© 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Clin Transplant 2015: 29: 499–505 DOI: 10.1111/ctr.12540
Clinical Transplantation
Clinical outcomes in sensitized heart transplant patients bridged with ventricular assist devices Kidambi S, Mohamedali B, Bhat G. Clinical outcomes in sensitized heart transplant patients bridged with ventricular assist devices.
Sumanth Kidambia, Burhan Mohamedalib and Geetha Bhata a
Abstract: Background: Left ventricular assist devices (LVADs) as a bridge to transplant (BTT) have been known to cause allosensitization, as measured by panel-reactive antibody (PRA) levels. The goal of this study was to measure the impact of this allosensitization on outcomes. Methods: Panel-reactive antibodies were analyzed in BTT patients, with sensitization defined as peak PRAs ≥ 10%. Baseline characteristics and outcomes in the two patient groups were evaluated using descriptive statistics, Kaplan–Meier, and regression analysis. Results: Thirty-eight patients were included in the study (17 sensitized vs. 21 non-sensitized). There were more women in the sensitized group (47% vs. 10%, p = 0.023). There was no difference in mean times to high-grade acute cellular rejection (ACR; 18.3 months in sensitized vs. 36.9 months in non-sensitized). Five patients in the sensitized groups developed antibody-mediated rejection (AMR) vs. 0 in the non-sensitized, and all five patients died (Kaplan–Meier log-rank p = 0.024). There was also a significant difference in the incidence of infection at the one- to six-month stage (52.9% vs. 19.0%, p = 0.03). Conclusion: Sensitization appears to have a negative effect on mortality. This mortality appears to be concentrated in patients with AMR, and we postulate that the development of AMR in a sensitized patient may be a predictor of mortality.
Division of Advanced Heart Failure and Clinical Transplantation, Advocate Christ Medical Center, Oak Lawn, IL, USA and bRush University, Chicago, IL, USA Key words: acute cellular rejection – allosensitization – antibody-mediated rejection – bridge to transplant – heart transplant – infection – left ventricular assist device – mechanical support – mortality – orthotopic heart transplant – outcomes – panel-reactive antibody Corresponding author: Geetha Bhat, PhD, MD, Medical Director, Center for Heart Transplant and Assist Devices Advocate Christ Medical Center, Heart Institute Oak Lawn, IL, USA. Tel: +1 708 684 7031; Fax: 1-708-520-1875; e-mail: [email protected] Conflict of interest: None. Accepted for publication 10 March 2015
Mechanical circulatory support has gained increasing popularity for patients with end-stage systolic heart failure (1). These devices often serve as a bridge to transplant (BTT), providing mechanical support while a donor heart becomes available. However, these devices have been shown to induce sensitization in that they are associated with the development of circulating anti-HLA antibodies with potential donor reactivity (2, 3). Early reports indicated that these elevated levels of antibodies were linked to poorer outcomes, notably graft rejection (4–6). More recent studies have indicated that this difference may not be as significant as previously stated (7–9). Other studies have noted a positive correlation between increased sensitization and antibody-mediated rejection (AMR) episodes (10, 11). This study sought to evaluate the association between elevated panel-reactive antibodies (PRAs) and its effects on outcomes, specifically: rejections (both acute cellular and antibody mediated), infections, and mortality.
Methods Patient population
This was a single-center study. It consisted of a retrospective analysis of consecutive patients who were bridged to transplant with ventricular assist devices (VADs) and who were transplanted between January 2008 and August 2013. This study was approved by the institutional review board. All patient demographic and outcome data were collected through a combination of chart review and internal databases. Data on peak PRA data were collected, defined as the maximum PRA level prior to transplant. Patients with peak PRA levels 10% were considered sensitized (n = 17), while
500
patient with PRA 10%)
0–6 7–12 13–18 19–24
0.24 0.20 0.09 0.17
0.44 (17) 0.41 (15) 0.302 (11) 0.41 (6)
Non-sensitized (PRA < 10%) 0.24 0.00 0.00 0.00
0.54 (21) 0.00 (19) 0.00 (9) 0.00 (7)
p-Value 0.986 0.164 0.380 0.363
A rejection episode was defined as International Society for Heart and Lung Transplantation 2R or greater. The data were organized by sensitized vs. non-sensitized patients and broken down by six-month time intervals. PRA, panel-reactive antibody.
Impact of VAD implantation on allosensitization
Of the 20 patients for whom pre-VAD PRA levels were available, three were sensitized prior to the LVAD implant (15%). Mean peak PRA for these three patients was 22.33%. There was a statistically significant increase in the PRA levels following VAD implantation with a mean increase of 15.25 19.36 following the implant. Median time to peak PRA following implant was 131.5 d (interquartile range: 33.75–449.5). Impact of elevated PRA levels on cross-match
In the sensitized subgroup, three of the 17 (17.6%) had a positive cross-match vs. two of 21 (9.6%) in the non-sensitized subgroup (p = 0.640). Outcomes
Acute cellular rejection. The mean number of highgrade ACR episodes (as defined as ISHLT grade ≥2R), was not seen to be significant during any sixmonth time interval for the two yr following transplantation (Table 2). However, there were no highgrade rejections among the non-sensitized patients beyond a six-month duration. Nevertheless, the difference in mean number of rejection episodes was not seen to be significant. The mean time to high-grade rejection in the sensitized group was 18.3 months vs. 36.9 months
Fig. 1. Time to development of high-grade rejections (International Society for Heart and Lung Transplantation ≥2R) after heart transplant in the sensitized and non-sensitized patients, as defined by peak panel-reactive antibody (PRA; log-rank p-value = 0.208).
in the non-sensitized group. However, this difference was not seen to be significant (log rank p value = 0.208; Fig. 1). The mean time to all rejections (as defined as ISHLT rejection grade ≥1R) in the sensitized group was 6.13 months vs. 5.04 months in the non-sensitized group. This difference was also not seen to be significant (p = 0.471). Antibody-mediated rejection. Five patients (13.1%) in the patient population experienced AMR episodes. All five were in the sensitized group. This difference was statistically significant (p = 0.012). Of the patients who developed AMR, the mean
501
Kidambi et al.
peak PRA was 51.7 31.3 vs. 11.3 14.8 in the patients without AMR (p = 0.024). Mortality. The survival analysis revealed that there was a significant difference in overall mortality (log rank p value = 0.024; Fig. 2). There was zero mortality among the non-sensitized groups vs. five (29.4%) in the sensitized group. Median survival in the sensitized group was 36.5 months. Of the five patients that died, all five had a history of AMR episodes (Table 3). The peak PRA of patients who expired was statistically significantly higher than the rest of the sensitized patients (62.00 20.61% vs. 26.92 13.24%; Fig. 3). PRA levels in the AMR-positive group ranged from 41% to 89%. Infections. The difference in the number of patients who contracted an infection in the sensitized vs.
Fig. 3. The difference in mean peak panel-reactive antibody (PRA) among (i) non-sensitized, (ii) alive sensitized patients, and (iii) dead sensitized patients. The numbers in parentheses represent the 95% confidence intervals. As can be seen, the difference between the means is statistically significant between the three groups.
non-sensitized groups was statistically significant at the 0–1 month (76.5% vs. 42.9%, p = 0.05) and at 1–6 months (52.9% vs. 19.0%, p = 0.029; Table 4). These differences were no longer significant when controlled for gender. The difference in the percentage of infected patients was not significant beyond six months. The difference in the mean number of infections between the sensitized group and the non-sensitized groups was statistically significant at the 1– 6 months (1.5 1.7 vs. 0.5 1.1, p = 0.04). This significance at 1–6 months existed even when controlling for gender. There was a larger percentage of sensitized patients who developed CMV infections (29.4% vs. 9.5%, p = 0.21). Table 4. Infections following transplants
Fig. 2. Probability of survival following heart transplant for sensitized and non-sensitized groups (log-rank p-value: 0.024). PRA, panel-reactive antibody. Table 3. Mortality, subcategorized by sensitization status as well as by a history of antibody-mediated rejection (AMR) episodes posttransplant
Alive AMR positive AMR negative Total alive Dead AMR positive AMR negative Total dead
502
Sensitized (n = 17)
Non-sensitized (n = 21)
0 12 12
0 21 21
5 0 5
0 0 0
Sensitized (PRA > 10%) Number of patients with infections (%) 0–1 month 76.5 1–6 months 52.9 >6 months 52.9 Mean number of infections SD 0–1 month 1.4 1.2 1–6 months 1.5 1.7 >6 months 1.6 2.3 Number of patients 29.4 with CMV infections (%)
Non-sensitized (PRA < 10%)
p-Value
42.9 19.0 38.1
0.05 0.03 0.36
0.9 1.4 0.5 1.1 0.5 0.7 9.5
0.30 0.04* 0.07 0.21
This table contains both the percentage and the number of infections seen in sensitized vs. non-sensitized patients. PRA, panel-reactive antibody. *p = 0.063 when we accounted for gender and whether the patient was transplanted due to an infected ventricular assist devices.
Post-transplant outcomes in allosensitized LVAD BTT patients Discussion Pathogenesis and risk factors
Pre-orthotopic heart transplant (OHT) LVAD implantation is known to increase pre-transplant PRA levels (2–7). Although it is well documented that women tend to have a higher pre-transplant PRA levels than men, LVAD-induced rise in PRA levels appears to affect both genders equally (3, 9). Increased sensitization in women is postulated to result from pregnancy and blood transfusions (9). However, multiple studies have shown that the avoidance of blood products did not decrease the production of HLA antibodies provided leukofiltered blood products are used (14, 15). Paradoxically, Drakos et al. (15) showed that transfusions may in fact be associated with less alloimmunization and may mitigate the sensitization. However, some studies have shown that there is an association between platelet transfusion (9) and plasma infusion (16) and allosensitization. Another mechanism that has been proposed is the interaction with the device biomaterials, specifically the textured chamber surface. Evidence of this comes from microscopic analysis of the pseudointima that develops inside the chamber, which was shown to have high concentrations of functionally activated cells of the monocyte/macrophage lineage (17). Other notable basic sciences research in this realm has shown increased aberrant T-cell activation on the LVAD surface, predominance of circulating TH2 cytokines, and polyclonal B-cell activation (18, 19). These TH2 defects were also seen to be better described in older generations (HM VE/ XVE). The theory that alloimmunization may be secondary to interaction with the biomaterials is further supported by studies that have reported an association between the type of device and the degree of alloimmunization. Studies have shown that newer devices such as the Heartmate II, which has axial flow, smaller surface chamber, and no chamber valves have lower degrees of alloimmunization compared to older, pulsatile devices (20). Other studies have noted similar degrees of alloimmunization but have reported greater degree of rejection in the older devices (21). Our data showed no statistically significant difference between the types of devices, quantity of blood products, or duration of support. In regard to the timing of the sensitization, Pagani et al. (22) showed that the median duration of support time to peak PRA was
Clin Transplant 2015: 29: 499–505 DOI: 10.1111/ctr.12540
Clinical Transplantation
Clinical outcomes in sensitized heart transplant patients bridged with ventricular assist devices Kidambi S, Mohamedali B, Bhat G. Clinical outcomes in sensitized heart transplant patients bridged with ventricular assist devices.
Sumanth Kidambia, Burhan Mohamedalib and Geetha Bhata a
Abstract: Background: Left ventricular assist devices (LVADs) as a bridge to transplant (BTT) have been known to cause allosensitization, as measured by panel-reactive antibody (PRA) levels. The goal of this study was to measure the impact of this allosensitization on outcomes. Methods: Panel-reactive antibodies were analyzed in BTT patients, with sensitization defined as peak PRAs ≥ 10%. Baseline characteristics and outcomes in the two patient groups were evaluated using descriptive statistics, Kaplan–Meier, and regression analysis. Results: Thirty-eight patients were included in the study (17 sensitized vs. 21 non-sensitized). There were more women in the sensitized group (47% vs. 10%, p = 0.023). There was no difference in mean times to high-grade acute cellular rejection (ACR; 18.3 months in sensitized vs. 36.9 months in non-sensitized). Five patients in the sensitized groups developed antibody-mediated rejection (AMR) vs. 0 in the non-sensitized, and all five patients died (Kaplan–Meier log-rank p = 0.024). There was also a significant difference in the incidence of infection at the one- to six-month stage (52.9% vs. 19.0%, p = 0.03). Conclusion: Sensitization appears to have a negative effect on mortality. This mortality appears to be concentrated in patients with AMR, and we postulate that the development of AMR in a sensitized patient may be a predictor of mortality.
Division of Advanced Heart Failure and Clinical Transplantation, Advocate Christ Medical Center, Oak Lawn, IL, USA and bRush University, Chicago, IL, USA Key words: acute cellular rejection – allosensitization – antibody-mediated rejection – bridge to transplant – heart transplant – infection – left ventricular assist device – mechanical support – mortality – orthotopic heart transplant – outcomes – panel-reactive antibody Corresponding author: Geetha Bhat, PhD, MD, Medical Director, Center for Heart Transplant and Assist Devices Advocate Christ Medical Center, Heart Institute Oak Lawn, IL, USA. Tel: +1 708 684 7031; Fax: 1-708-520-1875; e-mail: [email protected] Conflict of interest: None. Accepted for publication 10 March 2015
Mechanical circulatory support has gained increasing popularity for patients with end-stage systolic heart failure (1). These devices often serve as a bridge to transplant (BTT), providing mechanical support while a donor heart becomes available. However, these devices have been shown to induce sensitization in that they are associated with the development of circulating anti-HLA antibodies with potential donor reactivity (2, 3). Early reports indicated that these elevated levels of antibodies were linked to poorer outcomes, notably graft rejection (4–6). More recent studies have indicated that this difference may not be as significant as previously stated (7–9). Other studies have noted a positive correlation between increased sensitization and antibody-mediated rejection (AMR) episodes (10, 11). This study sought to evaluate the association between elevated panel-reactive antibodies (PRAs) and its effects on outcomes, specifically: rejections (both acute cellular and antibody mediated), infections, and mortality.
Methods Patient population
This was a single-center study. It consisted of a retrospective analysis of consecutive patients who were bridged to transplant with ventricular assist devices (VADs) and who were transplanted between January 2008 and August 2013. This study was approved by the institutional review board. All patient demographic and outcome data were collected through a combination of chart review and internal databases. Data on peak PRA data were collected, defined as the maximum PRA level prior to transplant. Patients with peak PRA levels 10% were considered sensitized (n = 17), while
500
patient with PRA 10%)
0–6 7–12 13–18 19–24
0.24 0.20 0.09 0.17
0.44 (17) 0.41 (15) 0.302 (11) 0.41 (6)
Non-sensitized (PRA < 10%) 0.24 0.00 0.00 0.00
0.54 (21) 0.00 (19) 0.00 (9) 0.00 (7)
p-Value 0.986 0.164 0.380 0.363
A rejection episode was defined as International Society for Heart and Lung Transplantation 2R or greater. The data were organized by sensitized vs. non-sensitized patients and broken down by six-month time intervals. PRA, panel-reactive antibody.
Impact of VAD implantation on allosensitization
Of the 20 patients for whom pre-VAD PRA levels were available, three were sensitized prior to the LVAD implant (15%). Mean peak PRA for these three patients was 22.33%. There was a statistically significant increase in the PRA levels following VAD implantation with a mean increase of 15.25 19.36 following the implant. Median time to peak PRA following implant was 131.5 d (interquartile range: 33.75–449.5). Impact of elevated PRA levels on cross-match
In the sensitized subgroup, three of the 17 (17.6%) had a positive cross-match vs. two of 21 (9.6%) in the non-sensitized subgroup (p = 0.640). Outcomes
Acute cellular rejection. The mean number of highgrade ACR episodes (as defined as ISHLT grade ≥2R), was not seen to be significant during any sixmonth time interval for the two yr following transplantation (Table 2). However, there were no highgrade rejections among the non-sensitized patients beyond a six-month duration. Nevertheless, the difference in mean number of rejection episodes was not seen to be significant. The mean time to high-grade rejection in the sensitized group was 18.3 months vs. 36.9 months
Fig. 1. Time to development of high-grade rejections (International Society for Heart and Lung Transplantation ≥2R) after heart transplant in the sensitized and non-sensitized patients, as defined by peak panel-reactive antibody (PRA; log-rank p-value = 0.208).
in the non-sensitized group. However, this difference was not seen to be significant (log rank p value = 0.208; Fig. 1). The mean time to all rejections (as defined as ISHLT rejection grade ≥1R) in the sensitized group was 6.13 months vs. 5.04 months in the non-sensitized group. This difference was also not seen to be significant (p = 0.471). Antibody-mediated rejection. Five patients (13.1%) in the patient population experienced AMR episodes. All five were in the sensitized group. This difference was statistically significant (p = 0.012). Of the patients who developed AMR, the mean
501
Kidambi et al.
peak PRA was 51.7 31.3 vs. 11.3 14.8 in the patients without AMR (p = 0.024). Mortality. The survival analysis revealed that there was a significant difference in overall mortality (log rank p value = 0.024; Fig. 2). There was zero mortality among the non-sensitized groups vs. five (29.4%) in the sensitized group. Median survival in the sensitized group was 36.5 months. Of the five patients that died, all five had a history of AMR episodes (Table 3). The peak PRA of patients who expired was statistically significantly higher than the rest of the sensitized patients (62.00 20.61% vs. 26.92 13.24%; Fig. 3). PRA levels in the AMR-positive group ranged from 41% to 89%. Infections. The difference in the number of patients who contracted an infection in the sensitized vs.
Fig. 3. The difference in mean peak panel-reactive antibody (PRA) among (i) non-sensitized, (ii) alive sensitized patients, and (iii) dead sensitized patients. The numbers in parentheses represent the 95% confidence intervals. As can be seen, the difference between the means is statistically significant between the three groups.
non-sensitized groups was statistically significant at the 0–1 month (76.5% vs. 42.9%, p = 0.05) and at 1–6 months (52.9% vs. 19.0%, p = 0.029; Table 4). These differences were no longer significant when controlled for gender. The difference in the percentage of infected patients was not significant beyond six months. The difference in the mean number of infections between the sensitized group and the non-sensitized groups was statistically significant at the 1– 6 months (1.5 1.7 vs. 0.5 1.1, p = 0.04). This significance at 1–6 months existed even when controlling for gender. There was a larger percentage of sensitized patients who developed CMV infections (29.4% vs. 9.5%, p = 0.21). Table 4. Infections following transplants
Fig. 2. Probability of survival following heart transplant for sensitized and non-sensitized groups (log-rank p-value: 0.024). PRA, panel-reactive antibody. Table 3. Mortality, subcategorized by sensitization status as well as by a history of antibody-mediated rejection (AMR) episodes posttransplant
Alive AMR positive AMR negative Total alive Dead AMR positive AMR negative Total dead
502
Sensitized (n = 17)
Non-sensitized (n = 21)
0 12 12
0 21 21
5 0 5
0 0 0
Sensitized (PRA > 10%) Number of patients with infections (%) 0–1 month 76.5 1–6 months 52.9 >6 months 52.9 Mean number of infections SD 0–1 month 1.4 1.2 1–6 months 1.5 1.7 >6 months 1.6 2.3 Number of patients 29.4 with CMV infections (%)
Non-sensitized (PRA < 10%)
p-Value
42.9 19.0 38.1
0.05 0.03 0.36
0.9 1.4 0.5 1.1 0.5 0.7 9.5
0.30 0.04* 0.07 0.21
This table contains both the percentage and the number of infections seen in sensitized vs. non-sensitized patients. PRA, panel-reactive antibody. *p = 0.063 when we accounted for gender and whether the patient was transplanted due to an infected ventricular assist devices.
Post-transplant outcomes in allosensitized LVAD BTT patients Discussion Pathogenesis and risk factors
Pre-orthotopic heart transplant (OHT) LVAD implantation is known to increase pre-transplant PRA levels (2–7). Although it is well documented that women tend to have a higher pre-transplant PRA levels than men, LVAD-induced rise in PRA levels appears to affect both genders equally (3, 9). Increased sensitization in women is postulated to result from pregnancy and blood transfusions (9). However, multiple studies have shown that the avoidance of blood products did not decrease the production of HLA antibodies provided leukofiltered blood products are used (14, 15). Paradoxically, Drakos et al. (15) showed that transfusions may in fact be associated with less alloimmunization and may mitigate the sensitization. However, some studies have shown that there is an association between platelet transfusion (9) and plasma infusion (16) and allosensitization. Another mechanism that has been proposed is the interaction with the device biomaterials, specifically the textured chamber surface. Evidence of this comes from microscopic analysis of the pseudointima that develops inside the chamber, which was shown to have high concentrations of functionally activated cells of the monocyte/macrophage lineage (17). Other notable basic sciences research in this realm has shown increased aberrant T-cell activation on the LVAD surface, predominance of circulating TH2 cytokines, and polyclonal B-cell activation (18, 19). These TH2 defects were also seen to be better described in older generations (HM VE/ XVE). The theory that alloimmunization may be secondary to interaction with the biomaterials is further supported by studies that have reported an association between the type of device and the degree of alloimmunization. Studies have shown that newer devices such as the Heartmate II, which has axial flow, smaller surface chamber, and no chamber valves have lower degrees of alloimmunization compared to older, pulsatile devices (20). Other studies have noted similar degrees of alloimmunization but have reported greater degree of rejection in the older devices (21). Our data showed no statistically significant difference between the types of devices, quantity of blood products, or duration of support. In regard to the timing of the sensitization, Pagani et al. (22) showed that the median duration of support time to peak PRA was
