- Email: [email protected]
Tricuspid valvulectomy without replacement
J
THORAC CARDIOVASC SURG
1991;102:917-22
Tricuspid valvulectomy without replacement Twenty years' experience Since September 1970, we have operated on 55 patients with intractable right-sided endocarditis. AU patients were addicted to heroin. Fifty-three underwent tricuspid valvulectomy without replacement and in addition two had pulmonic valve excision. Twenty-four patients (49 %) returned to their drug addiction. Six patients (11 %) required prosthetic heart valve insertion 2 days to 13 years later for medically refractory right-sided heart failure, and four of these died. OveraU, 16 patients (29%) died, six (11 %) within 45 days after the tricuspid valvulectomy. One (2 %) of these deaths was related to the operation and five were due to uncontroUable infection. Ten (18 %) deaths occurred 9 months to 13 years after the tricuspid valvulectomy. Nine were due to drug addiction and one to progressive right ventricular failure 2 months after prosthetic heart valve insertion and 10 years after the initial valve removal. Of the 39 patients who are alive, 37 (67%) have not required prosthetic heart valve insertion. From our observations we reached the foUowing conclusions: (1) Drug addiction is a recurrent and lethal disease. Among these patients, tricuspid valvulectomy without replacement is the operation of choice for the management of intractable right-sided endocarditis; (2) after tricuspid valvulectomy without replacement, only six of 55 patients (11 %) had required prosthetic heart valve insertion to control medicaUy refractory right-sided heart failure; (3) in a smaU percentage of patients the absence of the tricuspid valve may lead to severe and permanent impairment of right ventricular function.
Agustin Arbulu, MD, Robert J. Holmes, MD (by invitation), and Ingida Asfaw, MD (by invitation), Detroit, Mich.
Q
September 3, 1970, the first patient with intractable right-sided infective endocarditis caused by Pseudomonas aeruginosa underwent a tricuspid valvulectomy without insertion of a prosthesis. Since then, 55 patients have had a similar operation. This report reviews our experience in the management of this group of patients.
Patients and methods All the patients had been addicted to heroin for 1 to 20 years, average 8 years. Thirty-nine were men and 16 were women, a male/female ratio of 2.4:1. Their ages ranged from 20 to 53 years, with an average of 31 years. All patients had had unsuccessful medical treatment. The antibiotics were administered intravenously in various combinations for periods that ranged from 2 to 12 months. The operation was performed as a From Wayne State University, School of Medicine, Detroit, Mich. Read at the Seventy-first Annual Meeting of The American Association for Thoracic Surgery, Washington, D.C., May 6-8, 1991. Address for reprints: Agustin Arbulu, MD, 4160 John R. Ste. 829, Detroit, MI 48201.
12/6/32885
life-saving measure only. Table I shows the infecting microorganisms. Of the 55 patients, 29 had Pseudomonas aeruginosa infection. In four of these Candida albicans was also present. In 27 patients Staphylococcus aureus was the infecting organism. Nineteen (70%) of the gram-positive infections were resistant to methicillin, nafcillin, or both. Two additional patients (7%) had staphylococcal infection resistant to either vancomycin or cephalosporin. Five patients (19%) of the Staphylococcus aureus infection group also had Candida albicans. Eight patients had fungal infections, and of these only one (2%) had Candida albicans as the only infecting organism. Of the total population of patients, 22 (40%) had multiple organisms involved. The predominant microorganisms were significantly different when the 20-year period was divided into its two decades. During the decade of 1970 to 1980, gram-negative infections resulting from Pseudomonas aeruginosa dominated-2l (95%) of the 22 patients. In contrast, during the decade of 1981 to 1990, the dominant microorganism was Staphylococcus aureus-26 (79%) of33 patients. Only eight patients (24%) had Pseudomonas aeruginosa infections. Additionally, eight patients (24%) had fungal endocarditis. Also, in the second decade of this 20-year period, there were 22 patients (67%) who had two or more microorganisms (Table I). In all 55 patients, the diagnosis was made as follows: (I) history of intravenous drug addiction for periods that ranged from 1 to 20 years; (2) fever and prostration; (3) chest pain and chest
917
9I8
The Journal of Thoracic and Cardiovascular Surgery
Arbulu, Holmes, Asfaw
Table I. Microorganisms causing endocarditis Organism
No. ofpatients
1970-1980 (n = 22 patients) Pseudomonas aeruginasa Staphylococcus aureus* 1981-1990 (n = 33 patients) Pseudomonas aeruginosa Staphylococcus aureus* Fungir Staphylococcus epidermidis
Catheter position 21 I
Superiorvena cava
26 (8) 8 (7) I (I)
a-Hemolyticstreptococcus
I (I) I (I)
Figures in parentheses indicate mixed infections. 'Methicillin resistant, n = 12; nafcillin resistant, n = 3; methicillin and nafcillin resistant, n = 4; vancomycin resistant, n = I; and cephalosporin resistant, n = I.
= 2.
radiographic manifestations of septic embolization to the lung, such as infiltrate, pleural effusion (empyema), or both; and (4) positive blood cultures. Of 31 patients who had echocardiograms, 28 (90%) had a vegetation of various sizes involving the tricuspid valve. In two patients (6%) pulmonary embolism was diagnosed, and o~e of the two had a significant pericardial effusion. In two patients (6%) the echocardiogram was nondiagnostic, and in only one the echocardiogram failed to detect the vegetations. Fifty-three patients underwent a total excision of the tricuspid valve without replacement, and two underwent excision of the tricuspid and pulmonic valves without replacement. All patients received postoperative intravenous antibiotics for 2 to 6 weeks. Twenty-four of the 49 patients (49%) who survived the valvulectomy returned to using drugs after the operation. Six patients (11%) required prosthetic valve insertion from 2 days to 13 years after excision of valve(s). Five had the tricuspid valve excised and one had both the tricuspid and pulmonic valves excised. In all six of these patients the prosthetic valve insertion was necessary for medically refractory right-sided heart failure.
Results Sixteen patients died, for an overall mortality rate of 29%. Six (II %) of the deaths occurred within 45 days after excision of the tricuspid valve. Only one of these early deaths was related to the absence of the tricuspid valve. The other five early deaths were due to uncontrollable infection. Ten patients (18%) died 9 months to 13 years after tricuspid valve excision or tricuspid and pulmonic valve excision without replacement. In nine patients the death was related to the use of drugs, and one patient died of progressive and uncontrollable right ventricular failure 2 months after insertion of a prosthetic valve in the tricuspid position. The clinical course of this patient was as follows.
Case report A 32-year-old man who had abused intravenous drugs for 15 years was admitted to the hospital with the clinical manifesta-
Pressures (mmHg)
Saturation (%)
V= 30
58
A= 26
Right atrium
V= 33
Right ventricle Right pulmonary artery Left ventricle Aorta
35/5 (20) 30/15 (24) 120/8-4 120/70
8 (4)
Serratia marcescens
tCandida albicans, n = 6; Candida tropicalis, n
Table II. Cardiac catheterization results
A = 26
56
Figures in parentheses aremean values. V, V wave; A, A wave. tion of right-sided endocarditis. Blood cultures were positivefor Pseudomonas aeruginosa. Because all medical treatment had failed, he was referred to the surgical service. On May 5, 1973, the patient underwent a tricuspid valvulectomy without replacement. At the time of the operation the tricuspid valve was totally destroyed and replaced with florid vegetations. The cultures from the specimen grew Pseudomonas aeruginosa. After the operation, the patient received a 6-week course of intravenous antibiotics and was discharged home from the hospital on June 28, 1973, in good health. In the ensuing 6 years after his operation, he continued to use intravenous heroin. In 1979 he was admitted to the hospital with a groin infection and sepsis. He had pulsating neck veins and a large liver. The chest roentgenogram showed clear lungs and mild cardiomegaly. Blood cultures were positive for Staphylococcus aureus. He required drainage of a right groin abscess and intravenous antibiotics for 3 weeks. He was discharged home in improved condition. From 1980 to 1982 he was seen in the office at regular intervals. It was gratifying to see that he was finally free of his drug addiction. Toward the end of 1982 he showed signs of right ventricular failure not responsive to medical treatment. Early in 1983 ascites developed. Cardiac catheterization showed significant elevations of superior vena cava and right atrial pressures, slight elevation of the right ventricular and pulmonary artery pressures, and normal left ventricular and aortic pressures (Table 11). On June 2, 1983, he underwent a second cardiac operation. The right femoral artery and veins were cannulated in preparation for cardiopulmonary bypass. The heart was exposed through a midsternal thoracotomy. The right side of the heart, which was a single chamber with a dilated and hypertrophic right atrium, was moderately enlarged and adherent to the chest wall. The tricuspid anulus was a fibrotic line. A size 33 Carpentier-Edwards bioprosthetic valve (Baxter Healthcare Corp., Edwards Division, Irvine, Calif.) was placed in the tricuspid position. During the first 10 days after the operation his con?ition improved. He was extubated on the second postoperative day and resumed oral intake. However, lymph fluid began draining from the femoral artery and vein cannulation site. By the end of the third postoperative week the 1ymph drainage had increased to 800 to 1000 ml/24 hr. The patient's ascites had practically resolved but his nutritional status continued to deteriorate despite adequate oral alimentation. Noninvasive evaluation ofleft ventricular function showed no abnormalities. In an attempt to improve his nutritional status, we closed the lymphatic fistula with a sartorius muscle flap. The ascites reacum-
Volume 102 Number 6 December 1991
mulated after closure of the lymph fistula, and despite every possible treatment intervention the patient's nutritional status continued to deteriorate. Signs of right-sided heart failure developed, resulting in massive hepatomegaly and ascites. He died on August 3, 1983. Permission for autopsy was not granted by the family.
Six patients required tricuspid valve insertion because of intractable right-sided heart failure. Four of these patients died, one of low cardiac output on the fifth postoperative day. Another (the one whose case is described in detail) died of intractable right-sided heart failure 2 months after insertion of the tricuspid prosthesis. The remaining two deaths were due to continued drug abuse and were unrelated to the cardiac status of the patients. They were late deaths. The two patients who are alive and well required implantation of dual-chamber epicardial pacemakers. One of these patients also had an iatrogenic interventricular septal defect. This defect was created during debridement of the anulus during the tricuspid valvulectomy without replacement. It was repaired with pledget-supported sutures. This repair failed and the recurrent interventricular septal defect contributed to the medically refractory right-sided heart failure that necessitated the.second cardiac operation 2 years later. At this operation the defect was repaired with a Dacron patch, and a size 31 Carpentier-Edwards prosthesis was inserted in the tricuspid position. Thirty-nine patients are alive and well. One of the long-term survivors has no tricuspid or pulmonic valve. Two of these patients required tricuspid valve insertion. Thirty-seven patients (67%) had not required tricuspid valve insertion. Only two of the 37 patients (4%) had mild right ventricular failure and this was well controlled medically. Thirty-five (63%) are well and in stable clinical condition (New York Heart Association class 11). Twelve patients (22%) had 18 complications after the tricuspid valvulectomy without replacement (Table III). Study of liver function and its histologic changes. The liver was evaluated in 38 patients. These studies consisted of clinical assessment of the size of the liver and determination of total serum protein including the albumin and globulin ratio. Also, serial determinations of serum glutamic oxaloacetic transaminase (SGOT), serum glutamic pyruvic transaminase (SGPT), and alkaline phosphatase were done. In 30 patients (79%) the size of the liver was normal and in eight the liver was enlarged from 3.5 to 6 cm below the right costal margin. Only one patient had ascites. The alkaline phosphatase concentration was elevated in seven patients (I 8%). The total serum protein, albumin, and globulin ratios, as well as the SGOT and SGPT, were within the normal range. In addition, we studied the liver of two patients who
Tricuspid valvulectomy without replacement
9 19
Table III. Complications of tricuspid valvulectomy: 12 patients (22%)* Cardiac arrest Cardiac tamponade Clotted hemothorax Respiratory failure Ventricular septal defect Diffuse intravascular coagulation Low cardiac output Renal failure Central nervous system impairment Sternal wound infection
2
2 (I) 2 2 3
3 (2) I (I) 1
I 1
Figures in parentheses indicate early deaths. 'One patient had four complications. one had three complications. and one had two complications.
died 9 months and 13 months after the tricuspid valvulectomy without replacement. Both livers were moderately enlarged, weighing 1660 and 2200 gm, respectively. The central veins of the hepatic lobules were distended in both patients without changes suggestive of cardiac cirrhosis. Renal function and histologic findings. Renal function was normal in all patients. The gross autopsy studies of the kidneys from the two patients who had autopsies showed no abnormal changes. The microscopic structure of the kidneys was normal. Reproductive function. Two patients had normal pregnancies resulting in normal child birth. The overall postoperative follow-up in this population of patients ranged from 2 years to more than 20 years since the tricuspid valvulectomy without replacement. The actuarial survival rate is 63.8% at 20 years (Fig. 1).
Discussion This long-term experience in this large series of patients that underwent tricuspid valvulectomy without replacement now exceeds 20 years. All our patients were intravenous drug abusers. We have not seen this infection among patients not addicted to drugs, as it has been reported by others.v!'' Our recommendation that this operation is the treatment of choice in drug addicts with medically incurable right-sided endocarditis is supported by our results. In recommending an operation for patients with rightsided endocarditis, it is important to consider whether or not they are addicted to drugs. The fact that 49% of the patients in this series returned to the use of drugs after they were cured of their endocarditis indicates that drug dependency is a serious lethal disease. This condition is also seen clearly in the patient whose case is presented in detail in this article. The long-term follow-up of these patients indicates that almost half of the drug addicts
The JournaL of Thoracic and Cardiovascular
920 Arbulu, Holmes, Asfaw
Cases 55 53 Deaths
Surgery
52
51
49
48
45
43
1
1
2
1
3
2
2
Probability of Survival %
42
1 (1) (1) TVI TVI
25
13
1
1
11
832
100 81A
94.6
89.2
80
18A
74.6
82.0
69.6
16.5
12.9
60
63.B
40 20
o
o
5
8
Days
11
6
Wks
6
9
12
Months
2
4
6
8
10
13
14
16
18
20
Years Post Operation
Fig. 1. Actuarialsurvival curveshowing survival of patients 20 yearsafter tricuspidvalvulectomy without prosthetic replacement TV!, Tricuspid valve replacement
return to the use of drugs after they are cured of their heart infection. The drug addiction is the primary disease. The endocarditis is an intercurrent illness and its management must be planned accordingly. The application of therapeutic measures that are excellent in the management of patients not addicted to drugs, such as prosthetic heart valve replacement at the initial operation or the strict recommendation of reconstructive valvular surgery, appears unrealistic. In our experience, all of the patients had extensive involvement of the entire tricuspid valve by the infection. Perhaps this was related to the protracted ineffective use of intravenous antibiotics. In this series of patients, we were not able to save or repair any portion of the tricuspid valve. We think that these conservative operations are applicable to a small number of patients with tricuspid valve endocarditis with limited involvement of the valve apparatus. The anatomic characteristics and the extent of the infection are the most important determinants in the application of these conservative operations. 11, 12 The poor result after the late tricuspid valve insertion in our patient whose case is discussed in detail is worth further discussion. The first clinical course of this patient supports our contention that for drug addicts tricuspid valvulectomy without replacement is the indicated operation. This patient continued to use intravenous drugs for
more than 6 years after his valvulectomy. The absence of a foreign body in the tricuspid position prevented the development of recurrent right-sided endocarditis. It took 6 years for this patient to be cured of his drug' addiction. The development of intractable right-sided heart failure late in his clinical course suggests that there may have been changes in the pulmonary vascular circulation, such as an increment in the pulmonary resistance resulting from microembolization of the pulmonary capillaries. However, this theory was not supported by the cardiac catheterization findings before his second cardiac operation (Table II). He showed no evidence of pulmonary artery hypertension. Therefore we must conclude that the right ventricular decompensation was directly related to the 8- to lO-year absence of the tricuspid valve. The failure to control the right ventricular failure after insertion of a tricuspid prosthesis indicates that when medically refractory right heart failure develops in a patient who has been without a tricuspid valve for such a long time, the problem cannot be resolved with only insertion of a prosthetic valve. Perhaps, under these conditions, right ventricular function can be improved with the addition of a cardiomyoplastic operation to assist and improve right ventricular contractility. This suggestion is supported by the satisfactory experimental and clinical results reported in the literature when these operations have been used
Volume 102 Number 6 December 1991
to improve the function of severely impaired hearts. 13- 15 The prevalence of complications in this population of extremely ill patients was 22% (Table III). Medical treatment had failed in all of the patients and the operation was recommended as a last resort of treatment. Several patients were moribund at the time of the tricuspid valvulectomy without replacement. Four of these patients died early. In only one of these patients, the one with low cardiac output, was the death related to the absence of the tricuspid valve. In the other three, uncontrollable infection was the cause of death. It is encouraging to observe that among these patients the actuarial survival curve at 20 years is 64% (Fig. I). In calculating these results, we have excluded the two surviving patients who required insertion of a tricuspid prosthesis. This excellent survival rate at 20 years is due to the normal functional characteristics of the heart of each of these patients. The crucial ingredient for the success of this operation is normal function of the left side of the heart. This allows excision of the tricuspid valve without prosthetic valve replacement to be well tolerated for prolonged periods. From the technical viewpoint, tricuspid valvulectomy without replacement is a simple operation. It requires a short run of extracorporeal circulation. The intravenous fluid requirements after the operation are higher. These patients need a high right heart filling pressure during the initial 24 to 48 hours after excision of the tricuspid valve. I The long-term follow-up of these patients showed well-preserved liver function. The only consistent finding was an elevation of the alkaline phosphatase concentration in 18% of the 38 patients who underwent liver studies. This is probably related to the increased pressure in the central lobular vein of the hepatic parenchyma. 16 This experience also showed normal function of the renal and the endocrine systems. On the basis of our experience, we can make several conclusions: I. Drug addiction is a recurrent and lethal disease. 2. only six of 55 patients (11 %) who had tricuspid valvulectomy without replacement or tricuspid valvulectomy and pulmonic valvulectomy without replacement required prosthetic valve insertion at a later date, to control medically refractory right-sided heart failure. 3. tricuspid valvulectomy without replacement or tricuspid valvulectomy and pulmonic valvulectomy without replacement remain the operations of choice in the management of intractable right-sided endocarditis among drug addicts. The actuarial survival curve of 64% at 20 years compares well with long-term results of any cardiac operation. 4. A small percentage of patients without a tricuspid
Tricuspid valvulectomy without replacement
92I
valve will have intractable right-sided heart failure because of marked stretching and decompensation of the right side of the heart. Only one of our patients reached this advanced stage and prosthetic valve insertion in the tricuspid position was of no value. Perhaps under these conditions a cardiomyoplastic operation to improve the contractility of the right ventricle should be considered in addition to insertion of a prosthesis in the tricuspid valve position. We acknowledge the valuableassistanceof Roger F. Higgins, PhD, Director of Higgins Vascular Laboratory, 21701 West Eleven Mile, Southfield,MI 48076,in calculating and outlining the actuarial survival curve in this paper. REFERENCES I. Arbulu A, Asfaw I. Tricuspid valvulectomy without prosthetic replacement: ten years of clinical experience. J THORAC CARDIOVASC SURG 1981;82:684-91. 2. Vander Westhuizen NG, Rose AG. Right-sided valvular infective endocarditis: a clinicopathological study of 29 patients. S Afr Med J 1987;71 :25-7. 3. Kohli RS, Anand IS, Sivaram SV, Datta BN, Bidwai PS, Wahl PL. Isolated right-sidedendocarditis in non-addicts: a review of 10cases seen between 1967-79.Indian Heart J 1982;34:17-20. 4. Valla D, Parientd EA, Degott C, et al. Right-sided endocarditis complicating peritonea-venous shunting for ascites. Arch Intern Med 1983;143:1801-2. 5. Rowley KM, Clubb KS, Walker Smith GJ, Cabin HS. Right-sided infective endocarditis as a consequence of flow-directed pulmonary artery catheterization: a clinicopathologicalstudy of 55 autopsied patients. N Engl J Med 1984;311: 1152-6. 6. Chia BL, Yan PC, Ee BK, Chao MH, Lee CN, Tay MB. Right-sidedinfective endocarditis: a Singapore experience. Am Heart J 1988;116:568-71. 7. Mesa JM, Oliver J, Dominguez F, et al. Endocarditis infecciosa derecha. Experiencia en una serie de 35 pacientes. Rev Esp CardioI1990;43:142-52. 8. Chan P, Ogilby JD, Segal B. Tricuspid valveendocarditis. Am Heart J 1989;117:1140-6. 9. Naidao DP, Naicker S, Vythylingum,et al. Isolated tricuspid valve infective endocarditis: a report of 6 cases. S Afr Med J 1990;78:34-8. 10. Martino P, Micozzi A, Venditti M, et al. Catheter-related right-sidedendocarditisin bone marrow transplant recipients. Rev Infect Dis 1990; 12:250-7. II. Pruett TL, Rotstein on Anderson RW, Simmons RL. Tricuspidvalve Candida endocarditis: successful treatment withvalve-sparing debridement and antifungal chemotherapy in a multiorgan transplant recipient. Am J Med 1986;80: 116-8. 12. Yee ES, Khonsari S. Right-sided infective endocarditis: valvuloplasty, valvectomy or replacement. J Cardiovasc Surg 1989;30:744-8.
The Journatot
922
Arbulu, Holmes, Asfaw
13. Chiu RCJ. Biomechanical cardiac assist: cardiomyoplasty and muscle-powered devices. Mount Kisco, New York: Futura. 14. Koroteyev A, Pochettino A, Niinami H, Stephenson LW. Skeletal muscle: new techniques for treating heart failure. AORN J 1991;53:1005-19. 15. Anderson DR, Pochettino A, Hammond BA, et al. Autogenously lined skeletal muscle ventricles in circulation: up to nine months' experience. J THORAC CARDIOVASC SURG 1991;101:661-70. 16. Kaplan MM. Progress in hepatology: alkaline phosphatase. Gastroenterology 1972;62:452-68.
Discussion Dr. Delos M. Cosgrove (Cleveland, Ohio). Dr. Arbulu, is this the only operation for tricuspid endocarditis that you performed in this period of time? Dr. Arbulu. In this population of patients, Dr. Cosgrove, that was the only operation that we performed. In two patients we excised the pulmonic valve in addition to the tricuspid valve. All
Thoracic and Cardiovascular Surgery
the patients in this series had had unsuccessful medical treatment. Perhaps the extensive infection of the tricuspid valve was related to their prolonged medical management. Dr. Cosgrove. Did you excise the infected portion of the valve in any of your group? Dr. Arbulu. No. Dr. Verdi J. DiSesa (Philadelphia, Pa.). Do you have any data on pulmonary artery pressures in the patients, particularly those who had right-sided heart failure, and could particulate emboli from drug contaminants have created pulmonary hypertension in some of these patients and made the results of the tricuspid valve replacement as poor as you described? Dr. Arbulu. The patient who died 10 years after tricuspid valvulectomy and 2 months after insertion of the prosthesis underwent a complete cardiac catherization, and the right-sided pressures were slightly elevated. The pulmonary pressure was 30mm Hg. Dr. James L. Cox (St. Louis, Mo.). Dr. Arbulu, did you ever have a patient in whom you excised the tricuspid valve and were unable then to wean that patient from bypass because of absence of the tricuspid valve? I noticed that all of your patient deaths occurred after you had weaned them from bypass. Dr. Arbulu. We did not have that experience, Dr. Cox.
THORAC CARDIOVASC SURG
1991;102:917-22
Tricuspid valvulectomy without replacement Twenty years' experience Since September 1970, we have operated on 55 patients with intractable right-sided endocarditis. AU patients were addicted to heroin. Fifty-three underwent tricuspid valvulectomy without replacement and in addition two had pulmonic valve excision. Twenty-four patients (49 %) returned to their drug addiction. Six patients (11 %) required prosthetic heart valve insertion 2 days to 13 years later for medically refractory right-sided heart failure, and four of these died. OveraU, 16 patients (29%) died, six (11 %) within 45 days after the tricuspid valvulectomy. One (2 %) of these deaths was related to the operation and five were due to uncontroUable infection. Ten (18 %) deaths occurred 9 months to 13 years after the tricuspid valvulectomy. Nine were due to drug addiction and one to progressive right ventricular failure 2 months after prosthetic heart valve insertion and 10 years after the initial valve removal. Of the 39 patients who are alive, 37 (67%) have not required prosthetic heart valve insertion. From our observations we reached the foUowing conclusions: (1) Drug addiction is a recurrent and lethal disease. Among these patients, tricuspid valvulectomy without replacement is the operation of choice for the management of intractable right-sided endocarditis; (2) after tricuspid valvulectomy without replacement, only six of 55 patients (11 %) had required prosthetic heart valve insertion to control medicaUy refractory right-sided heart failure; (3) in a smaU percentage of patients the absence of the tricuspid valve may lead to severe and permanent impairment of right ventricular function.
Agustin Arbulu, MD, Robert J. Holmes, MD (by invitation), and Ingida Asfaw, MD (by invitation), Detroit, Mich.
Q
September 3, 1970, the first patient with intractable right-sided infective endocarditis caused by Pseudomonas aeruginosa underwent a tricuspid valvulectomy without insertion of a prosthesis. Since then, 55 patients have had a similar operation. This report reviews our experience in the management of this group of patients.
Patients and methods All the patients had been addicted to heroin for 1 to 20 years, average 8 years. Thirty-nine were men and 16 were women, a male/female ratio of 2.4:1. Their ages ranged from 20 to 53 years, with an average of 31 years. All patients had had unsuccessful medical treatment. The antibiotics were administered intravenously in various combinations for periods that ranged from 2 to 12 months. The operation was performed as a From Wayne State University, School of Medicine, Detroit, Mich. Read at the Seventy-first Annual Meeting of The American Association for Thoracic Surgery, Washington, D.C., May 6-8, 1991. Address for reprints: Agustin Arbulu, MD, 4160 John R. Ste. 829, Detroit, MI 48201.
12/6/32885
life-saving measure only. Table I shows the infecting microorganisms. Of the 55 patients, 29 had Pseudomonas aeruginosa infection. In four of these Candida albicans was also present. In 27 patients Staphylococcus aureus was the infecting organism. Nineteen (70%) of the gram-positive infections were resistant to methicillin, nafcillin, or both. Two additional patients (7%) had staphylococcal infection resistant to either vancomycin or cephalosporin. Five patients (19%) of the Staphylococcus aureus infection group also had Candida albicans. Eight patients had fungal infections, and of these only one (2%) had Candida albicans as the only infecting organism. Of the total population of patients, 22 (40%) had multiple organisms involved. The predominant microorganisms were significantly different when the 20-year period was divided into its two decades. During the decade of 1970 to 1980, gram-negative infections resulting from Pseudomonas aeruginosa dominated-2l (95%) of the 22 patients. In contrast, during the decade of 1981 to 1990, the dominant microorganism was Staphylococcus aureus-26 (79%) of33 patients. Only eight patients (24%) had Pseudomonas aeruginosa infections. Additionally, eight patients (24%) had fungal endocarditis. Also, in the second decade of this 20-year period, there were 22 patients (67%) who had two or more microorganisms (Table I). In all 55 patients, the diagnosis was made as follows: (I) history of intravenous drug addiction for periods that ranged from 1 to 20 years; (2) fever and prostration; (3) chest pain and chest
917
9I8
The Journal of Thoracic and Cardiovascular Surgery
Arbulu, Holmes, Asfaw
Table I. Microorganisms causing endocarditis Organism
No. ofpatients
1970-1980 (n = 22 patients) Pseudomonas aeruginasa Staphylococcus aureus* 1981-1990 (n = 33 patients) Pseudomonas aeruginosa Staphylococcus aureus* Fungir Staphylococcus epidermidis
Catheter position 21 I
Superiorvena cava
26 (8) 8 (7) I (I)
a-Hemolyticstreptococcus
I (I) I (I)
Figures in parentheses indicate mixed infections. 'Methicillin resistant, n = 12; nafcillin resistant, n = 3; methicillin and nafcillin resistant, n = 4; vancomycin resistant, n = I; and cephalosporin resistant, n = I.
= 2.
radiographic manifestations of septic embolization to the lung, such as infiltrate, pleural effusion (empyema), or both; and (4) positive blood cultures. Of 31 patients who had echocardiograms, 28 (90%) had a vegetation of various sizes involving the tricuspid valve. In two patients (6%) pulmonary embolism was diagnosed, and o~e of the two had a significant pericardial effusion. In two patients (6%) the echocardiogram was nondiagnostic, and in only one the echocardiogram failed to detect the vegetations. Fifty-three patients underwent a total excision of the tricuspid valve without replacement, and two underwent excision of the tricuspid and pulmonic valves without replacement. All patients received postoperative intravenous antibiotics for 2 to 6 weeks. Twenty-four of the 49 patients (49%) who survived the valvulectomy returned to using drugs after the operation. Six patients (11%) required prosthetic valve insertion from 2 days to 13 years after excision of valve(s). Five had the tricuspid valve excised and one had both the tricuspid and pulmonic valves excised. In all six of these patients the prosthetic valve insertion was necessary for medically refractory right-sided heart failure.
Results Sixteen patients died, for an overall mortality rate of 29%. Six (II %) of the deaths occurred within 45 days after excision of the tricuspid valve. Only one of these early deaths was related to the absence of the tricuspid valve. The other five early deaths were due to uncontrollable infection. Ten patients (18%) died 9 months to 13 years after tricuspid valve excision or tricuspid and pulmonic valve excision without replacement. In nine patients the death was related to the use of drugs, and one patient died of progressive and uncontrollable right ventricular failure 2 months after insertion of a prosthetic valve in the tricuspid position. The clinical course of this patient was as follows.
Case report A 32-year-old man who had abused intravenous drugs for 15 years was admitted to the hospital with the clinical manifesta-
Pressures (mmHg)
Saturation (%)
V= 30
58
A= 26
Right atrium
V= 33
Right ventricle Right pulmonary artery Left ventricle Aorta
35/5 (20) 30/15 (24) 120/8-4 120/70
8 (4)
Serratia marcescens
tCandida albicans, n = 6; Candida tropicalis, n
Table II. Cardiac catheterization results
A = 26
56
Figures in parentheses aremean values. V, V wave; A, A wave. tion of right-sided endocarditis. Blood cultures were positivefor Pseudomonas aeruginosa. Because all medical treatment had failed, he was referred to the surgical service. On May 5, 1973, the patient underwent a tricuspid valvulectomy without replacement. At the time of the operation the tricuspid valve was totally destroyed and replaced with florid vegetations. The cultures from the specimen grew Pseudomonas aeruginosa. After the operation, the patient received a 6-week course of intravenous antibiotics and was discharged home from the hospital on June 28, 1973, in good health. In the ensuing 6 years after his operation, he continued to use intravenous heroin. In 1979 he was admitted to the hospital with a groin infection and sepsis. He had pulsating neck veins and a large liver. The chest roentgenogram showed clear lungs and mild cardiomegaly. Blood cultures were positive for Staphylococcus aureus. He required drainage of a right groin abscess and intravenous antibiotics for 3 weeks. He was discharged home in improved condition. From 1980 to 1982 he was seen in the office at regular intervals. It was gratifying to see that he was finally free of his drug addiction. Toward the end of 1982 he showed signs of right ventricular failure not responsive to medical treatment. Early in 1983 ascites developed. Cardiac catheterization showed significant elevations of superior vena cava and right atrial pressures, slight elevation of the right ventricular and pulmonary artery pressures, and normal left ventricular and aortic pressures (Table 11). On June 2, 1983, he underwent a second cardiac operation. The right femoral artery and veins were cannulated in preparation for cardiopulmonary bypass. The heart was exposed through a midsternal thoracotomy. The right side of the heart, which was a single chamber with a dilated and hypertrophic right atrium, was moderately enlarged and adherent to the chest wall. The tricuspid anulus was a fibrotic line. A size 33 Carpentier-Edwards bioprosthetic valve (Baxter Healthcare Corp., Edwards Division, Irvine, Calif.) was placed in the tricuspid position. During the first 10 days after the operation his con?ition improved. He was extubated on the second postoperative day and resumed oral intake. However, lymph fluid began draining from the femoral artery and vein cannulation site. By the end of the third postoperative week the 1ymph drainage had increased to 800 to 1000 ml/24 hr. The patient's ascites had practically resolved but his nutritional status continued to deteriorate despite adequate oral alimentation. Noninvasive evaluation ofleft ventricular function showed no abnormalities. In an attempt to improve his nutritional status, we closed the lymphatic fistula with a sartorius muscle flap. The ascites reacum-
Volume 102 Number 6 December 1991
mulated after closure of the lymph fistula, and despite every possible treatment intervention the patient's nutritional status continued to deteriorate. Signs of right-sided heart failure developed, resulting in massive hepatomegaly and ascites. He died on August 3, 1983. Permission for autopsy was not granted by the family.
Six patients required tricuspid valve insertion because of intractable right-sided heart failure. Four of these patients died, one of low cardiac output on the fifth postoperative day. Another (the one whose case is described in detail) died of intractable right-sided heart failure 2 months after insertion of the tricuspid prosthesis. The remaining two deaths were due to continued drug abuse and were unrelated to the cardiac status of the patients. They were late deaths. The two patients who are alive and well required implantation of dual-chamber epicardial pacemakers. One of these patients also had an iatrogenic interventricular septal defect. This defect was created during debridement of the anulus during the tricuspid valvulectomy without replacement. It was repaired with pledget-supported sutures. This repair failed and the recurrent interventricular septal defect contributed to the medically refractory right-sided heart failure that necessitated the.second cardiac operation 2 years later. At this operation the defect was repaired with a Dacron patch, and a size 31 Carpentier-Edwards prosthesis was inserted in the tricuspid position. Thirty-nine patients are alive and well. One of the long-term survivors has no tricuspid or pulmonic valve. Two of these patients required tricuspid valve insertion. Thirty-seven patients (67%) had not required tricuspid valve insertion. Only two of the 37 patients (4%) had mild right ventricular failure and this was well controlled medically. Thirty-five (63%) are well and in stable clinical condition (New York Heart Association class 11). Twelve patients (22%) had 18 complications after the tricuspid valvulectomy without replacement (Table III). Study of liver function and its histologic changes. The liver was evaluated in 38 patients. These studies consisted of clinical assessment of the size of the liver and determination of total serum protein including the albumin and globulin ratio. Also, serial determinations of serum glutamic oxaloacetic transaminase (SGOT), serum glutamic pyruvic transaminase (SGPT), and alkaline phosphatase were done. In 30 patients (79%) the size of the liver was normal and in eight the liver was enlarged from 3.5 to 6 cm below the right costal margin. Only one patient had ascites. The alkaline phosphatase concentration was elevated in seven patients (I 8%). The total serum protein, albumin, and globulin ratios, as well as the SGOT and SGPT, were within the normal range. In addition, we studied the liver of two patients who
Tricuspid valvulectomy without replacement
9 19
Table III. Complications of tricuspid valvulectomy: 12 patients (22%)* Cardiac arrest Cardiac tamponade Clotted hemothorax Respiratory failure Ventricular septal defect Diffuse intravascular coagulation Low cardiac output Renal failure Central nervous system impairment Sternal wound infection
2
2 (I) 2 2 3
3 (2) I (I) 1
I 1
Figures in parentheses indicate early deaths. 'One patient had four complications. one had three complications. and one had two complications.
died 9 months and 13 months after the tricuspid valvulectomy without replacement. Both livers were moderately enlarged, weighing 1660 and 2200 gm, respectively. The central veins of the hepatic lobules were distended in both patients without changes suggestive of cardiac cirrhosis. Renal function and histologic findings. Renal function was normal in all patients. The gross autopsy studies of the kidneys from the two patients who had autopsies showed no abnormal changes. The microscopic structure of the kidneys was normal. Reproductive function. Two patients had normal pregnancies resulting in normal child birth. The overall postoperative follow-up in this population of patients ranged from 2 years to more than 20 years since the tricuspid valvulectomy without replacement. The actuarial survival rate is 63.8% at 20 years (Fig. 1).
Discussion This long-term experience in this large series of patients that underwent tricuspid valvulectomy without replacement now exceeds 20 years. All our patients were intravenous drug abusers. We have not seen this infection among patients not addicted to drugs, as it has been reported by others.v!'' Our recommendation that this operation is the treatment of choice in drug addicts with medically incurable right-sided endocarditis is supported by our results. In recommending an operation for patients with rightsided endocarditis, it is important to consider whether or not they are addicted to drugs. The fact that 49% of the patients in this series returned to the use of drugs after they were cured of their endocarditis indicates that drug dependency is a serious lethal disease. This condition is also seen clearly in the patient whose case is presented in detail in this article. The long-term follow-up of these patients indicates that almost half of the drug addicts
The JournaL of Thoracic and Cardiovascular
920 Arbulu, Holmes, Asfaw
Cases 55 53 Deaths
Surgery
52
51
49
48
45
43
1
1
2
1
3
2
2
Probability of Survival %
42
1 (1) (1) TVI TVI
25
13
1
1
11
832
100 81A
94.6
89.2
80
18A
74.6
82.0
69.6
16.5
12.9
60
63.B
40 20
o
o
5
8
Days
11
6
Wks
6
9
12
Months
2
4
6
8
10
13
14
16
18
20
Years Post Operation
Fig. 1. Actuarialsurvival curveshowing survival of patients 20 yearsafter tricuspidvalvulectomy without prosthetic replacement TV!, Tricuspid valve replacement
return to the use of drugs after they are cured of their heart infection. The drug addiction is the primary disease. The endocarditis is an intercurrent illness and its management must be planned accordingly. The application of therapeutic measures that are excellent in the management of patients not addicted to drugs, such as prosthetic heart valve replacement at the initial operation or the strict recommendation of reconstructive valvular surgery, appears unrealistic. In our experience, all of the patients had extensive involvement of the entire tricuspid valve by the infection. Perhaps this was related to the protracted ineffective use of intravenous antibiotics. In this series of patients, we were not able to save or repair any portion of the tricuspid valve. We think that these conservative operations are applicable to a small number of patients with tricuspid valve endocarditis with limited involvement of the valve apparatus. The anatomic characteristics and the extent of the infection are the most important determinants in the application of these conservative operations. 11, 12 The poor result after the late tricuspid valve insertion in our patient whose case is discussed in detail is worth further discussion. The first clinical course of this patient supports our contention that for drug addicts tricuspid valvulectomy without replacement is the indicated operation. This patient continued to use intravenous drugs for
more than 6 years after his valvulectomy. The absence of a foreign body in the tricuspid position prevented the development of recurrent right-sided endocarditis. It took 6 years for this patient to be cured of his drug' addiction. The development of intractable right-sided heart failure late in his clinical course suggests that there may have been changes in the pulmonary vascular circulation, such as an increment in the pulmonary resistance resulting from microembolization of the pulmonary capillaries. However, this theory was not supported by the cardiac catheterization findings before his second cardiac operation (Table II). He showed no evidence of pulmonary artery hypertension. Therefore we must conclude that the right ventricular decompensation was directly related to the 8- to lO-year absence of the tricuspid valve. The failure to control the right ventricular failure after insertion of a tricuspid prosthesis indicates that when medically refractory right heart failure develops in a patient who has been without a tricuspid valve for such a long time, the problem cannot be resolved with only insertion of a prosthetic valve. Perhaps, under these conditions, right ventricular function can be improved with the addition of a cardiomyoplastic operation to assist and improve right ventricular contractility. This suggestion is supported by the satisfactory experimental and clinical results reported in the literature when these operations have been used
Volume 102 Number 6 December 1991
to improve the function of severely impaired hearts. 13- 15 The prevalence of complications in this population of extremely ill patients was 22% (Table III). Medical treatment had failed in all of the patients and the operation was recommended as a last resort of treatment. Several patients were moribund at the time of the tricuspid valvulectomy without replacement. Four of these patients died early. In only one of these patients, the one with low cardiac output, was the death related to the absence of the tricuspid valve. In the other three, uncontrollable infection was the cause of death. It is encouraging to observe that among these patients the actuarial survival curve at 20 years is 64% (Fig. I). In calculating these results, we have excluded the two surviving patients who required insertion of a tricuspid prosthesis. This excellent survival rate at 20 years is due to the normal functional characteristics of the heart of each of these patients. The crucial ingredient for the success of this operation is normal function of the left side of the heart. This allows excision of the tricuspid valve without prosthetic valve replacement to be well tolerated for prolonged periods. From the technical viewpoint, tricuspid valvulectomy without replacement is a simple operation. It requires a short run of extracorporeal circulation. The intravenous fluid requirements after the operation are higher. These patients need a high right heart filling pressure during the initial 24 to 48 hours after excision of the tricuspid valve. I The long-term follow-up of these patients showed well-preserved liver function. The only consistent finding was an elevation of the alkaline phosphatase concentration in 18% of the 38 patients who underwent liver studies. This is probably related to the increased pressure in the central lobular vein of the hepatic parenchyma. 16 This experience also showed normal function of the renal and the endocrine systems. On the basis of our experience, we can make several conclusions: I. Drug addiction is a recurrent and lethal disease. 2. only six of 55 patients (11 %) who had tricuspid valvulectomy without replacement or tricuspid valvulectomy and pulmonic valvulectomy without replacement required prosthetic valve insertion at a later date, to control medically refractory right-sided heart failure. 3. tricuspid valvulectomy without replacement or tricuspid valvulectomy and pulmonic valvulectomy without replacement remain the operations of choice in the management of intractable right-sided endocarditis among drug addicts. The actuarial survival curve of 64% at 20 years compares well with long-term results of any cardiac operation. 4. A small percentage of patients without a tricuspid
Tricuspid valvulectomy without replacement
92I
valve will have intractable right-sided heart failure because of marked stretching and decompensation of the right side of the heart. Only one of our patients reached this advanced stage and prosthetic valve insertion in the tricuspid position was of no value. Perhaps under these conditions a cardiomyoplastic operation to improve the contractility of the right ventricle should be considered in addition to insertion of a prosthesis in the tricuspid valve position. We acknowledge the valuableassistanceof Roger F. Higgins, PhD, Director of Higgins Vascular Laboratory, 21701 West Eleven Mile, Southfield,MI 48076,in calculating and outlining the actuarial survival curve in this paper. REFERENCES I. Arbulu A, Asfaw I. Tricuspid valvulectomy without prosthetic replacement: ten years of clinical experience. J THORAC CARDIOVASC SURG 1981;82:684-91. 2. Vander Westhuizen NG, Rose AG. Right-sided valvular infective endocarditis: a clinicopathological study of 29 patients. S Afr Med J 1987;71 :25-7. 3. Kohli RS, Anand IS, Sivaram SV, Datta BN, Bidwai PS, Wahl PL. Isolated right-sidedendocarditis in non-addicts: a review of 10cases seen between 1967-79.Indian Heart J 1982;34:17-20. 4. Valla D, Parientd EA, Degott C, et al. Right-sided endocarditis complicating peritonea-venous shunting for ascites. Arch Intern Med 1983;143:1801-2. 5. Rowley KM, Clubb KS, Walker Smith GJ, Cabin HS. Right-sided infective endocarditis as a consequence of flow-directed pulmonary artery catheterization: a clinicopathologicalstudy of 55 autopsied patients. N Engl J Med 1984;311: 1152-6. 6. Chia BL, Yan PC, Ee BK, Chao MH, Lee CN, Tay MB. Right-sidedinfective endocarditis: a Singapore experience. Am Heart J 1988;116:568-71. 7. Mesa JM, Oliver J, Dominguez F, et al. Endocarditis infecciosa derecha. Experiencia en una serie de 35 pacientes. Rev Esp CardioI1990;43:142-52. 8. Chan P, Ogilby JD, Segal B. Tricuspid valveendocarditis. Am Heart J 1989;117:1140-6. 9. Naidao DP, Naicker S, Vythylingum,et al. Isolated tricuspid valve infective endocarditis: a report of 6 cases. S Afr Med J 1990;78:34-8. 10. Martino P, Micozzi A, Venditti M, et al. Catheter-related right-sidedendocarditisin bone marrow transplant recipients. Rev Infect Dis 1990; 12:250-7. II. Pruett TL, Rotstein on Anderson RW, Simmons RL. Tricuspidvalve Candida endocarditis: successful treatment withvalve-sparing debridement and antifungal chemotherapy in a multiorgan transplant recipient. Am J Med 1986;80: 116-8. 12. Yee ES, Khonsari S. Right-sided infective endocarditis: valvuloplasty, valvectomy or replacement. J Cardiovasc Surg 1989;30:744-8.
The Journatot
922
Arbulu, Holmes, Asfaw
13. Chiu RCJ. Biomechanical cardiac assist: cardiomyoplasty and muscle-powered devices. Mount Kisco, New York: Futura. 14. Koroteyev A, Pochettino A, Niinami H, Stephenson LW. Skeletal muscle: new techniques for treating heart failure. AORN J 1991;53:1005-19. 15. Anderson DR, Pochettino A, Hammond BA, et al. Autogenously lined skeletal muscle ventricles in circulation: up to nine months' experience. J THORAC CARDIOVASC SURG 1991;101:661-70. 16. Kaplan MM. Progress in hepatology: alkaline phosphatase. Gastroenterology 1972;62:452-68.
Discussion Dr. Delos M. Cosgrove (Cleveland, Ohio). Dr. Arbulu, is this the only operation for tricuspid endocarditis that you performed in this period of time? Dr. Arbulu. In this population of patients, Dr. Cosgrove, that was the only operation that we performed. In two patients we excised the pulmonic valve in addition to the tricuspid valve. All
Thoracic and Cardiovascular Surgery
the patients in this series had had unsuccessful medical treatment. Perhaps the extensive infection of the tricuspid valve was related to their prolonged medical management. Dr. Cosgrove. Did you excise the infected portion of the valve in any of your group? Dr. Arbulu. No. Dr. Verdi J. DiSesa (Philadelphia, Pa.). Do you have any data on pulmonary artery pressures in the patients, particularly those who had right-sided heart failure, and could particulate emboli from drug contaminants have created pulmonary hypertension in some of these patients and made the results of the tricuspid valve replacement as poor as you described? Dr. Arbulu. The patient who died 10 years after tricuspid valvulectomy and 2 months after insertion of the prosthesis underwent a complete cardiac catherization, and the right-sided pressures were slightly elevated. The pulmonary pressure was 30mm Hg. Dr. James L. Cox (St. Louis, Mo.). Dr. Arbulu, did you ever have a patient in whom you excised the tricuspid valve and were unable then to wean that patient from bypass because of absence of the tricuspid valve? I noticed that all of your patient deaths occurred after you had weaned them from bypass. Dr. Arbulu. We did not have that experience, Dr. Cox.