World J. Surg. 23, 1137–1147, 1999

WORLD Journal of

SURGERY © 1999 by the Socie´te´ Internationale de Chirurgie

Surgery for Chronic Thromboembolic Pulmonary Hypertension Renee S. Hartz, M.D. Department of Surgery, Division of Cardiothoracic Surgery, Tulane University Medical School, 1430 Tulane Avenue, New Orleans, Louisiana 70112, USA Abstract. The modern era of surgery for chronic thromboembolic pulmonary hypertension (CTEPH) began just over 10 years ago. Until that time pulmonary thromboendarterectomy (PTE) was performed infrequently and essentially at a single medical center (University of California at San Diego—UCSD). It posed a formidable technical challenge and was associated with both high operative mortality (> 20%) and excessive morbidity due to respiratory and multiorgan system failure. Currently PTE is performed at numerous medical centers throughout the world, largely due to the pioneering efforts of those surgeons who developed and perfected the operation at UCSD. Operative mortality rates have fallen, and postoperative complications have become less common. Although no longer simply an autopsy curiosity, CTEPH continues to be an underdiagnosed condition. Increased awareness and better diagnosis will lead to curative surgery in more patients worldwide.

Because most episodes are clinically silent, the true incidence of pulmonary embolism (PE) is unknown. In the United States more than half a million patients are afflicted annually, and approximately 10% die within the first hour of the event. According to Dalen (1975), accurate diagnosis is made in only one-third of the survivors so that at least 400,000 cases of pulmonary embolus are missed annually in the United States alone [1] (Fig. 1). Because complete resolution of the thrombus depends on adequate anticoagulant therapy, it is likely that many patients in the subgroup with missed diagnosis go on to develop the chronic, progressively obstructive form of the disease known as chronic thromboembolic pulmonary hypertension (CTEPH). Previously considered by many to represent a rare and aberrant outcome of acute PE, CTEPH may actually represent a logical physiologic outcome of untreated or recurrent PE: organization of the thrombi, incorporation into the wall of the pulmonary artery, occlusion, and recanalization, all occurring in repeated cycles and eventually leading to pulmonary hypertension and right heart failure. Pathophysiology The pulmonary hypertension observed in CTEPH is due to chronic obstruction and subsequent medial proliferation in the central pulmonary arteries (main, lobar, segmental). With progressive elevation of pulmonary artery pressure and pulmonary vascular resistance, right heart failure and hypoxemia occur. At this advanced stage, the prognosis without surgical therapy is poor.

Microscopically, chronic emboli are fibrous, organized, and adherent to the pulmonary artery intima. In large vessels, recanalization frequently occurs with resultant vascular bands and webs composed of dense fibrous tissue traversing some lumens. Medial hypertrophy also occurs in the larger pulmonary arteries. In the distal vessels, organized lesions resulting from ingrowth of collagen and elastin often mimic the plexiform changes seen with primary pulmonary hypertension. These peripheral lesions are not surgically removable. CTEPH is a dynamic process with repeated cycles of thrombosis and recanalization occurring throughout the pulmonary arterial circulation. It tends to be more pronounced on the right side and has been found to be present in up to 3% of autopsy series [2]. Although etiologic factors have not been completely elucidated, it is clear that adequate anticoagulation for acute PE prevents its occurrence completely. Both Wilhelmsen et al. [3] and Sabiston et al. [4] proposed that recurrent embolization was responsible for CTEPH, and Sabiston et al. thought that inadequate thrombolysis was also involved. Reidel et al. subsequently stated that occult PE is responsible for virtually all cases of chronic pulmonary hypertension [5]. Chitwood et al. concurred but proposed several contributing factors including embolization of previously organized thrombi into more distal arterial branches [6]. Rich’s group postulated active intravascular thrombosis after they demonstrated increased levels of fibrinopeptide A in the sera of their patients with CTEPH. These levels fell to normal after institution of heparin therapy [7]. Given these considerations, it seems unnecessary to invoke abnormal pulmonary endothelium as the cause for this condition. Rather, if the diagnosis of deep venous thrombosis is missed, any subsequent emboli will tend to be more organized and less likely to resolve spontaneously with intrinsic fibrinolysis. When a shower of such organized thrombi occurs, or if repeated small emboli are thrown to the lungs, the stage is set for the process described by Presti et al. [8], who performed postmortem examinations on numerous patients who died of CTEPH; the thrombi are incorporated into the wall of the pulmonary artery, and endothelial proliferation occurs at the margins of the thrombi. Fresh thrombus is continually deposited because the patient is not anticoagulated appropriately. Fibrovascular organization and recanalization occur in some vessels, whereas fresh thrombosis occurs in others. The vessel wall underlying the thrombus has cleft-like

1138

Fig. 1. Incidence of pulmonary embolism per year in the United States. (From Dalen and Alpert [1], with permission.)

spaces similar to those seen in cystic medial necrosis, the thrombi are made more adherent by varying degrees of neovascularization and fibrous attachment, and the lumens are filled with thrombus of varying age. If anticoagulation therapy is initiated at this point, the vicious cycle of thrombosis–recanalization can be stopped, but resolution of fibrous tissue and of obstructive thickened intima and media cannot occur without surgical therapy. Predisposing Factors Until recently, it was thought that most patients with CTEPH did not have a prothrombotic state, and that only a small percentage of patients (0.1– 0.2%) with PE went on to develop the chronic form. Given the fact that the diagnosis of acute PE is missed in most cases, any prothrombotic state should increase the likelihood of the development of CTEPH. The most common coagulation abnormality seen is the presence of lupus-like anticoagulant. Reported to be present in only 10% of patients at the University of California at San Diego (UCSD) medical center [9], Sandoval et al. noted an incidence of 20% in a small series in Mexico [10], and Simonneau et al. saw a 30% incidence in France [11]. These authors believe that CTEPH should be added to the list of known vascular complications seen in the antiphosphlipid syndrome. Other coagulation abnormalities occur as well and include protein C, protein S, and antithrombin III deficiency [11]. Therefore, even though most patients with CTEPH do not have a coagulation disorder, it can no longer be stated that the incidence of such disorders in CTEPH is not higher than in the general population. Other conditions that have been associated with chronic thromboembolic disease include malignancy, the presence of indwelling venous catheters, and atrial septal defect [12, 13]. Clinical Presentation Progressive dyspnea on exertion is the hallmark of CTEPH. As with mitral stenosis, the course is often so insidious that patients do not remember the onset of the illness. Eventually, however, when more than 50% of the pulmonary circulation becomes obstructed, pulmonary artery pressure rises rapidly, and right heart failure and hypoxemia intervene. The patient’s symptoms then

World J. Surg. Vol. 23, No. 11, November 1999

progress to class III or IV over weeks to months. When cardiac output is extremely low, angina (indistinguishable from that seen with obstructive coronary artery disease) and syncope may also occur. The diagnosis is missed in most patients despite the inexorable progression of symptoms, largely because the medical community has only recently become aware that it is so common. In addition, only half of the patients with CTEPH have a history suggestive of phlebitis, and 20% have no recollection whatever of an inciting event [8]. It also appears that there is a “honeymoon period” between the initial thromboembolic event(s) and the development of dyspnea, lasting months to years, which masks the true nature of the illness in most patients. Indeed, “doctor shopping” is common, and most patients with CTEPH have seen several physicians before they are correctly diagnosed. The physical examination in CTEPH patients is that of a patient with pulmonary hypertension and right heart failure with no characteristics to distinguish it from primary pulmonary hypertension except for the occasional presence of low murmurs over the lung fields. Described by Moser et al. [14], these are due to partial obstruction at the pulmonary arteries. Patients typically have a prominent right ventricular impulse, loud second heart sound, murmur of tricuspid regurgitation, engorged liver and neck veins, peripheral edema, and cyanosis. Rarely do they have stigmata of deep venous disease. Diagnosis The diagnosis of pulmonary hypertension should be straightforward for any student of cardiopulmonary disease; but because a thromboembolic etiology is rarely considered, the diagnosis of CTEPH is often delayed or made only at autopsy. The chest roentgenogram may show hilar fullness (due to enlarged central pulmonary arteries), clear or oligemic lung fields, and right ventricular enlargement. The electrocardiogram usually demonstrates right ventricular hypertrophy and strain. Neither of these tests is definitive, however, and it is the ventilation perfusion (V/Q) lung scan that is the single most important test in the differential diagnosis of primary versus obstructive pulmonary hypertension. A completely normal lung scan essentially precludes the diagnosis of CTEPH, but even a single V/Q mismatch should be suspicious and is distinctly different than the mottled, patchy, or “plexiform” pattern seen with primary pulmonary hypertension. When the patchy pattern is seen, confirmatory angiography is unnecessary for confirming the diagnosis of primary pulmonary hypertension, whereas any segmental or subsegmental perfusion defect should be followed with pulmonary arteriography. Long considered to be the “gold standard” for the diagnosis of CTEPH, angiography is the examination that also determines technical operability [15]. In addition, it has been repeatedly demonstrated that the lung scan underestimates the severity of the disease, and angiography provides a much better estimate of the obstructive process [15, 16]. Auger and others have demonstrated the safety of pulmonary arteriography in patients with obstructive pulmonary hypertension. In a series of 250 patients undergoing the examination, Auger et al. observed no mortality, serious arrhythmia, or hemodynamic compromise using a standard approach involving a single injection of nonionic contrast material into each main pulmonary artery. Based on this series, they described the typical angiographic features of CTEPH: “pouching” defects at the site of abrupt occlusions, webs or bands across lumens due to fibrovas-

Hartz: Chronic Pulmonary Hypertension

cular organization, abrupt vascular narrowings, and complete obstruction [17] (Fig. 2). As can be seen in Figure 2, oligemia or absence of perfusion in some or all pulmonary segments was universally present distal to the obstructed vessels. At the time of pulmonary arteriography, hemodynamic data should be obtained (cardiac output, pulmonary artery pressure, pulmonary vascular resistance), as this information is needed to determine whether the patient should undergo surgery to relieve the obstructions. Generally, a patient who is severely symptomatic, who has no other life-threatening illness, and whose pulmonary vascular resistance is greater than 300 dynes/s/cm25 is a candidate for pulmonary thromboendarterectomy, presuming there is evidence of obstructive disease in the central pulmonary arteries. Some degree of tricuspid regurgitation is universally present with this disease and may be evaluated at right heart catheterization, as may be the presence or absence of a patent foramen ovale. Left heart catheterization should be reserved for patients with suspected coronary or valvular heart disease. Echocardiography is becoming an increasingly useful screening procedure in all forms of cardiovascular disease, and CTEPH is no exception. Right heart dimensions, degree of tricuspid insufficiency, presence or absence of a patent formaen ovale, and an estimate of the pulmonary artery pressure can be readily obtained and used as a baseline for postoperative studies. Dittrich et al. performed pre- and post-operative right heart studies and echocardiograms on 30 patients undergoing pulmonary thromboendarterectomy (PTE) and documented the marked changes noted with successful operation [18] (Tables 1, 2). Recently, computed tomographic (CT) scanning has become important in the diagnosis of CTEPH, especially outside the United States. At University Hospital in Mainz, Germany, where the second largest series of PTE has been reported, both unenhanced and enhanced scans were performed in 75 patients, 63 of whom subsequently underwent PTE. The authors described both vascular and parenchymal changes in CTEPH. Visualization of thrombus in the central pulmonary arteries, present in 53 of 75 patients, was the most important direct criterion for the diagnosis (Fig. 3); vascular findings were 77% sensitive in determining the technical operability in this group of patients. Lung parenchymal changes included inhomogeneous areas of hyperattenuation in underperfused areas. These areas were nodular or wedge-shaped and pleura-based, occurred in patients with central thrombi, and were unchanged after injection of contrast medium. Areas of “mosaic oligemia” (sharply demarcated areas of hyper- and hypoattenuation) disappeared on postoperative scans in 20 of 23 patients who had serial examinations [19]. King and others at UCSD did not find CT scans useful for vascular evaluation of CTEPH. In a group of five patients studied, they used the scan for the parenchymal findings and single-photon emission computed tomography (SPECT) scanning for the vascular examination [20]. Rich et al., on the other hand, found the cine-CT scan particularly useful for delineating a central thrombus [15]. Finally, definitive diagnosis cannot be made in a small percentage of patients with the above examinations. Shure and others [21] have described the used of fiberoptic angioscopy in this subset. In summary, the V/Q lung scan is the key examination for the differential diagnosis of thrombotic pulmonary hypertension. Either pulmonary angiography or CT scanning by an experienced

1139

radiographer should be used to verify the extent of obstruction and to determine technical operability. Treatment The natural history of pulmonary hypertension secondary to thromboembolic disease was described by Dalen and Alpert [1] (Fig. 4). When the mean pulmonary artery pressure is 30 mmHg, survival at 5 years is 30%; when it reaches 50 mmHg, the 5-year survival is only 10%. Survival is unaffected by anticoagulants or vasodilator therapy [22, 23]. For these reasons, Jamieson, who has performed the largest number of pulmonary thromboendarterectomies to date, recommends PTE for almost all patients with documented severe CTEPH (more than 90% of patients referred to UCSD for PTE undergo the procedure). On the other hand, some groups prefer lung transplantation in select patients. Of 72 patients with CTEPH evaluated at Antoine Beclere Hospital in France, only 11 (15%) underwent PTE, and 8 had lung transplantation [11]. The authors thought that most patients referred to their institution were not candidates for PTE because of angiographic demonstration of severe distal disease. Similarly, only 45% of 75 patients with CTEPH evaluated at the University of Illinois Hospital underwent PTE; the remainder were thought to have inoperable distal disease [23]. The reported number of patients who have undergone PTE is now just under 1000 cases, almost all performed since 1984. The evolution of the procedure can be divided into three historical eras: The introductory era (1958 –1976); the first successful procedure was performed by Synder et al. in 1958 [24]. By 1976 a total of 18 cases had been reported with an overall operative mortality of 22% [4]. During this era, the operation was not standardized, many were performed without the use of cardiopulmonary bypass or median sternotomy, and the procedure was frequently unilateral. Crucial in this era was the observation by Sabiston that “good back bleeding” predicted a successful outcome after PTE. The photograph of one of his operative specimens clearly established the fact that the procedure was a thromboendarterectomy even though he referred to it as a thromboembolectomy [4]. This operative description and good results obtained in five of the six patients operated on at Duke University ushered in the second era of PTE; the developmental era (1976 –1984). The number of reported procedures went from 22 to approximately 90, with the overall mortality still in the range of 22% [6]. During this era, surgeons at Duke continued to talk about thoracic incisions, avoidance of cardiopulmonary bypass, and “counterincisions” in the pulmonary arteries to ensure removal of thrombus. More importantly during this era, Moser established a referral base for patients with thrombotic hypertension at UCSD, allowing surgeons there to perfect the techniques. In 1982 Utley et al. reported a series of 10 patients operated on with profound hypothermia and circulatory arrest, all with bilateral PTE performed through central pulmonary arteriotomies [25]. Although there was only one death in this series, excessive morbidity occurred. This series led to the modem era of PTE (1986 –present). The surgical groups at UCSD, led by Daily and then Jamieson, progressively refined the operation to its present state. There were 700 PTEs performed at UCSD by early 1996, and the mortality rate had fallen to 6.6% for the last 300 patients operated on (personal communication). Simultaneously, surgeons from around the world traveled to San Diego, and Jamieson traveled extensively to

1140

World J. Surg. Vol. 23, No. 11, November 1999

Fig. 2. Angiographic findings in CTEPH. A. Pouch-like defect (arrows) eventually gives rise to more distal vessels. B. Small arrow points out a fibrous band. Large arrows depict vascular irregularity that proved to represent a large amount of thrombus at the time of surgery. C. Numerous vascular webs and vessel narrowings are shown at arrows. Note oligemia in the lung periphery. (From Auger et al. [17], with permission.)

Hartz: Chronic Pulmonary Hypertension

1141

Table 1. Improvements noted immediately postoperatively in hemodynamic values in 30 patients who underwent pulmonary thromboendarterectomy. Measurement

Preoperative

p

Postoperative

RA mean (mmHg) RV systolic (mmHg) PA systolic (mmHg) PA diastolic (mmHg) PA mean (mmHg) PA wedge (mmHg) Cardiac index (L/min/m2) PVR (dynes z s z cm25)

13 6 8 76 6 20 76 6 20 34 6 11 48 6 12 965 2.0 6 0.5 935 6 620

NS * * * * NS * *

11 6 5 47 6 15 46 6 14 18 6 7 28 6 8 11 6 4 2.9 6 0.6 278 6 252

Reprinted with permission from the American College of Cardiology from Dittrick et al. [18]. All values are expressed as mean 6 standard deviation. PA: pulmonary artery; PVR: pulmonary vascular resistance; RA: right atrial; RV: right ventricular. *p , 0.001.

teach the procedure in other countries. Several moderate-sized series were reported outside of UCSD. Table 3 is a list of the reported series in the modern era [9 –11, 23, 26, 27]. Details of the Procedure The indications for operation have already been described and have been adopted worldwide. The patient should be severely symptomatic [generally New York Heart Association (NYHA) class III or IV], have no other life-threatening illness, have a pulmonary vascular resistance greater than 300 dynes/s/cm25, and should have documented disease of the central pulmonary arteries. It should be noted that Jamieson has liberalized criteria in younger patients and is currently operating on patients with less severe symptoms and more distal disease. Criteria for inoperability in terms of distal disease have not been established by him. Rather, it is his opinion that any patient who manifests central disease should undergo PTE no matter how far distally the disease progresses (personal communication). Aside from the extensive diagnostic evaluation detailed above, there is little in the way of preoperative preparation. All patients are on warfarin therapy at the time of diagnosis, and some phy-

Table 2. Improvements noted immediately postoperatively in echocardiographic values in 30 patients who underwent pulmonary thromboendarterectomy.

Measurement Diameter (cm) PA LA IVC RA long axis RV long axis RV short axis Area (cm2) RA end-systole RV end-diastole LV end-systole LV end-diastole Eccentricity index

Preoperative

p

Postoperative

Normal subjects (n 5 15)

2.8 6 0.3 3.7 6 0.6 2.9 6 0.6 6.8 6 1.5 8.7 6 0.9 4.5 6 0.8

* ** * * * *

2.4 6 0.4 4.0 6 0.7 2.2 6 0.4 5.9 6 1.5 8.1 6 0.9 3.7 6 0.8

1.9 6 0.4 3.2 6 0.4 1.8 6 0.3 4.4 6 0.5 6.9 6 0.7 3.2 6 0.6

31 6 12 33 6 7 14 6 5 24 6 8 1.26 6 0.20

* * NS *** *

24 6 8 24 6 8 16 6 4 28 6 6 0.94 6 0.12

14 6 4 21 6 3 13 6 3 31 6 6 0.86 6 0.08

Reprinted with permission from the American College of Cardiology from Dittrick et al. [18]. All values are expressed as mean 6 standard deviation. IVC: inferior vena cava; LA: left atrium; LV: left ventricle; other abbreviations as in Table 1. *p , 0.001; **p , 0.05, ***p , 0.01.

sicians may want to admit the patient for conversion to heparin. We have found this unnecessary and treat patients with subcutaneous heparin at home for a few days prior to surgery. They are then admitted either the night before or morning of the planned procedure. It is important not to treat the patient with vasodilators preoperatively. These drugs are contraindicated because the pulmonary artery pressure cannot be lowered until the mechanical obstruction is relieved. Preoperative vasodilators may complicate the postoperative care of the patient. At UCSD all patients have a vena cava filter placed preoperatively (unless the thrombi have come from arm veins), but we place filters only in those patients who have documented thrombi in their leg veins at the time of the diagnostic evaluation. Because the best published description of the operative procedure is that in Jamieson et al.’s 1993 article [9] this information is

Fig. 3. Thrombus in the proximal right pulmonary artery on CT scan preoperatively (arrow) (A) is absent after successful pulmonary thromboendarterectomy (B). (From Schwickert et al. [19], with permission.)

1142

World J. Surg. Vol. 23, No. 11, November 1999

Fig. 4. Cumulative survival curves according to the initial mean pulmonary artery pressure in pulmonary hypertension secondary to thromboembolic disease. (From Dalen and Alpert [1], with permission.)

summarized with comments made on variations used by my group. All patients are extensively monitored. The arterial line, SwanGanz thermodilution catheter, bladder and nasopharyngeal temperature probes, and transesophageal echocardiography are routine. We place the pulmonary artery catheter at the beginning of the operation, but remove the balloon from the tip before closing the pulmonary arteriotomies so wedge determinations cannot be made postoperatively. Although Jamieson recommended routine electroencephalographic monitoring we have discontinued this practice because we try to minimize or avoid circulatory arrest. The patients are usually polycythemic because of their prolonged hypoxia, and it is usually possible to remove two units of autologous blood after the monitoring lines are in place. This is particularly important because avoidance of homologous blood products decreases the morbidity of the operation [29]. PTE is always performed on cardiopulmonary bypass with ascending aortic and bicaval cannulation. Left ventricular venting through the right superior pulmonary vein is crucial. Jamieson also prefers to vent the pulmonary artery temporarily during cooling, but we simply vent through the pulmonary arteriotomies. Gradual cooling is begun immediately, and both pulmonary arteries are extensively dissected. This can be done with electrocautery so long as care is taken to avoid the phrenic nerves. In

addition, extensive mobilization of the superior vena cava is undertaken to provide better exposure of the entire proximal right pulmonary artery. As the temperature is lowered, pump flows are appropriately decreased. At 23°C the aorta is clamped (we clamp as soon as the heart fibrillates), and a large dose of antegrade blood cardioplegia is administered. Retrograde cardioplegia does not add to right ventricular protection and is avoided if the coronary arteries are unobstructed. We prefer to repeat the antegrade dose every 20 to 30 minutes, but Jamieson relies on systemic and topical hypothermia after the first dose. A topical cooling jacket is used for external myocardial cooling. Since Utley’s publication in 1982, all surgical groups at UCSD have stated that circulatory arrest is mandatory for the performance of this operation. Although this principle may not be inviolate, circulatory arrest is certainly desirable during the learning curve of this difficult operation. The right pulmonary endarterectomy is usually performed first. After the caval tapes have been tightened, an arteriotomy is begun just to the right of the aorta and carried to the division of the lower lobe branches (Fig. 5a). Mobilization of the superior vena cava medially is recommended and is greatly facilitated by the hinged, blunt-tipped Weitlander retractor shown in Figure 5b. Establishing the correct endarterectomy plane is the most crucial phase of the operation, as a too-superficial plane results in removal of only thrombus, and a too-deep plane causes inadvertent perforation of the pulmonary artery. After loose thrombus is extracted, the endarterectomy plane is started directly posteriorly (Fig. 5c), rather than at the cut edges of the arteriotomy where it is tempting to do so. There are several reasons for this approach. First, it is easy to develop the wrong plane at the arteriotomy edge because the layers of the pulmonary artery tend to separate here. Second, it is desirable to leave a rim of thickened arteriotomy for subsequent closure of the pulmonary arteriotomy. Finally, when the plane is begun posteriorly and very proximally, the specimen starts out quite thin but gets thicker as the surgeon progresses farther distally into the arterial branches. This automatically results in development of the correct plane while making it possible to keep the specimen intact. It is easy to tell when the plane is too deep because visualization of pinkish tissue indicates that the adventitia has been exposed. A new plane should be begun elsewhere. Intense focus, good lighting, and proper instruments are mandatory for a thorough endarterectomy. An array of blunt-tipped

Table 3. Series reported during the modern era of pulmonary thromboendarterectomy. Study

Year

Location

Simonneau [11] Mayer [26] Sandoval [10] Alfieri [27] Hartz [23]

1995 1996 1996 1995 1996

Daily [28] Jamieson [9]a

1990 1993

France Germany Mexico Italy University of Illinois UCSD UCSD

No. of patients

Mean PAP (mmHg)

Mean PVR (dynes)

Operative mortality (%)

11 119 3 15 34

39 491 NA 56 54

823 1015 NA NA 1094

18 24 33 20 23

127 150

46 48

813 937

12.6 8.7

From Jamieson et al. [9] with permission. NA: not available. PAP: pulmonary artery pressure; UCSD: University of California San Diego; PVR: pulmonary vascular resistance. a Number of cases now. Operative mortality in last 500 cases 5 6.4% (personal communication).

Hartz: Chronic Pulmonary Hypertension

1143

Fig. 5. Details of pulmonary thromboendarterectomy. A. Exposure of the right pulmonary artery is improved with retractor lateral to the aorta and superior vena cava rather than between the great vessels. SVC: superior vena cava; Ao: aorta. B. Blunt-tipped retractor aids in maintaining exposure and pulmonary arteriotomy is made. RA, LA: right and left atria; RLL: right lower lobe; RUL: right upper lobe. C. Note that the endarter-

ectomy plane is begun posteriorly rather than at the cut edges of the arteriotomy. D. The left pulmonary arteriotomy should also be quite proximal (large arrows) and affords better visualization than a more distal approach (small arrow). LPA: left pulmonary artery; PT: pulmonary truncus. (From Jamieson et al. [9], with permission.)

suction-dissectors have been designed at UCSD [29] (Fig. 6). They are available in an assortment of lengths and angles, but it is possible to use a single straight one on every case or do the procedure entirely without special instruments. Because it is highly desirable that the specimen be kept intact on each side, a hand-over-hand technique is used, and the assistant must be ready to exchange the sucker-dissector with a blunt forceps repeatedly so the surgeon need never release the specimen. In essence, the pulmonary artery is peeled off the specimen with a gentle motion of the dissector, rather than vice versa. The fact that the specimen is quite elastic and does not fragment as easily as arteriosclerotic plaques seen in the systemic circulation facilitates the procedure. When the subsegmental branches are reached, the feathered ends of the specimen literally “pop” out of the arteries. If not, excessive traction must not be placed on the specimen. Rather, it should be sharply amputated to avoid rupture of the pulmonary artery distally where repair is impossible. The entire endarterectomy can be performed through the single arteriotomy, and counterincisions are unnecessary. Jamieson limits circulatory arrest periods to 20

minutes and can generally perform the entire right procedure during that period of time. If not, reperfusion for a minimum of 10 minutes is carried out before continuing. We prefer to avoid circulatory arrest and perform as much as possible of the endarterectomy at flows of 500 to 1000 ml/min. After completion of the right side, the pulmonary arteriotomy can be closed immediately or attention can be turned to the left side (the open right arteriotomy can be used for venting the pulmonary artery). As mentioned previously, the disease is usually less extensive on the left side. For this reason the plane may be more difficult to develop and the specimen less organized. The arteriotomy begins proximally on the left pulmonary artery and again extends to the lower lobe branch takeoff (Fig. 5d). The endarterectomy then proceeds in a fashion similar to that on the right. After completion of the endarterectomies, rewarming is immediately begun. During this time the pulmonary arteriotomies are closed with running polypropylene suture. Jamieson routinely inspects the interatrial septum for a patent foramen ovale at this point in the procedure. Conversely, the

1144

World J. Surg. Vol. 23, No. 11, November 1999

Fig. 6. Blunt-tipped suction-dissectors designed at UCSD for pulmonary thromboendartectomy. (Reprinted with permission from the Society of Thoracic Surgeons from Daily et al. [29].)

results of an intraoperative contrast echocardiography study, performed by the anesthesiologist or echocardiographer, can be used to determine the presence or absence of an atrial shunt. Ancillary cardiac procedures can also be performed at this point, but it is not necessary to address the tricuspid regurgitation. If desired, a dose of warm, substrate-enhanced cardioplegia can be administered over 15 to 20 minutes before removing the crossclamp. This provides repletion of energy stores in the hypertrophied right ventricle while raising the temperature of the systemic perfusate. During separation from cardiopulmonary bypass, nitroglycerine and prostaglandin E1 should be available. Inotropes for low cardiac output may also be necessary especially when the pulmonary artery pressure has not been lowered to the normal range. In general, however, the patient is dramatically improved and can be separated from bypass with no difficulty once normothermia is reached. The patient’s autologous blood is then reinfused and the heparin reversed. If there is no bleeding, the activated clotting time need not be completely normalized. Postoperatively, anticoagulation with heparin is begun as soon as possible. At UCSD subcutaneous heparin is given on the night of surgery. Alternatively, a low-dose heparin infusion (400 –500 U/ph) can be initiated as soon as the patient has stabilized and bleeding from the chest tubes is minimal. In any case, some form of early heparinization is imperative so fresh thrombosis of the freshly endarterectomized pulmonary arteries does not occur. The Swan-Ganz catheter is left in place for at least 24 hours and until it is clear that the patient is not developing severe reperfusion pulmonary edema. As during the operation, homologous blood and blood products should be avoided. Results of Surgery Operative mortality has been previously discussed and is detailed for the modern era of PTE in Table 3. Death usually occurs from either severe right heart failure (when sufficient reduction in pulmonary artery pressure is not achieved) or from pulmonary failure due to reperfusion pulmonary edema. Virtually all patients who require right ventricular assist devices expire intraoperatively

Fig. 7. Operative mortality for PTE at the University of Illinois was six times greater in patients with PVR of .1100 dynes (A) and five times greater in patients with mean pulmonary artery pressure (PAP) . 50 mmHg (B). (Reprinted with permission from the Society of Thoracic Surgeons from Hartz et al. [23].)

[23]; those with respiratory failure due to reperfusion pulmonary edema expire within a few days. An occasional patient with multiorgan system failure may linger for days to weeks. Thus the course is usually characterized by “rapid recovery” or “swift demise.” The mortality has dropped from the 22% observed during the earlier era to less than 10% at UCSD. At the University of Illinois, we found that operative mortality was six times greater in patients with a preoperative pulmonary vascular resistance (PVR) greater than 1100 dynes and five times greater in patients with a preoperative mean pulmonary artery pressure (PAP) greater than 50 mmHg (Fig. 7) [23]. Mortality in our patients with PVR less than 1100 dynes was 5.8%. In addition to the usual array of complications seen after any surgery performed on cardiopulmonary bypass, two problems are particularly prominent. Reperfusion pulmonary edema occurred in all of Utley et al.’s 10 patients [25] and was present in 10% of patients in Jamieson et al.’s recent series [9]. It is characterized by sustained arterial hypoxemia, radiographic infiltrates in regions distal to endarterectomized vessels, and normal left-sided filling pressures [22]. It is treated with fluid restriction, the lowest possible inspired oxygen concentration to maintain the arterial saturation above 90%, and mechanical ventilation and positive endexpiratory pressure when severe. Delerium is also common after PTE and according to Wragg et al. is associated with deep hypothermia and the circulatory arrest

Hartz: Chronic Pulmonary Hypertension

1145

Table 4. Preoperative and postoperative hemodynamics in 150 patients operated on by Jamieson et al. at UCSD. Measurement

Preop.

Postop.

p

CO (L/min) PAP systolic (mmHg) PAP mean (mmHg) PVR (dyne z s z cm25)

3.8 6 1.2 78.8 6 22.3 48.5 6 13.7 936.9 6 44.5

5.6 6 1.5 47.0 6 16.9 28.9 6 10.8 299.4 6 16.0

, 0.0001 , 0.0001 , 0.0001 , 0.0001

From Jamieson et al. [9], with permission. Cardiac output (CO), pulmonary artery systolic and mean pressures (PAP), and PVR before and after PTE. The postoperative values were those recorded just before removal of the thermodilution catheter, usually on postoperative day 2 or 3.

time. A total arrest time greater than 55 minutes was 82% sensitive and 80% specific for delerium [30]. It occurred in 77% of patients operated on at UCSD up to 1987 and peaked 72 hours postoperatively. At the University of Illinois, where every attempt has been made to minimize or avoid circulatory arrest, we have seen only one case of severe postoperative delerium. All series report immediate and gratifying relief of symptoms in most patients. The favorable hemodynamic and echocardiographic results demonstrated immediately postoperatively [18] (Tables 1, 2) have been shown by most surgical groups, and Jamieson showed highly significant improvements in cardiac output, peak systolic PAP, mean PAP, and PVR [9] (Table 4). In addition, Mayer et al. have reported on midterm results after PTE [26]. Of their 119 patients who underwent the procedure, 65 had a hemodynamic evaluation 13 to 48 months after surgery (mean 27 months). Sixty-two of the patients were in NYHA class III–IV preoperatively, and 62 were in class I–II postoperatively. In addition, sustained improvements in cardiac output, PAP, and PVR were noted (Fig. 8). Comments Although it is unlikely that pulmonary thromboendarterectomy can be performed in more than a handful of centers in the United States for the foreseeable future, it is clear from a review of the world literature that chronic thromboembolic pulmonary hypertension is a common illness. As increased recognition occurs, it is obvious that in any country or region where large numbers of cardiac surgical procedures are performed at least one surgical group should become familiar with the disease and its surgical treatment. Because results of surgery are clearly dependent on proper patient selection and it is unclear what makes the patient technically inoperable, it is important that the surgeon embarking on a PTE program choose patients who are least ill early in their series. These are relatively young patients with only moderately elevated pulmonary artery pressure and pulmonary vascular resistance who also have clearly demonstrated proximal thrombus angiographically or by CT scan. In this group, an operative mortality less than 10% can be readily achieved. As more experience is gained, the surgeon will be able to offer the procedure to sicker patients including those with more distal disease. Fortunately, the serious morbidity seen in the early days of PTE has practically disappeared. Multiorgan system failure with prolonged stays in the intensive care unit are rare, and the incidence of ischemic complications (neurologic impairment and prolonged

Fig. 8. Midterm results after PTE (mean follow-up at 27 months). Note the sustained improvements in (A) mean pulmonary artery pressure (mPAP), (B) cardiac index (CI), and (C) PVR. (Reprinted with permission from the Society of Thoracic Surgeons from Mayer et al. [26].)

pulmonary failure) have diminished as the circulatory arrest time has been minimized. It is still doubtful whether the entire thromboendarterectomy can be performed without circulatory arrest, especially when bronchial backbleeding is profuse. However, an occasional patient can be operated on without arresting the circulation (Figure 9 shows a specimen removed by us with no circulatory arrest.) Chronic thromboembolic pulmonary hypertension is a fascinating illness that is not uncommon. Although lung transplantation is occasionally considered an option, the true cure of the disease is with surgical thromboendarterectomy followed by lifelong anticoagulation. The results of surgery are gratifying, and a low operative mortality can be achieved after a relatively short learning curve.

1146

World J. Surg. Vol. 23, No. 11, November 1999

poco diagnosticada. Su mejor conocimiento y diagno ´stico permitira´ una cirugı´a curativa en un mayor nu ´mero de pacientes repartidos por todo el mundo; por ello, hemos realizado esta revisio ´n.

References

Fig. 9. PTE specimen removed at the University of Illinois without circulatory arrest. (Reprinted with permission from the Society of Thoracic Surgeons from Hartz et al. [23].)

Re´sume´ L’e`re moderne de la chirurgie pour hypertension pulmonaire en rapport avec la maladie thromboembolique chronique (HTPTEC) a commence´ il y a un peu plus de dix ans. Jusqu’a` cette ´epoque, la thromboendarte ´riectomie pulmonaire (TAP) n’e´tait re ´alise´e que rarement et essentiellement dans un seul centre me´dical (Universite ´ de Californie `a San Diego-UCSD). Cette intervention pose un challenge technique formidable; elle est associe´e `a une mortalite´ ope´ratoire ´eleve´e (.20%) et une morbidite´ excessive en rapport avec l’insuffisance respiratoire et multivisce´rale. A pre ´sent, la TAP est re ´alise´e dans de nombreux centres me´dicaux dans le monde, essentiellement gra ˆce aux efforts des chirurgiens pionniers qui ont de´veloppe ´ et perfectionne´ l’intervention `a UCSD. La mortalite´ ope´ratoire a chute´ et les complications postope´ratoires sont moins fre´quentes. Bien que l’HTPTEC ne soit plus une simple curiosite´ autopsique, le diagnostic n’est pas toujours fait. Le but de cette revue a ´ete´ de souligner qu’une meilleure connaissance de la maladie et un diagnostic plus pre´coce pourrait nous conduire vers une chirurgie `a vise´e curative chez plus de patients dans le monde. Resumen El tratamiento actual de la hipertensio ´n pulmonar tromboembo ´lica cro ´nica (CTEPH) se inicio ´ hace ahora 10 an ˜os. Hasta entonces, la tromboendoarteriectomı´a pulmonar (PTE) era una operacio ´n infrecuente que se efectuaba, pra´cticamente, en un solo hospital (el de la Universidad de California de San Diego UCSD). La intervencio ´n constituı´a un formidable reto te´cnico y se acompan ˜aba de una alta mortalidad operatoria (.20%) ası´ como de una excesiva morbilidad debida no so ´lo al fracaso respiratorio sino tambie´n, al fallo multiorga´nico. Actualmente, la PTE se efectu ´a en numerosos hospitales de todo el mundo, debido sobre todo a los esfuerzos pioneros de aquellos cirujanos que desarrollaron y perfeccionaron la te´cnica quiru ´rgica en la UCSD. La mortalidad operatoria ha disminuido y las complicaciones postoperatorias son mucho menos frecuentes. Aunque la CTEPH no constituye, desde hace tiempo, una mera curiosidad necro ´psica, continu ´a siendo una entidad nosolo ´gica

1. Benotti, J.R., Dalen, J.E.: The natural history of pulmonary embolism. Chest 5:403, 1984 2. Tartulier, M., Boutarin, J., Ritz, B.: Chronic pulmonary thromboembolism. G. Ital. Cardiol. 84(Suppl. I):13, 1984 3. Wilhelmsen, L., Hagman, M., Werko, L.: Recurrent pulmonary embolism: incidence, predisposing factors and prognosis. Acta Med. Scand. 192:565, 1972 4. Sabiston, D.C., Wolfe, W.G., Oldham, H.N., Wechsler, A.S., Crawford, F.A., Jones, K.W., Robert, H.J.: Surgical management of chronic pulmonary embolism. Ann. Surg. 185:699, 1977 5. Reidel, M., Stanek, V., Widimsky, J., Prerovsky, I.: Longterm follow-up of patients with pulmonary thromboembolism. Chest 81:151, 1982 6. Chitwood, W.R., Sabiston, D.C., Wechsler, A.S.: Surgical treatment of chronic unresolved pulmonary embolism. Clin. Chest Med. 5:507, 1984 7. Eisenberg, P.R., Lucore, C., Kaufman, L., Sobel, B.E., Jaffe, A.S., Rich, S.: Fibrinopeptide A levels indicative of pulmonary vascular thrombosis in patients with primary pulmonary hypertension. Circulation 82:841, 1990 8. Presti, B., Berthrong, M., Sherwin, R.: Chronic thrombosis of major pulmonary arteries. Hum. Pathol. 21:601, 1990 9. Jamieson, S.W., Augur, W.R., Fedullo, P.F., Channick, R.N., Kriett, J.M., Tarazi, R.Y., Moser, K.M.: Experience and results of 150 pulmonary thromboendarterectomy operations over a 29 month period. J. Thorac. Cardiovasc. Surg. 106:116, 1993 10. Sandoval, J., Amigo, M.C., Barragan, R., Izaguirre, R., Reyes, P.A., Martinez-Guerra, M.L., Palomar, A., Gomez, A., Garcia-Torres, R.: Primary antiphospholipid syndrome presenting as chronic thromboembolic pulmonary hypertension. J. Rheumatol. 23:772, 1996 11. Simonneau, G., Azarian, R., Brenot, F., Dartevelle, P.G., Musset, D., Duroux, P.: Surgical management of unresolved pulmonary embolism. Chest 107:52S, 1995 12. Woodruff, W.W., Merten, D.F., Wagner, M.L., Kirks, D.E.: Chronic pulmonary embolism in children. Radiology 159:511, 1986 13. Schamroth, C.L., Sarelli, P., Pocock, W.A., Davidoff, R., King, J., Reinach, G., Barlow, J.B.: Pulmonary arterial thrombosis in secundum atrial septal defect. Am. J. Cardiol. 60:1152, 1987 14. Moser, K.M., Auger, W.R., Fedullo, P.F., Jamieson, S.W.: Chronic thromboembolic pulmonary hypertension: clinical picture and surgical treatment. Eur. Respir. J. 5:334, 1992 15. Rich, S., Levitsky, S., Brundage, B.H.: Pulmonary hypertension from chronic pulmonary thromboembolism. Ann. Intern. Med. 198:425, 1988 16. Ryan, K.L., Fedullo, P.F., Davis, G.B., Vasquez, T.E., Moser, K.M.: Perfusion scan findings understate the severity of angiographic and hemodynamic compromise in chronic thromboembolic pulmonary hypertension. Chest 93:1180, 1988 17. Auger, W.R., Fedullo, P.F., Moser, K.M., Buchbinder, M., Peterson, K.L.: Chronic major-vessel thromboembolic pulmonary artery obstruction: appearance at angiography. Radiology 182:393, 1992 18. Dittrick, H.C., Nicod, P.H., Chow, L.C., Chappuis, F.P., Moser, K.M., Peterson, K.L.: Early changes of right heart geometry after pulmonary thromboendarterectomy. J. Am. Coll. Cardiol. 11:937, 1988 19. Schwickert, H.C., Schweden, F., Schild, H.H., Piepenburg, R., Duber, C., Kauczor, H., Renner, C., Iversen, S., Thelen, M.: Pulmonary arteries and lung parenchyma in chronic pulmonary embolism: preoperative and postoperative CT findings. Radiology 191:351, 1994 20. King, M.A., Bergin, C.J., Yeung, D.W.C., Belezzouli, E.E., Olson, L.K., Ashburn, W.L., Auger, W.R., Moser, K.M.: Chronic pulmonary thromboembolism: detection of regional hypoperfusion with CT. Radiology 191:59, 1994 21. Shure, D., Gregoratos, G., Moser, K.M.: Fiberoptic angioscopy: role

Hartz: Chronic Pulmonary Hypertension

22. 23. 24. 25. 26.

in the diagnosis of pulmonary arterial obstruction. Ann. Intern. Med. 103:844, 1985 Levinson, R.M., Shure, D., Moser, K.M.: Reperfusion pulmonary edema after pulmonary artery thromboendarterectomy. Am. Rev. Respir. Dis. 134:1241, 1986 Hartz, R.S., Byme, J.G., Levitsky, S., Park, J., Rich, S.: Predictors of mortality in pulmonary thromboendarterectomy. Am. Thorac. Surg. 62:1255, 1996 Synder, W.H., Kent, D.C., Baesch, B.F.: Successful endarterectomy of chronically occluded pulmonary artery; clinical report and physiologic. J. Thorac. Surg. 45:482, 1964 Utley, J.R., Spragg, R.G., Long, W.B., Moser, K.M.: Pulmonary endarterectomy for chronic thromboembolic obstruction: recent surgical experience. Surgery 92:1096, 1982 Mayer, E., Manfred, D., Hake, U., Schmid, F., Pitton, M., Kupferwas-

1147

27.

28. 29. 30.

ser, I., Iversen, S., Oelert, H.: Mid-term results of pulmonary thromboendarterectomy for chronic thromboembolic pulmonary hypertension. Ann. Thorac. Surg. 61:1788, 1996 Alfieri, O., Toracca, L., Pardini, A., Ferrari, G., Santoli, C., Vigano, M., Pellegrini, A.: La tromboendoarterectomia polmonare nel trattamento chirurgico dell’ipertensione polmonare cronica tromboembolica: prime esperienze Italiane. Cardiologia 40:561, 1995 Daily, P.O., Dembitsky, W.P., Daily, R.P.: Dissectors for pulmonary thromboendarterectomy. Ann. Thorac. Surg. 51:842, 1991 Daily, P.O., Dembitsky, W.P., Iversen, S., Moser, K.M., Auger, W.: Risk factors for pulmonary thromboendarterectomy. J. Thorac. Cardiovasc. Surg. 99:670, 1990 Wragg, R.E., Dimsdale, T.E., Moser, K.M., Daily, P.O., Dembitsky, W.P., Archibald, C.: Operative predictors of delirium after pulmonary thromboendarterectomy. J. Thorac. Cardiovasc. Surg. 96:514, 1988