| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2003;23:1007-1016
© 2003 Elsevier Science NL
a Department of Cardiovascular Surgery, Hospital Clínico, Institut d'Investigacions Biomediques August Pi i Sunyer (IDIBAPS), University of Barcelona, Barcelona, Spain
b Infectious Diseases Service-Institut Clinic Infeccions Inmunologia (ICII), Hospital Clínic, Institut d'Investigacions Biomediques August Pi i Sunyer (IDIBAPS), University of Barcelona, Barcelona, Spain
c Hospital de Basurto, Bilbao, Spain
Received 27 October 2002; received in revised form 10 March 2003; accepted 11 March 2003.
* Corresponding author. Tel.: +34-93-227-5515; fax: +34-93-451-4898
e-mail: cmestres{at}clinic.ub.es
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Material and methods3. Results4. Discussion5. ConclusionsAppendix A. Conference...References |
|---|
Key Words: Cardiac surgery • Human immunodeficiency virus • Endocarditis • Intravenous drug abusers • Homosexuality • Antiretroviral therapy
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Material and methods3. Results4. Discussion5. ConclusionsAppendix A. Conference...References |
|---|
This population is defined by young age at the time of operation and a majority of patients suffering from acute infective endocarditis (AIE) as a result of parenteral drug addiction which is associated with complications being infectious the most common and AIE the most severe [2]. High surgical mortality could be expected related to preoperative AIE and patient condition associated to intraoperative factors but there are scanty data on late attrition. It seems that the incidence of AIE in HIV-1-infected patients is slightly decreasing [3,4]. There is an increase in the number of patients with other valvular non-infectious conditions and patients with coronary artery disease (CAD) because of lypodistrophy, hyperlipidemia, glucose intolerance and metabolic abnormalities induced by highly-active antiretroviral therapy (HAART) together with other factors common to the general population linked to higher survival of these patients (older age, nutritional profile, lack of exercising, smoking, etc.) [5–7]. Some changes in HIV-1-infected patients requiring cardiac surgery have been noticed.
A handful of papers on HIV-1-infected patients undergoing cardiac surgery produced some data regarding acute mortality but there is lack of information on follow-up [1,8–12]. Major changes in the diagnosis and treatment of HIV-1 infection have been introduced like viral load and CD4 count to monitor the infection, effective HAART regimens and prophylaxis of latent infections [13]. Off-pump coronary artery bypass has emerged as an alternative to surgery with ECC. After 15 years we thought worth reporting the data from a single institution bringing additional light to the late behavior of HIV-infected patients undergoing cardiac surgery.
|
2. Material and methods |
|---|
|
TopAbstract1. Introduction2. Material and methods3. Results4. Discussion5. ConclusionsAppendix A. Conference...References |
|---|
2.2. Definitions
2.2.1. Reporting
Results were reported following the guidelines for reporting morbidity and mortality for cardiac valvular operations as published by Edmunds et al. [15].
2.2.2. Infective endocarditis
The diagnosis of infective endocarditis was made according to the 1994 Durack criteria [16].
2.2.3. HIV infection
HIV infection was accepted if patients had two positive ELISA test or a positive ELISA test plus a positive Western blot test. HIV infection was classified according to the 1993 Center for Disease Control (CDC) recommendations [17].
2.2.4. Type of operation
An operation was considered elective when the patient was admitted following an admission in which the diagnosis and indication were made. It was considered urgent when the condition of the patient deteriorate and surgery was performed during the same admission. It was considered an emergency when the condition of the patient deteriorated and surgery found to be mandatory within 24 h.
2.3. Precautions
Most institutions have adopted common precautions to decrease the risk of exposure of the staff working in the operating room to viral agents. Double gloving, use of goggles and extreme care at the time of handling sharp instruments through specific trays in the surgical field have been incorporated by our team as a routine [18].
2.4. Statistics
Data are presented as actual values and the mean. The survival has been calculated following the actuarial method. The Wilcoxon test has been used for non-parametric estimates.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Material and methods3. Results4. Discussion5. ConclusionsAppendix A. Conference...References |
|---|
|
3.1.2. Sexual status
Five patients (16.12%) were homosexuals and one practising bisexual.
3.1.3. Other factors
One patient (3.22%) acquired HIV-1 infection through hemodialysis. Three patients (9.67%) were heterosexual and had no other risk factors.
3.1.4. Diagnosis of HIV infection
In 17 patients, the HIV-1 infection was diagnosed at the time of the admission in which the operation was performed. In the remaining 14 cases, the duration of HIV-1 infection ranged from 1 to 15 years (mean 5.64).
3.1.5. Immunological status
CD4 is available in 19 patients for 21 operations. Mean preoperative CD4 count was 289 cells/µL (6–600 cells/µL).
3.2. Preoperative diagnosis
Three groups of patients were considered (Table 1).
3.2.1. Acute infective endocarditis
Twenty-one patients (67.74%) had an episode of AIE at the time of admission. In ten cases involved the left valves (eight aortic, one mitral, one aortic+mitral), in six the tricuspid valve and in five it was mixed right and left (two aortic-tricuspid, one aortic-mitral, one mitral-tricuspid).
3.2.2. Non-infective valve disease (NIVD)
This group included five patients (16.12%). Two had rheumatic mitral and aortic disease, one had aortic insufficiency, one tricuspid insufficiency following an episode of Brucella endocarditis 9 years earlier and one had calcific dysfunction of a Liotta porcine mitral bioprosthesis implanted 5 years earlier.
3.2.3. Coronary artery disease (CAD)
Five patients (16.12%) had CAD. Four had unstable angina.
3.3. Preoperative pathology
For the 35 operations, preoperative pathology is listed in Table 2. CAD patients had one-(1), two-(1) and triple-vessel disease (3). Eight cases were reoperations (22.9%). Previous operations included aortic valve replacement (AVR) in five patients (homograft 3, mechanical 1, bioprosthesis 1), mitral valve replacement (MVR) and AVR with bioprostheses in 1, MVR with bioprosthesis in 1 and closure of aorta-right atrial fistula in 1.
|
3.5. Surgical profile
The 31 patients underwent 35 operations. For the whole group, aortic cross-clamping time ranged from 45 to 207 min (mean 76.44) and ECC from 40 to 285 (mean 107.47). Two coronary patients were operated without ECC. Corresponding data and surgical procedures are shown in Table 3.
|
3.5.2. Coronary operations
Three patients received a double, triple and quadruple bypass operation, respectively using the left internal mammary artery for the left anterior coronary artery supplemented with saphenous vein grafts. Mean aortic cross-clamping time was 40 min (20–60) and mean ECC time 55.67 min (40–67). The remaining two patients received an off-pump single and triple bypass operation using the left internal mammary artery for the left anterior descending coronary artery.
3.5.3. Replacement or repair devices
The involved valves were replaced using cryopreserved mitral or aortic homografts, bioprostheses or mechanical valves. Flexible rings were used for repair of the mitral and tricuspid valves in three cases. These devices are listed in Table 3.
3.5.4. Type of operation
Twelve operations (34.28%) were elective, 21 (60%) urgent and two (5.71%) emergencies.
These patients have been followed at the outpatient clinic, by telephone interview or through their referring physicians. Follow-up was closed on July 31st, 2002 and is 100% complete.
3.6. In-hospital mortality and morbidity
Table 4 summarizes morbidity, mortality and length of stay for each group of patients.
|
3.6.2. Cause of early death
In three cases, multiple organ failure was the cause of death. Low cardiac output syndrome, S. enteritidis sepsis, rupture of aortotomy line in a patient with Salmonella and Candida endocarditis and intractable arrhythmia were the cause of deaths in the remaining patients. Postmortem examination was performed in three patients. The one who died from aortotomy line disruption in the ward was reopened on an emergency basis. This desperate surgical manoeuver was considered as a postmortem examination as it confirmed the diagnosis.
3.6.3. Major morbidity
Two patients required reoperation for persistent bleeding (5.71%) one of them requiring delayed sternal closure on the 5th postoperative day. One patient had postoperative bleeding because of coagulopathy not requiring surgical exploration. One patient was reoperated because of paravalvular leak 30 days after the initial operation. One patient had a postoperative cerebrovascular accident from which he recovered (2.85%). Two patients had postoperative pneumothorax treated with chest tube. Two patients required permanent VDD pacing because of complete AV block. Two patients, one coronary and one valve, had persistent fever with negative cultures that eventually subsided. No case of mediastinitis, wound infection nor sternal dehiscence was recorded.
3.7. Length of stay
Postoperative length of stay and overall interval between admission and discharge were calculated. Overall length of stay for the series was 32 days (range 1–88) and postoperative length of stay was 17.62 (4–55). Data are shown in Table 4.
3.8. Follow-up
3.8.1. Antiretroviral therapy
First, an important part of our patients, 17 (58.06%) were not diagnosed before the admission during which surgery was performed and were not on antiretrovirals before the operation. The remaining 14 patients were included in methadone detoxication programmes. All survivors were on antiretrovirals at discharge.
3.8.2. Survival
For the entire series, 1-year survival rate is 64.99±8.19%. After 3.4 years the survival remains stable at 42.25±9.33% extending up to 163.93 months. The mean survival is 77.23±13.84 months (confidence interval 95%, 50.11–104.35) and the median survival is 38.50±10.20 months (confidence interval 95%, 18.50–58.50). The actuarial survival curve is shown in Fig. 1
. The group of CAD patients is five patients and no patient died being the longest follow-up 57 months. The same applies to NIVD patients although one patient in this group is one of the longest survivors. Regarding the AIE group of patients, the 1-year survival is 64.93±10.01%. After 3.4 years the survival remains stable at 35.33±10.60% up to 129.90 months, with a mean survival of 55.35±12.25 months (confidence interval 95%, 31.35–79.36) and a median survival of 27.23±11.62 (confidence interval 95%, 4.46–50.01). For the NIVD patients the follow-up extends up to 163.93 months with a mean survival of 72.19±30.04 (confidence interval 95%, 13.31–30.04) and a median survival of 6.97±1.66 (confidence interval 95%, 3.72–10.22). The actuarial survival curves for these groups are shown in Fig. 2
.
|
|
3.8.4. Progression of the disease
Three patients advanced from CDC class A to C. For the nine patients with available pre- and postoperative CD4 count there has been a significant change. Preoperative CD4 was 185.33 cells/µL and rose up to 396.55 cells/µL during the follow-up (P=0.043).
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Material and methods3. Results4. Discussion5. ConclusionsAppendix A. Conference...References |
|---|
Our experience indicates that there have been some changes over the years with regards to the initial experiences including the Spanish survey. The first has been a steady increase in the number of HIV-1-infected patients sent for surgical treatment. Breaking this 17-year lapse in periods of 5 years it is clear that HIV-1-infected patients are sent for surgery more frequently now than in the beginning of our experience. In the first period only three patients (8.57%) underwent surgical treatment however in the most recent since 1996, 19 patients (54.28%) were operated, more than half of the experience. This could reflect a gradual change in the management of HIV-1-infected patients and an acceptance of them as potential candidates for improvement through a combination of antiretroviral therapy and surgery regardless of their preoperative immune status. In the beginning there was a strong reluctance to accept these immunodepressed patients as there was not enough information to support an aggressive treatment like surgery. In addition, we could expect that the surgical risk was high considering the poor preoperative condition and immune system. Although some contributions tried to establish the indications for surgery [19–21], there was no consensus.
The second change observed has been the modification in the type of disease. Over the first 10 years of our experience, AIE was the only diagnosis. Only one patient had NIVD although he was a PDA. This patient had a dysfunction of a mitral porcine valve which was initially implanted to treat an episode of AIE 5 years earlier. It has been after 1997 when we have operated patients for NIVD and, probably more important, for CAD. We found therefore worth consider different groups. We believe that this may be an extremely important change and it has to be confirmed in the years to come. The association between CAD and antiretroviral therapy has already been established in the literature, especially since the HAART has been introduced. The contributions of Martínez et al. [5] and García-Viejo et al. [22] confirm that in addition to a still relatively poor understanding of the risk factors for lipodystrophy in HIV-1-infected patients treated with HAART containing HIV-1 protease inhibitors, lipodystrophy itself can be considered as a prevalent problem for HVI-1-infected patients [6]. Considering the amount of patients currently on HAART and that HIV infection is considered a chronic disease, the chances of an increased number of them developing CAD in the future are not to be neglected.
4.1. Patients with AIE
These patients have a specific profile, a very young mean age and continued drug addiction which we found in 95.2% of them. Their immune status is usually better than patients suffering from NIVD as they have a long history of PDA and HIV infection was on the average diagnosed earlier allowing to start medical treatment. This is confirmed by a higher CD4 count before the operation that reflects into a less advanced CDC class as 80.95% of class A patients were PDAs. However in CDC classes B and C, PDAs are less than their counterparts from other groups. The still small figures may also influence on the distribution of the patients according to the CDC.
The clinical characteristics of patients in this AIE group are not different from the PDAs in general. The most commonly pathogens in PDAs undergoing surgery are S. aureus and S. viridans and Candida [3,7]. They are infected patients with frequently locally aggressive infection with systemic consequences like sepsis, they need more frequently an urgent or emergency operation. Hospital mortality has been high, 28.57% for the 31 patients and 24% for the 35 operations performed. The accumulated mortality during the follow-up has been higher than in the remaining two groups, reaching up to 50%, however the actuarial survival is 35.33% after 3.4 years and has remained like this up to 129.90 months. For the eight patients who died in this group we found no evidence of any death related to the immunity as five died from complications of continued addiction like overdose and cerebral complications. PDA by itself must not be a contraindication for surgery. Other studies have also shown that cardiac surgery improves the outlook for early and late survival of IVDA with IE in whom surgery is indicated. Mathew et al. [23], studied a cohort of 80 PDAs (the HIV status was not known) who underwent several types of operations for AIE. The probability of survival at 3 and 5 years was 74 and 70%, respectively and these figures are comparable to the general population underwent surgery for AIE. Arbulu et al. [24], in a cohort of 54 PDAs undergoing surgery (the HIV status was not known) for right-sided AIE, found a survival rate of 64% at 22 years. There are obviously some differences between right- and left-sided AIE.
4.2. Patients with NIVD
We started our experience with this type of patients 5 years ago although one of our very early patients had a NIVD in the form of a dysfunctioning mitral bioprosthesis implanted for an initial episode of AIE. They are a group in which different factors are melted together as we have seen a case like the aforementioned one, cases of rheumatic disease [2], degenerative disease [1] and the consequences of a previous episode of AIE 9 years earlier. As we only had five patients we should wait to gain more experience with this particular subgroup.
4.3. Patients with CAD
We had the chance of operating only five patients for coronary bypass and this precludes consistent particular conclusions however our belief is that the medical community has to focus attention on this particular group for the reasons stated earlier in this section. The profile is different from the AIE. In our series, the risk factors for HIV infection were homosexuality or heterosexuality with bisexual practices. Their immune status was worse than the other groups as 60% of them were in CDC class B. Their preoperative CD4 was below 200 cells/µL in 3. Conversely to their counterparts they all received antiretrovirals before the operation. Three patients were operated with ECC and two without ECC. There is almost no experience reported in the literature as only Flum, Tyras and Wallack reported four cases [25]. It is of interest in these patients the probable relationship with lypodystrophy induced by antiretroviral therapy [5,6,22] as they had an abnormal lipid profile. After the operation they all continue on HAART and antiplatelet agents. All patients are alive with a maximum follow-up of 57 months.
Patients have been routinely operated on ECC, however the advent of off-pump coronary artery bypass surgery may eventually change the profile of these patients. According to previously reported information [1] and from our own data ECC does not accelerate the progression of HIV infection. We do not know if off-pump surgery may influence the outcomes of these patients in the future as it is less aggressive than surgery with ECC.
4.4. Limitations of the study
This study has some limitations that preclude at best our ability to draw substantial conclusions. First, its retrospective nature. Second, its small numbers although our series could probably be considered one of the largest single institutional experiences reported until now. Third, the inability to perform controlled studies for obvious reasons. Finally, the changes observed in the population and therapeutic regimes over the years.
4.5. Current institutional policy
After reviewing our experience with this small series, we are now taking all these HIV-1-infected patients for surgery regardless of their CDC classification unless there is an absolute contraindication for medical or surgical reasons. Active drug addiction is not a contraindication for surgery. The surgical procedure is performed by a team familiar with these patients which, from the surgical point of view have no practical differences from non-infected patients. We believe that there are other issues that may influence on possible underdiagnosis of HIV infection, namely confidentiality precluding routine testing as correctly pointed out by Paone and Silverman [12]. Until now we have not denied surgery to any patient based on bioethical issues.
|
5. Conclusions |
|---|
|
TopAbstract1. Introduction2. Material and methods3. Results4. Discussion5. ConclusionsAppendix A. Conference...References |
|---|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
Appendix A. Conference discussion |
|---|
|
TopAbstract1. Introduction2. Material and methods3. Results4. Discussion5. ConclusionsAppendix A. Conference...References |
|---|
I am sure this is due to better retroviral medication, and in view of these results my question is, what is your advice, or the advice of your group, with this better prognosis for HIV patients as far as performing surgery in an AIDS patient?
Dr Mestres: I think you are right. We can't say anything else. We just take these cases in today's days. I think you made a major point because a number of things have changed the scope of treatment of AIDS patients, basically the establishment of immunological markers for detection like CD4 and viral count, and obviously the tremendous changes in antiretroviral therapy, especially with the HAART regimens. What we have seen over the years is that these patients, exception made of those with acute complicated endocarditis and sepsis, are coming in better condition, not only the cardiac patients but the vascular as well, which are not the issue in this presentation.
So I think we have seen a trend towards a better quality of patients, and you see that we are getting more and more patients today, slowly picking up, because people are aware that the survival of the patients with retroviral therapy is going to be better, and definitely it is.
Dr Aris: I have an explanation for you operating on more patients: everybody is sending you the patients in Barcelona for surgery. The second question is what is your advice for the audience in this room. If you have a patient with AIDS, would you operate on him or her, and what is the general rule? We want to walk out of here knowing well if we can say no or we should say yes. That is the important point.
Dr Mestres: As I said, it is a very important question even from an ethical point of view, because at the end what happens is that each and every one of us know perhaps someone around us who got some HIV disease. The question is, are you going to deny treatment to a young person which is an active addict and has a B stage and that maybe could be your son? So I think it is a very important question. So far, unless the patient is actually dying, we are taking these patients to the operating room, obviously considering that there might be some risks for the surgical team. But so far now we don't have evidence not to operate on the patient based just on HIV status.
If you remember, the CDC classification shows that we even had C class patients in the beginning, and one of them was a patient with AIDS that is still alive I think it is 8 or 9 years after surgery.
Dr C. Yankah (Berlin, Germany): We are studying also a small number of patients, about 18, with HIV positive and some with AIDS. In our Institution like yours we do not refuse surgery for patients with HIV positive or AIDS. My question is related to your assessment of the risk factor for perioperative mortality with regard to viral load and CD4 count and when do you start the postoperative retroviral drug medication of these patients?
My second question is related to the selection of the type of valves you used for those patients with endocarditis. Being aware that their life expectancy is limited and they are non-compliant to medication we prefer biological valves for these patients. What is your surgical strategy for these patients and what is your biological valve of choice for endocarditis with annular abscess in such moribund patients?
Dr Mestres: As I said, there are a number of limitations in a retrospective study like this, covering almost 17 years. As far as your second question is concerned, the type of valve is left to the discretion of the surgeon, although in recent years just a handful of us have been personally taking care of those patients. We are basically using today biological tissue and most of the patients implanted with mechanical valves belong to the early experience. We are using a bioprosthesis unless there is annular destruction. In these cases we try to use homografts in the aortic position, and sometimes in the mitral as well, as you have seen.
Regarding immunological marking before the operation, obviously in today's days, and only over the past 5 years and a half, we had viral load and CD4 count to be sure that the patient is viable from an immunological point of view, and the retroviral therapy is started immediately after the operation if the patient has not had previous treatment, as it happens, because most of the time the diagnosis of HIV has been made at the admission.
|
References |
|---|
|
TopAbstract1. Introduction2. Material and methods3. Results4. Discussion5. ConclusionsAppendix A. Conference...References |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
