- Email: [email protected]
Reduction in driveline infection rates: Results from the HeartMate II Multicenter Driveline Silicone Skin Interface (SSI) Registry
http://www.jhltonline.org
ORIGINAL CLINICAL SCIENCE
Reduction in driveline infection rates: Results from the HeartMate II Multicenter Driveline Silicone Skin Interface (SSI) Registry David Dean, MD,a Faouzi Kallel, PhD,b Gregory A. Ewald, MD,c Antony Tatooles, MD,d Brett C. Sheridan, MD,e Robert J. Brewer, MD,f Christian Caldeira, MD,g David J. Farrar, PhD,b and Shahab A. Akhter, MD,h on Behalf of the SSI Registry Investigators From the aPiedmont Healthcare, Atlanta, Georgia; bThoratec Corporation, Pleasanton, California; cWashington University School of Medicine, St Louis, Missouri; dAdvocate Christ Hospital Medical Center, Oak Lawn, Illinois; e University of North Carolina, Chapel Hill, North Carolina; fHenry Ford Hospital, Detroit, Michigan; gTampa General Hospital, Tampa, Florida; and the hUniversity of Wisconsin, Madison, Wisconsin.
KEYWORDS: MCS; VAD; HeartMate II; driveline infection; heart failure
BACKGROUND: During left ventricular assist device implantation, a surgical tunneling technique to keep the entire driveline (DL) velour portion in the subcutaneous tunnel, resulting in a silicone-skin interface (SSI) at the exit site, has been adopted by many centers. To assess long-term freedom from DL infection associated with this technique, a multicenter SSI registry was initiated. It was hypothesized that the modified tunneling technique is associated with at least 50% reduction in DL infection at 1 year post-implant compared with the velour-to-skin method used in the HeartMate II (HMII) Destination Therapy (DT) trial. METHODS: SSI is a retrospective and prospective registry of patients who have received the HMII device. Results are reported from the retrospective cohort, which consists of 200 patients who were implanted during the period 2009–2012 with the SSI tunneling method and on HMII support for at least 10 months at the time of enrollment. The prevalence and incidence of DL infection after left ventricular assist device implantation in the SSI retrospective cohort were determined and compared with a control group of 201 patients also on HMII support for at least 10 months from the HMII DT clinical trial who were implanted during the period 2007–2009 using the traditional method in which a small section of the velour portion of the DL was externalized. RESULTS: The 1-year and 2-year prevalence rates of DL infection were 9% and 19% in the SSI patient group compared with 23% and 35% in the control group (hazard ratio 0.49, 95% confidence interval 0.33– 0.73, p o 0.001). The event-per-patient year was 0.11 and 0.22 for the SSI and control groups, respectively (p o 0.001). Based on a multivariate analysis, age and DL exit side were the only independent variables associated with DL infection. Effects of management changes over the eras were not studied and could have contributed to the findings. CONCLUSIONS: These results suggest that leaving the entire DL velour portion below the skin is associated with 50% reduction in DL infection compared with results from the HMII DT trial. J Heart Lung Transplant ]]]];]:]]]–]]] r 2015 International Society for Heart and Lung Transplantation. All rights reserved.
Reprint requests: David Dean, MD, Piedmont Healthcare, 95 Collier Road NW, Suite 5015, Atlanta, GA 30309. Telephone: 404-605-5699. Fax: 404-355-4235. E-mail address: [email protected] 1053-2498/$ - see front matter r 2015 International Society for Heart and Lung Transplantation. All rights reserved. http://dx.doi.org/10.1016/j.healun.2014.11.021
2
The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]
In the United States, 45 million people have heart failure, which accounts for 41 million hospitalizations and 4280,000 deaths each year.1 The HeartMate II (HMII) left ventricular assist device (LVAD)/pump was approved by the U.S. Food and Drug Administration more recently for use as bridge to transplantation (BTT) and destination therapy (DT) for treatment of patients with end-stage heart failure. The DT patients are generally not eligible for heart transplantation and require LVAD support for an indefinite duration. Although the HMII has provided a significant survival benefit for patients with end-stage heart failure, serious adverse events continue to be a major cause of morbidity.2 Device-related infections are among the most commonly occurring serious adverse events, with reported incidences of 0.34 events per patient-year (19%) in BTT patients and 0.48 events per patient-year (35%) in DT patients.3,4 They are also a leading cause of rehospitalization, which increases the overall costs associated with LVAD therapy. The 3 main LVAD components that are at risk for infections are the driveline (DL), the pump and its surrounding pocket, and the internal (blood-containing) flow tract. Among these 3 components, the DL is the most susceptible to infection at the site where it exits the patient. DL stabilization and exit site management are of paramount importance in the prevention of DL infections. As more experience has been obtained with the device, DL exit management has improved resulting in a reduction in device-related infections in DT patients from 0.47 events per patient-year in the early part of the HeartMate II Multicenter trial to 0.27 later in the trial.5 The HMII percutaneous lead or DL consists of a single cable that extends from the implanted LVAD through the skin to the external environment and carries motor drive power to the LVAD. The DL is approximately 102 cm long and contains a Silastic sleeve with a polyester velour-covered portion (32– 33 cm from pump to end of velour) to promote tissue ingrowth. The diameter of the velour portion of the DL is 8 mm. During HMII implantation, surgeons in the trial externalized at least 2 cm of the velour portion of the DL outside the body resulting in a velour-skin interface (VSI) that was believed to promote tissue ingrowth and optimize DL stability at the exit site. To reduce DL infections further, the Allegheny group proposed a change in the surgical technique for DL placement.6 Instead of leaving a portion of the velourcoated part outside the skin, they kept the entire velour length in the subcutaneous tunnel, resulting in a siliconeskin interface (SSI) at the exit site. They reported that this technique resulted in a significant reduction in DL exit site infection. Similar results were duplicated independently at 3 different centers.7–10 However, there are not enough patients who have been followed for sufficient durations in these preliminary studies to make a final determination on the overall multicenter infection rate with this technique. A large multicenter registry was designed and conducted.
Methods Study design The SSI registry is a prospective and retrospective, multicenter, nonrandomized, observational study designed to evaluate freedom from
DL infection events in subjects with HMII devices in whom the full length of the velour-coated portion of the DL is kept under the skin. The secondary objectives of the study were to collect additional data on the surgical technique and exit site management used at the different sites and to determine the variables associated with DL infection. Data were compared with a control group from the DT clinical trial in which a velour portion of the DL was externalized in all patients resulting in a VSI. It is hypothesized that the SSI technique is associated with at least 50% reduction in the incidence of DL infection at 1 year after HMII implantation compared with the control group. The retrospective and prospective cohorts were powered independently to detect this difference in the 1-year incidence of DL infection. Based on their prior experience with the new DL tunneling surgical technique, 15 U.S. centers were selected to participate in the SSI registry (Appendix).
Study subjects Only results from the retrospective cohort are reported here. All patients 418 years old who had been on HMII support for at least 10 months at participating centers were reviewed. Each site received institutional review board (IRB) approval, and subjects provided written informed consent. Consecutive patients who underwent the SSI technique during implantation and consented to participate in the study were enrolled. Patients who had a sternal reentry for bleeding or other complications were excluded. Only subjects who signed the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) and the SSI registry informed consent forms were considered for enrollment. Between June 2012 and July 2013, 200 patients in whom the HMII device was implanted between June 2009 and August 2012 were enrolled in the retrospective cohort from 14 sites. The number of patients enrolled per site ranged from 5 to 30 with a median of 12 patients per site.
Control group The control group consisted of patients in the HMII DT mid-trial cohort, which consisted of 281 patients enrolled under the DT continuous access protocol between May 5, 2007, and March 31, 2009. To match the inclusion criteria of SSI further, only patients who were on HMII support for at least 10 months (201 patients) were included in the control group. Per the DT trial protocol, a velour portion of the DL was externalized in all patients resulting in a VSI.
Data collection and validation Patient demographics, medical history including comorbidities, and infection adverse events were entered and collected per the SSI study protocol using the INTERMACS registry database. Additional data on tunneling technique and exit site management were collected using supplemental case report forms. Each site was asked to confirm the occurrence of DL infection at 12 months postimplant for each patient enrolled in the registry. This information was used for comparison with data obtained for the study subjects from the INTERMACS database. Inconsistencies in DL infection events were addressed and corrected, if necessary, for each subject. Before locking the database for this analysis, all sites were asked to confirm all the DL infection events they had reported and to report any missing events.
Dean et al.
Reduction in Driveline Infection Rates
Adverse event definitions The INTERMACS definitions of infection-related adverse events were used. DL and/or pocket infection was defined as a positive culture from the skin and/or tissue surrounding the DL or from the tissue surrounding the external housing of a pump implanted within the body, coupled with the need to treat with antimicrobial therapy and clinical evidence of infection such as pain, fever, drainage, or leukocytosis. A localized non-device infection was defined as infection localized to any organ system or region (e.g., mediastinitis) without evidence of systemic involvement (see sepsis definition), ascertained by standard clinical methods and associated with evidence of bacterial, viral, fungal, and protozoal infection and/or requiring empirical treatment. Internal pump component or inflow or outflow tract infection was defined as infection of blood-contacting surfaces of the LVAD documented by positive site culture. Sepsis was defined as evidence of systemic involvement by infection, manifested by positive blood cultures and/or hypotension.
the multivariate Cox regression analyses. The correlation between categorical variables is assessed using the chi-square and odds ratio tests. A p-value o 0.05 was considered statistically significant for all comparisons. Statistical analysis was conducted using SigmaPlot 12.5 software (Systat Software, Inc, San Jose, CA).
Results Table 1 compares the characteristics of the patients enrolled in the SSI registry with the control group. The median age of the SSI registry patients was 60.4 years, 159 (79.5%) were male, 85 (45.5%) had an ischemic etiology of heart failure, and 97 (49%) were implanted as a bridge to transplantation. The average duration of support was 2.03 years. The SSI registry patients were relatively younger and had a larger body mass index and body surface area compared with the control group, which included DT patients only.
Infection events and outcomes
Statistical analysis Continuous variables are reported as mean ⫾ SD or median (25th, 75th percentile), and categorical variables are reported as frequencies. Fisher’s exact test or chi-square test was used to compare categorical variables for univariate analysis. Depending on the data distribution, Student’s t-test or Mann-Whitney test was used for continuous variables. Comparison of the 1-year prevalence of DL infection in the SSI and control groups was based on the relative risk test. The actuarial freedom from DL infection after LVAD implantation was determined with the Kaplan-Meier method with patients censored for transplantation, recovery, and pump exchange or ongoing on device support, and comparison was made using the log-rank test. Univariate and multivariate analyses based on the Cox proportional hazard models were used to determine association of pre-operative, intraoperative, and post-operative factors with DL infection. For the purpose of the analysis, age was dichotomized to 460 or r60 years. Continuous and categorical variables exhibiting relatively significant difference defined as p o 0.1 in univariate comparisons between the infection and the non-infection groups were used in Table 1
3
As shown in Table 2, at 1 year post-implant, 18 (9%) patients in the SSI group and 46 (23%) patients in the control group experienced a DL infection (relative risk 0.44, 95% confidence interval 0.26–0.74, p ¼ 0.002). At last follow-up, corresponding to average support durations of 2.03 and 2.38 years, the overall prevalence of DL infection was 19% and 35% in the SSI and control groups, respectively (relative risk 0.61, 95% confidence interval 0.43–0.87, p ¼ 0.007). Among all patients on HMII support for at least 10 months, the median time to first event was 13.0 months for the SSI group and 9.9 months for the control group (p ¼ 0.04). There were 47 DL infection events in 38 patients in the SSI registry group, and 30 of these patients experienced 1 event, 5 patients had 2 events, and 3 patients had 3 events. There were 106 events in the control group in 46 patients, and 42 of these patients experienced 1 event, 23 patients had 2 events, and 6 patients had 3 events. The event-per-patient year was
Baseline Characteristics of SSI and Control Group Patients
Variables
SSI group (n ¼ 200)
Control group (n ¼ 201)
p-value
Implant date period Age, yearsa DT, no. (%) Male, no. (%) Ischemic etiology, no. (%) NYHA IV class, no. (%) Body surface area, m2 BMI, kg/m2 Creatinine, mg/dl BUN, mg/dl Albumin, g/dl Hemoglobin, g/dl Cholesterol, mg/dl Total bilirubin, mg/dl
June 2009–August 2012 60.4 (51.9, 67.5) 103 (51%) 159 (79.5) 85 (45.5) 133 (68) 2.05 (1.8, 2.2) 28.3 (25.1, 32.0) 1.3 (1.0, 1.7) 24 (18.0, 33.7) 3.5 (3.1, 3.9) 11.0 (9.9, 12.3) 120.0 (98.5, 151.0) 0.9 (0.6, 1.4)
May 2007–March 2009 66 (56.2, 72.0) 201 (100%) 162 (80.6) 117 (58.2) 125 (65) 1.96 (1.8, 2.1) 25.9 (22.4, 30.5) 1.35 (1.1, 1.8) 29 (20.0, 42.0) 3.5 (3.0, 3.9) 11.3 (9.9, 12.9) 118 (97.0, 140.0) 1.0 (0.6, 1.4)
NA o0.01 o0.01 0.881 0.02 0.74 o0.01 o0.01 0.24 0.02 0.56 0.17 0.43 0.75
BMI, body mass index; BUN, blood urea nitrogen; DT, destination therapy; NA, not applicable; NYHA, New York Heart Association; SSI, silicone-skin interface. Values in parentheses for variables are (25th, 75th percentile). a Median (25th, 75th percentile).
The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]
4 Table 2
DL Infection Prevalence and Incidence for SSI Patients Compared with Control Group SSI (n ¼ 200)
Cumulative follow-up, patient-years 407 Average follow-up duration, months 24.7 Patients with DL infection at 1 year post-implant, no. (%) 18 (9) Patients with DL infection at last follow-up, no. (%) 38 (19) Median months to first event 13.0 (8.3, 22.1) Range months to first event 0.85–31.50 DL infection events 47 Event per patient year 0.11
Control group (n ¼ 201) Risk ratio (CI) 478 28.5 46 (23) 71 (35) 9.9 (5.43, 13.75) 0.30–33.75 106 0.22
p-value
— — — — 0.44 (0.26–0.74) 0.002 0.61 (0.43–0.87) 0.007 — 0.04 — — — — — o 0.001
CI, confidence interval; DL, driveline; SSI, silicone-skin interface. Values in parentheses for variables are (25th, 75th percentile).
0.11 and 0.22 for the SSI and control groups, respectively (p o 0.001). As shown in Figure 1, the actuarial freedom from DL infection for the SSI group (91% and 83% at 12 and 24 months, respectively) was significantly lower (p o 0.001, log-rank) compared with the control group (77% and 65%). The Cox regression hazard ratio was 0.49 (95% confidence interval 0.33–0.73, p o 0.001). Among the SSI patients, there were 3 patients who had pump pocket infections that were not associated with a DL infection, 1 patient had an infection of the pump interior, 36 patients had blood-borne infections (72 events), and 5 patients had sepsis (6 events).
shape in 153 of 187 patients (82%) with a large curve in 110 of 153 patients (72%). The DL was looped in the chest in 96 of 193 patients (50%). In 135 of 191 patients (71%), o5 cm of the silicone portion of the DL was internalized. The DL exited the skin on the right side in 122 of 199 patients (61%). The exit site was in the right upper quadrant in 99 of 199 patients (50%), in the right lower quadrant in 23 of 199 patients (11%), in the left upper quadrant in 53 of 199 patients (27%), and in the left lower quadrant in 24 of 199 patients (12%). Anchoring sutures were used in 120 of 198 patients (61%). A monofilament suture was used in 90 patients, and a braided suture was used in 30 patients.
Surgical technique and exit site management
Exit site management
Tables 3 and 4 summarize data related to the surgical implanting technique of the DL and exit site management for the SSI group. The DL was tunneled abdominally in 80 of 194 patients (41%) and through the abdominal wall in 58 of 83 patients (70%). The DL was tunneled in a “U”
The DL was immobilized in 179 of 199 patients (90%). The most frequently used immobilization systems consisted of the Thoratec support binder (Thoratec Corporation; n ¼ 72 patients), a Hollister anchor (Hollister, Inc; n ¼ 29 patients), and a Centurion Foley anchor (Centurion Medical Products; n ¼ 17 patients). Dressing changes were performed in a sterile fashion in most patients (180/200). Topical antibiotic was used immediately post-operatively in only 3 of 199 patients (1%) and after DL infection in 6 of 199 patients (3%). The frequency of dressing changes before discharge was daily for most patients (155/200). After discharge, 108 of 197 patients (55%) had daily dressing changes; the remaining patients had various dressing change frequencies including every other day,
Table 3
Surgical Tunneling Technique
Internalized silicone portion o5 cm 45 cm DL exit location LLQ LUQ RLQ RUQ DL tunneled intraabdominally Anchoring suture DL immobilized Figure 1 Comparison of the actuarial freedom from DL infection for the SSI registry and control group up to 24 months after HMII implant.
71% 29% 12% 27% 11% 50% 41% 61% 90%
DL, driveline; LLQ, left lower quadrant; LUQ, left upper quadrant; RLQ, right lower quadrant; RUQ, right upper quadrant.
Dean et al. Table 4
Reduction in Driveline Infection Rates Exit Site Management Technique
Frequency of dressing changes in hospital Daily Per-drainage Other Frequency of dressing changes post-discharge Daily Weekly Other Occlusive DL dressing change Sterile DL dressing change Topical antibiotic at exit site
80% 6% 14% 55% 11% 34% 57% 90% 5%
DL, driveline.
every 3 days, every 5 days, and every week. The DL dressing was performed in a sterile fashion in 180 of 200 patients (90%). The dressing was occlusive in 112 of 198 patients (57%) and non-occlusive in 86 of 198 patients (43%). A variety of occlusive patches and dressing agents were used.
Analysis of DL infection risk factors The baseline characteristics of the SSI patients with and without infection (Table 5) were similar except for age. Patients with DL infection were slightly younger with a median age of 57.5 years compared with 61.8 years for patients without infection (p ¼ 0.02). Table 6 compares the surgical techniques and DL exit site management used in patients with and without DL infection; there were no statistically significant differences in the management of the exit site. However, surgical techniques were different between patients with and without DL infection. The proportion of patients in whom the DL exited on the right side was significantly higher in patients with DL infection (p ¼ 0.02), and an anchoring suture was used more frequently in patients with infection (p ¼ 0.05). There was also a trend for the silicone length internalized to be relatively shorter in patients with DL infection, but the
Table 5
5 difference did not reach statistical significance (p ¼ 0.06). Although the surgical implantation technique varied between patients with and without DL infection (Table 6), the prevalence of DL infection events also varied from center to center, but the difference did not reach statistical significance (Table 7). Among the different centers, the 1-year prevalence of DL infection ranged from 0% to 20%. A shown in Table 7 and Figure 2, only age and DL exit side were variables that appeared to be associated with DL infection and were significant in univariate and multivariate analysis. Figure 3 shows the freedom from DL infection stratified by age (Figure 3A) and by DL exit side (Figure 3B).
Discussion The results from the retrospective cohort of the SSI registry show that the SSI technique resulted in significant reduction in the incidence and prevalence of DL infection. The 1-year prevalence of DL infection in the SSI registry was 9% compared with 23% in the control group from the HMII clinical trial, resulting in a 61% reduction in DL infection. DL infection at 2 years after implant in the SSI patients remained significantly lower compared with the control group. These multicenter registry results are consistent with the results obtained from different single centers, confirming the impact of the SSI technique on reducing the risk of DL infections.6–9 The exact underlying mechanisms of the SSIassociated reduction in the incidence of DL infections are unknown. As reported by Akhter et al10 the SSI technique resulted in the formation of a “sinus” at the exit site, which may have reduced the risk of trauma-related injuries at the exit site. Based on the structural and ultrastructural analysis conducted by Ledford et al,7 the SSI technique resulted in less dermal inflammation compared with the VSI technique, which resulted in a faster incorporation of the skin. Based on univariate and multivariate analyses, age was determined to be significantly associated with the risk of DL infection. Older age was associated with lower risk of DL
Baseline Characteristics of SSI Patients With DL Infection and Patients Without DL Infection
Variables
SSI patients with DL infection (n ¼ 38)
SSI patients without DL infection (n ¼ 162)
p-value
Age, yearsa Male, no. (%) Ischemic etiology, no. (%) DT, no. (%) Body surface area, m2 BMI, kg/m2 Creatinine, mg/dl BUN, mg/dl Albumin, g/dl Hemoglobin, g/dl Cholesterol, mg/dl Total bilirubin, mg/dl
57.5 30 11 18 2.0 30.0 1.2 24.0 3.5 11.1 133.0 1.1
61.8 129 74 85 2.0 28.1 1.3 25.0 3.5 11.0 118.0 0.9
0.02 0.97 0.28 0.86 0.74 0.31 0.53 0.47 0.34 0.36 0.46 0.54
(46.3, 63.0) (81) (29) (45) (1.9, 2.2) (26.2, 34.2) (1.0, 1.6) (16.5, 32.5) (3.2, 4.0) (10.3, 12.8) (101.0, 151.0) (0.6, 1.8)
(52.9, 68.3) (79.1) (45.7) (52) (1.8, 2.2) (24.9, 31.7) (1.0, 1.7) (18.0, 34.0) (3.1, 3.9) (9.8, 12.3) (96.5, 152.0) (0.6, 1.3)
BMI, body mass index; BUN, blood urea nitrogen; DL, driveline; DT, destination therapy; SSI, silicone-skin interface. Values in parentheses for variables are (25th, 75th percentile). a Median (25th, 75th percentile).
The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]
6
Table 6 Comparison of Surgical and DL Exit Site Management Techniques Used in Patients With DL Infection and in Patients Without DL Infection
Variables Surgical technique variables DL tunneled abdominally (%) DL looped (%) Silicone length internalized o5 cm (%) Anchoring sutured used (%) DL exit on the right side (%) Exit site management variables Occlusive DL dressing (%) Daily post-operative dressing changes (%) Daily post-discharge dressing changes (%) DL immobilized (%)
SSI patients with DL infection (n ¼ 38)
SSI patients without DL infection (n ¼ 162)
p-value
17 22 31 29 30
(44.8) (57.9) (81.6) (76.3) (78.9)
61 74 104 91 92
(38.0) (45.7) (64.2) (56.2) (56.8)
0.53 0.88 0.06 0.05 0.02
21 30 21 32
(55.3) (78.9) (60.5) (84.2)
91 121 85 147
(56.2) (74.7) (52.5) (90.7)
0.88 0.97 0.80 0.87
DL, driveline; SSI, silicone-skin interface.
infection, which is consistent with previous studies.11 Because young patients tend to be more active than older patients, they are more likely to experience trauma at the DL exit site, disrupting the integrity of the DL-integument barrier. None of the variables related to post-implant exit site management were associated with DL infection. Among the surgical variables, the DL exit site on the left side was found to be associated with fewer DL infections. When the DL exits the skin on the left side, the relative risk of DL infection appears to be reduced by 40%. It is possible that right-handed patients may be at lower risk of trauma from DL tugging. Additional data being collected from the prospective cohort are expected to help confirm these findings. Goldstein et al11 analyzed the DL infections reported to the INTERMACS registry from patients implanted with a continuous flow LVAD between June 2006 and September 2010. Among the 2,006 patients included in the analysis, the 1-year prevalence of DL infection was 19%, which is 2 times the DL infection prevalence in the SSI registry. The mean time for the first DL infection was 6.6 months (range 0.2–22.5 months). Based on a multivariate analysis, these authors showed that younger age was the only predictor of DL infection.
Table 7
Gordon et al12 reported on results from a multicenter prospective study designed to identify and characterize the microbiology, epidemiology, risk factors, and outcomes of ventricular assist device infections. Among the 150 patients enrolled in the study between 2006 and 2008, 86 patients received a HMII device, 50 patients received a HeartMate I device, and the remaining 10 patients received an IVAD (Thoratec Corporation; n ¼ 5), VentrAssist (Ventracor Limited; n ¼ 3), and Novacor (Baxter Healthcare; n ¼ 2). By 1 year post-implant, 33 of 150 patients (22%) developed a ventricular assist device–related infection with an incidence rate of 0.36 per patient-year. Most of the infections (28/33) were related to the DL corresponding to 1-year prevalence of 18.7%. The median time to infection was 68 days. Staphylococci were the most common pathogen (47%), but Pseudomonas or other gram-negative bacteria caused 32% of infections. A history of depression and elevated baseline serum creatinine were independent predictors of ventricular assist device infection. According to the study results, the HMII was not associated with a decreased risk of infections. Stulak et al13 studied the impact of prophylactic antibiotics on the prevention of DL infection. Their study included patients implanted with a HMII at 2 different
Univariate and Multivariate Cox Proportional Hazard Analyses of Variables Associated With DL infection
Covariate Univariate analysis using Cox regression Age 460 years Anchoring sutured used 4 5 cm silicone internalized Center Exit left side Multivariate analysis using Cox regression Age 4 60 years Exit left side CI, confidence interval; DL, driveline.
Relative risk of DL infection
95% CI
p-value
0.26 1.10 0.49 0.24 0.45
0.13–0.52 0.57–2.12 0.21–1.11 0.05–1.18 0.21–0.96
o0.001 0.78 0.07 0.29 0.03
0.23 0.40
0.11–0.47 0.19–0.86
o0.001 0.016
Dean et al.
Figure 2
Reduction in Driveline Infection Rates
7
Univariate and multivariate Cox proportional hazard of variables associated with DL infection. *p o 0.05; **p o 0.001.
institutions between 2007 and 2011. Although all patients received peri-operative antibiotic prophylaxis and educated and trained to perform daily DL exit site care using Hibiclens cleansing under sterile conditions, only 1 institution added daily administration of oral doxycycline and levofloxacin (Abx cohort). The cumulative support duration was 172 years for 140 patients in the Abx cohort and 146 years for 127 patients in the No-Abx cohort. At last follow-up, a DL infection occurred in 25 of 140 patients (18%) in the Abx group and 16 of 127 patients (13%) in the No-Abx group (p ¼ 0.15). The freedom from DL infection for the entire cohort was 90% at 1 year, 83% at 2 years, and 73% at 3 years. The authors concluded that other factors may be more important than prophylactic antibiotics in preventing DL infections. This study has several limitations. This was not a randomized study but used a historical control group from patients enrolled in the HMII DT trial. Because the SSI patients were implanted 2 to 3 years later than the patients of the control group, it is possible that the incidence of DL infection is lower because centers may have improved exit site management protocol over time. To minimize the impact of the difference in the implant time periods, the
more recently implanted DT midtrial patients were used as the control group. A randomized study is required to control for all confounding variables and center-to-center variability in the surgical technique and exit site management protocols. However, it is clear that DL infection rates are significantly less than they were in the original clinical trial. One interesting finding from this study is the significant reduction of DL infection rate in patients in whom the DL exited the skin on the left side. Although based on univariate and multivariate analyses, only the exit side among the variables associated with the surgical technique and exit site management protocols was the independent predictor of infection, it is possible that other variables not included in the study may have a stronger association with DL infection. In the retrospective cohort, only patients on support for at least 10 months were enrolled. The rationale was to focus on patients with sufficiently long support durations to be at risk for DL infection. However, this restriction may have excluded some patients with short-term DL infection events; to control for this, the same survival requirement was used in the control group. Finally, the current analysis was based on data collected
8
The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]] support from Thoratec Corporation. F.K. and D.J.F. are employees of Thoratec Corporation. The remaining authors have no conflicts of interest to disclose. This material was presented at the International Society for Heart and Lung Transplantation, April 2014, San Diego, California.
Appendix. and Sites
Participating SSI Investigators
A. Mangi (Yale–New Haven Hospital, New Haven Connecticut), B. Reid (Intermountain Medical Center, Salt Lake City, Utah), B. Sheridan (University of North Carolina, Chapel Hill, North Carolina), B. Sun (Abbott Northwestern, Minneapolis, Minnesota), C. Caldeira (Tampa General Hospital, Tampa, Florida), R. Brewer (Henry Ford Hospital, Detroit, Michigan), D. Dean (Piedmont Hospital, Atlanta, Georgia), D. Tang (Virginia Commonwealth University, Richmond, Virginia), J. Goerbig-Campbell (University of Iowa, Iowa City, Iowa), M. Walsh (St. Vincent Hospital, Indianapolis, Indiana), G. Ewald (Barnes-Jewish Hospital, St Louis, Missouri), V. Jeevanandam (University of Chicago Medical Center, Chicago, Illinois), S. Hall (Baylor University Medical Center, Dallas, Texas), T. Icenogle (Sacred Heart Medical Center, Spokane, Washington), T. Tatooles (Advocate Christ Hospital and Medical Center, Oak Lawn, Illinois).
References
Figure 3 Actuarial freedom from DL infection stratified by age (A) and by DL exit side (B).
retrospectively. Future analysis based on pooling data from the patients enrolled in the retrospective cohort and the patients enrolled in the prospective cohort and being followed will help confirm the current findings. In conclusion, the results of the retrospective cohort of the SSI multicenter registry demonstrate a 50% reduction in DL infection when the velour portion of the DL was fully buried under the skin.
Disclosure statement The SSI Registry was sponsored and conducted by Thoratec Corporation. D.D., G.A.E., A.T. and C.C. have received consulting fees from Thoratec Corporation. S.A.A. received research grant
1. Miller LW, Pagani FD, Russell SD, et al. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med 2007;357: 885-896. 2. Kirklin JK, Naftel DC, Kormos RL, et al. Fifth INTERMACS annual report: risk factor analysis from more than 6,000 mechanical circulatory support patients. J Heart Lung Transplant 2013;32:141-56. 3. Starling RC, Naka Y, Boyle AJ, et al. Results of the post-U.S. Food and Drug Administration-approval study with a continuous flow left ventricular assist device as a bridge to heart transplantation: a prospective study using the INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support). J Am Coll Cardiol 2011;57:1890-8. 4. Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med 2009;361:2241-51. 5. Park SJ, Milano CA, Tatooles AJ, et al. Outcomes in advanced heart failure patients with left ventricular assist devices for destination therapy. Circ Heart Fail 2012;5:241-8. 6. Pelz GB, Hashmi RJ, Moraca RJ, et al. Achilles’ heel of left ventricular assist devices: a novel technique to reduce drive line infections. Abstract presented at: International Society for Heart and Lung Transplantation; April 21-24, 2010; Chicago, IL. 7. Ledford ID, Miller DV, Mason NO, et al. Differential infection rates between velour versus silicone interface at the HeartMate II driveline exit site: structural and ultrastructural insight into possible causes. Abstract presented at: International Society for Heart and Lung Transplantation; April 13-16, 2011; San Diego, CA. 8. Tatooles AJ, Gallagher CT, Pappas PS, Bresticker MA. Avoiding driveline infections in patients with left ventricular assist devices. Abstract presented at: International Society for Heart and Lung Transplantation; April 13-16, 2011; San Diego, CA. 9. Akhter S, Valeroso T, Pisarski B, et al. Technique of HeartMate II percutaneous lead externalization is associated with incidence of infection. Abstract presented at: International Society for Heart and Lung Transplantation; April 13-16, 2011; San Diego, CA.
Dean et al.
Reduction in Driveline Infection Rates
10. Akhter S, Russo MJ, Singh A, et al. Modified HeartMateII driveline externalization technique significantly decreases incidence of infection and improves long term survival. Abstract presented at: International Society for Heart and Lung Transplantation; April 18-21, 2012; Prague, Czech Republic. 11. Goldstein DJ, Naftel D, Holman W, et al. Continuous-flow devices and percutaneous site infections: clinical outcomes. J Heart Lung Transplant 2012;31:1151-7.
9 12. Gordon RJ, Weinberg AD, Pagani FD, et al. Ventricular Assist Device Infection Study Group. Prospective, multicenter study of ventricular assist device infections. Circulation 2013;127:691-702. 13. Stulak JM, Maltais S, Cowger J, et al. Prevention of percutaneous driveline infection after left ventricular assist device implantation: prophylactic antibiotics are not necessary. ASAIO J 2013;59: 570-574.
ORIGINAL CLINICAL SCIENCE
Reduction in driveline infection rates: Results from the HeartMate II Multicenter Driveline Silicone Skin Interface (SSI) Registry David Dean, MD,a Faouzi Kallel, PhD,b Gregory A. Ewald, MD,c Antony Tatooles, MD,d Brett C. Sheridan, MD,e Robert J. Brewer, MD,f Christian Caldeira, MD,g David J. Farrar, PhD,b and Shahab A. Akhter, MD,h on Behalf of the SSI Registry Investigators From the aPiedmont Healthcare, Atlanta, Georgia; bThoratec Corporation, Pleasanton, California; cWashington University School of Medicine, St Louis, Missouri; dAdvocate Christ Hospital Medical Center, Oak Lawn, Illinois; e University of North Carolina, Chapel Hill, North Carolina; fHenry Ford Hospital, Detroit, Michigan; gTampa General Hospital, Tampa, Florida; and the hUniversity of Wisconsin, Madison, Wisconsin.
KEYWORDS: MCS; VAD; HeartMate II; driveline infection; heart failure
BACKGROUND: During left ventricular assist device implantation, a surgical tunneling technique to keep the entire driveline (DL) velour portion in the subcutaneous tunnel, resulting in a silicone-skin interface (SSI) at the exit site, has been adopted by many centers. To assess long-term freedom from DL infection associated with this technique, a multicenter SSI registry was initiated. It was hypothesized that the modified tunneling technique is associated with at least 50% reduction in DL infection at 1 year post-implant compared with the velour-to-skin method used in the HeartMate II (HMII) Destination Therapy (DT) trial. METHODS: SSI is a retrospective and prospective registry of patients who have received the HMII device. Results are reported from the retrospective cohort, which consists of 200 patients who were implanted during the period 2009–2012 with the SSI tunneling method and on HMII support for at least 10 months at the time of enrollment. The prevalence and incidence of DL infection after left ventricular assist device implantation in the SSI retrospective cohort were determined and compared with a control group of 201 patients also on HMII support for at least 10 months from the HMII DT clinical trial who were implanted during the period 2007–2009 using the traditional method in which a small section of the velour portion of the DL was externalized. RESULTS: The 1-year and 2-year prevalence rates of DL infection were 9% and 19% in the SSI patient group compared with 23% and 35% in the control group (hazard ratio 0.49, 95% confidence interval 0.33– 0.73, p o 0.001). The event-per-patient year was 0.11 and 0.22 for the SSI and control groups, respectively (p o 0.001). Based on a multivariate analysis, age and DL exit side were the only independent variables associated with DL infection. Effects of management changes over the eras were not studied and could have contributed to the findings. CONCLUSIONS: These results suggest that leaving the entire DL velour portion below the skin is associated with 50% reduction in DL infection compared with results from the HMII DT trial. J Heart Lung Transplant ]]]];]:]]]–]]] r 2015 International Society for Heart and Lung Transplantation. All rights reserved.
Reprint requests: David Dean, MD, Piedmont Healthcare, 95 Collier Road NW, Suite 5015, Atlanta, GA 30309. Telephone: 404-605-5699. Fax: 404-355-4235. E-mail address: [email protected] 1053-2498/$ - see front matter r 2015 International Society for Heart and Lung Transplantation. All rights reserved. http://dx.doi.org/10.1016/j.healun.2014.11.021
2
The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]
In the United States, 45 million people have heart failure, which accounts for 41 million hospitalizations and 4280,000 deaths each year.1 The HeartMate II (HMII) left ventricular assist device (LVAD)/pump was approved by the U.S. Food and Drug Administration more recently for use as bridge to transplantation (BTT) and destination therapy (DT) for treatment of patients with end-stage heart failure. The DT patients are generally not eligible for heart transplantation and require LVAD support for an indefinite duration. Although the HMII has provided a significant survival benefit for patients with end-stage heart failure, serious adverse events continue to be a major cause of morbidity.2 Device-related infections are among the most commonly occurring serious adverse events, with reported incidences of 0.34 events per patient-year (19%) in BTT patients and 0.48 events per patient-year (35%) in DT patients.3,4 They are also a leading cause of rehospitalization, which increases the overall costs associated with LVAD therapy. The 3 main LVAD components that are at risk for infections are the driveline (DL), the pump and its surrounding pocket, and the internal (blood-containing) flow tract. Among these 3 components, the DL is the most susceptible to infection at the site where it exits the patient. DL stabilization and exit site management are of paramount importance in the prevention of DL infections. As more experience has been obtained with the device, DL exit management has improved resulting in a reduction in device-related infections in DT patients from 0.47 events per patient-year in the early part of the HeartMate II Multicenter trial to 0.27 later in the trial.5 The HMII percutaneous lead or DL consists of a single cable that extends from the implanted LVAD through the skin to the external environment and carries motor drive power to the LVAD. The DL is approximately 102 cm long and contains a Silastic sleeve with a polyester velour-covered portion (32– 33 cm from pump to end of velour) to promote tissue ingrowth. The diameter of the velour portion of the DL is 8 mm. During HMII implantation, surgeons in the trial externalized at least 2 cm of the velour portion of the DL outside the body resulting in a velour-skin interface (VSI) that was believed to promote tissue ingrowth and optimize DL stability at the exit site. To reduce DL infections further, the Allegheny group proposed a change in the surgical technique for DL placement.6 Instead of leaving a portion of the velourcoated part outside the skin, they kept the entire velour length in the subcutaneous tunnel, resulting in a siliconeskin interface (SSI) at the exit site. They reported that this technique resulted in a significant reduction in DL exit site infection. Similar results were duplicated independently at 3 different centers.7–10 However, there are not enough patients who have been followed for sufficient durations in these preliminary studies to make a final determination on the overall multicenter infection rate with this technique. A large multicenter registry was designed and conducted.
Methods Study design The SSI registry is a prospective and retrospective, multicenter, nonrandomized, observational study designed to evaluate freedom from
DL infection events in subjects with HMII devices in whom the full length of the velour-coated portion of the DL is kept under the skin. The secondary objectives of the study were to collect additional data on the surgical technique and exit site management used at the different sites and to determine the variables associated with DL infection. Data were compared with a control group from the DT clinical trial in which a velour portion of the DL was externalized in all patients resulting in a VSI. It is hypothesized that the SSI technique is associated with at least 50% reduction in the incidence of DL infection at 1 year after HMII implantation compared with the control group. The retrospective and prospective cohorts were powered independently to detect this difference in the 1-year incidence of DL infection. Based on their prior experience with the new DL tunneling surgical technique, 15 U.S. centers were selected to participate in the SSI registry (Appendix).
Study subjects Only results from the retrospective cohort are reported here. All patients 418 years old who had been on HMII support for at least 10 months at participating centers were reviewed. Each site received institutional review board (IRB) approval, and subjects provided written informed consent. Consecutive patients who underwent the SSI technique during implantation and consented to participate in the study were enrolled. Patients who had a sternal reentry for bleeding or other complications were excluded. Only subjects who signed the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) and the SSI registry informed consent forms were considered for enrollment. Between June 2012 and July 2013, 200 patients in whom the HMII device was implanted between June 2009 and August 2012 were enrolled in the retrospective cohort from 14 sites. The number of patients enrolled per site ranged from 5 to 30 with a median of 12 patients per site.
Control group The control group consisted of patients in the HMII DT mid-trial cohort, which consisted of 281 patients enrolled under the DT continuous access protocol between May 5, 2007, and March 31, 2009. To match the inclusion criteria of SSI further, only patients who were on HMII support for at least 10 months (201 patients) were included in the control group. Per the DT trial protocol, a velour portion of the DL was externalized in all patients resulting in a VSI.
Data collection and validation Patient demographics, medical history including comorbidities, and infection adverse events were entered and collected per the SSI study protocol using the INTERMACS registry database. Additional data on tunneling technique and exit site management were collected using supplemental case report forms. Each site was asked to confirm the occurrence of DL infection at 12 months postimplant for each patient enrolled in the registry. This information was used for comparison with data obtained for the study subjects from the INTERMACS database. Inconsistencies in DL infection events were addressed and corrected, if necessary, for each subject. Before locking the database for this analysis, all sites were asked to confirm all the DL infection events they had reported and to report any missing events.
Dean et al.
Reduction in Driveline Infection Rates
Adverse event definitions The INTERMACS definitions of infection-related adverse events were used. DL and/or pocket infection was defined as a positive culture from the skin and/or tissue surrounding the DL or from the tissue surrounding the external housing of a pump implanted within the body, coupled with the need to treat with antimicrobial therapy and clinical evidence of infection such as pain, fever, drainage, or leukocytosis. A localized non-device infection was defined as infection localized to any organ system or region (e.g., mediastinitis) without evidence of systemic involvement (see sepsis definition), ascertained by standard clinical methods and associated with evidence of bacterial, viral, fungal, and protozoal infection and/or requiring empirical treatment. Internal pump component or inflow or outflow tract infection was defined as infection of blood-contacting surfaces of the LVAD documented by positive site culture. Sepsis was defined as evidence of systemic involvement by infection, manifested by positive blood cultures and/or hypotension.
the multivariate Cox regression analyses. The correlation between categorical variables is assessed using the chi-square and odds ratio tests. A p-value o 0.05 was considered statistically significant for all comparisons. Statistical analysis was conducted using SigmaPlot 12.5 software (Systat Software, Inc, San Jose, CA).
Results Table 1 compares the characteristics of the patients enrolled in the SSI registry with the control group. The median age of the SSI registry patients was 60.4 years, 159 (79.5%) were male, 85 (45.5%) had an ischemic etiology of heart failure, and 97 (49%) were implanted as a bridge to transplantation. The average duration of support was 2.03 years. The SSI registry patients were relatively younger and had a larger body mass index and body surface area compared with the control group, which included DT patients only.
Infection events and outcomes
Statistical analysis Continuous variables are reported as mean ⫾ SD or median (25th, 75th percentile), and categorical variables are reported as frequencies. Fisher’s exact test or chi-square test was used to compare categorical variables for univariate analysis. Depending on the data distribution, Student’s t-test or Mann-Whitney test was used for continuous variables. Comparison of the 1-year prevalence of DL infection in the SSI and control groups was based on the relative risk test. The actuarial freedom from DL infection after LVAD implantation was determined with the Kaplan-Meier method with patients censored for transplantation, recovery, and pump exchange or ongoing on device support, and comparison was made using the log-rank test. Univariate and multivariate analyses based on the Cox proportional hazard models were used to determine association of pre-operative, intraoperative, and post-operative factors with DL infection. For the purpose of the analysis, age was dichotomized to 460 or r60 years. Continuous and categorical variables exhibiting relatively significant difference defined as p o 0.1 in univariate comparisons between the infection and the non-infection groups were used in Table 1
3
As shown in Table 2, at 1 year post-implant, 18 (9%) patients in the SSI group and 46 (23%) patients in the control group experienced a DL infection (relative risk 0.44, 95% confidence interval 0.26–0.74, p ¼ 0.002). At last follow-up, corresponding to average support durations of 2.03 and 2.38 years, the overall prevalence of DL infection was 19% and 35% in the SSI and control groups, respectively (relative risk 0.61, 95% confidence interval 0.43–0.87, p ¼ 0.007). Among all patients on HMII support for at least 10 months, the median time to first event was 13.0 months for the SSI group and 9.9 months for the control group (p ¼ 0.04). There were 47 DL infection events in 38 patients in the SSI registry group, and 30 of these patients experienced 1 event, 5 patients had 2 events, and 3 patients had 3 events. There were 106 events in the control group in 46 patients, and 42 of these patients experienced 1 event, 23 patients had 2 events, and 6 patients had 3 events. The event-per-patient year was
Baseline Characteristics of SSI and Control Group Patients
Variables
SSI group (n ¼ 200)
Control group (n ¼ 201)
p-value
Implant date period Age, yearsa DT, no. (%) Male, no. (%) Ischemic etiology, no. (%) NYHA IV class, no. (%) Body surface area, m2 BMI, kg/m2 Creatinine, mg/dl BUN, mg/dl Albumin, g/dl Hemoglobin, g/dl Cholesterol, mg/dl Total bilirubin, mg/dl
June 2009–August 2012 60.4 (51.9, 67.5) 103 (51%) 159 (79.5) 85 (45.5) 133 (68) 2.05 (1.8, 2.2) 28.3 (25.1, 32.0) 1.3 (1.0, 1.7) 24 (18.0, 33.7) 3.5 (3.1, 3.9) 11.0 (9.9, 12.3) 120.0 (98.5, 151.0) 0.9 (0.6, 1.4)
May 2007–March 2009 66 (56.2, 72.0) 201 (100%) 162 (80.6) 117 (58.2) 125 (65) 1.96 (1.8, 2.1) 25.9 (22.4, 30.5) 1.35 (1.1, 1.8) 29 (20.0, 42.0) 3.5 (3.0, 3.9) 11.3 (9.9, 12.9) 118 (97.0, 140.0) 1.0 (0.6, 1.4)
NA o0.01 o0.01 0.881 0.02 0.74 o0.01 o0.01 0.24 0.02 0.56 0.17 0.43 0.75
BMI, body mass index; BUN, blood urea nitrogen; DT, destination therapy; NA, not applicable; NYHA, New York Heart Association; SSI, silicone-skin interface. Values in parentheses for variables are (25th, 75th percentile). a Median (25th, 75th percentile).
The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]
4 Table 2
DL Infection Prevalence and Incidence for SSI Patients Compared with Control Group SSI (n ¼ 200)
Cumulative follow-up, patient-years 407 Average follow-up duration, months 24.7 Patients with DL infection at 1 year post-implant, no. (%) 18 (9) Patients with DL infection at last follow-up, no. (%) 38 (19) Median months to first event 13.0 (8.3, 22.1) Range months to first event 0.85–31.50 DL infection events 47 Event per patient year 0.11
Control group (n ¼ 201) Risk ratio (CI) 478 28.5 46 (23) 71 (35) 9.9 (5.43, 13.75) 0.30–33.75 106 0.22
p-value
— — — — 0.44 (0.26–0.74) 0.002 0.61 (0.43–0.87) 0.007 — 0.04 — — — — — o 0.001
CI, confidence interval; DL, driveline; SSI, silicone-skin interface. Values in parentheses for variables are (25th, 75th percentile).
0.11 and 0.22 for the SSI and control groups, respectively (p o 0.001). As shown in Figure 1, the actuarial freedom from DL infection for the SSI group (91% and 83% at 12 and 24 months, respectively) was significantly lower (p o 0.001, log-rank) compared with the control group (77% and 65%). The Cox regression hazard ratio was 0.49 (95% confidence interval 0.33–0.73, p o 0.001). Among the SSI patients, there were 3 patients who had pump pocket infections that were not associated with a DL infection, 1 patient had an infection of the pump interior, 36 patients had blood-borne infections (72 events), and 5 patients had sepsis (6 events).
shape in 153 of 187 patients (82%) with a large curve in 110 of 153 patients (72%). The DL was looped in the chest in 96 of 193 patients (50%). In 135 of 191 patients (71%), o5 cm of the silicone portion of the DL was internalized. The DL exited the skin on the right side in 122 of 199 patients (61%). The exit site was in the right upper quadrant in 99 of 199 patients (50%), in the right lower quadrant in 23 of 199 patients (11%), in the left upper quadrant in 53 of 199 patients (27%), and in the left lower quadrant in 24 of 199 patients (12%). Anchoring sutures were used in 120 of 198 patients (61%). A monofilament suture was used in 90 patients, and a braided suture was used in 30 patients.
Surgical technique and exit site management
Exit site management
Tables 3 and 4 summarize data related to the surgical implanting technique of the DL and exit site management for the SSI group. The DL was tunneled abdominally in 80 of 194 patients (41%) and through the abdominal wall in 58 of 83 patients (70%). The DL was tunneled in a “U”
The DL was immobilized in 179 of 199 patients (90%). The most frequently used immobilization systems consisted of the Thoratec support binder (Thoratec Corporation; n ¼ 72 patients), a Hollister anchor (Hollister, Inc; n ¼ 29 patients), and a Centurion Foley anchor (Centurion Medical Products; n ¼ 17 patients). Dressing changes were performed in a sterile fashion in most patients (180/200). Topical antibiotic was used immediately post-operatively in only 3 of 199 patients (1%) and after DL infection in 6 of 199 patients (3%). The frequency of dressing changes before discharge was daily for most patients (155/200). After discharge, 108 of 197 patients (55%) had daily dressing changes; the remaining patients had various dressing change frequencies including every other day,
Table 3
Surgical Tunneling Technique
Internalized silicone portion o5 cm 45 cm DL exit location LLQ LUQ RLQ RUQ DL tunneled intraabdominally Anchoring suture DL immobilized Figure 1 Comparison of the actuarial freedom from DL infection for the SSI registry and control group up to 24 months after HMII implant.
71% 29% 12% 27% 11% 50% 41% 61% 90%
DL, driveline; LLQ, left lower quadrant; LUQ, left upper quadrant; RLQ, right lower quadrant; RUQ, right upper quadrant.
Dean et al. Table 4
Reduction in Driveline Infection Rates Exit Site Management Technique
Frequency of dressing changes in hospital Daily Per-drainage Other Frequency of dressing changes post-discharge Daily Weekly Other Occlusive DL dressing change Sterile DL dressing change Topical antibiotic at exit site
80% 6% 14% 55% 11% 34% 57% 90% 5%
DL, driveline.
every 3 days, every 5 days, and every week. The DL dressing was performed in a sterile fashion in 180 of 200 patients (90%). The dressing was occlusive in 112 of 198 patients (57%) and non-occlusive in 86 of 198 patients (43%). A variety of occlusive patches and dressing agents were used.
Analysis of DL infection risk factors The baseline characteristics of the SSI patients with and without infection (Table 5) were similar except for age. Patients with DL infection were slightly younger with a median age of 57.5 years compared with 61.8 years for patients without infection (p ¼ 0.02). Table 6 compares the surgical techniques and DL exit site management used in patients with and without DL infection; there were no statistically significant differences in the management of the exit site. However, surgical techniques were different between patients with and without DL infection. The proportion of patients in whom the DL exited on the right side was significantly higher in patients with DL infection (p ¼ 0.02), and an anchoring suture was used more frequently in patients with infection (p ¼ 0.05). There was also a trend for the silicone length internalized to be relatively shorter in patients with DL infection, but the
Table 5
5 difference did not reach statistical significance (p ¼ 0.06). Although the surgical implantation technique varied between patients with and without DL infection (Table 6), the prevalence of DL infection events also varied from center to center, but the difference did not reach statistical significance (Table 7). Among the different centers, the 1-year prevalence of DL infection ranged from 0% to 20%. A shown in Table 7 and Figure 2, only age and DL exit side were variables that appeared to be associated with DL infection and were significant in univariate and multivariate analysis. Figure 3 shows the freedom from DL infection stratified by age (Figure 3A) and by DL exit side (Figure 3B).
Discussion The results from the retrospective cohort of the SSI registry show that the SSI technique resulted in significant reduction in the incidence and prevalence of DL infection. The 1-year prevalence of DL infection in the SSI registry was 9% compared with 23% in the control group from the HMII clinical trial, resulting in a 61% reduction in DL infection. DL infection at 2 years after implant in the SSI patients remained significantly lower compared with the control group. These multicenter registry results are consistent with the results obtained from different single centers, confirming the impact of the SSI technique on reducing the risk of DL infections.6–9 The exact underlying mechanisms of the SSIassociated reduction in the incidence of DL infections are unknown. As reported by Akhter et al10 the SSI technique resulted in the formation of a “sinus” at the exit site, which may have reduced the risk of trauma-related injuries at the exit site. Based on the structural and ultrastructural analysis conducted by Ledford et al,7 the SSI technique resulted in less dermal inflammation compared with the VSI technique, which resulted in a faster incorporation of the skin. Based on univariate and multivariate analyses, age was determined to be significantly associated with the risk of DL infection. Older age was associated with lower risk of DL
Baseline Characteristics of SSI Patients With DL Infection and Patients Without DL Infection
Variables
SSI patients with DL infection (n ¼ 38)
SSI patients without DL infection (n ¼ 162)
p-value
Age, yearsa Male, no. (%) Ischemic etiology, no. (%) DT, no. (%) Body surface area, m2 BMI, kg/m2 Creatinine, mg/dl BUN, mg/dl Albumin, g/dl Hemoglobin, g/dl Cholesterol, mg/dl Total bilirubin, mg/dl
57.5 30 11 18 2.0 30.0 1.2 24.0 3.5 11.1 133.0 1.1
61.8 129 74 85 2.0 28.1 1.3 25.0 3.5 11.0 118.0 0.9
0.02 0.97 0.28 0.86 0.74 0.31 0.53 0.47 0.34 0.36 0.46 0.54
(46.3, 63.0) (81) (29) (45) (1.9, 2.2) (26.2, 34.2) (1.0, 1.6) (16.5, 32.5) (3.2, 4.0) (10.3, 12.8) (101.0, 151.0) (0.6, 1.8)
(52.9, 68.3) (79.1) (45.7) (52) (1.8, 2.2) (24.9, 31.7) (1.0, 1.7) (18.0, 34.0) (3.1, 3.9) (9.8, 12.3) (96.5, 152.0) (0.6, 1.3)
BMI, body mass index; BUN, blood urea nitrogen; DL, driveline; DT, destination therapy; SSI, silicone-skin interface. Values in parentheses for variables are (25th, 75th percentile). a Median (25th, 75th percentile).
The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]
6
Table 6 Comparison of Surgical and DL Exit Site Management Techniques Used in Patients With DL Infection and in Patients Without DL Infection
Variables Surgical technique variables DL tunneled abdominally (%) DL looped (%) Silicone length internalized o5 cm (%) Anchoring sutured used (%) DL exit on the right side (%) Exit site management variables Occlusive DL dressing (%) Daily post-operative dressing changes (%) Daily post-discharge dressing changes (%) DL immobilized (%)
SSI patients with DL infection (n ¼ 38)
SSI patients without DL infection (n ¼ 162)
p-value
17 22 31 29 30
(44.8) (57.9) (81.6) (76.3) (78.9)
61 74 104 91 92
(38.0) (45.7) (64.2) (56.2) (56.8)
0.53 0.88 0.06 0.05 0.02
21 30 21 32
(55.3) (78.9) (60.5) (84.2)
91 121 85 147
(56.2) (74.7) (52.5) (90.7)
0.88 0.97 0.80 0.87
DL, driveline; SSI, silicone-skin interface.
infection, which is consistent with previous studies.11 Because young patients tend to be more active than older patients, they are more likely to experience trauma at the DL exit site, disrupting the integrity of the DL-integument barrier. None of the variables related to post-implant exit site management were associated with DL infection. Among the surgical variables, the DL exit site on the left side was found to be associated with fewer DL infections. When the DL exits the skin on the left side, the relative risk of DL infection appears to be reduced by 40%. It is possible that right-handed patients may be at lower risk of trauma from DL tugging. Additional data being collected from the prospective cohort are expected to help confirm these findings. Goldstein et al11 analyzed the DL infections reported to the INTERMACS registry from patients implanted with a continuous flow LVAD between June 2006 and September 2010. Among the 2,006 patients included in the analysis, the 1-year prevalence of DL infection was 19%, which is 2 times the DL infection prevalence in the SSI registry. The mean time for the first DL infection was 6.6 months (range 0.2–22.5 months). Based on a multivariate analysis, these authors showed that younger age was the only predictor of DL infection.
Table 7
Gordon et al12 reported on results from a multicenter prospective study designed to identify and characterize the microbiology, epidemiology, risk factors, and outcomes of ventricular assist device infections. Among the 150 patients enrolled in the study between 2006 and 2008, 86 patients received a HMII device, 50 patients received a HeartMate I device, and the remaining 10 patients received an IVAD (Thoratec Corporation; n ¼ 5), VentrAssist (Ventracor Limited; n ¼ 3), and Novacor (Baxter Healthcare; n ¼ 2). By 1 year post-implant, 33 of 150 patients (22%) developed a ventricular assist device–related infection with an incidence rate of 0.36 per patient-year. Most of the infections (28/33) were related to the DL corresponding to 1-year prevalence of 18.7%. The median time to infection was 68 days. Staphylococci were the most common pathogen (47%), but Pseudomonas or other gram-negative bacteria caused 32% of infections. A history of depression and elevated baseline serum creatinine were independent predictors of ventricular assist device infection. According to the study results, the HMII was not associated with a decreased risk of infections. Stulak et al13 studied the impact of prophylactic antibiotics on the prevention of DL infection. Their study included patients implanted with a HMII at 2 different
Univariate and Multivariate Cox Proportional Hazard Analyses of Variables Associated With DL infection
Covariate Univariate analysis using Cox regression Age 460 years Anchoring sutured used 4 5 cm silicone internalized Center Exit left side Multivariate analysis using Cox regression Age 4 60 years Exit left side CI, confidence interval; DL, driveline.
Relative risk of DL infection
95% CI
p-value
0.26 1.10 0.49 0.24 0.45
0.13–0.52 0.57–2.12 0.21–1.11 0.05–1.18 0.21–0.96
o0.001 0.78 0.07 0.29 0.03
0.23 0.40
0.11–0.47 0.19–0.86
o0.001 0.016
Dean et al.
Figure 2
Reduction in Driveline Infection Rates
7
Univariate and multivariate Cox proportional hazard of variables associated with DL infection. *p o 0.05; **p o 0.001.
institutions between 2007 and 2011. Although all patients received peri-operative antibiotic prophylaxis and educated and trained to perform daily DL exit site care using Hibiclens cleansing under sterile conditions, only 1 institution added daily administration of oral doxycycline and levofloxacin (Abx cohort). The cumulative support duration was 172 years for 140 patients in the Abx cohort and 146 years for 127 patients in the No-Abx cohort. At last follow-up, a DL infection occurred in 25 of 140 patients (18%) in the Abx group and 16 of 127 patients (13%) in the No-Abx group (p ¼ 0.15). The freedom from DL infection for the entire cohort was 90% at 1 year, 83% at 2 years, and 73% at 3 years. The authors concluded that other factors may be more important than prophylactic antibiotics in preventing DL infections. This study has several limitations. This was not a randomized study but used a historical control group from patients enrolled in the HMII DT trial. Because the SSI patients were implanted 2 to 3 years later than the patients of the control group, it is possible that the incidence of DL infection is lower because centers may have improved exit site management protocol over time. To minimize the impact of the difference in the implant time periods, the
more recently implanted DT midtrial patients were used as the control group. A randomized study is required to control for all confounding variables and center-to-center variability in the surgical technique and exit site management protocols. However, it is clear that DL infection rates are significantly less than they were in the original clinical trial. One interesting finding from this study is the significant reduction of DL infection rate in patients in whom the DL exited the skin on the left side. Although based on univariate and multivariate analyses, only the exit side among the variables associated with the surgical technique and exit site management protocols was the independent predictor of infection, it is possible that other variables not included in the study may have a stronger association with DL infection. In the retrospective cohort, only patients on support for at least 10 months were enrolled. The rationale was to focus on patients with sufficiently long support durations to be at risk for DL infection. However, this restriction may have excluded some patients with short-term DL infection events; to control for this, the same survival requirement was used in the control group. Finally, the current analysis was based on data collected
8
The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]] support from Thoratec Corporation. F.K. and D.J.F. are employees of Thoratec Corporation. The remaining authors have no conflicts of interest to disclose. This material was presented at the International Society for Heart and Lung Transplantation, April 2014, San Diego, California.
Appendix. and Sites
Participating SSI Investigators
A. Mangi (Yale–New Haven Hospital, New Haven Connecticut), B. Reid (Intermountain Medical Center, Salt Lake City, Utah), B. Sheridan (University of North Carolina, Chapel Hill, North Carolina), B. Sun (Abbott Northwestern, Minneapolis, Minnesota), C. Caldeira (Tampa General Hospital, Tampa, Florida), R. Brewer (Henry Ford Hospital, Detroit, Michigan), D. Dean (Piedmont Hospital, Atlanta, Georgia), D. Tang (Virginia Commonwealth University, Richmond, Virginia), J. Goerbig-Campbell (University of Iowa, Iowa City, Iowa), M. Walsh (St. Vincent Hospital, Indianapolis, Indiana), G. Ewald (Barnes-Jewish Hospital, St Louis, Missouri), V. Jeevanandam (University of Chicago Medical Center, Chicago, Illinois), S. Hall (Baylor University Medical Center, Dallas, Texas), T. Icenogle (Sacred Heart Medical Center, Spokane, Washington), T. Tatooles (Advocate Christ Hospital and Medical Center, Oak Lawn, Illinois).
References
Figure 3 Actuarial freedom from DL infection stratified by age (A) and by DL exit side (B).
retrospectively. Future analysis based on pooling data from the patients enrolled in the retrospective cohort and the patients enrolled in the prospective cohort and being followed will help confirm the current findings. In conclusion, the results of the retrospective cohort of the SSI multicenter registry demonstrate a 50% reduction in DL infection when the velour portion of the DL was fully buried under the skin.
Disclosure statement The SSI Registry was sponsored and conducted by Thoratec Corporation. D.D., G.A.E., A.T. and C.C. have received consulting fees from Thoratec Corporation. S.A.A. received research grant
1. Miller LW, Pagani FD, Russell SD, et al. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med 2007;357: 885-896. 2. Kirklin JK, Naftel DC, Kormos RL, et al. Fifth INTERMACS annual report: risk factor analysis from more than 6,000 mechanical circulatory support patients. J Heart Lung Transplant 2013;32:141-56. 3. Starling RC, Naka Y, Boyle AJ, et al. Results of the post-U.S. Food and Drug Administration-approval study with a continuous flow left ventricular assist device as a bridge to heart transplantation: a prospective study using the INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support). J Am Coll Cardiol 2011;57:1890-8. 4. Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med 2009;361:2241-51. 5. Park SJ, Milano CA, Tatooles AJ, et al. Outcomes in advanced heart failure patients with left ventricular assist devices for destination therapy. Circ Heart Fail 2012;5:241-8. 6. Pelz GB, Hashmi RJ, Moraca RJ, et al. Achilles’ heel of left ventricular assist devices: a novel technique to reduce drive line infections. Abstract presented at: International Society for Heart and Lung Transplantation; April 21-24, 2010; Chicago, IL. 7. Ledford ID, Miller DV, Mason NO, et al. Differential infection rates between velour versus silicone interface at the HeartMate II driveline exit site: structural and ultrastructural insight into possible causes. Abstract presented at: International Society for Heart and Lung Transplantation; April 13-16, 2011; San Diego, CA. 8. Tatooles AJ, Gallagher CT, Pappas PS, Bresticker MA. Avoiding driveline infections in patients with left ventricular assist devices. Abstract presented at: International Society for Heart and Lung Transplantation; April 13-16, 2011; San Diego, CA. 9. Akhter S, Valeroso T, Pisarski B, et al. Technique of HeartMate II percutaneous lead externalization is associated with incidence of infection. Abstract presented at: International Society for Heart and Lung Transplantation; April 13-16, 2011; San Diego, CA.
Dean et al.
Reduction in Driveline Infection Rates
10. Akhter S, Russo MJ, Singh A, et al. Modified HeartMateII driveline externalization technique significantly decreases incidence of infection and improves long term survival. Abstract presented at: International Society for Heart and Lung Transplantation; April 18-21, 2012; Prague, Czech Republic. 11. Goldstein DJ, Naftel D, Holman W, et al. Continuous-flow devices and percutaneous site infections: clinical outcomes. J Heart Lung Transplant 2012;31:1151-7.
9 12. Gordon RJ, Weinberg AD, Pagani FD, et al. Ventricular Assist Device Infection Study Group. Prospective, multicenter study of ventricular assist device infections. Circulation 2013;127:691-702. 13. Stulak JM, Maltais S, Cowger J, et al. Prevention of percutaneous driveline infection after left ventricular assist device implantation: prophylactic antibiotics are not necessary. ASAIO J 2013;59: 570-574.