Late-Onset Driveline Infections: The Achilles Heel of Prolonged Left Ventricular Assist Device Support

Transcription

1 Late-Onset Driveline s: The Achilles Heel of Prolonged Left Ventricular Assist Device Support Andreas Zierer, MD, Spencer J. Melby, MD, Rochus K. Voeller, MD, Tracey J. Guthrie, RN, Gregory A. Ewald, MD, Kim Shelton, RN, Michael K. Pasque, MD, Marc R. Moon, MD, Ralph J. Damiano, Jr, MD, and Nader Moazami, MD Divisions of Cardiothoracic Surgery and Cardiology, Washington University School of Medicine, Barnes-Jewish Hospital, St. Louis, Missouri Background. A successful left ventricular assist device (LVAD) long-term support in an outpatient setting demands that device-related complications are reduced to a minimum. We hypothesized that late onset driveline infections have serious implications on the anticipated application of LVAD as permanent therapy. Methods. Between 1996 and 2005, 73 patients were implanted with the Novacor (World Heart Corp, Ottawa, Ontario, Canada; n 35) or the HeartMate (Thoratec Corp, Pleasanton, CA; n 38) as either bridge to transplantation (n 44) or destination therapy (n 29). Our analysis focused on patients with late-onset infection (>30 days) of the driveline exit site with prior clinical healing of all incisions. Results. Late driveline infections developed in 17 patients (23%) at a median of 158 days (intraquartile range [IQR]: 68 to 213 days) after implantation. The median duration of support in this subgroup was 400 days (IQR, 283 to 849 days). Despite an aggressive treatment algorithm, repeat surgical revision was needed in 12 patients, up to six times in 2 individuals. In 6 patients, the infection progressed to pump pocket infections that led to urgent heart transplantation (n 4) or explantation (n 2). The individual risk that a driveline infection would develop dramatically increased with the duration of support, reaching 94% at 1 year. Multivariate analysis identified duration of support (p < 0.001) and documented trauma at the driveline exit site (p < 0.001) as independent predictors of infection. Number and duration of readmissions to the hospital significantly increased (p < 0.001), and long-term follow-up for survival ( years, 100% complete) showed a trend towards impaired outcome after driveline infection (5-year survival: 41% versus 70%, p 0.10). Conclusions. Long-term LVAD support in the current series was jeopardized by late-onset driveline infections, which occurred in all patients with support duration longer than 1 year. Once driveline infections developed, they were difficult to control and significantly increased morbidity. (Ann Thorac Surg 2007;84:515 21) 2007 by The Society of Thoracic Surgeons Heart failure remains a leading cause of death in Western countries, affecting approximately 5 million individuals and causing more than 50,000 deaths per year in the United States alone [1]. An estimated 15,000 patients with end-stage heart failure could potentially benefit from cardiac transplantation each year. Yet, only 2500 hearts are donated annually, and approximately one third of patients awaiting heart transplantation die each year [2, 3]. Mechanical circulatory support, primarily in the form of left ventricular assist devices (LVAD), has become an attractive treatment option for many patients with severe congestive heart failure. The successful clinical introduction of LVAD has resulted in a shift in the management of end-stage heart failure, one that will undoubtedly Accepted for publication March 27, Address correspondence to Dr Moazami, Division of Cardiothoracic Surgery, Washington University School of Medicine, 660 S Euclid Ave, Campus Box 8234, St. Louis, MO 62236; moazamin@wustl.edu. affect and include a vast proportion of future patients. The use of implantable LVADs is also associated with possible complications, however, such as hemorrhage, thromboembolism, device failure, arrhythmias, organ system failure, and infection [4 6]. New technology used in LVAD successfully addressed some of the previous clinical problems, and the development of portable, vented, electric LVADs facilitated the transfer of patient care into the outpatient setting [7 9]. Owing to the shortage of available hearts for transplantation and the technical improvements in the currently available devices, centers have looked to LVAD as longterm therapy to entirely bypass cardiac transplantation [5]. The use of LVAD as a long-term or even permanent therapy demands that the risk of device-related complications, including infection, be reduced such that a reasonable patient self-care protocol is possible. Late-onset driveline infections may be a particularly difficult obstacle to overcome. The mechanisms by which 2007 by The Society of Thoracic Surgeons /07/$32.00 Published by Elsevier Inc doi: /j.athoracsur

2 516 ZIERER ET AL Ann Thorac Surg LATE LVAD DRIVE-LINE INFECTIONS 2007;84: these infections typically occur have previously been described by our unit [10]. However, despite a brief mention in the Randomized Evaluation Of Mechanical Assistance Therapy As An Alternative In Congestive Heart Failure (REMATCH) trial [11], the true impact of late-onset driveline infections and its implication on the anticipated application of LVAD as permanent therapy remain ill defined. Therefore, the purpose of the current investigation was to address this specific issue by evaluating its incidence, cause, and clinical course during a 10-year period at our institution. Material and Methods This retrospective review includes 73 consecutive patients between October 1996 and July 2005 who were implanted at Washington University School of Medicine (Barnes-Jewish Hospital) with the Novacor (World Heart Corporation, Ottawa, Ontario Canada; n 35) or the HeartMate (Thoratec Corporation, Pleasanton, CA; n 38) LVAD as either bridge to transplantation (n 44) or destination therapy (n 29). The study was approved by the Institutional Review Board, and informed consent and permission for the release of information were obtained from each patient. There were 47 men (64%) and 26 women (36%) with a mean age of years at the time of implantation. Among other important clinical characteristics, traumatic Table 1. Demographic Characteristics of Study Population Characteristic a No Late (Group A, n 56) Late (Group B, n 17) p Value Age (years) Gender (male) 44 (78.6) 11 (64.7) 0.34 History of tobacco use 24 (42.9) 10 (58.8) 0.28 COPD 12 (21.4) 3 (17.6) 1.00 Diabetes 17 (30.4) 3 (17.6) 0.37 Chronic renal insufficiency 10 (17.9) 3 (17.6) 1.00 Peripheral vascular disease 6 (10.7) 4 (23.5) 0.23 Ischemic etiology of heart 24 (42.9) 6 (35.3) 0.78 failure Previous cardiac surgery 11 (19.6) 2 (11.8) 0.72 Type of LVAD device 0.15 Novacor b 24 (42.9) 11 (64.7) HeartMate c 32 (57.1) 6 (35.3) Indication for LVAD 0.58 Destination therapy 21 (37.5) 8 (47.1) Bridge to heart transplant 35 (62.5) 9 (52.9) Hospital stay immediately before LVAD (IRQ) 12 (6 23) 14 (4 30) 0.42 a Categoric variables listed as n (%); mean standard deviation for normally distributed continuous variables or median with intra-quartile range (IQR) listed for nonnormally distributed continuous variables where appropriate. b World Heart Corporation, Ottawa, Ontario, Canada. c Thoratec Corporation, Pleasanton, California. COPD chronic obstructive pulmonary disease; LVAD left ventricular assist device. events to the driveline exit site were prospectively collected in our LVAD database from each patient during routine follow-up visits and whenever patients presented at our institution with device-related complications. Patient demographics are summarized in Table 1. Smoking history, chronic obstructive pulmonary disease (COPD), diabetes mellitus, chronic renal insufficiency, and peripheral vascular disease were the most common preoperative comorbidities. All patients undergoing LVAD insertion received a standardized antimicrobial prophylaxis consisting of intravenous vancomycin, aztreonam, and fluconazole, which was initiated during the operation and continued for 3 days after the procedure. The cutaneous driveline exit site was created by making a small stab incision in the skin approximately 75% of the diameter of the driveline. The skin was then gently stretched with a clamp to make a sufficient opening for the passage of the driveline. In our experience, this maneuver has helped establish a tight seal around the driveline. We have not found skin excision to be necessary for creation of the driveline exit site. Immobilization of the percutaneous lead of the LVAD occurred immediately after device placement and before the patient was transferred from the cardiothoracic operating room. Initially the lead was secured by using a Hollister Drain/Tube Attachment Device (Hollister Global, Melbourne, Australia). If the implanted device used a filter attachment, the drain/tube attachment device was placed distal to this attachment. When it was determined that the patient was hemodynamically stable and turning could be tolerated in the immediate postoperative phase, an abdominal binder was then applied. Hospital staff and the supported LVAD recipient were aggressively taught about the importance of wearing both the abdominal binder and drain/tube attachment device at all times and to ensure proper placement. During in-hospital treatment, all driveline exit sites underwent daily antiseptic cleaning with hydrogen peroxide and povidone-iodine (Betadine, Purdue Pharma, Stanford, CT) solutions and were dressed with sterile gauze according to our standardized protocol. Data Analysis Our analysis focused on patients with late-onset infection of the driveline exit site and their comparison with patients who did not present with this complication. Late onset driveline infections ( 30 days after LVAD implantation with prior clinical healing of all incisions) were defined as erythema, drainage, or purulence at the driveline exit site in the presence of a positive culture result. Operative mortality included any death that occurred during the initial hospitalization or within 30 days after operation for discharged patients. Cumulative survival rates were calculated using Kaplan-Meier analysis, and survival curves were compared using the log-rank test. Mean follow up was years and was 100% complete. The cumulative risk of developing a driveline infection was calculated using hazard analysis. Continuous data are reported as mean one standard deviation (SD), or

3 Ann Thorac Surg ZIERER ET AL 2007;84: LATE LVAD DRIVE-LINE INFECTIONS 517 median with intraquartile range (IQR), if appropriate, and compared using the Student t test. Categoric variables were analyzed using the 2 test or the Fisher exact tests, as appropriate. Odds ratios (OR) are reported with 95% confidence intervals (CI). Multivariate analysis (stepwise backward regression) was used to determine risk factors that were significant, independent predictors of late onset driveline infections (SigmaStat 2.03, SPSS Inc, Chicago, IL). The 21 variables analyzed were age, gender, body mass index (BMI), history of tobacco use, COPD, diabetes mellitus, chronic renal insufficiency, peripheral vascular disease, ischemic etiology of heart failure, previous cardiac surgery, preoperative intraaortic balloon pump (IAPB), indication for LVAD placement, length of hospital stay immediately before LVAD placement, year of implant, type of LVAD implanted (Novacor versus HeartMate), reexploration for bleeding, postoperative hemodialysis, length of mechanical ventilation, length of stay at the intensive care unit (ICU), length of LVAD support, and history of documented trauma to the driveline exit site. These variables were initially screened by performing univariate logistic regression on all possible explanatory variables with the dependent variable of late onset driveline infection. This was screened to include any variable with a value of p 0.20 for further examination. The variables that remained were then subjected to a crosstab analysis to look for any strong correlations between Table 2. Outcome After Left Ventricular Assist Device Placement Outcome a No Late (Group A, n 56) Late (Group B, n 17) p Value Reexploration for bleeding 8 (14.2) 3 (17.6) 0.83 Post-op hemodialysis 10 (17.9) 4 (23.5) 0.73 Length of mechanical ventilation, hours 65 (20 167) 38 (19 190) 0.71 Length of stay in ICU, hours 240 ( ) 198 ( ) 0.93 Operative mortality 13 (23.2) 0 NA LVAD support, days 35 (10 144) 400 ( ) LVAD support (days), excluding hospital mortality 86 (14 281) 400 ( ) Documented driveline 2 (3.6) 12 (70.6) trauma Number of readmissions/ patient Duration of readmissions/ patient, days Days readmitted/month/ patient a Categoric variables listed as n (%); mean standard deviation for normally distributed continuous variables or median with intraquartile range (IQR) listed for nonnormally distributed continuous variables where appropriate. ICU intensive care unit; LVAD left ventricular assist device; NA not applicable. Fig 1. Cumulative hazard of developing a late-onset driveline infection with increased duration of mechanical support. The numbers of patients at risk at 200, 400, and 600 days of left ventricular assist device (LVAD) support are indicated. pairs of variables, and none were found. These variables were then forced into a logistic regression model and this model was bootstrapped with 1000 repetitions. The resulting bootstrap regression models were aggregated by counting the number of models in which each potential explanatory variable was found to be significant with a p Results Overall, operative mortality in our series was 18% (13/73). Late localized exit site infections developed in 23% of patients (17/73) and were equally distributed among patients who were implanted with the Novacor or the HeartMate LVAD (p 0.15). No preoperative clinical characteristics that might put the patients at a higher risk of late-onset driveline infections could be identified. Similarly, postoperative morbidities, including reexploration for bleeding (p 0.83), need for hemodialysis (p 0.73), ventilation time (p 0.71), and ICU stay (p 0.93) were equally distributed between patients without (group A) and with (group B) late infection (Table 2). Median duration of LVAD support was 35 days (IQR, 20 to 167 days) for group A and 400 days (IQR, 283 to 849 days) for group B (p 0.001). Support duration exceeded 1 year in 10 group B patients. Excluding operative mortality from group A, median duration of support was 86 days (IQR, 14 to 281 days; p versus B). Within group A, 3 patients presented with early ( 30 days after implantation) onset infections at the driveline exit site (n 2) or the pump pocket (n 1). In 71% patients (12/17) of group B, the infection could be attributed to a documented trauma to the driveline exit site. The median onset of late infection was 158 days (IQR, 68 to 213 days). Multivariate analysis identified duration of support (p 0.001) and documented trauma at the driveline exit site (p 0.001) as the only independent predictors of late infection. The cumulative hazard of developing a driveline infection dramatically increased with the duration of support (Fig 1). Within 1 year of mechanical support,

4 518 ZIERER ET AL Ann Thorac Surg LATE LVAD DRIVE-LINE INFECTIONS 2007;84: Fig 4. Distribution of late driveline infections by year of left ventricular assist device (LVAD) insertion. Patients without late driveline infections (grey bars) were compared with patients who presented with this complication (black bars). Fig 2. Kaplan-Meier survival after left ventricular assist device (LVAD) insertion conditional on surviving 30 days in patients who did (gray line) and did not (black line) have late driveline infection. The numbers of patients at risk at 1, 3, and 5 years are indicated. Fig 3. Kaplan-Meier survival of left ventricular assist device patients after subsequent heart transplantation in those who did (gray line) and did not (black line) have late driveline infection. The numbers of patients at risk at 1, 3, and 5 years are indicated. the individual risk of a late driveline infection developing was 94%. All patients who were readmitted to the hospital because of late localized infections underwent a standardized aggressive treatment algorithm, which included a course of specific intravenous antibiotics tailored to the individual culture results, total immobilization of the driveline exit site, and a nuclear medicine-labeled white blood cell scan to predict the tract anatomy and the necessity and extent of a surgical approach. Organisms identified included Staphylococcus aureus in 8 patients, S epidermidis in 6, Corynebacterium jejunii in 5, Enterobacter faecalis in 4, and Klebsiella pneumoniae in 4; some patients had multiple positive results. Excisional débridement and wide drainage were necessary in 14 patients. The surgical goal was to excise the exit site back to a place on the driveline where the tissues were clearly adherent circumferentially. Despite this aggressive approach, repeat revision was needed in 12 of the 14 patients, up to six times in 2 individuals. Despite surgical débridement, the infection progressed to the pump pocket in 6 patients, which led to urgent heart transplantation in 4 or explantation in 2. Permanent healing of the infected exit site could only be achieved in 3 patients. This complicated clinical course is also reflected in a significant increase in the number of readmissions to the hospital and length of stay (p versus group A; Table 2). The presence of late driveline infections did not influence the frequency of subsequent heart transplantation. In group B, 53% (9/17) underwent cardiac transplantation compared with 63% of patients (35/56) in group A (p 0.58). There was a trend for overall long-term survival (Fig 2) and survival after subsequent heart transplantation (Fig 3) to be diminished with late onset infections. Late driveline infections during this 10-year period in our series were equally distributed, without significant improvement in more recent years (Fig 4). Comment In its early days, LVAD support was designed for an inpatient setting, and the average duration of mechanical support was rather short. Accordingly, infections that occurred early in the course of LVAD therapy built the focus of interest. Antimicrobial coatings with chlorhexidine and silver sulfadiazine have been shown to reduce central venous catheter colonization and were subsequently used to impregnate LVAD drivelines [12, 13]. This helped to prevent early infections and facilitated initial tissue in-growth with success [14]. Decreasing the number of transcutaneous accesses by the use of larger single-lead drivelines was another attempt to reduce the risk of infections [15]. Researchers also paid special attention to the systemic nonspecific immunoregulatory dysfunction caused by LVAD materials, which may in turn contribute to an increased susceptibility to local infections [16]. With these advances in the field, some of the problems with early infections were successfully addressed. The achieved improvement was confirmed by our data, which

5 Ann Thorac Surg ZIERER ET AL 2007;84: LATE LVAD DRIVE-LINE INFECTIONS 519 showed that only 15% of local infections (3/20) occurred within the first 30 days after device implantation. More recently, there was a paradigm shift towards an outpatient setting, and late-onset infections came to the forefront of interest. In the current series, we therefore focused on late driveline infections, which turned out to be a major cause of morbidity. Furthermore, late driveline infections were associated with a trend towards diminished long-term survival. Their incidence significantly increased with the duration of support. Not a single patient had a support duration of more than one year without presenting with this complication. For the individual patient presenting with late driveline infections, the number of days readmitted to the hospital per month of mechanical circulatory support increased ninefold compared with patients from group A. Despite recent technical and medical improvements, the initiating mechanism for these driveline infections was most frequently a rather simple mechanical problem: a documented trauma to the driveline exit site. We therefore propose that driveline trauma remains the most common initiator of late driveline exit site infections. In our experience, this trauma usually takes the form of a shearing traction or torsion injury, commonly initiated by events such as dropping the controller and battery pack, moving without picking up the controller and battery pack, or accidentally hooking the driveline on a passing object. Because the goal of placing these devices is to allow patient mobility and the chance to live as normal a life as possible, all such trauma to the driveline exit site can certainly not be avoided. Another important finding of our study was that 71% of patients from group B needed repeated surgical revisions, consistent with a very challenging clinical course. The clinical course of LVAD related infections has been previously investigated by other authors. Argenziano and coworkers [18] reported that pump-pocket infections were best managed by device explantation or transplantation. Prendergast and associates [19] confirmed these findings by showing that transplantation in the face of LVAD infection was an effective treatment option. In their series, at 17 9 months after cardiac transplantation, 80% of patients (8/10) without previous LAVD infection were alive compared with 88% of patients (7/8) who underwent transplantation in the setting of a devicerelated infection. In contrast, current follow-up of patients undergoing subsequent heart transplantation in the presence of a late driveline infection did show a trend towards impaired outcome (5-year survival 36% versus 85%). Yet, the number of patients in group B was probably too small to show a statistical significance. In addition to important differences to the current series, Prendergast and colleagues [19] acknowledged in their report that patients with infections at time of transplantation often were in less stable condition postoperatively. We concur in that highurgency heart transplantation and emergency device explantation are both extreme measures and can certainly not be considered a standard therapeutic approach to treat device-related infections. Axial flow pumps have been introduced into clinical practice and have since gained increased interest and acceptance, at least for medium-term support. So far, more than 200 MicroMed DeBakey VADs (MicroMed Cardiovascular, Houston, TX) and more than 50 Jarvik 2000 axial flow pumps (Jarvik Heart Inc, New York, NY) have been implanted worldwide. Hetzer and colleagues [20] reported their initial clinical experience with a new magnetically suspended axial flow LVAD in 24 patients. The longest individual time of support was 12.6 months, which is shorter than the median duration of support in group B of our series. It is still impressive that there was not a single case of driveline infection. These excellent results coincided well with the experience from Siegenthaler and associates with the Jarvik 2000 LVAD [21]. In this report the incidence of local infection was shown to be significantly lower when compared with the HeartMate LVAD. In a different report of 11 patients who received a DeBakey ventricular assist device axial flow pump for bridge to transplantation, no device, pocket, or driveline infections occurred [22]. Mean duration of support was again rather short (51 49 days, range 11 days to 4.7 months). Given these substantial differences in the time of support, we should not draw any final conclusions from these preliminary data. However, the incidence of infection in all these series was drastically lower than that observed with pulsatile pumps in the current series or in the REMATCH trial [10]. Potential advantages of axial flow pumps include the smaller diameter and the higher flexibility of the percutaneous driveline. Additional experiences with larger numbers of patients and longer duration of assist are warranted to judge the role of axial flow pumps not only as bridge to transplant but possibly destination therapy. In summary, this investigation showed that duration of LVAD support could be limited by late-onset driveline infections most commonly initiated by traumatic events. The cumulative hazard of developing this complication dramatically increased over time and reached almost 100% at 1 year of circulatory support. Furthermore, current data suggest that once late-driveline infections occurred, they were extremely difficult to control, despite our ongoing attempts to precociously treat this complication. Consequently, during the last decade, our hospital staff and the supported LVAD recipients were aggressively taught about the importance of avoiding any activities that might expose the driveline to any abnormal mechanical stress. From these data and the anticipated use of LVAD as permanent therapy, we also believe that it is imperative that we reinforce to manufacturers of available pulsatile assist devices the importance of continuously improving the design of extracorporal device accessories to minimize torsion and further protect the driveline. Dr Zierer was supported by a DFG-Research Fellowship of the German Research Foundation.

6 520 ZIERER ET AL Ann Thorac Surg LATE LVAD DRIVE-LINE INFECTIONS 2007;84: References 1. Thom T, Haase N, Rosamond W, et al. Heart disease and stroke statistics-2006 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation 2006;113:e Stevenson LW. When is heart failure a surgical disease? In: Rose EA, Stevenson LW, eds. Management of end-stage heart disease. 1st ed. Philadelphia, PA: Lipincott-Raven; 1998: Goldstein DJ, Rose EA. Cardiac allotransplantation. In: Rose EA, Stevenson LW, eds. Management of end-stage heart disease. 1st ed. Philadelphia, PA: Lipincott-Raven; 1998: Kormos RL, Borovetz HS, Armitage JM, Hardesty RL, Marrone GC, Griffith BP. Evolving experience with mechanical circulatory support. Ann Surg 1991;214: Myers TJ, McGee MG, Zeluff BJ, Radovancevic B, Frazier OH. Frequency and significance of infections in patients receiving prolonged LVAD support. ASAIO 1991;37:M Petri WA. s in heart transplant recipients. Clin Infect Dis 1994;18: Catanese KA, Goldstein DJ, Williams DL, et al. Outpatient left ventricular assist device support: a destination rather than a bridge. Ann Thorac Surg 1996;62: Frazier OH. Outpatient LVAD: its time has arrived. Ann Thorac Surg 1994;58: DeRose JJ, Umana JP, Argenziano M, et al. Implantable left ventricular assist devices provide an excellent outpatient bridge to transplantation and recovery. J Am Coll Cardiol 1997;30: Pasque MK, Hanselman T, Shelton K, et al. Surgical management of Novacor drive-line exit site infections. Ann Thorac Surg 2002;74: Rose EA, Gelijns AC, Moskowitz AJ, et al. Longterm mechanical left ventricular assistance for end-stage heart failure. N Engl J Med 2001;345: Maki DG, Stolz SM, Wheeler S, Mermel LA. Prevention of central venous catheter-related bloodstream infection by use of an antiseptic-impregnated catheter. A randomized, controlled trial. Ann Intern Med 1997;127: Greenfeld JI, Sampath L, Popilskis SJ, Brunnert SR, Stylianos S, Modak S. Decreased bacterial adherence and biofilm formation on chlorhexidine and silver sulfadiazineimpregnated central venous catheters implanted in swine. Crit Care Med 1995;23: Choi L, Choudhri AF, Pillarisetty VG, et al. Development of an infection-resistant LVAD drive-line: a novel approach to the prevention of device-related infections. J Heart Lung Transplant 1999;18: Sun BC, Catanese KA, Spanier TB, et al. 100 long-term implantable left ventricular assist devices: the Columbia Presbyterian interim experience. Ann Thorac Surg 1999;68: Gristina AG, Giridhar G, Gabriel BL, Naylor PT, Myrvik QN. Cell biology and molecular mechanisms in artificial device infections. Int J Artif Organs 1993;16: Sinha P, Chen JM, Flannery M, Scully BE, Oz MC. Edwards NM. s during left ventricular assist device support do not affect posttransplant outcomes. Circulation 2000;102: III Argenziano M, Catanese KA, Moazami N, et al. The influence of infection on survival and successful transplantation in patients with left ventricular assist devices. J Heart Lung Transplant 1997;16: Prendergast TW, Todd BA, Beyer AJ, et al. Management of left ventricular assist device infection with heart transplantation. Ann Thorac Surg 1997;64: Hetzer R, Weng Y, Potapov EV, et al. First experiences with a novel magnetically suspended axial flow left ventricular assist device. Eur J Cardiothoracic Surg 2004;25: Siegenthaler MP, Martin J, Pernice K, et al. The Jarvik 2000 is associated with less infections than the HeartMate left ventricular assist device. Eur J Cardiothorac Surg 2003;23: Vitali E, Lanfranconi M, Ribera E, et al. Successful experience in bridging patients to heart transplantation with the MicroMed DeBakey ventricular assist device. Ann Thorac Surg 2003;75: INVITED COMMENTARY has been an ongoing major concern in patients supported by implantable ventricular assist devices (VADs). Numerous reports have been published during the last 3 decades documenting infection rates that may include septicemias and nondevice-related infections, as well as more surgically specific infections, including local wound, drive-line, and pocket infections. In most studies describing the general experiences with VADs, it is often difficult to tease out the importance of drive-line infections. The article by Zierer and colleagues [1] from Washington University focuses on late onset drive-line infections that are relatively prevalent in the long-term device patients and very difficult to treat. Although efforts are made to optimize surgical technique for the creation of the percutaneous drive-line, its postoperative care, and immobilization, the authors provide evidence showing that a majority of drive-line infections are preceded by trauma in the form of traction or torsion of the drive-line. How this can be prevented remains a problem because carrying and manipulating heavy batteries and controllers will eventually result in an inadvertent drop or fumble, which leads to stress at the percutaneous attachment site. The only real solution seems to be the creation of a drive-line that is thinner, more flexible, or elastic, and could decrease the stress at the percutaneous site, or the use of transcutaneous energy transmission and communication. A change in drive-line for the current generation devices might be difficult due to regulatory and materials issues; however axial flow devices are smaller in size and have thinner, more flexible drive-lines. There is some indication that these differences may be important with recently reported incidences of drive-line infection being less in the Jarvik 2000 (Jarvik Heart Inc, New York, NY) devices compared with HeartMate I (Thoratec Corp, Pleasanton, CA) in patients generally supported for less than 1 year [2]. Similarly, the HeartMate II (Thoratec Corp) and debakey MicroMed (MicroMed Cardiovascular Inc, Houston, TX) devices have been shown to have lower rates of wound and pocket infections, but with a nonstatistically significant lower rate of drive-line infections [3]. The move to a transcutaneous power and communications system has many obstacles with a significant number of engineering issues to solve regarding methods of power transmission, interference between transcutane by The Society of Thoracic Surgeons /07/$32.00 Published by Elsevier Inc doi: /j.athoracsur