients may require supplemental oxygen at the time of discharge, nearly all patients were off of oxygen at follow-up 6 to 8 weeks later.
CONCLUSIONS
In the past, the surgical option for patients with CTEPH was often lung transplantation. If thorough evaluation demonstrates that pulmonary thromboendarterectomy is appropriate, this can obviate the need for chronic long-term immunosuppression and its associated complications. Furthermore these operations are not limited by donor organ supply.
The timing of operation and selection of patients present difficult issues. However, acceptable morbidity and mortality have been achieved in a limited number of centers largely due to the experience and skill of multidisciplinary teams that work together during patient evaluation, surgery, and postoperative care. Survival in patients without operation is severely limited. With acceptable perioperative mortality, pulmonary endarterectomy can convey a marked survival benefit to appropriately selected patients.
Recommended References and Readings
Benotti JR, Ockene IS, Alpert JS, et al. The clinical profile of unresolved pulmonary embolism. Chest. 1983;84:669–678.
Bergin CJ, Sirlin CB, Hauschildt JP, et al. Chronic thromboembolism: Diagnosis with helical CT and MR imaging with angiographic and surgical correlation. Radiology. 1997;204:695–702.
Chow LC, Dittrich HC, Hoit BD, et al. Doppler assessment of changes in right-sided cardiac hemodynamics after pulmonary thromboendarterectomy. Am J Cardiol. 1988;61:1092–1097.
Dalan JE, Alpert JS. Natural history of pulmonary empolism. Prog Cardiovasc Dis. 1975;17:257–270.
Dartevelle P, Fadel E, Chapelier A, et al. Pulmonary thromboendarterectomy with video-angioscopy and circulatory arrest: An alternative to cardiopulmonary transplantation and post-embolism pulmonary artery hypertension. Chirurgie. 1998;123:32–40.
de Perrot M, McRae K, Shargall Y, et al. Pulmonary endarterectomy for chronic thromboembolic pulmonary hypertension: The Toronto experience. Can J Cardiol. 2011;27:692–697.
Hoeper MM, Mayer E, Simonneau G, et al. Chronic thromboembolic pulmonary hypertension. Circulation. 2006;113:2011–2020.
Jamieson SW, Auger WR, Fedullo PF, et al. Experience and results of 150 pulmonary thromboendarterectomy operations over a 29 month period. J Thorac Cardiovasc Surg. 1993;106:116–127.
Jamieson SW. Pulmonary Thromboendarterectomy (editorial). Heart. 1998;79:118–120.
Jamieson SW, Kapelanski DP. Pulmonary Endarterectomy. Curr Prob Surg. 2000;37:165–252.
Levinson RM, Shure D, Moser KM. Repertusion pulmonary edema after pulmonary artery thromboendarterectomy. Am Rev Respir Dis. 1986;134:1241–1245.
Madani MM, Jamieson SW. Technical advances of pulmonary endarterectomy for chronic thromboembolic pulmonary hypertension. Semin Thorac Cardiovasc Surg. 2006;18:243–249.
Mayer E, Jenkins D, Lindner J, et al. Surgical management and outcome of patients with chronic thromboembolic pulmonary hypertension: Results from an international prospective registry. J Thorac Cardiovasc Surg.2011;141:702–710.
Moser KM, Auger WR, Fedullo PF. Chronic major-vessel thromboembolic pulmonary hypertension. Circulation. 1990;81:1735–1743.
Moser KM, Auger WR, Fedullo PF, et al. Chronic thromboembolic pulmonary hypertension: Clinical picture and surgical treatment. Eur Respir J. 1992;5:334–342.
Moser KM, Page GT, Ashburn WL, et al. Perfusion lung scans provide a guide to which patients with apparent primary pulmonary hypertension merit angiography. West J Med. 1988; 148:167–170.
Olman MA, Auger WR, Fedullo PF, et al. Pulmonary vascular steal in chronic thromboembolic pulmonary hypertension. Chest. 1990; 98:1430–1434.
Opitz I, De Perrot M. Technique of Pulmonary Thromboarterectomy. Operative Techniques in Thoracic and Cardiovascular Surgery: A Comparative Atlas. 2012;17(3):168–180.
Thistlethwaite PA, Kaneko K, Madani MM, et al. Technique and outcomes of pulmonary endarterectomy surgery. Ann Thorac Cardiovasc Surg. 2008;14:274–282.