- Email: [email protected]
Reoperative pulmonary thromboendarterectomy
Reoperative Pulmonary Thromboendarterectomy Makoto Mo, MD, David P. Kapelanski, MD, Surindra N. Mitruka, MD, William R. Auger, MD, Peter F. Fedullo, MD, Richard N. Channick, MD, Kim Kerr, MD, Carol Archibald, RN, and Stuart W. Jamieson, FRCS Division of Cardiothoracic Surgery and Division of Pulmonary and Critical Care Medicine, University of California, San Diego, California
Background. Recurrent symptomatic pulmonary hypertension is uncommon after primary pulmonary thromboendarterectomy (PTE). We reviewed our experience with patients undergoing repeat PTE to determine the risk factors for recurrent disease, and the selection criteria, relative risks, and functional outcomes of reoperative PTE. Methods. Since 1990, 13 of 870 (1.5%) patients underwent reoperative PTE at our institution. These 7 men and 6 women (mean age 38.6 years) were contrasted with the most recent 225 patients (111 men, 114 women, mean age 52.7 years) who underwent primary PTE for whom complete hemodynamic data are available. The preoperative evaluation of all patients was similar. Pulmonary hemodynamic data and outcome measures were compared between groups. Results. Of 13 reoperated patients: 69% (9/13) had their primary operation at another institution, 54% (7/13) initially underwent unilateral PTE, 38% (5/13) had identifiable coagulation disorders, 38% (5/13) had ineffective caval filtration, 31% (4/13) had suboptimal anticoagulation management, and 31% (4/13) had complete unilateral
pulmonary artery obstruction. The mean interval to reoperation was 5.2 years (range 0.7 to 10.9 years). All control patients underwent bilateral PTE using hypothermic circulatory arrest. Operative mortality was 7.7% (1/13) with reoperation vs 8.4% (19/225) in controls. No difference (p ⴝ NS) was observed between groups in the preoperative pulmonary artery pressure (PAP) or pulmonary vascular resistance; however, the control group had a significantly (p < 0.05) greater reduction in the postoperative PAP (46/19, mean 28 mm Hg vs 59/23, mean 35 mm Hg) and PVR (271 ⴞ 172 vs 399 ⴞ 154 dynes/s/ cmⴚ5) compared with the redo group. No substantial difference in morbidity or functional outcomes was observed between groups. Conclusions. Reoperative PTE can be performed with a perioperative risk comparable with primary PTE, although the improvement in pulmonary hemodynamics is not as favorable. Bilateral primary operation, effective caval filtration, and vigilant anticoagulant management would prevent the need for most reoperative PTEs. (Ann Thorac Surg 1999;68:1770– 7) © 1999 by The Society of Thoracic Surgeons
C
been substantial, with a mortality of 6.4% in the last 500 cases performed, and with 95% of surviving patients in New York Heart Association (NYHA) class I or II at 1 year postoperatively [10]. The natural history of patients with chronic thromboembolic pulmonary hypertension after PTE remains unclear, although our experience suggests that relatively few patients develop recurrent disease. Despite 5%–10% of patients having a defined coagulation abnormality, such as lupus anticoagulant, protein C deficiency, antithrombin III deficiency, or heparininduced platelet antibody, the vast majority of cases of thromboembolic pulmonary hypertension are due to “spontaneous” thromboembolism. Additionally, although we have previously documented a greater risk for recurrent thromboembolic pulmonary hypertension with unilateral PTE and unilateral PA obstruction [11], the risk factors for recurrent disease and the selection criteria for reoperation remain poorly defined. Likewise, analogous to other reoperative cardiac operation procedures [12, 13], it might be anticipated that a second PTE entails greater risk than the initial operation, although the extent of that risk remains unclear. Similarly, it can not be presumed that normal pulmonary blood flow and resis-
hronic thromboembolic pulmonary hypertension is an insidious and infrequently diagnosed disease. The prognosis for patients with untreated pulmonary hypertension is poor, with a 5-year mortality of 90% if the mean pulmonary artery pressures (PAP) are higher than 50 mm Hg [1, 2]. Medical therapy for thromboembolic pulmonary hypertension, using anticoagulants, vasodilators, or thrombolytic agents, is seldom effective [3, 4]. Surgical intervention offers the only consistent and definitive treatment for this disease [5, 6]. Our institution has been engaged in a program for the surgical management of chronic thromboembolic pulmonary hypertension since 1970. In that interval, over 1,100 pulmonary thromboendarterectomies (PTE) have been performed, although the majority (n ⫽ 910) have been performed by two surgeons (S.W.J. and D.P.K.) at our facility within the past decade [5–9]. The success with surgical intervention in this group of difficult patients has Presented at the Thirty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 25–27, 1999. Address reprint requests to Dr Jamieson, Division of Cardiothoracic Surgery (8892), University of California, San Diego, 200 West Arbor Dr, San Diego, CA 92103-8892; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
0003-4975/99/$20.00 PII S0003-4975(99)01043-7
Ann Thorac Surg 1999;68:1770 –7
MO ET AL REOPERATIVE PTE
1771
tance, ie, the functional outcome, would be restored in all instances, although this is currently unsubstantiated. The purpose of this study was to analyze our experience with patients undergoing repeat PTE to determine the specific risk factors for the development of recurrent disease, the selection criteria for reoperation, the relative risk for reoperation, and the functional outcome after reoperative PTE.
Patients and Methods Between November 1989 and December 1998, 870 consecutive PTEs for chronic thromboembolic pulmonary hypertension were performed by two surgeons at the University of California, San Diego Medical Center. During this period, 13 patients (1.5%) underwent repeat PTE. Reoperated patients were contrasted with a consecutive cohort of 225 control patients who underwent primary PTE between March 1996 and December 1997, and for whom complete hemodynamic data were available. Our routine preoperative evaluation for all patients with suspected chronic thromboembolic pulmonary hypertension has been reported previously [5–10]. In brief, this assessment entails a comprehensive medical history and physical examination; chest radiograph and electrocardiogram, pulmonary function studies, and arterial blood gas analysis; transthoracic echocardiography; radionuclide perfusion and ventilation scans; screening for recognized coagulation abnormalities; and right heart catheterization and pulmonary angiography [5, 6]. Although magnetic resonance imaging, computed tomography, and bronchial angiography can contribute important diagnostic information, at present, they lack sufficient resolution to detect disease originating at the origin of the segmental pulmonary artery branches. Accordingly, these modalities are not routinely utilized in our institution to determine operative candidacy. Fiberoptic pulmonary angioscopy is reserved for those patients (approximately 20%) with equivocal angiographic features of chronic thromboembolic disease [14]. In the absence of a precedent concerning repeat PTE, we reasoned that if the current angiogram demonstrated limited disease proximal to the origin of the segmental pulmonary arteries, deteriorating pulmonary hemodynamics could only be attributed to the progression of distal small vessel disease not accessible at reoperation. Accordingly, we believed that the most useful assessment when considering repeat PTE was a qualitative comparison of the current pulmonary angiogram with that preceding the primary operation (Fig 1). Patients excluded following this evaluation were considered for pulmonary transplantation. Our technique for primary PTE, which utilizes profound hypothermia and circulatory arrest, has been reported previously [9]. Although no major modifications were necessary for reoperation, particular care was taken during repeat sternotomy to avoid injury to the dilated and often adherent right atrium or ventricle. Once adequate exposure of the right atrium and ascending aorta was achieved, additional dissection was readily per-
Fig 1. A qualitative comparison of the pulmonary angiogram preceding the primary operation (A) with the current angiogram (B) preceding the reoperation.
formed during the induction of hypothermia while on cardiopulmonary bypass. Identification and distal dissection in the appropriate thromboendarterectomy plane was generally more demanding at reoperation due to the intense and often inflammatory adherence of recurrent thromboemboli to the previously operated pulmonary arteries. With care, this dissection was effectively extended at least two branch divisions distal to the origin of the segmental arteries (Fig 2). Hemostasis was readily accomplished in all cases, without the routine use of platelets, fresh frozen plasma, epsilon aminocaproic acid, or aprotinin. Intravenous heparin anticoagulation was initiated once operative bleeding had ceased, followed by warfarin, titrated to prolong the international normalized ratio (INR) beyond 3.0. Final postoperative hemodynamic determinations were obtained in the absence of inotropic support or intravenous vasodilator agents. Patient follow-up was performed at the outpatient clinic or by telephone contact with the patient and their primary physician. The mean ⫾ standard deviation are tabulated for all continuous variables. The unpaired Student’s t test,
1772
MO ET AL REOPERATIVE PTE
Ann Thorac Surg 1999;68:1770 –7
Table 1. Reoperative Pulmonary Thromboendarterectomy Patient Profile Patient/ Gender
Age (years)
Previous PTE
Coagulation Disorder
1M 2F 3M 4F 5M 6F 7F 8F
64 34 32 20 30 38 36 25
L thor R thor L thor R stern R stern L stern L stern Bilateral
9F 10 M 11 M 12 M 13 M
35 24 64 41 59
Bilateral Bilateral Lupus anticoagulant Bilateral ⫻ 2 Bilateral Bilateral
Anticoagulation
Caval Filtration
None None Poor D/C 2 years None
Anticardiolipin Ab HIPA Anticardiolipin Ab HIPA Congenital dysfibrinogenemia Poor Poor D/C 1 year HIPA lupus anticoagulant
13 total 39 ⫾ 15 7 unilateral 5/13 (38%) 7 male 6 bilateral 6 female
PA Obstruction
Yes Yes
Occluded
Yes
Tilted Yes Poor 4/13 (31%)
5/13 (38%)
4/13 (31%)
Preop PAP
Postop PAP
72/30 (48) 51/17 (28) 85/27 (46) 32/15 (21) 95/40 (60) 52/15 (27) 90/42 (60) 64/26 (39) 85/33 (50) 43/21 (28) 98/30 (53) 100/37 (58) 60/28 (40) 59/23 (35) 75/35 (50) 84/38 (52)
92/25 (52) 70/30 (43) 100/45 (60) 70/28 (41) 70/20 (35)
51/24 (33) 43/15 (27) 55/20 (32) 58/18 (31) 70/29 (43)
82/31 (49) 59/23 (35)
paired Student’s t test, and 2 test were utilized for statistical analysis. A p value less than 0.05 was considered significant. This study was performed under the auspices of the Institutional Review Board of the University of California, San Diego.
Results Patient Demographics
Fig 2. Gross specimens removed from the pulmonary arteries of a patient after the primary operation (A) and after the repeat operation (B). The “tails” represent thrombus removed from the segmental pulmonary arteries.
Preoperative and postoperative data for redo and primary patients are displayed in Tables 1 and 2. Seven reoperated patients were men and 6 were women. This distribution did not differ from the control group (111 men and 114 women). Reoperated patients were younger (39 ⫾ 15 years, range 20 to 69 years) than the reference population (53 ⫾ 15 years, range 16 to 85 years) ( p ⬍ 0.01). Nine of the 13 (69%) redo patients received their initial PTE operation elsewhere. This operation was unilateral in 7 of those 9 patients, of whom 3 underwent the first PTE by thoracotomy. The remaining 5 patients, as well as the 4 who had received their primary operation at the UCSD Medical Center, previously underwent bilateral PTE utilizing our current technique [15]. One of these 6 patients underwent two reoperations. None of the reoperated patients experienced a documented deep venous thrombosis or clinically recognized pulmonary embolism after their initial operation, although patient 3, with tricuspid endocarditis, suffered a paradoxical cerebral embolism. Despite documented chronic thromboembolic pulmonary disease, 5 of 13 (38%) redo patients had ineffective caval filtration. Caval filtration was never performed in patients 1, 2, or 3 before the recurrence of symptoms, while in patient 11, the caval
Ann Thorac Surg 1999;68:1770 –7
1773
MO ET AL REOPERATIVE PTE
Table 1. Continued Preop PVR
Postop PVR
1013 1013 887 1124 557 906 650 1152
357 322 190 610 313 669 319 621
1539 883 634 489 354
480 295 229 330 456
Perfusion Scan of Occluded PA
Morbidity SVT SVT Reperfusion injury
No flow Min flow
Reperfusion injury, trach med bldng Reperfusion injury, trach Reperfusion, ECCO2R Med bldng, pneumonia
Good flow
Sinus node dysfunction pacemaker Reperfusion injury
No flow Reperfusion injury SVT
862 ⫾ 332 399 ⫾ 154 Reperfusion 1/4 (25%) Reperfusion injury 6/13 dysrhythmia 4/13 med bldng 2/13
Follow-up (years)
Current Status
Cause of Death
? 4 5.6 7.5 1.3 6.7 4.5 2.2
Dead Alive Alive Alive Alive Alive Dead Alive
?
3 3.1 1.6 1.3 6 days
Alive Alive Alive Alive Dead
3.4 years
R heart failure
Asystole
Hospital mortality 1/13 (7.7%)
ECCO2R ⫽ extracorporeal carbon dioxide removal; HIPA ⫽ heparin-induced platelet antibody; med bldng ⫽ mediastinal bleeding; PA ⫽ pulmonary artery; PAP ⫽ pulmonary artery pressure (in mm Hg); PTE ⫽ pulmonary thromboendarterectomy; PVR ⫽ pulmonary vascular ⫺5 resistance in dynes/s/cm ; SVT ⫽ supraventricular tachycardia.
filter was malpositioned, and in patient 9, occluded by thrombus. Four of 13 (31%) redo patients had suboptimal anticoagulation management after the primary PTE, at variance with our established policy of lifelong anticoagulation of patients with this disease. Warfarin was discontinued within 2 years of operation in 2 patients (nos. 3 and 6), while in 2 additional patients (nos. 5 and 13), therapeutic prolongation of the INR was not adequately sustained. Five of 13 patients (38%) in the reoperated cohort had a defined abnormality of coagulation. While this incidence is higher than in the reference group (22%), this difference was not statistically significant, nor was there a predominant coagulation defect among reoperated patients. Notably, 4 patients (31%) in the reoperated series had complete unilateral occlusion of the pulmonary arteries, compared with only 16 of 225 patients (7%) in the control group, which represented a significantly higher incidence ( p ⬍ 0.01). Postoperative radionuclide perfusion scans demonstrated variable improvement in the 12 surviving reoperated patients. No flow was reestablished in patients 6 or 12, minimal flow in patient 7, whereas only patient 9 demonstrated a significant improvement in flow to the unilaterally occluded lung after reoperation. The low incidence (25%) of reperfusion in these 4 patients is comparable with our previously reported experience of absent reflow (8 of 16, 50%) among primarily operated patients presenting with complete unilateral pulmonary artery obstruction [11].
Survival The 7.7% (1 of 13) operative mortality among redo patients did not differ ( p ⬎ 0.05) from the 8.4% (19 of 225)
mortality among primary patients in this series, although the previously reported operative mortality in a larger cohort of 500 primary PTE patients was 6.4% (10). The single death among reoperated patients (patient 13) occurred on postoperative day 6, when hypotension from refractory supraventricular tachycardia (SVT) precipitated cardiac arrest. The 1- and 3-year actuarial survival, at a mean follow-up of 3.4 years after reoperation, was 92% and 82%, respectively. One patient (patient 1) died 1.3 years after reoperation of progressive respiratory failure due to preexistent idiopathic pulmonary fibrosis. A second patient (patient 7) with lupus anticoagulant antibody, complete unilateral pulmonary artery obstruction, and minimal reperfusion after reoperation died of progressive right heart failure 4.5 years after repeat PTE. The 1- and 3-year actuarial survival among primarily operated patients was 90% and 81%, respectively.
Pulmonary Hemodynamics The preoperative pulmonary artery pressures (pulmonary artery systolic/diastolic; mean PAP in mm Hg) were equivalent (p ⫽ NS) between the redo patients (82 ⫾ 13/31 ⫾ 6; 49 ⫾ 8) and the reference cohort (80 ⫾ 22/31 ⫾ 10; 48 ⫾ 14). Likewise, the preoperative pulmonary vascular resistance (PVR) (dynes/s/cm⫺5) was similar (p ⫽ NS) between the reoperative PTE patients (862 ⫾ 332) and the primary PTE patients (842 ⫾ 431) (Figs 3 and 4; Table 2). In each patient, pulmonary angiography depicted significant obstruction of the central arteries (Fig 1). When measured in the intensive care unit after withdrawal of vasoactive and inotropic medications, the reoperated patients demonstrated a significant reduction in PAP (82 ⫾ 13/31 ⫾ 6; 49 ⫾ 8 preoperative vs 59 ⫾ 18/23 ⫾
1774
MO ET AL REOPERATIVE PTE
Ann Thorac Surg 1999;68:1770 –7
Table 2. Redo Versus Primary PTE Patient Outcomes Redo (n ⫽ 13) Mortality 1/13 (7.7%) Morbidity Reperfusion injury 6/13 (46%) Atrial arrhythmia 4/13 (31%) Mediastinal bleeding 2/13 (15%) Ventilator days 9.8 ⫾ 12.3 ICU length of stay 11 ⫾ 12.5 Pulmonary hemodynamics Preoperative PAP 82/31 (49) Postoperative PAP 59/23 (35) Preoperative PVR 862 ⫾ 332 Postoperative PVR 399 ⫾ 154
Control (n ⫽ 225)
p Value
20/225 (8.9%)
NS
75/25 (33%) 29/225 (13%) 5/225 (2%) 5.1 ⫾ 12.1 6.6 ⫾ 12.7
NS ⬍ 0.05 ⬍ 0.01 NS NS
80/31 (48) 46/19 (28) 842 ⫾ 431 271 ⫾ 172
NS ⬍ 0.01 NS ⬍ 0.01
ICU ⫽ intensive care unit; NS ⫽ not significant; PAP ⫽ pulmonary artery pressure (in mm Hg); PVR ⫽ pulmonary vascular resistance (in dynes/s/cm⫺5).
8; 35 ⫾ 11 postoperative) and PVR (862 ⫾ 332 preoperative vs 399 ⫾ 154 postoperative) ( p ⬍ 0.01 for all comparisons) (Figs 3 and 4). Primarily operated patients likewise demonstrated a significant reduction in PAP (80 ⫾ 22/31 ⫾ 10; 48 ⫾ 14 preoperative vs 46 ⫾ 16/19 ⫾ 7; 28 ⫾ 9 postoperative) and PVR (842 ⫾ 431 preoperative vs 271 ⫾ 172 postoperative) ( p ⬍ 0.01 for all comparisons) (Figs 3 and 4). The relative reduction in postoperative PAP between redo patients (59 ⫾ 18/23 ⫾ 8; 35 ⫾ 11) and primary PTE patients (46 ⫾ 16/19 ⫾ 7; 28 ⫾ 9) was greater in the reference cohort ( p ⬍ 0.01). Similarly, the reduction in PVR observed in the primary PTE patients (271 ⫾ 172) was greater ( p ⬍ 0.01) than that observed in the redo
Fig 3. The preoperative and postoperative mean PAPs in the redo (49 vs 35 mm Hg) and primary (48 vs 28 mm Hg) PTE cohorts. As demonstrated, there was a greater reduction in the mean PAP among patients undergoing a primary PTE (p ⬍ 0.01).
Fig 4. The preoperative and postoperative PVR in the redo (862 vs 399 dynes/s/cm⫺5) and primary (842 vs 271 dynes/s/cm⫺5) PTE cohorts. As indicated, there was a greater reduction in PVR among patients undergoing primary PTE (p ⬍ 0.01).
PTE patients (399 ⫾ 154) postoperatively (Figs 3 and 4; Table 2).
Functional Outcome Eleven of 13 redo patients experienced a functional benefit after their initial PTE, but developed recurrent symptoms of pulmonary hypertension from 6 months to 10 years after the primary operation (mean 2.7 ⫾ 2.8 years). The mean interval to reoperative PTE was 5.2 years ⫾ 2.7 years (range 8 months to 10.9 years). Immediately before the second PTE, the majority of patients were NYHA class III (11 of 13, 85%) or class IV (1 of 13, 8%). The exception was the patient with paradoxical embolism, who was in NYHA class II (Fig 5).
Fig 5. The preoperative and postoperative functional NYHA class for the 13 reoperative patients.
Ann Thorac Surg 1999;68:1770 –7
The majority of surviving patients (10 of 12, 83%) reported symptomatic improvement after repeat PTE. Postoperatively, 5 patients were in NYHA class I, 5 in class II, and 2 in class III (Fig 5). Of the 2 patients remaining in class III, patient 2 reported no functional benefit despite a 70% reduction in the preoperative PVR from 1,013 to 322 dynes/s/cm⫺5, whereas patient 12 had unilateral PA obstruction with no reflow demonstrated by perfusion scan. Notably, patient 9 has deteriorated from NYHA class I to class IV 2 years after reoperation.
Surgical Management Indicative of the greater difficulties encountered during repeat PTE, the cumulative duration of circulatory arrest (45 ⫾ 19 vs 35 ⫾ 14 minutes, p ⬍ 0.05), and the aortic cross-clamp time (127 ⫾ 14 vs 97 ⫾ 41 minutes, p ⬍ 0.01) were prolonged in the reoperated patients. The duration of cardiopulmonary bypass was also prolonged (249 ⫾ 51 vs 225 ⫾ 50 minutes), although not significantly ( p ⬎ 0.05). Notwithstanding the increased technical challenge of repeat PTE, substantial chronic thromboembolic material was recovered in each patient (Fig 2). The type of disease encountered in reoperative cases was different than that typically encountered in our primarily operated series. Among redo patients, 5 of 13 (38%) had type 1 disease (partially organized central thrombus proximal to organized distal thrombus), and 8 of 13 patients (62%) had type 2 disease (organized thrombus proximal to the origin of the segmental artery). In contrast, although type 2 disease predominates (70%) among primarily operated patients, only 10% historically demonstrate type 1 disease [15].
Complications Despite the inherent risks of reoperative cardiac procedures, the incidence and nature of complications encountered in this reoperative series was comparable with those experienced after primary operation (Table 2). Two of 13 reoperated patients (15%) required reexploration for perioperative hemorrhage, significantly greater than the 2% (5 of 225) incidence of reexploration among primarily operated patients ( p ⬍ 0.01). Four reoperated patients (31%) experienced significant perioperative rhythm disturbances (sinus node dysfunction in 1, and supraventricular tachycardia in 3 patients), nearly threefold greater than the incidence of arrhythmias experienced in the reference population (29 of 225, 13%; p ⬍ 0.05). Additionally, the incidence of reperfusion injury (6 of 13; 46% vs 75 of 225; 33%), the duration of mechanical ventilation (9.8 ⫾ 12.3 days vs 5.1 ⫾ 12.1 days), and the length of intensive care unit stay (11 ⫾ 12.5 days vs 6.6 ⫾ 12.7 days) was greater in reoperated patients, although none of these differences was statistically significant ( p ⬎ 0.05 for all comparisons). Although 3 reoperated patients required more than 20 days of assisted ventilation, including patient 8, who was supported with extracorporeal carbon dioxide removal (ECCO2R) for 6 days, no patient in the reoperated series succumbed to respiratory failure.
MO ET AL REOPERATIVE PTE
1775
Comment Acute pulmonary thromboembolism is a more common condition than is generally appreciated, and in many cases, is asymptomatic until end-stage debilitating pulmonary hypertension develops. In the majority of patients, spontaneous resolution of acute pulmonary embolism is the rule. However, a small but uncertain percentage develop chronic thromboembolic pulmonary hypertension for which medical therapy is ineffective [3, 4], leaving surgical intervention the only definitive option [7–9, 16, 17]. The team at the University of California, San Diego Medical Center has benefited from an institutional experience of more than 1,100 PTE cases since the inception of this program in 1970 [7–9, 18]. While accumulating this experience, we have had the opportunity to consider reoperation for patients initially managed here as well as at other institutions. The purpose of this study was to determine the risk factors for recurrent thromboembolic disease, the selection criteria for reoperation, the relative risk of repeat PTE, and the functional benefit derived from reoperative PTE. Generally, reoperative cardiac surgery is accompanied by a higher perioperative risk and reduced functional outcome compared with the equivalent primary procedure [12, 13]. Accordingly, an appreciation of the factors, technical and others, that may contribute to an unsatisfactory outcome after the primary operation can only serve to improve the results of primary operation and limit the likelihood that subsequent procedures will be necessary. Additionally, because PTE is almost invariably lethal whenever a significant improvement in pulmonary blood flow is not achieved, the selection criteria for reoperative PTE warrants particular scrutiny. Seven of the 13 patients in this series initially underwent a unilateral primary operation, yet at reoperation, each of these 7 patients had significant bilateral chronic thromboembolic disease, as we found substantial thromboembolic material in the contralateral side of the prior surgery in these redo patients. In the comparison cohort of 225 patients, in whom bilateral pulmonary artery exploration was performed regardless of the preoperative findings, only 6 of 225 patients (2.7%) exhibited true unilateral disease. Accordingly, we believe that unilateral PTE for bilateral disease poses a substantial risk for early recurrence of symptomatic pulmonary hypertension. Less certain were the possible causes of recurrent symptomatic pulmonary hypertension in those 6 patients who initially underwent bilateral PTE using our present technique, which employs profound hypothermia and circulatory arrest. While none of those patients had experienced a clinical episode consistent with recurrent deep venous thrombosis or pulmonary embolism, it must be recalled that thromboembolic events are clinically occult in the majority of patients [19]. Additionally, the possibility of in situ thrombosis of portions of the previously endarterectomized proximal vascular bed cannot be excluded by present means [20]. Each of the patients in the reoperated series, except patient 8, had at least one
1776
MO ET AL REOPERATIVE PTE
recognized risk factor for recurrent disease, eg, coagulation disorder, poor anticoagulation management, ineffective caval filtration, or complete unilateral PA obstruction, that could have predisposed to a thrombotic or embolic event. While it can be argued that differentiation between these two possibilities is of academic interest only, the essential observation remains that scrupulous prophylaxis may have eliminated the need for reoperation in these patients. The validity of our initially empiric selection criteria, based predominantly on the pulmonary angiogram, appears to be verified by the outcomes achieved in this small sample of patients. In each instance, at operation, we were able to identify a suitable plane of dissection and extract significant chronic thromboembolic material from both lungs, even when the prior operation was bilateral (Fig 2). Though the improvement in pulmonary blood flow was not as great as could be accomplished in primarily operated patients (Figs 3 and 4), the majority of reoperated patients experienced some sustained increment in functional class (Fig 5). Notwithstanding the difficulties encountered at reoperation, and despite higher perioperative morbidity, these benefits were attained without incurring a substantially greater mortality than in primarily operated patients. While the benefits identified among those patients with complete unilateral pulmonary artery obstruction suggest that this cohort should undergo reoperation with only limited expectations, we believe that reoperation with this caveat is preferable to either pulmonary transplantation or inexorable right heart failure and premature death.
References 1. Dalen JE, Alpert JS. Natural history of pulmonary embolism. Prog Cardiovasc Dis 1975;17:259–70. 2. Riedel M, Stanek V, Widimsky J, Prerovsky I. Long term follow-up of patients with pulmonary thromboembolism. Late prognosis and evolution of hemodynamic and respiratory data. Chest 1982;81:151– 8. 3. Dash H, Ballentine N, Zelis R. Vasodilators ineffective in secondary pulmonary hypertension. N Eng J Med 1980;303:1062–3. 4. Dantzker DR, Bower JS. Partial reversibility of chronic pulmonary hypertension caused by pulmonary thromboembolic disease. Am Rev Respir Dis 1981;124:129–33. 5. Moser KM, Auger WR, Fedullo PF. Chronic major vessel
Ann Thorac Surg 1999;68:1770 –7
6. 7. 8.
9.
10. 11.
12.
13. 14. 15. 16. 17.
18. 19. 20.
thromboembolic pulmonary hypertension. Circulation 1990; 81:1735– 43. Fedullo PF, Auger WR, Channick RN, Moser KM, Jamieson SW. Chronic thromboembolic pulmonary hypertension. Clin Chest Med 1995;16:353–74. Utley JR, Spragg RG, Long WB, Moser KM. Pulmonary endarterectomy for chronic thromboembolic obstruction: recent surgical experience. Surgery 1982;92:1096 –102. Daily PO, Dembitsky WP, Iversen S, Moser KM, Auger WR. Current early results of pulmonary thromboendarterectomy for chronic pulmonary embolism. Eur J Cardiothorac Surg 1990;4:117–21. Jamieson SW, Auger WR, Fedullo PF, et al. Experience and results with 150 pulmonary thromboendarterectomy operations over a 29 month period. J Thorac Cardiovasc Surg 1993; 106:116–26. Moser KM, Auger WR, Fedullo PF, Jamieson SW. Chronic thromboembolic pulmonary hypertension: clinical picture and surgical treatment. Eur Resp J 1992;5:334– 42. Hirsch AM, Moser KM, Auger WR, Channick RN, Fedullo PF. Unilateral pulmonary artery thrombotic occlusion: is distal arteriopathy a consequence? Am J Respir Crit Care Med 1996;154:491– 6. Salomon NW, Page US, Bigelow JC, Krause AH, Okies JE, Metzdorff MT. Reoperative coronary surgery. Comparative analysis of 6591 patients undergoing primary bypass and 508 patients undergoing reoperative coronary artery bypass. J Thorac Cardiovasc Surg 1990;100:250–9. Pansini S, Ottino G, Forsennati PG, et al. Reoperations on heart valve prostheses: an analysis of operative risks and late results. Ann Thorac Surg 1990;50:590– 6. Shure D, Gregoratos G, Moser KM. Fiberoptic angioscopy: role in the diagnosis of chronic pulmonary arterial obstruction. Ann Intern Med 1985;103:844–50. Jamieson SW. Pulmonary thromboendarterectomy. In: Franco KL, Putnam JB, eds. Advanced therapy in thoracic surgery. Hamilton: B.C. Decker, Inc., 1998:310– 8. Chitwood WR, Sabiston DC, Wechsler AS. Surgical treatment of chronic unresolved pulmonary embolism. Clin Chest Med 1984;5:507–36. Mayer E, Dahm M, Hake U, et al. Mid-term results of pulmonary thromboendarterectomy for chronic thromboembolic pulmonary hypertension. Ann Thorac Surg 1996;61: 1788 –90. Moser KM, Braunwald NS. Successful surgical intervention in severe chronic thromboembolic pulmonary hypertension. Chest 1973;64:29–35. Cranley JJ, Canos AJ, Sull WJ. The diagnosis of deep venous thrombosis. Fallibility of clinical symptoms and signs. Arch Surg 1976;111:34– 6. Moser KM, Fedullo PF, Finkbeiner WE, Golden J. Do patients with primary pulmonary hypertension develop extensive central thrombi? Circulation 1995;91:741–5.
DISCUSSION DR CHRISTOPHER G. A. McGREGOR (Rochester, MN): Dr Anderson, Dr Pairolero, ladies, and gentlemen. It is a great pleasure to discuss this very well-presented paper on redo pulmonary thromboendarterectomy (PTE) from the UC, San Diego group. This presentation represents 13 of 870 patients undergoing PTE at San Diego, a total number that probably represents about 80% of the world experience. Doctor Jamieson and colleagues are to be congratulated on their leadership in this area and their outstanding results in this series. They have an operative mortality of 7.7% for redos and 8.4% for the reference population of first-time cases. The few studies of any numbers in
the literature of PTE performed elsewhere have reported a mortality generally in excess of 20%. In our own initial experience at Mayo, in the first 29 elective PTEs, the mortality has been at 10%, somewhat similar to the San Diego experience. And I would like to thank the San Diego group for the opportunity to visit them and learn from their experience. In today’s report, the redo patients were younger than the reference population, and only 2 of 13 had no identifiable risk factors for recurrent thrombotic problems. These risk factors may be coagulation disorders, suboptimal anticoagulation, an occluded or malpositioned IVC filter, or previous unilateral
Background. Recurrent symptomatic pulmonary hypertension is uncommon after primary pulmonary thromboendarterectomy (PTE). We reviewed our experience with patients undergoing repeat PTE to determine the risk factors for recurrent disease, and the selection criteria, relative risks, and functional outcomes of reoperative PTE. Methods. Since 1990, 13 of 870 (1.5%) patients underwent reoperative PTE at our institution. These 7 men and 6 women (mean age 38.6 years) were contrasted with the most recent 225 patients (111 men, 114 women, mean age 52.7 years) who underwent primary PTE for whom complete hemodynamic data are available. The preoperative evaluation of all patients was similar. Pulmonary hemodynamic data and outcome measures were compared between groups. Results. Of 13 reoperated patients: 69% (9/13) had their primary operation at another institution, 54% (7/13) initially underwent unilateral PTE, 38% (5/13) had identifiable coagulation disorders, 38% (5/13) had ineffective caval filtration, 31% (4/13) had suboptimal anticoagulation management, and 31% (4/13) had complete unilateral
pulmonary artery obstruction. The mean interval to reoperation was 5.2 years (range 0.7 to 10.9 years). All control patients underwent bilateral PTE using hypothermic circulatory arrest. Operative mortality was 7.7% (1/13) with reoperation vs 8.4% (19/225) in controls. No difference (p ⴝ NS) was observed between groups in the preoperative pulmonary artery pressure (PAP) or pulmonary vascular resistance; however, the control group had a significantly (p < 0.05) greater reduction in the postoperative PAP (46/19, mean 28 mm Hg vs 59/23, mean 35 mm Hg) and PVR (271 ⴞ 172 vs 399 ⴞ 154 dynes/s/ cmⴚ5) compared with the redo group. No substantial difference in morbidity or functional outcomes was observed between groups. Conclusions. Reoperative PTE can be performed with a perioperative risk comparable with primary PTE, although the improvement in pulmonary hemodynamics is not as favorable. Bilateral primary operation, effective caval filtration, and vigilant anticoagulant management would prevent the need for most reoperative PTEs. (Ann Thorac Surg 1999;68:1770– 7) © 1999 by The Society of Thoracic Surgeons
C
been substantial, with a mortality of 6.4% in the last 500 cases performed, and with 95% of surviving patients in New York Heart Association (NYHA) class I or II at 1 year postoperatively [10]. The natural history of patients with chronic thromboembolic pulmonary hypertension after PTE remains unclear, although our experience suggests that relatively few patients develop recurrent disease. Despite 5%–10% of patients having a defined coagulation abnormality, such as lupus anticoagulant, protein C deficiency, antithrombin III deficiency, or heparininduced platelet antibody, the vast majority of cases of thromboembolic pulmonary hypertension are due to “spontaneous” thromboembolism. Additionally, although we have previously documented a greater risk for recurrent thromboembolic pulmonary hypertension with unilateral PTE and unilateral PA obstruction [11], the risk factors for recurrent disease and the selection criteria for reoperation remain poorly defined. Likewise, analogous to other reoperative cardiac operation procedures [12, 13], it might be anticipated that a second PTE entails greater risk than the initial operation, although the extent of that risk remains unclear. Similarly, it can not be presumed that normal pulmonary blood flow and resis-
hronic thromboembolic pulmonary hypertension is an insidious and infrequently diagnosed disease. The prognosis for patients with untreated pulmonary hypertension is poor, with a 5-year mortality of 90% if the mean pulmonary artery pressures (PAP) are higher than 50 mm Hg [1, 2]. Medical therapy for thromboembolic pulmonary hypertension, using anticoagulants, vasodilators, or thrombolytic agents, is seldom effective [3, 4]. Surgical intervention offers the only consistent and definitive treatment for this disease [5, 6]. Our institution has been engaged in a program for the surgical management of chronic thromboembolic pulmonary hypertension since 1970. In that interval, over 1,100 pulmonary thromboendarterectomies (PTE) have been performed, although the majority (n ⫽ 910) have been performed by two surgeons (S.W.J. and D.P.K.) at our facility within the past decade [5–9]. The success with surgical intervention in this group of difficult patients has Presented at the Thirty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 25–27, 1999. Address reprint requests to Dr Jamieson, Division of Cardiothoracic Surgery (8892), University of California, San Diego, 200 West Arbor Dr, San Diego, CA 92103-8892; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
0003-4975/99/$20.00 PII S0003-4975(99)01043-7
Ann Thorac Surg 1999;68:1770 –7
MO ET AL REOPERATIVE PTE
1771
tance, ie, the functional outcome, would be restored in all instances, although this is currently unsubstantiated. The purpose of this study was to analyze our experience with patients undergoing repeat PTE to determine the specific risk factors for the development of recurrent disease, the selection criteria for reoperation, the relative risk for reoperation, and the functional outcome after reoperative PTE.
Patients and Methods Between November 1989 and December 1998, 870 consecutive PTEs for chronic thromboembolic pulmonary hypertension were performed by two surgeons at the University of California, San Diego Medical Center. During this period, 13 patients (1.5%) underwent repeat PTE. Reoperated patients were contrasted with a consecutive cohort of 225 control patients who underwent primary PTE between March 1996 and December 1997, and for whom complete hemodynamic data were available. Our routine preoperative evaluation for all patients with suspected chronic thromboembolic pulmonary hypertension has been reported previously [5–10]. In brief, this assessment entails a comprehensive medical history and physical examination; chest radiograph and electrocardiogram, pulmonary function studies, and arterial blood gas analysis; transthoracic echocardiography; radionuclide perfusion and ventilation scans; screening for recognized coagulation abnormalities; and right heart catheterization and pulmonary angiography [5, 6]. Although magnetic resonance imaging, computed tomography, and bronchial angiography can contribute important diagnostic information, at present, they lack sufficient resolution to detect disease originating at the origin of the segmental pulmonary artery branches. Accordingly, these modalities are not routinely utilized in our institution to determine operative candidacy. Fiberoptic pulmonary angioscopy is reserved for those patients (approximately 20%) with equivocal angiographic features of chronic thromboembolic disease [14]. In the absence of a precedent concerning repeat PTE, we reasoned that if the current angiogram demonstrated limited disease proximal to the origin of the segmental pulmonary arteries, deteriorating pulmonary hemodynamics could only be attributed to the progression of distal small vessel disease not accessible at reoperation. Accordingly, we believed that the most useful assessment when considering repeat PTE was a qualitative comparison of the current pulmonary angiogram with that preceding the primary operation (Fig 1). Patients excluded following this evaluation were considered for pulmonary transplantation. Our technique for primary PTE, which utilizes profound hypothermia and circulatory arrest, has been reported previously [9]. Although no major modifications were necessary for reoperation, particular care was taken during repeat sternotomy to avoid injury to the dilated and often adherent right atrium or ventricle. Once adequate exposure of the right atrium and ascending aorta was achieved, additional dissection was readily per-
Fig 1. A qualitative comparison of the pulmonary angiogram preceding the primary operation (A) with the current angiogram (B) preceding the reoperation.
formed during the induction of hypothermia while on cardiopulmonary bypass. Identification and distal dissection in the appropriate thromboendarterectomy plane was generally more demanding at reoperation due to the intense and often inflammatory adherence of recurrent thromboemboli to the previously operated pulmonary arteries. With care, this dissection was effectively extended at least two branch divisions distal to the origin of the segmental arteries (Fig 2). Hemostasis was readily accomplished in all cases, without the routine use of platelets, fresh frozen plasma, epsilon aminocaproic acid, or aprotinin. Intravenous heparin anticoagulation was initiated once operative bleeding had ceased, followed by warfarin, titrated to prolong the international normalized ratio (INR) beyond 3.0. Final postoperative hemodynamic determinations were obtained in the absence of inotropic support or intravenous vasodilator agents. Patient follow-up was performed at the outpatient clinic or by telephone contact with the patient and their primary physician. The mean ⫾ standard deviation are tabulated for all continuous variables. The unpaired Student’s t test,
1772
MO ET AL REOPERATIVE PTE
Ann Thorac Surg 1999;68:1770 –7
Table 1. Reoperative Pulmonary Thromboendarterectomy Patient Profile Patient/ Gender
Age (years)
Previous PTE
Coagulation Disorder
1M 2F 3M 4F 5M 6F 7F 8F
64 34 32 20 30 38 36 25
L thor R thor L thor R stern R stern L stern L stern Bilateral
9F 10 M 11 M 12 M 13 M
35 24 64 41 59
Bilateral Bilateral Lupus anticoagulant Bilateral ⫻ 2 Bilateral Bilateral
Anticoagulation
Caval Filtration
None None Poor D/C 2 years None
Anticardiolipin Ab HIPA Anticardiolipin Ab HIPA Congenital dysfibrinogenemia Poor Poor D/C 1 year HIPA lupus anticoagulant
13 total 39 ⫾ 15 7 unilateral 5/13 (38%) 7 male 6 bilateral 6 female
PA Obstruction
Yes Yes
Occluded
Yes
Tilted Yes Poor 4/13 (31%)
5/13 (38%)
4/13 (31%)
Preop PAP
Postop PAP
72/30 (48) 51/17 (28) 85/27 (46) 32/15 (21) 95/40 (60) 52/15 (27) 90/42 (60) 64/26 (39) 85/33 (50) 43/21 (28) 98/30 (53) 100/37 (58) 60/28 (40) 59/23 (35) 75/35 (50) 84/38 (52)
92/25 (52) 70/30 (43) 100/45 (60) 70/28 (41) 70/20 (35)
51/24 (33) 43/15 (27) 55/20 (32) 58/18 (31) 70/29 (43)
82/31 (49) 59/23 (35)
paired Student’s t test, and 2 test were utilized for statistical analysis. A p value less than 0.05 was considered significant. This study was performed under the auspices of the Institutional Review Board of the University of California, San Diego.
Results Patient Demographics
Fig 2. Gross specimens removed from the pulmonary arteries of a patient after the primary operation (A) and after the repeat operation (B). The “tails” represent thrombus removed from the segmental pulmonary arteries.
Preoperative and postoperative data for redo and primary patients are displayed in Tables 1 and 2. Seven reoperated patients were men and 6 were women. This distribution did not differ from the control group (111 men and 114 women). Reoperated patients were younger (39 ⫾ 15 years, range 20 to 69 years) than the reference population (53 ⫾ 15 years, range 16 to 85 years) ( p ⬍ 0.01). Nine of the 13 (69%) redo patients received their initial PTE operation elsewhere. This operation was unilateral in 7 of those 9 patients, of whom 3 underwent the first PTE by thoracotomy. The remaining 5 patients, as well as the 4 who had received their primary operation at the UCSD Medical Center, previously underwent bilateral PTE utilizing our current technique [15]. One of these 6 patients underwent two reoperations. None of the reoperated patients experienced a documented deep venous thrombosis or clinically recognized pulmonary embolism after their initial operation, although patient 3, with tricuspid endocarditis, suffered a paradoxical cerebral embolism. Despite documented chronic thromboembolic pulmonary disease, 5 of 13 (38%) redo patients had ineffective caval filtration. Caval filtration was never performed in patients 1, 2, or 3 before the recurrence of symptoms, while in patient 11, the caval
Ann Thorac Surg 1999;68:1770 –7
1773
MO ET AL REOPERATIVE PTE
Table 1. Continued Preop PVR
Postop PVR
1013 1013 887 1124 557 906 650 1152
357 322 190 610 313 669 319 621
1539 883 634 489 354
480 295 229 330 456
Perfusion Scan of Occluded PA
Morbidity SVT SVT Reperfusion injury
No flow Min flow
Reperfusion injury, trach med bldng Reperfusion injury, trach Reperfusion, ECCO2R Med bldng, pneumonia
Good flow
Sinus node dysfunction pacemaker Reperfusion injury
No flow Reperfusion injury SVT
862 ⫾ 332 399 ⫾ 154 Reperfusion 1/4 (25%) Reperfusion injury 6/13 dysrhythmia 4/13 med bldng 2/13
Follow-up (years)
Current Status
Cause of Death
? 4 5.6 7.5 1.3 6.7 4.5 2.2
Dead Alive Alive Alive Alive Alive Dead Alive
?
3 3.1 1.6 1.3 6 days
Alive Alive Alive Alive Dead
3.4 years
R heart failure
Asystole
Hospital mortality 1/13 (7.7%)
ECCO2R ⫽ extracorporeal carbon dioxide removal; HIPA ⫽ heparin-induced platelet antibody; med bldng ⫽ mediastinal bleeding; PA ⫽ pulmonary artery; PAP ⫽ pulmonary artery pressure (in mm Hg); PTE ⫽ pulmonary thromboendarterectomy; PVR ⫽ pulmonary vascular ⫺5 resistance in dynes/s/cm ; SVT ⫽ supraventricular tachycardia.
filter was malpositioned, and in patient 9, occluded by thrombus. Four of 13 (31%) redo patients had suboptimal anticoagulation management after the primary PTE, at variance with our established policy of lifelong anticoagulation of patients with this disease. Warfarin was discontinued within 2 years of operation in 2 patients (nos. 3 and 6), while in 2 additional patients (nos. 5 and 13), therapeutic prolongation of the INR was not adequately sustained. Five of 13 patients (38%) in the reoperated cohort had a defined abnormality of coagulation. While this incidence is higher than in the reference group (22%), this difference was not statistically significant, nor was there a predominant coagulation defect among reoperated patients. Notably, 4 patients (31%) in the reoperated series had complete unilateral occlusion of the pulmonary arteries, compared with only 16 of 225 patients (7%) in the control group, which represented a significantly higher incidence ( p ⬍ 0.01). Postoperative radionuclide perfusion scans demonstrated variable improvement in the 12 surviving reoperated patients. No flow was reestablished in patients 6 or 12, minimal flow in patient 7, whereas only patient 9 demonstrated a significant improvement in flow to the unilaterally occluded lung after reoperation. The low incidence (25%) of reperfusion in these 4 patients is comparable with our previously reported experience of absent reflow (8 of 16, 50%) among primarily operated patients presenting with complete unilateral pulmonary artery obstruction [11].
Survival The 7.7% (1 of 13) operative mortality among redo patients did not differ ( p ⬎ 0.05) from the 8.4% (19 of 225)
mortality among primary patients in this series, although the previously reported operative mortality in a larger cohort of 500 primary PTE patients was 6.4% (10). The single death among reoperated patients (patient 13) occurred on postoperative day 6, when hypotension from refractory supraventricular tachycardia (SVT) precipitated cardiac arrest. The 1- and 3-year actuarial survival, at a mean follow-up of 3.4 years after reoperation, was 92% and 82%, respectively. One patient (patient 1) died 1.3 years after reoperation of progressive respiratory failure due to preexistent idiopathic pulmonary fibrosis. A second patient (patient 7) with lupus anticoagulant antibody, complete unilateral pulmonary artery obstruction, and minimal reperfusion after reoperation died of progressive right heart failure 4.5 years after repeat PTE. The 1- and 3-year actuarial survival among primarily operated patients was 90% and 81%, respectively.
Pulmonary Hemodynamics The preoperative pulmonary artery pressures (pulmonary artery systolic/diastolic; mean PAP in mm Hg) were equivalent (p ⫽ NS) between the redo patients (82 ⫾ 13/31 ⫾ 6; 49 ⫾ 8) and the reference cohort (80 ⫾ 22/31 ⫾ 10; 48 ⫾ 14). Likewise, the preoperative pulmonary vascular resistance (PVR) (dynes/s/cm⫺5) was similar (p ⫽ NS) between the reoperative PTE patients (862 ⫾ 332) and the primary PTE patients (842 ⫾ 431) (Figs 3 and 4; Table 2). In each patient, pulmonary angiography depicted significant obstruction of the central arteries (Fig 1). When measured in the intensive care unit after withdrawal of vasoactive and inotropic medications, the reoperated patients demonstrated a significant reduction in PAP (82 ⫾ 13/31 ⫾ 6; 49 ⫾ 8 preoperative vs 59 ⫾ 18/23 ⫾
1774
MO ET AL REOPERATIVE PTE
Ann Thorac Surg 1999;68:1770 –7
Table 2. Redo Versus Primary PTE Patient Outcomes Redo (n ⫽ 13) Mortality 1/13 (7.7%) Morbidity Reperfusion injury 6/13 (46%) Atrial arrhythmia 4/13 (31%) Mediastinal bleeding 2/13 (15%) Ventilator days 9.8 ⫾ 12.3 ICU length of stay 11 ⫾ 12.5 Pulmonary hemodynamics Preoperative PAP 82/31 (49) Postoperative PAP 59/23 (35) Preoperative PVR 862 ⫾ 332 Postoperative PVR 399 ⫾ 154
Control (n ⫽ 225)
p Value
20/225 (8.9%)
NS
75/25 (33%) 29/225 (13%) 5/225 (2%) 5.1 ⫾ 12.1 6.6 ⫾ 12.7
NS ⬍ 0.05 ⬍ 0.01 NS NS
80/31 (48) 46/19 (28) 842 ⫾ 431 271 ⫾ 172
NS ⬍ 0.01 NS ⬍ 0.01
ICU ⫽ intensive care unit; NS ⫽ not significant; PAP ⫽ pulmonary artery pressure (in mm Hg); PVR ⫽ pulmonary vascular resistance (in dynes/s/cm⫺5).
8; 35 ⫾ 11 postoperative) and PVR (862 ⫾ 332 preoperative vs 399 ⫾ 154 postoperative) ( p ⬍ 0.01 for all comparisons) (Figs 3 and 4). Primarily operated patients likewise demonstrated a significant reduction in PAP (80 ⫾ 22/31 ⫾ 10; 48 ⫾ 14 preoperative vs 46 ⫾ 16/19 ⫾ 7; 28 ⫾ 9 postoperative) and PVR (842 ⫾ 431 preoperative vs 271 ⫾ 172 postoperative) ( p ⬍ 0.01 for all comparisons) (Figs 3 and 4). The relative reduction in postoperative PAP between redo patients (59 ⫾ 18/23 ⫾ 8; 35 ⫾ 11) and primary PTE patients (46 ⫾ 16/19 ⫾ 7; 28 ⫾ 9) was greater in the reference cohort ( p ⬍ 0.01). Similarly, the reduction in PVR observed in the primary PTE patients (271 ⫾ 172) was greater ( p ⬍ 0.01) than that observed in the redo
Fig 3. The preoperative and postoperative mean PAPs in the redo (49 vs 35 mm Hg) and primary (48 vs 28 mm Hg) PTE cohorts. As demonstrated, there was a greater reduction in the mean PAP among patients undergoing a primary PTE (p ⬍ 0.01).
Fig 4. The preoperative and postoperative PVR in the redo (862 vs 399 dynes/s/cm⫺5) and primary (842 vs 271 dynes/s/cm⫺5) PTE cohorts. As indicated, there was a greater reduction in PVR among patients undergoing primary PTE (p ⬍ 0.01).
PTE patients (399 ⫾ 154) postoperatively (Figs 3 and 4; Table 2).
Functional Outcome Eleven of 13 redo patients experienced a functional benefit after their initial PTE, but developed recurrent symptoms of pulmonary hypertension from 6 months to 10 years after the primary operation (mean 2.7 ⫾ 2.8 years). The mean interval to reoperative PTE was 5.2 years ⫾ 2.7 years (range 8 months to 10.9 years). Immediately before the second PTE, the majority of patients were NYHA class III (11 of 13, 85%) or class IV (1 of 13, 8%). The exception was the patient with paradoxical embolism, who was in NYHA class II (Fig 5).
Fig 5. The preoperative and postoperative functional NYHA class for the 13 reoperative patients.
Ann Thorac Surg 1999;68:1770 –7
The majority of surviving patients (10 of 12, 83%) reported symptomatic improvement after repeat PTE. Postoperatively, 5 patients were in NYHA class I, 5 in class II, and 2 in class III (Fig 5). Of the 2 patients remaining in class III, patient 2 reported no functional benefit despite a 70% reduction in the preoperative PVR from 1,013 to 322 dynes/s/cm⫺5, whereas patient 12 had unilateral PA obstruction with no reflow demonstrated by perfusion scan. Notably, patient 9 has deteriorated from NYHA class I to class IV 2 years after reoperation.
Surgical Management Indicative of the greater difficulties encountered during repeat PTE, the cumulative duration of circulatory arrest (45 ⫾ 19 vs 35 ⫾ 14 minutes, p ⬍ 0.05), and the aortic cross-clamp time (127 ⫾ 14 vs 97 ⫾ 41 minutes, p ⬍ 0.01) were prolonged in the reoperated patients. The duration of cardiopulmonary bypass was also prolonged (249 ⫾ 51 vs 225 ⫾ 50 minutes), although not significantly ( p ⬎ 0.05). Notwithstanding the increased technical challenge of repeat PTE, substantial chronic thromboembolic material was recovered in each patient (Fig 2). The type of disease encountered in reoperative cases was different than that typically encountered in our primarily operated series. Among redo patients, 5 of 13 (38%) had type 1 disease (partially organized central thrombus proximal to organized distal thrombus), and 8 of 13 patients (62%) had type 2 disease (organized thrombus proximal to the origin of the segmental artery). In contrast, although type 2 disease predominates (70%) among primarily operated patients, only 10% historically demonstrate type 1 disease [15].
Complications Despite the inherent risks of reoperative cardiac procedures, the incidence and nature of complications encountered in this reoperative series was comparable with those experienced after primary operation (Table 2). Two of 13 reoperated patients (15%) required reexploration for perioperative hemorrhage, significantly greater than the 2% (5 of 225) incidence of reexploration among primarily operated patients ( p ⬍ 0.01). Four reoperated patients (31%) experienced significant perioperative rhythm disturbances (sinus node dysfunction in 1, and supraventricular tachycardia in 3 patients), nearly threefold greater than the incidence of arrhythmias experienced in the reference population (29 of 225, 13%; p ⬍ 0.05). Additionally, the incidence of reperfusion injury (6 of 13; 46% vs 75 of 225; 33%), the duration of mechanical ventilation (9.8 ⫾ 12.3 days vs 5.1 ⫾ 12.1 days), and the length of intensive care unit stay (11 ⫾ 12.5 days vs 6.6 ⫾ 12.7 days) was greater in reoperated patients, although none of these differences was statistically significant ( p ⬎ 0.05 for all comparisons). Although 3 reoperated patients required more than 20 days of assisted ventilation, including patient 8, who was supported with extracorporeal carbon dioxide removal (ECCO2R) for 6 days, no patient in the reoperated series succumbed to respiratory failure.
MO ET AL REOPERATIVE PTE
1775
Comment Acute pulmonary thromboembolism is a more common condition than is generally appreciated, and in many cases, is asymptomatic until end-stage debilitating pulmonary hypertension develops. In the majority of patients, spontaneous resolution of acute pulmonary embolism is the rule. However, a small but uncertain percentage develop chronic thromboembolic pulmonary hypertension for which medical therapy is ineffective [3, 4], leaving surgical intervention the only definitive option [7–9, 16, 17]. The team at the University of California, San Diego Medical Center has benefited from an institutional experience of more than 1,100 PTE cases since the inception of this program in 1970 [7–9, 18]. While accumulating this experience, we have had the opportunity to consider reoperation for patients initially managed here as well as at other institutions. The purpose of this study was to determine the risk factors for recurrent thromboembolic disease, the selection criteria for reoperation, the relative risk of repeat PTE, and the functional benefit derived from reoperative PTE. Generally, reoperative cardiac surgery is accompanied by a higher perioperative risk and reduced functional outcome compared with the equivalent primary procedure [12, 13]. Accordingly, an appreciation of the factors, technical and others, that may contribute to an unsatisfactory outcome after the primary operation can only serve to improve the results of primary operation and limit the likelihood that subsequent procedures will be necessary. Additionally, because PTE is almost invariably lethal whenever a significant improvement in pulmonary blood flow is not achieved, the selection criteria for reoperative PTE warrants particular scrutiny. Seven of the 13 patients in this series initially underwent a unilateral primary operation, yet at reoperation, each of these 7 patients had significant bilateral chronic thromboembolic disease, as we found substantial thromboembolic material in the contralateral side of the prior surgery in these redo patients. In the comparison cohort of 225 patients, in whom bilateral pulmonary artery exploration was performed regardless of the preoperative findings, only 6 of 225 patients (2.7%) exhibited true unilateral disease. Accordingly, we believe that unilateral PTE for bilateral disease poses a substantial risk for early recurrence of symptomatic pulmonary hypertension. Less certain were the possible causes of recurrent symptomatic pulmonary hypertension in those 6 patients who initially underwent bilateral PTE using our present technique, which employs profound hypothermia and circulatory arrest. While none of those patients had experienced a clinical episode consistent with recurrent deep venous thrombosis or pulmonary embolism, it must be recalled that thromboembolic events are clinically occult in the majority of patients [19]. Additionally, the possibility of in situ thrombosis of portions of the previously endarterectomized proximal vascular bed cannot be excluded by present means [20]. Each of the patients in the reoperated series, except patient 8, had at least one
1776
MO ET AL REOPERATIVE PTE
recognized risk factor for recurrent disease, eg, coagulation disorder, poor anticoagulation management, ineffective caval filtration, or complete unilateral PA obstruction, that could have predisposed to a thrombotic or embolic event. While it can be argued that differentiation between these two possibilities is of academic interest only, the essential observation remains that scrupulous prophylaxis may have eliminated the need for reoperation in these patients. The validity of our initially empiric selection criteria, based predominantly on the pulmonary angiogram, appears to be verified by the outcomes achieved in this small sample of patients. In each instance, at operation, we were able to identify a suitable plane of dissection and extract significant chronic thromboembolic material from both lungs, even when the prior operation was bilateral (Fig 2). Though the improvement in pulmonary blood flow was not as great as could be accomplished in primarily operated patients (Figs 3 and 4), the majority of reoperated patients experienced some sustained increment in functional class (Fig 5). Notwithstanding the difficulties encountered at reoperation, and despite higher perioperative morbidity, these benefits were attained without incurring a substantially greater mortality than in primarily operated patients. While the benefits identified among those patients with complete unilateral pulmonary artery obstruction suggest that this cohort should undergo reoperation with only limited expectations, we believe that reoperation with this caveat is preferable to either pulmonary transplantation or inexorable right heart failure and premature death.
References 1. Dalen JE, Alpert JS. Natural history of pulmonary embolism. Prog Cardiovasc Dis 1975;17:259–70. 2. Riedel M, Stanek V, Widimsky J, Prerovsky I. Long term follow-up of patients with pulmonary thromboembolism. Late prognosis and evolution of hemodynamic and respiratory data. Chest 1982;81:151– 8. 3. Dash H, Ballentine N, Zelis R. Vasodilators ineffective in secondary pulmonary hypertension. N Eng J Med 1980;303:1062–3. 4. Dantzker DR, Bower JS. Partial reversibility of chronic pulmonary hypertension caused by pulmonary thromboembolic disease. Am Rev Respir Dis 1981;124:129–33. 5. Moser KM, Auger WR, Fedullo PF. Chronic major vessel
Ann Thorac Surg 1999;68:1770 –7
6. 7. 8.
9.
10. 11.
12.
13. 14. 15. 16. 17.
18. 19. 20.
thromboembolic pulmonary hypertension. Circulation 1990; 81:1735– 43. Fedullo PF, Auger WR, Channick RN, Moser KM, Jamieson SW. Chronic thromboembolic pulmonary hypertension. Clin Chest Med 1995;16:353–74. Utley JR, Spragg RG, Long WB, Moser KM. Pulmonary endarterectomy for chronic thromboembolic obstruction: recent surgical experience. Surgery 1982;92:1096 –102. Daily PO, Dembitsky WP, Iversen S, Moser KM, Auger WR. Current early results of pulmonary thromboendarterectomy for chronic pulmonary embolism. Eur J Cardiothorac Surg 1990;4:117–21. Jamieson SW, Auger WR, Fedullo PF, et al. Experience and results with 150 pulmonary thromboendarterectomy operations over a 29 month period. J Thorac Cardiovasc Surg 1993; 106:116–26. Moser KM, Auger WR, Fedullo PF, Jamieson SW. Chronic thromboembolic pulmonary hypertension: clinical picture and surgical treatment. Eur Resp J 1992;5:334– 42. Hirsch AM, Moser KM, Auger WR, Channick RN, Fedullo PF. Unilateral pulmonary artery thrombotic occlusion: is distal arteriopathy a consequence? Am J Respir Crit Care Med 1996;154:491– 6. Salomon NW, Page US, Bigelow JC, Krause AH, Okies JE, Metzdorff MT. Reoperative coronary surgery. Comparative analysis of 6591 patients undergoing primary bypass and 508 patients undergoing reoperative coronary artery bypass. J Thorac Cardiovasc Surg 1990;100:250–9. Pansini S, Ottino G, Forsennati PG, et al. Reoperations on heart valve prostheses: an analysis of operative risks and late results. Ann Thorac Surg 1990;50:590– 6. Shure D, Gregoratos G, Moser KM. Fiberoptic angioscopy: role in the diagnosis of chronic pulmonary arterial obstruction. Ann Intern Med 1985;103:844–50. Jamieson SW. Pulmonary thromboendarterectomy. In: Franco KL, Putnam JB, eds. Advanced therapy in thoracic surgery. Hamilton: B.C. Decker, Inc., 1998:310– 8. Chitwood WR, Sabiston DC, Wechsler AS. Surgical treatment of chronic unresolved pulmonary embolism. Clin Chest Med 1984;5:507–36. Mayer E, Dahm M, Hake U, et al. Mid-term results of pulmonary thromboendarterectomy for chronic thromboembolic pulmonary hypertension. Ann Thorac Surg 1996;61: 1788 –90. Moser KM, Braunwald NS. Successful surgical intervention in severe chronic thromboembolic pulmonary hypertension. Chest 1973;64:29–35. Cranley JJ, Canos AJ, Sull WJ. The diagnosis of deep venous thrombosis. Fallibility of clinical symptoms and signs. Arch Surg 1976;111:34– 6. Moser KM, Fedullo PF, Finkbeiner WE, Golden J. Do patients with primary pulmonary hypertension develop extensive central thrombi? Circulation 1995;91:741–5.
DISCUSSION DR CHRISTOPHER G. A. McGREGOR (Rochester, MN): Dr Anderson, Dr Pairolero, ladies, and gentlemen. It is a great pleasure to discuss this very well-presented paper on redo pulmonary thromboendarterectomy (PTE) from the UC, San Diego group. This presentation represents 13 of 870 patients undergoing PTE at San Diego, a total number that probably represents about 80% of the world experience. Doctor Jamieson and colleagues are to be congratulated on their leadership in this area and their outstanding results in this series. They have an operative mortality of 7.7% for redos and 8.4% for the reference population of first-time cases. The few studies of any numbers in
the literature of PTE performed elsewhere have reported a mortality generally in excess of 20%. In our own initial experience at Mayo, in the first 29 elective PTEs, the mortality has been at 10%, somewhat similar to the San Diego experience. And I would like to thank the San Diego group for the opportunity to visit them and learn from their experience. In today’s report, the redo patients were younger than the reference population, and only 2 of 13 had no identifiable risk factors for recurrent thrombotic problems. These risk factors may be coagulation disorders, suboptimal anticoagulation, an occluded or malpositioned IVC filter, or previous unilateral