- Email: [email protected]
Thrombosis of bileaflet tricuspid valve prosthesis: Clinical spectrum and the role of nonsurgical treatment
Thrombosis of bileaflet tricuspid valve prosthesis: Clinical spectrum and the role of nonsurgical treatment Yaron Shapira, MD, Alex Sagie, MD, Ruth Jortner, MD, Yehuda Adler, MD, and Rafael Hirsch, MD Petah-Tiqva and Tel Aviv, Israel
Background Thrombosis of a mechanical tricuspid valve prosthesis is a potentially hazardous event. This study aimed to explore the incidence and the clinical presentation of tricuspid valve thrombosis occurring in bileaflet valves and to evaluate the diagnostic and the therapeutic approach.
Methods and Results Tricuspid valve thrombosis was sought in 22 late survivors with the CarboMedics valve in a follow-up period of 36.0 ± 20.8 months. Limited leaflet motion and/or a visible thrombus were considered diagnostic of valve thrombosis. Eight episodes of tricuspid valve thrombosis were diagnosed among 5 patients (12.1 episodes per 100 patient-years). Anticoagulation was inadequate in 3 patients and fair in 2. Florid right heart failure occurred in 3 episodes. Common physical findings included increased jugular venous pulse (5 patients), diastolic tricuspid murmur (4 patients), and peripheral edema (4 patients). The diagnosis was suspected in all clinically and by transthoracic echocardiography and confirmed by fluoroscopy and/or transesophageal echocardiography. In 4 patients, both leaflets were involved. No thrombi were visualized. Three patients received thrombolytic therapy in 4 episodes (complete success in 3, partial success in 1) without hemorrhagic or embolic complications. One patient responded to aggressive anticoagulant therapy. One patient required an emergent repeat surgery. In 1 patient, valve thrombosis recurred thrice.
Conclusions In patients with fair or poor anticoagulation, a bileaflet valve in the tricuspid position is associated with a high incidence of valve thrombosis. Hinge entrapment requires only a small amount of thrombotic material. Valve thrombosis may be asymptomatic. Involvement of both leaflets is usually required to produce symptoms. A nonsurgical approach (thrombolysis or intensified anticoagulation) is usually successful. Patients should be instructed about heralding signs of valve thrombosis. (Am Heart J 1999;137:721-5.)
Thrombosis of a mechanical tricuspid valve prosthesis is an important and not infrequently seen complication of such a device. Early experience with monoleaflet valves was associated with near 4% annual incidence of tricuspid valve thrombosis,1 but recent experience with bileaflet valves is more favorable, with an annual incidence approaching 1%.2-4 Bileaflet mechanical valves show promise in the reduction of valve thrombosis because of their better hemodynamic profile as compared with older valve models. Because valve implantation in the tricuspid position makes up <2% of all mechanical valve implantations,5-7 experience with tricuspid valve thrombosis is limited. In the current study, we present our experi-
From the Sheingarten Echocardiography Unit, the Department of Cardiology, and the Grown-Up Congenital Heart Disease Unit, Rabin Medical Center, Beilinson Campus; and Sackler Faculty of Medicine, Tel-Aviv University. Submitted February 23, 1998; accepted July 2, 1998. Reprint requests: Yaron Shapira, MD, Department of Cardiology, Rabin Medical Center, Beilinson Campus, Petah Tiqva, Israel, 49100. Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 + 0 4/1/93408
ence with tricuspid valve thrombosis occurring in CarboMedics valves over a 6-year period, with emphasis on the clinical presentation, the diagnostic approach, and the therapeutic approach.
Methods During a 6-year period (January 1991 to December 1996), 30 adult patients underwent tricuspid valve replacement with mechanical bileaflet prostheses in our institution. Eight patients died (7 within 3 months).Twenty-two late survivors (19 with the CarboMedics model, 2 with the Sorin Bicabon model, and 1 with the St Jude Medical model), were monitored for a mean period of 36.0 ± 20.8 months (range 4 to 67 months).The follow-up included history and physical examination on a regular basis (at least annually) and an echocardiogram that preceded the physical examination by ≤4 weeks. Patients were examined by experienced cardiologists. Patient history included questions oriented at possible valve malfunction, for example, muffled valve clicks, fatigue, fluid retention, distended neck veins, or syncope.The anticoagulation status was recorded.Valve clicks (intensity and quality), right-sided murmurs (inspiratory augmentation), neck veins, and signs of fluid retention were specifically sought in physical examination. Patients in whom stuck valve was suspected on clinical
722 Shapira et al
grounds underwent transthoracic echocardiography (TTE) within the same day. All patients’ files and the computerized database of the valvular clinic, the echocardiography laboratory, the congenital heart disease unit, and the cardiothoracic surgery department were screened for evidence of thrombosed tricuspid valve up to September 1997.Valve thrombosis was inferred from at least 1 of the following: (1) restricted leaflet motion, as seen by TTE, transesophageal echocardiography (TEE), or fluoroscopy; (2) a thrombus attached to the valve (seen by TEE); and (3) surgical evidence of valve thrombosis.
Echocardiographic evaluation All patients were evaluated by an experienced echocardiographer, with a Hewlett-Packard Sonos 1000 or Sonos 2000 echocardiography machine, with a 2.5- or 3.5-MHz probe.TTE was performed in all patients, with all standard views.TEE (biplane or multiplane) was performed on an individual basis, according to the physician in charge of the patient. Patients were considered for TEE evaluation if TTE or fluoroscopy showed thrombus or impaired leaflet motion suggestive of a thrombosed valve. The following were considered suggestive of valve thrombosis in TTE (≥1 of the listed criteria): (a) high transvalvular gradients: mean ≥6 mm Hg, peak ≥15 mm Hg; (b) transvalvular gradients ≥50% higher than observed before; (c) visible thrombus attached to the valve; and (d) a wide intravalvular jet of tricuspid regurgitation. Inability to demonstrate 2 different mobile echoes (representing the valve leaflets) in a high-quality image was considered supportive of valve thrombosis in the appropriate setting but could not serve as a single criterion. The diagnostic criterion for valve thrombosis in TEE was limited leaflet motion and/or a visible thrombus.
Fluoroscopy A fluoroscopic study was performed in all the episodes of suspected stuck valve (excluding during pregnancy). Multiple acquisition angles were used to determine leaflet mobility, including the side (pivot) view, with the disks parallel to the xray beams. Inability to fully open or close was considered indicative of stuck valve.
Hemodynamic evaluation The inflow velocities were recorded by continuous wave Doppler from the apical 4-chamber view. Peak and mean transvalvular gradients were calculated from measured velocities with the modified Bernoulli equation, averaging 5 beats in sinus rhythm and 7 in atrial fibrillation. The presence of prosthetic valve regurgitation was assessed by color flow mapping.The severity of tricuspid regurgitation was evaluated in multiple views and graded from 0 to 4+ (severe) according to jet penetration into the right atrium.
Results Among the 22 late survivors, we diagnosed 8 episodes of prosthetic tricuspid valve thrombosis in 5 patients with the CarboMedics model (1 patient with 3 episodes, 1 patient with 2 episodes, and 3
American Heart Journal April 1999
patients with a single episode each).The linearized ratio of valve thrombosis was 12.1 episodes per 100 patient-years. Four patients were treated medically (intensified anticoagulants or thrombolysis).An additional patient had both mitral and tricuspid valve thrombosis 4 months after surgery and had anasarca, cardiac tamponade, and low cardiac output state, which necessitated an emergency successful repeat operation. There were 4 women and 1 man.The mean age was 35.4 ± 7.4 years (range 25 to 45 years). Four patients had rheumatic heart disease and had mechanical mitral valve prosthesis in addition. One patient had Ebstein’s anomaly. One patient was asymptomatic (only indicating muffled valve sounds).Three patients had mild to moderate symptoms, including fatigue and signs of fluid retention. One patient (with stuck mitral valve in addition) was in class IV heart failure.The symptom duration ranged from 1 week to 6 months and was shorter as the functional class was worse. Only the patient with a single prosthetic heart valve indicated muffled valve clicks. In 2 patients, the admission international normalized ratio (INR) was 2.8. In the other 3, the admission INR was far below 2.0. The physical examination revealed prominent jugular venous pulse waves in all.A diastolic murmur of presumed tricuspid origin was heard in 4 and a systolic murmur in 1. Muffled valve clicks were noted in only 1 patient (who did not have mechanical mitral valve in addition). Peripheral edema was noted in 4 patients, of whom 1 had anasarca. One patient was in sinus rhythm, 2 had a paced rhythm, and 2 were in atrial fibrillation. TTE revealed peak diastolic tricuspid gradients of 19 to 24 mm Hg in 4 patients, whereas the fifth had a peak gradient reaching 47 mm Hg.The mean diastolic tricuspid gradients were 9 to 16 mm Hg in 4 patients and 31 mm Hg in the fifth. In this last patient, a forward flow through the tricuspid valve was noted even during systole. Severe tricuspid regurgitation was observed in 3 cases, whereas in 2 it was absent or mild. It was impossible to delineate both leaflets. No thrombi were visualized. All patients underwent fluoroscopy (excluding 1 episode during pregnancy).The asymptomatic patient had only 1 leaflet involved (it was stuck in a closed position). In the other patients both leaflets were involved, usually in a semiopen position. TEE was performed in all cases but 2:The patient with class IV heart failure and additional stuck mitral valve was immediately sent to a repeat surgery after TTE and fluoroscopy.The pregnant patient was followed by TTE.TEE confirmed the diagnosis of stuck valve in all. No thrombi were visualized in TEE as well. One patient responded to enhanced oral anticoagu-
American Heart Journal Volume 137, Number 4, Part 1
Shapira et al 723
Table I. Experience with mechanical prosthetic valves in tricuspid position Valve model* Reference 1 2 3 4 17
No. of patients
Cage ball and other old models
67 78 64 39 25
27 43 (4) 17 (3) — —
Single leaflet 40 23 (3) 18 (4) — most (1)
Follow-up Double leaflet (patient-years) — 12 (1) 29 (2) 39 (1) ≤4
399 NA 681 149 NA
Tricuspid valve thrombosis (% per patient-year) 9 (2.26%) 8 9 (1.32%) 1 (0.67%) 1 (0.37%)
NA, Not available. *Valve thrombosis episodes in parentheses.
lation but had a second thrombotic during early gestation, responding to intensified intravenous anticoagulation after 3 weeks. Eventually, she had a third episode (2 months after the second), requiring valve replacement with bioprosthesis.A third patient was given a 2-day intravenous heparin course, with a partial response, complemented by thrombolysis.All the other episodes were treated by thrombolysis (streptokinase, urokinase or tissue plasminogen activator). Complete success was achieved in all except for 1 patient with partial success who was discharged with intensified oral anticoagulation. One month later, fluoroscopy revealed complete resolution of the leaflet motion.
Discussion There is an ongoing debate in the medical literature regarding replacement of the tricuspid valve with a mechanical prosthesis, even in the era of new low-profile models.The low velocity of blood across the valve makes it prone to thrombosis. Bileaflet models have the advantage of independent function of both leaflets, reducing the risk of complete obstruction. However, the delicate mechanism is more susceptible to functional deterioration, even with a relatively small clot. Our report is confined to the bileaflet CarboMedics valve in the tricuspid position, which is the valve most commonly used in our institution and for which there is little information in the literature.The following discussion corresponds to data from the current series and from fully detailed reports.8-16 Contrary to the florid appearance of left-sided heart failure in patients with stuck left-sided valves, signs of right heart failure associated with stuck tricuspid valve may be quite subtle.The disappearance of formerly audible prosthetic valve clicks in a patient with an isolated tricuspid valve prosthesis may be the first clue leading to the diagnosis of stuck valve. Indeed, in all the cases in which the tricuspid valve was the only prosthesis implanted, the valve clicks were either muffled or absent and self-appreciated by most of the adult patients. On the other
hand, none of the 5 patients with an additional mechanical mitral valve prosthesis noted any change in valve sounds, nor was it noted by the physicians examining the patients. It is prudent to instruct patients to pay attention to the quality of their valve clicks and report any change. Patients should also be instructed to pay attention to warning signs such as distended neck veins or evidence of fluid retention (pedal edema, increased abdominal girth, unexplained weight gain). In all the cases, excluding 1, both valve leaflets were involved. It may be that involvement of both leaflets is needed to cause symptomatic physiologic derangement.Additionally, the close proximity of the hinge points makes it easier for a clot to involve both leaflets. Bedside physical examination can be diagnostic or at least raise clues to the diagnosis of stuck valve. Useful findings include muffled or absent valve clicks, alterations in jugular venous pulse (tall systolic waves, slow Y descent), and diastolic murmur augmented during the inspiratory phase.8-16 Fluid retention is common. Nevertheless, auscultatory findings may be hampered by an associated prosthetic mitral valve, and fluid retention may be associated with various other mechanisms, including predominant leftsided heart failure. TTE was suggestive of valve thrombosis in all cases. The mean transvalvular gradients in the current study and in the literature, derived from Doppler velocities, was at least 8 mm Hg. In each of our cases, the 2dimensional echocardiography showed indirect evidence of abnormal leaflet motion, although the leaflets themselves could not be delineated. Fluoroscopy is the most widely used method for making the diagnosis of stuck valve.TEE, on the other hand, was rarely used for this purpose.15 We used TEE more commonly because most patients had mitral prosthesis in addition, and precise information on that valve was needed.A visible thrombus was found in only 3 cases in the literature and in none in our study.The absence of a visible thrombus may be explained by
American Heart Journal April 1999
724 Shapira et al
Table II. Experience with CarboMedics valves in tricuspid position
Reference 5 6 7
Total valves implanted
Valves in tricuspid position
569 612 277
2 (0.4%) 9 (1.5%) 2 (0.7%)
Tricuspid valve thrombosis 2 1 NA
NA, Not available.
the small amount of thrombotic material needed to catch the hinges of bileaflet valves. In terms of prosthetic valve thrombosis, mechanical heart prostheses in the tricuspid position are not viewed favorably. Many authors cite the 20% incidence of valve thrombosis observed by Thorburn et al1 in monoleaflet models (mainly Bjork-Shiley), corresponding to an incidence of 4% per patient-year. Later reports, corresponding to a mixture of valve models, show some more favorable results (≤1% per patientyear), as shown in Table I.1-4,17 Despite a large core of data on the CarboMedics heart prostheses in left-sided positions, experience with this valve model in the tricuspid position is limited, as shown in Table II.Among 1458 patients with the CarboMedics valve reported in 3 large series, only 13 (0.9%) were implanted in the tricuspid position.5-7 Three patients had documented episodes of valve thrombosis: 2 responded to thrombolytic treatment (streptokinase)5 and 1 was referred to surgery.6 Three out of 4 pediatric patients had prosthetic valve thrombosis 1 to 6 months after surgery, which responded promptly to 12- to 96-hour streptokinase courses.18 Thrombolysis is a valid alternative to surgery in these patients. It is highly effective, and side effects are usually tolerable (minor bleeding or an allergic reaction). Pulmonary embolism was reported only once,10 and this may be attributable to the small amount of thrombotic material needed to disturb leaflet mobility. Long-standing symptoms (>1 month) should not make one reluctant to use thrombolysis in the stuck tricuspid valve, as shown in 8 successfully treated patients (including 1 in our series). Intensifying anticoagulation alone may be effective in selected cases, such as mildly symptomatic patients, as shown twice in 1 patient in our series. Such a treatment may also be beneficial in patients at high risk who partially respond to thrombolysis or when thrombolysis is contraindicated. Reddy et al19 achieved delayed success by such a strategy in 9 of 21 patients with left-sided prostheses in whom the initial result of thrombolysis was only partial. The high incidence of valve thrombosis in the tricus-
pid position may be partially attributed to poor anticoagulant therapy. However, in 2 of our patients, admission INR values were in the therapeutic range (INR = 2.8), considering current general recommendations on anticoagulant therapy.20-23 The INR levels required for prosthetic tricuspid valves may be higher than those needed for mitral valves, especially in a prothrombotic milieu (atrial fibrillation, coexisting prosthetic valve, and so forth).We advocate target INR levels of 3.5 to 4.0 for patients with tricuspid prostheses.The addition of aspirin may further contribute to the prevention of valve thrombosis.24
Conclusions (1) In patients with fair or poor anticoagulation, the CarboMedics bileaflet valve in the tricuspid position is associated with a high incidence of valve thrombosis. (2) The clinical presentation of tricuspid valve thrombosis is frequently mild and even symptomless. (3) The amount of thrombotic material needed to cause restriction of the tricuspid valve leaflets is small and therefore a thrombus is rarely visualized, even in TEE. (4) Involvement of both leaflets is usually required to produce symptoms. (5) Tricuspid valve thrombosis may occasionally respond to intensified anticoagulant therapy. (6) Thrombolysis is almost always successful and is the first line of treatment in these patients, even when thrombosis is not recent. Surgery should be reserved for critically ill patients. (7) Asymptomatic patients with a partially stuck leaflet, who do not respond to thrombolysis, may be discharged from the hospital with long-term intensified anticoagulant therapy and close follow-up.The leaflet motion can be fully restored later.
References 1. Thorburn CW, Morgan JJ, Shanahan MX, et al. Long-term results of tricuspid valve replacement and the problem of prosthetic valve thrombosis. Am J Cardiol 1983;51:1128-32. 2. Van Nooten GJ, Caes F, Taeymans Y, et al. Tricuspid valve replacement: postoperative and long-term results. J Thorac Cardiovasc Surg 1995;110:672-9. 3. Horstkotte D, Burckardt D. Prosthetic valve thrombosis. J Heart Valve Dis 1995;4:141-53. 4. Nakano K, Koyanagi H, Hashimoto A, et al. Tricuspid valve replacement with the St. Jude Medical valve prosthesis. J Thorac Cardiovasc Surg 1994;108:888-92. 5. Fiane AE, Saatvedt K, Svennevig JL, et al. The CarboMedics valve: midterm follow-up with analysis of risk factors. Ann Thorac Surg 1995;60:1053-8. 6. Nistal JF, Hurle A, Revuelta JM, et al. Clinical experience with the CarboMedics valve: early results with a new bileaflet mechanical prosthesis. J Thorac Cardiovasc Surg 1996;112:59-68. 7. Aagaard J, Hansen CN, Tingleff J, et al. Seven-and-a-half years clin-
American Heart Journal Volume 137, Number 4, Part 1
8.
9.
10.
11.
12.
13.
14.
15.
ical experience with the CarboMedics prosthetic heart valve. J Heart Valve Dis 1995;4:628-33. Joyce LD, Boucek M, McGough EC. Urokinase therapy for thrombosis of tricuspid prosthetic valve. J Thorac Cardiovasc Surg 1983; 85:935-7. Kurzrok S, Singh AK, Most AS, et al. Thrombolytic therapy for prosthetic cardiac valve thrombosis. J Am Coll Cardiol 1987; 9:592-8. Mehan VK, Dalvi BY, Kale PA. Thrombosed prosthetic valve in tricuspid position: successful therapy with intravenous streptokinase. Chest 1992;102:1599-600. Tischler MC, Lee RT, Kirshenbaum JM. Successful treatment of prosthetic tricuspid valve thrombosis with short-course recombinant tissue-type plasminogen activator. Am Heart J 1990; 120:975-8. Cohen ML, Barzilai B, Gutierrez F, et al. Treatment of tricuspid valve thrombosis with low-dose tissue plasminogen activator. Am Heart J 1990;120:978-80. Prieto Palomino MA, Ruiz de Elvira MJ, Sanchea Llorente F, et al. Successful thrombolysis on a mechanical tricuspid prosthesis. Eur Heart J 1989;10:1115-7. Fyfe DA, Taylor AB, Kline CH, et al. Doppler echocardiographic evaluation of streptokinase lysis of thrombosed right-sided St. Jude Medical valves in patients with congenital heart defects. Am Heart J 1991;121:1156-60. Hurrell DG, Hartzell VS, Tajik JA. Thrombolytic therapy for
Shapira et al 725
16. 17.
18.
19.
20.
21. 22.
23. 24.
obstruction of mechanical prosthetic valves. Mayo Clin Proc 1996;71:605-13. Delacretaz E, Beer J, Fleisch M, et al. Low-dose thrombolysis for thrombosed prosthetic heart valve. Chest 1996;110:574-5. Scully HE, Armstrong CS. Tricuspid valve replacement: 15 years of experience with mechanical prostheses and bioprostheses. J Thorac Cardiovasc Surg 1995;109:1035-41. Fiane AE, Seem E, Geiran O, et al. CarboMedics valve in congenital heart disease. Scand J Thorac Cardiovasc Surg 1994; 28:123-6. Reddy NK, Padmanabhan TN, Singh S, et al. Thrombolysis in leftsided prosthetic valve occlusion: immediate and follow-up results. Ann Thorac Surg 1994;58:462-71. Gohlke-Barwolf C, Acar J, Oakley C, et al. Guidelines for prevention of thromboembolic events in valvular heart disease. Eur Heart J 1995;16:1320-30. Israel DH, Sharma SK, Fuster V. Antithrombotic therapy in prosthetic valve replacement. Am Heart J 1994;127:400-11. Stein PD, Alpert JS, Copeland J, et al. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves. Chest 1995;108(suppl):371S-9S. Vongpatanasin W, Hillis LD, Lange RA. Prosthetic heart valves. N Engl J Med 1996;335:407-16. Circulation 1992;86(suppl I):I-806. Turpie AGG, Gent M, Laupacis A, et al. A comparison of aspirin with placebo in patients treated with warfarin after heart-valve replacement. N Engl J Med 1993;329:524-9.
Background Thrombosis of a mechanical tricuspid valve prosthesis is a potentially hazardous event. This study aimed to explore the incidence and the clinical presentation of tricuspid valve thrombosis occurring in bileaflet valves and to evaluate the diagnostic and the therapeutic approach.
Methods and Results Tricuspid valve thrombosis was sought in 22 late survivors with the CarboMedics valve in a follow-up period of 36.0 ± 20.8 months. Limited leaflet motion and/or a visible thrombus were considered diagnostic of valve thrombosis. Eight episodes of tricuspid valve thrombosis were diagnosed among 5 patients (12.1 episodes per 100 patient-years). Anticoagulation was inadequate in 3 patients and fair in 2. Florid right heart failure occurred in 3 episodes. Common physical findings included increased jugular venous pulse (5 patients), diastolic tricuspid murmur (4 patients), and peripheral edema (4 patients). The diagnosis was suspected in all clinically and by transthoracic echocardiography and confirmed by fluoroscopy and/or transesophageal echocardiography. In 4 patients, both leaflets were involved. No thrombi were visualized. Three patients received thrombolytic therapy in 4 episodes (complete success in 3, partial success in 1) without hemorrhagic or embolic complications. One patient responded to aggressive anticoagulant therapy. One patient required an emergent repeat surgery. In 1 patient, valve thrombosis recurred thrice.
Conclusions In patients with fair or poor anticoagulation, a bileaflet valve in the tricuspid position is associated with a high incidence of valve thrombosis. Hinge entrapment requires only a small amount of thrombotic material. Valve thrombosis may be asymptomatic. Involvement of both leaflets is usually required to produce symptoms. A nonsurgical approach (thrombolysis or intensified anticoagulation) is usually successful. Patients should be instructed about heralding signs of valve thrombosis. (Am Heart J 1999;137:721-5.)
Thrombosis of a mechanical tricuspid valve prosthesis is an important and not infrequently seen complication of such a device. Early experience with monoleaflet valves was associated with near 4% annual incidence of tricuspid valve thrombosis,1 but recent experience with bileaflet valves is more favorable, with an annual incidence approaching 1%.2-4 Bileaflet mechanical valves show promise in the reduction of valve thrombosis because of their better hemodynamic profile as compared with older valve models. Because valve implantation in the tricuspid position makes up <2% of all mechanical valve implantations,5-7 experience with tricuspid valve thrombosis is limited. In the current study, we present our experi-
From the Sheingarten Echocardiography Unit, the Department of Cardiology, and the Grown-Up Congenital Heart Disease Unit, Rabin Medical Center, Beilinson Campus; and Sackler Faculty of Medicine, Tel-Aviv University. Submitted February 23, 1998; accepted July 2, 1998. Reprint requests: Yaron Shapira, MD, Department of Cardiology, Rabin Medical Center, Beilinson Campus, Petah Tiqva, Israel, 49100. Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 + 0 4/1/93408
ence with tricuspid valve thrombosis occurring in CarboMedics valves over a 6-year period, with emphasis on the clinical presentation, the diagnostic approach, and the therapeutic approach.
Methods During a 6-year period (January 1991 to December 1996), 30 adult patients underwent tricuspid valve replacement with mechanical bileaflet prostheses in our institution. Eight patients died (7 within 3 months).Twenty-two late survivors (19 with the CarboMedics model, 2 with the Sorin Bicabon model, and 1 with the St Jude Medical model), were monitored for a mean period of 36.0 ± 20.8 months (range 4 to 67 months).The follow-up included history and physical examination on a regular basis (at least annually) and an echocardiogram that preceded the physical examination by ≤4 weeks. Patients were examined by experienced cardiologists. Patient history included questions oriented at possible valve malfunction, for example, muffled valve clicks, fatigue, fluid retention, distended neck veins, or syncope.The anticoagulation status was recorded.Valve clicks (intensity and quality), right-sided murmurs (inspiratory augmentation), neck veins, and signs of fluid retention were specifically sought in physical examination. Patients in whom stuck valve was suspected on clinical
722 Shapira et al
grounds underwent transthoracic echocardiography (TTE) within the same day. All patients’ files and the computerized database of the valvular clinic, the echocardiography laboratory, the congenital heart disease unit, and the cardiothoracic surgery department were screened for evidence of thrombosed tricuspid valve up to September 1997.Valve thrombosis was inferred from at least 1 of the following: (1) restricted leaflet motion, as seen by TTE, transesophageal echocardiography (TEE), or fluoroscopy; (2) a thrombus attached to the valve (seen by TEE); and (3) surgical evidence of valve thrombosis.
Echocardiographic evaluation All patients were evaluated by an experienced echocardiographer, with a Hewlett-Packard Sonos 1000 or Sonos 2000 echocardiography machine, with a 2.5- or 3.5-MHz probe.TTE was performed in all patients, with all standard views.TEE (biplane or multiplane) was performed on an individual basis, according to the physician in charge of the patient. Patients were considered for TEE evaluation if TTE or fluoroscopy showed thrombus or impaired leaflet motion suggestive of a thrombosed valve. The following were considered suggestive of valve thrombosis in TTE (≥1 of the listed criteria): (a) high transvalvular gradients: mean ≥6 mm Hg, peak ≥15 mm Hg; (b) transvalvular gradients ≥50% higher than observed before; (c) visible thrombus attached to the valve; and (d) a wide intravalvular jet of tricuspid regurgitation. Inability to demonstrate 2 different mobile echoes (representing the valve leaflets) in a high-quality image was considered supportive of valve thrombosis in the appropriate setting but could not serve as a single criterion. The diagnostic criterion for valve thrombosis in TEE was limited leaflet motion and/or a visible thrombus.
Fluoroscopy A fluoroscopic study was performed in all the episodes of suspected stuck valve (excluding during pregnancy). Multiple acquisition angles were used to determine leaflet mobility, including the side (pivot) view, with the disks parallel to the xray beams. Inability to fully open or close was considered indicative of stuck valve.
Hemodynamic evaluation The inflow velocities were recorded by continuous wave Doppler from the apical 4-chamber view. Peak and mean transvalvular gradients were calculated from measured velocities with the modified Bernoulli equation, averaging 5 beats in sinus rhythm and 7 in atrial fibrillation. The presence of prosthetic valve regurgitation was assessed by color flow mapping.The severity of tricuspid regurgitation was evaluated in multiple views and graded from 0 to 4+ (severe) according to jet penetration into the right atrium.
Results Among the 22 late survivors, we diagnosed 8 episodes of prosthetic tricuspid valve thrombosis in 5 patients with the CarboMedics model (1 patient with 3 episodes, 1 patient with 2 episodes, and 3
American Heart Journal April 1999
patients with a single episode each).The linearized ratio of valve thrombosis was 12.1 episodes per 100 patient-years. Four patients were treated medically (intensified anticoagulants or thrombolysis).An additional patient had both mitral and tricuspid valve thrombosis 4 months after surgery and had anasarca, cardiac tamponade, and low cardiac output state, which necessitated an emergency successful repeat operation. There were 4 women and 1 man.The mean age was 35.4 ± 7.4 years (range 25 to 45 years). Four patients had rheumatic heart disease and had mechanical mitral valve prosthesis in addition. One patient had Ebstein’s anomaly. One patient was asymptomatic (only indicating muffled valve sounds).Three patients had mild to moderate symptoms, including fatigue and signs of fluid retention. One patient (with stuck mitral valve in addition) was in class IV heart failure.The symptom duration ranged from 1 week to 6 months and was shorter as the functional class was worse. Only the patient with a single prosthetic heart valve indicated muffled valve clicks. In 2 patients, the admission international normalized ratio (INR) was 2.8. In the other 3, the admission INR was far below 2.0. The physical examination revealed prominent jugular venous pulse waves in all.A diastolic murmur of presumed tricuspid origin was heard in 4 and a systolic murmur in 1. Muffled valve clicks were noted in only 1 patient (who did not have mechanical mitral valve in addition). Peripheral edema was noted in 4 patients, of whom 1 had anasarca. One patient was in sinus rhythm, 2 had a paced rhythm, and 2 were in atrial fibrillation. TTE revealed peak diastolic tricuspid gradients of 19 to 24 mm Hg in 4 patients, whereas the fifth had a peak gradient reaching 47 mm Hg.The mean diastolic tricuspid gradients were 9 to 16 mm Hg in 4 patients and 31 mm Hg in the fifth. In this last patient, a forward flow through the tricuspid valve was noted even during systole. Severe tricuspid regurgitation was observed in 3 cases, whereas in 2 it was absent or mild. It was impossible to delineate both leaflets. No thrombi were visualized. All patients underwent fluoroscopy (excluding 1 episode during pregnancy).The asymptomatic patient had only 1 leaflet involved (it was stuck in a closed position). In the other patients both leaflets were involved, usually in a semiopen position. TEE was performed in all cases but 2:The patient with class IV heart failure and additional stuck mitral valve was immediately sent to a repeat surgery after TTE and fluoroscopy.The pregnant patient was followed by TTE.TEE confirmed the diagnosis of stuck valve in all. No thrombi were visualized in TEE as well. One patient responded to enhanced oral anticoagu-
American Heart Journal Volume 137, Number 4, Part 1
Shapira et al 723
Table I. Experience with mechanical prosthetic valves in tricuspid position Valve model* Reference 1 2 3 4 17
No. of patients
Cage ball and other old models
67 78 64 39 25
27 43 (4) 17 (3) — —
Single leaflet 40 23 (3) 18 (4) — most (1)
Follow-up Double leaflet (patient-years) — 12 (1) 29 (2) 39 (1) ≤4
399 NA 681 149 NA
Tricuspid valve thrombosis (% per patient-year) 9 (2.26%) 8 9 (1.32%) 1 (0.67%) 1 (0.37%)
NA, Not available. *Valve thrombosis episodes in parentheses.
lation but had a second thrombotic during early gestation, responding to intensified intravenous anticoagulation after 3 weeks. Eventually, she had a third episode (2 months after the second), requiring valve replacement with bioprosthesis.A third patient was given a 2-day intravenous heparin course, with a partial response, complemented by thrombolysis.All the other episodes were treated by thrombolysis (streptokinase, urokinase or tissue plasminogen activator). Complete success was achieved in all except for 1 patient with partial success who was discharged with intensified oral anticoagulation. One month later, fluoroscopy revealed complete resolution of the leaflet motion.
Discussion There is an ongoing debate in the medical literature regarding replacement of the tricuspid valve with a mechanical prosthesis, even in the era of new low-profile models.The low velocity of blood across the valve makes it prone to thrombosis. Bileaflet models have the advantage of independent function of both leaflets, reducing the risk of complete obstruction. However, the delicate mechanism is more susceptible to functional deterioration, even with a relatively small clot. Our report is confined to the bileaflet CarboMedics valve in the tricuspid position, which is the valve most commonly used in our institution and for which there is little information in the literature.The following discussion corresponds to data from the current series and from fully detailed reports.8-16 Contrary to the florid appearance of left-sided heart failure in patients with stuck left-sided valves, signs of right heart failure associated with stuck tricuspid valve may be quite subtle.The disappearance of formerly audible prosthetic valve clicks in a patient with an isolated tricuspid valve prosthesis may be the first clue leading to the diagnosis of stuck valve. Indeed, in all the cases in which the tricuspid valve was the only prosthesis implanted, the valve clicks were either muffled or absent and self-appreciated by most of the adult patients. On the other
hand, none of the 5 patients with an additional mechanical mitral valve prosthesis noted any change in valve sounds, nor was it noted by the physicians examining the patients. It is prudent to instruct patients to pay attention to the quality of their valve clicks and report any change. Patients should also be instructed to pay attention to warning signs such as distended neck veins or evidence of fluid retention (pedal edema, increased abdominal girth, unexplained weight gain). In all the cases, excluding 1, both valve leaflets were involved. It may be that involvement of both leaflets is needed to cause symptomatic physiologic derangement.Additionally, the close proximity of the hinge points makes it easier for a clot to involve both leaflets. Bedside physical examination can be diagnostic or at least raise clues to the diagnosis of stuck valve. Useful findings include muffled or absent valve clicks, alterations in jugular venous pulse (tall systolic waves, slow Y descent), and diastolic murmur augmented during the inspiratory phase.8-16 Fluid retention is common. Nevertheless, auscultatory findings may be hampered by an associated prosthetic mitral valve, and fluid retention may be associated with various other mechanisms, including predominant leftsided heart failure. TTE was suggestive of valve thrombosis in all cases. The mean transvalvular gradients in the current study and in the literature, derived from Doppler velocities, was at least 8 mm Hg. In each of our cases, the 2dimensional echocardiography showed indirect evidence of abnormal leaflet motion, although the leaflets themselves could not be delineated. Fluoroscopy is the most widely used method for making the diagnosis of stuck valve.TEE, on the other hand, was rarely used for this purpose.15 We used TEE more commonly because most patients had mitral prosthesis in addition, and precise information on that valve was needed.A visible thrombus was found in only 3 cases in the literature and in none in our study.The absence of a visible thrombus may be explained by
American Heart Journal April 1999
724 Shapira et al
Table II. Experience with CarboMedics valves in tricuspid position
Reference 5 6 7
Total valves implanted
Valves in tricuspid position
569 612 277
2 (0.4%) 9 (1.5%) 2 (0.7%)
Tricuspid valve thrombosis 2 1 NA
NA, Not available.
the small amount of thrombotic material needed to catch the hinges of bileaflet valves. In terms of prosthetic valve thrombosis, mechanical heart prostheses in the tricuspid position are not viewed favorably. Many authors cite the 20% incidence of valve thrombosis observed by Thorburn et al1 in monoleaflet models (mainly Bjork-Shiley), corresponding to an incidence of 4% per patient-year. Later reports, corresponding to a mixture of valve models, show some more favorable results (≤1% per patientyear), as shown in Table I.1-4,17 Despite a large core of data on the CarboMedics heart prostheses in left-sided positions, experience with this valve model in the tricuspid position is limited, as shown in Table II.Among 1458 patients with the CarboMedics valve reported in 3 large series, only 13 (0.9%) were implanted in the tricuspid position.5-7 Three patients had documented episodes of valve thrombosis: 2 responded to thrombolytic treatment (streptokinase)5 and 1 was referred to surgery.6 Three out of 4 pediatric patients had prosthetic valve thrombosis 1 to 6 months after surgery, which responded promptly to 12- to 96-hour streptokinase courses.18 Thrombolysis is a valid alternative to surgery in these patients. It is highly effective, and side effects are usually tolerable (minor bleeding or an allergic reaction). Pulmonary embolism was reported only once,10 and this may be attributable to the small amount of thrombotic material needed to disturb leaflet mobility. Long-standing symptoms (>1 month) should not make one reluctant to use thrombolysis in the stuck tricuspid valve, as shown in 8 successfully treated patients (including 1 in our series). Intensifying anticoagulation alone may be effective in selected cases, such as mildly symptomatic patients, as shown twice in 1 patient in our series. Such a treatment may also be beneficial in patients at high risk who partially respond to thrombolysis or when thrombolysis is contraindicated. Reddy et al19 achieved delayed success by such a strategy in 9 of 21 patients with left-sided prostheses in whom the initial result of thrombolysis was only partial. The high incidence of valve thrombosis in the tricus-
pid position may be partially attributed to poor anticoagulant therapy. However, in 2 of our patients, admission INR values were in the therapeutic range (INR = 2.8), considering current general recommendations on anticoagulant therapy.20-23 The INR levels required for prosthetic tricuspid valves may be higher than those needed for mitral valves, especially in a prothrombotic milieu (atrial fibrillation, coexisting prosthetic valve, and so forth).We advocate target INR levels of 3.5 to 4.0 for patients with tricuspid prostheses.The addition of aspirin may further contribute to the prevention of valve thrombosis.24
Conclusions (1) In patients with fair or poor anticoagulation, the CarboMedics bileaflet valve in the tricuspid position is associated with a high incidence of valve thrombosis. (2) The clinical presentation of tricuspid valve thrombosis is frequently mild and even symptomless. (3) The amount of thrombotic material needed to cause restriction of the tricuspid valve leaflets is small and therefore a thrombus is rarely visualized, even in TEE. (4) Involvement of both leaflets is usually required to produce symptoms. (5) Tricuspid valve thrombosis may occasionally respond to intensified anticoagulant therapy. (6) Thrombolysis is almost always successful and is the first line of treatment in these patients, even when thrombosis is not recent. Surgery should be reserved for critically ill patients. (7) Asymptomatic patients with a partially stuck leaflet, who do not respond to thrombolysis, may be discharged from the hospital with long-term intensified anticoagulant therapy and close follow-up.The leaflet motion can be fully restored later.
References 1. Thorburn CW, Morgan JJ, Shanahan MX, et al. Long-term results of tricuspid valve replacement and the problem of prosthetic valve thrombosis. Am J Cardiol 1983;51:1128-32. 2. Van Nooten GJ, Caes F, Taeymans Y, et al. Tricuspid valve replacement: postoperative and long-term results. J Thorac Cardiovasc Surg 1995;110:672-9. 3. Horstkotte D, Burckardt D. Prosthetic valve thrombosis. J Heart Valve Dis 1995;4:141-53. 4. Nakano K, Koyanagi H, Hashimoto A, et al. Tricuspid valve replacement with the St. Jude Medical valve prosthesis. J Thorac Cardiovasc Surg 1994;108:888-92. 5. Fiane AE, Saatvedt K, Svennevig JL, et al. The CarboMedics valve: midterm follow-up with analysis of risk factors. Ann Thorac Surg 1995;60:1053-8. 6. Nistal JF, Hurle A, Revuelta JM, et al. Clinical experience with the CarboMedics valve: early results with a new bileaflet mechanical prosthesis. J Thorac Cardiovasc Surg 1996;112:59-68. 7. Aagaard J, Hansen CN, Tingleff J, et al. Seven-and-a-half years clin-
American Heart Journal Volume 137, Number 4, Part 1
8.
9.
10.
11.
12.
13.
14.
15.
ical experience with the CarboMedics prosthetic heart valve. J Heart Valve Dis 1995;4:628-33. Joyce LD, Boucek M, McGough EC. Urokinase therapy for thrombosis of tricuspid prosthetic valve. J Thorac Cardiovasc Surg 1983; 85:935-7. Kurzrok S, Singh AK, Most AS, et al. Thrombolytic therapy for prosthetic cardiac valve thrombosis. J Am Coll Cardiol 1987; 9:592-8. Mehan VK, Dalvi BY, Kale PA. Thrombosed prosthetic valve in tricuspid position: successful therapy with intravenous streptokinase. Chest 1992;102:1599-600. Tischler MC, Lee RT, Kirshenbaum JM. Successful treatment of prosthetic tricuspid valve thrombosis with short-course recombinant tissue-type plasminogen activator. Am Heart J 1990; 120:975-8. Cohen ML, Barzilai B, Gutierrez F, et al. Treatment of tricuspid valve thrombosis with low-dose tissue plasminogen activator. Am Heart J 1990;120:978-80. Prieto Palomino MA, Ruiz de Elvira MJ, Sanchea Llorente F, et al. Successful thrombolysis on a mechanical tricuspid prosthesis. Eur Heart J 1989;10:1115-7. Fyfe DA, Taylor AB, Kline CH, et al. Doppler echocardiographic evaluation of streptokinase lysis of thrombosed right-sided St. Jude Medical valves in patients with congenital heart defects. Am Heart J 1991;121:1156-60. Hurrell DG, Hartzell VS, Tajik JA. Thrombolytic therapy for
Shapira et al 725
16. 17.
18.
19.
20.
21. 22.
23. 24.
obstruction of mechanical prosthetic valves. Mayo Clin Proc 1996;71:605-13. Delacretaz E, Beer J, Fleisch M, et al. Low-dose thrombolysis for thrombosed prosthetic heart valve. Chest 1996;110:574-5. Scully HE, Armstrong CS. Tricuspid valve replacement: 15 years of experience with mechanical prostheses and bioprostheses. J Thorac Cardiovasc Surg 1995;109:1035-41. Fiane AE, Seem E, Geiran O, et al. CarboMedics valve in congenital heart disease. Scand J Thorac Cardiovasc Surg 1994; 28:123-6. Reddy NK, Padmanabhan TN, Singh S, et al. Thrombolysis in leftsided prosthetic valve occlusion: immediate and follow-up results. Ann Thorac Surg 1994;58:462-71. Gohlke-Barwolf C, Acar J, Oakley C, et al. Guidelines for prevention of thromboembolic events in valvular heart disease. Eur Heart J 1995;16:1320-30. Israel DH, Sharma SK, Fuster V. Antithrombotic therapy in prosthetic valve replacement. Am Heart J 1994;127:400-11. Stein PD, Alpert JS, Copeland J, et al. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves. Chest 1995;108(suppl):371S-9S. Vongpatanasin W, Hillis LD, Lange RA. Prosthetic heart valves. N Engl J Med 1996;335:407-16. Circulation 1992;86(suppl I):I-806. Turpie AGG, Gent M, Laupacis A, et al. A comparison of aspirin with placebo in patients treated with warfarin after heart-valve replacement. N Engl J Med 1993;329:524-9.