THE MONITORING OF IMMUNOSUPPRESSIVE THERAPY IN TRANSPLANT RECIPIENTS
Transplant recipients generally require prolonged treatment with immunosuppressive agents to prevent rejection of the allograft. Immunosuppressants deplete the entire immune system, so that the body has difficulty fighting off infections. In addition, over time, these potent drugs can damage the kidneys and raise the risk for diabetes, high blood pressure, cancer, and cardiovascular disease. Correct dosing of immunosuppressives is critical. Higher dosing than required may leave the individual susceptible to infections or lead to drug toxicity, while under-dosing can lead to graft rejection and graft loss due to the immune response to the transplanted tissue. Although levels of immunosuppressive drugs are routinely monitored, they do not always correlate with the degree of immunosuppression. Immunosuppressive drugs can have different effects in different individuals despite drug concentrations and dosages that are the same. Infection and organ rejection remain the major causes of death in the majority of transplant recipients.
Monitoring of immunosuppression attempts to balance the dual risks of infection and rejection. It is proposed that individual immune profiles, such as an immune cell function assay, will help assess the immune function of the transplant recipient and individualize the immunosuppressive therapy. Currently, immunosuppression is determined by well-established testing for clinical toxicity and by therapeutic drug monitoring (TDM). The value of TDM is the avoidance of toxic levels and the monitoring of an individual's compliance with the prescribed immunosuppressive therapy. Monitoring of graft function may include a biopsy with histopathology of the transplanted organ, imaging modalities and B-type natriuretic peptide, and/or genomic markers of rejection.
IMMUNE CELL FUNCTION ASSAY
The immune cell function assay (e.g., ImmuKnow®) measures the concentration of adenosine triphosphate (ATP) from circulating CD4+ T lymphocyte cells following in vitro stimulation with phytohemagglutinin (PHA) as an indicator of immune cell function. This measurement is made on heparin anticoagulated whole blood using a luminometer and luciferin/luciferase. The assay is used for the detection of cell-mediated immune response in populations undergoing immunosuppressive therapy for organ transplant. The immune cell function assay detects cell-mediated immunity in an individual's whole blood sample after a 15- to 18-hour incubation with stimulant. During incubation, increased ATP synthesis occurs within the cells that respond to the stimulant PHA. Concurrently, whole blood is incubated in the absence of the stimulant for the purpose of assessing basal ATP activity. Anti-CD4+ monoclonal antibody–coated magnetic particles are added to immunoselect CD4+ cells from both the stimulated and nonstimulated wells. After washing the selected CD4+ cells on a magnet tray, lysis reagent is added to release intracellular ATP. Addition of luminescence reagent (luciferin/luciferase) to the released ATP produces light. The amount of light is then measured by a luminometer, which is proportional to the concentration of ATP. The concentration of ATP is calculated from a calibration curve and compared to ATP level ranges to characterize the cellular immune function of the sample. Immune cell function assay values are suggested to reflect more closely the activity of immune cells in vivo.
In April 2002, Cylex obtained 510(k) clearance from the US Food and Drug Administration (FDA) to market the immune cell function assay (Cylex, Inc., Columbia, MD) based on substantial equivalence of two flow cytometry reagents (“predicate devices”) manufactured by Becton Dickinson: the TriTestTM CD4 FITC/CD8 PE/CD3 PerCP reagent and the MultiTestTM CD3 FITC/CD8 PE/CD45 PerCP/CD4 APC reagent. These reagents are used to determine CD4+ T-lymphocyte counts in immunocompromised patients. The FDA-indicated use of the Cylex immune cell function assay is for the detection of cell-mediated immunity in individuals undergoing immunosuppressive therapy after solid organ transplantation. The FDA's approval for 510(k) clearance was based on results submitted by Cylex from a multicenter study of 44 healthy adults and 127 solid organ transplant recipients (kidney, liver, pancreas, and simultaneous kidney and pancreas). ImmuKnow® assays were performed less than 1 month to greater than 4 years after transplant. Ninety-two percent of the transplant patients had CD4+ ATP values greater than 225 ng/mL. From these data, three zones of individual immune response were defined: 1.) an ATP level at or less than 225 ng/mL indicates that the individual's circulating immune cells are showing a low response to PHA stimulation, and suggests that the individual may be at increased risk of infection; 2.) an ATP level at or greater than 535 ng/mL indicates that the individual's circulating immune cells are showing a strong response to PHA, and that the individual may be at increased risk of rejection; 3.) a moderate ATP level between 225 and 525 ng/mL is the proposed ideal response to PHA stimulation.
In a subsequent manufacturer-supported meta-analysis, Kowalski et al/ (2006) presented data of 504 immunosuppressed solid organ transplant recipients (kidney, liver, heart, and small bowel) from centers throughout the United States where ImmuKnow® assay was used. The prospective data were collected and combined with the data used to support the FDA 510(k) clearance for immune cell function assay. The study reported that only 5% of individuals with ATP levels between 130 ng/mL and 450 mg/mL, and a median ATP value of 249 ng/mL, demonstrated either rejection or infection, so the authors proposed this as the ATP targeted range for immunosuppressed transplant recipients. These ATP values and resulting targeted immunological response zone were different than those proposed in the earlier analysis.The authors suggested that this target immunological response zone was the ideal response to PHA stimulation in immunosuppressed transplant recipients, and that these data may be useful to assess relative risks of infection and rejection. Prospective controlled, double-blind outcome studies are needed to confirm this proposed range.
Cylex proposes that the information yielded by the assay can then be used to help guide professional providers in administering the correct amount of immunosuppressant drugs to prevent organ rejection while avoiding infection.
Pleximmune™ received FDA approval through the humanitarian device exemption process in 2014. The test measures the CD 154 expression on T-cytotoxic Memory cells (TcM) in the recipients. It is a qualitative prognostic test in individuals younger than 21 years old with liver or small bowel transplantation to aid in the assessment of the risk of acute cellular rejection (ACR) and must be used in conjunction with biopsy, standard clinical assessment, and other laboratory information.
Pleximmune™ is intended to be used during pretransplantation and/or early and late posttransplantation. During the pretransplantation period, the blood samples are collected prior to transplantation and the test is used to predict the risk of transplant rejection within 60 days after transplantation. During the early and late posttransplantation period, the blood samples are collected within 60 days (early) after transplantation, and for blood samples collected at 200 or more days (late) after transplantation, the test is used to predict the risk of transplant rejection within 60 days after sampling.
The test measures the risk of rejection by calculating an Immunoreactivity Index (IR). The recipient's TcM response is measured via flow cytometry of marker CD154 after exposure to the donor cells. The recipient's response to “third-party" reference is also calculated. The IR is calculated by forming a ratio between the donor-response and the third-party (reference) response. If the donor-induced response is larger than the “third-party" reference response, then the recipient is considered at increased risk of rejection. If the donor-response is smaller than the “third-party" reference response, then the recipient is considered at a decreased risk of rejection. For pretransplant samples, an IR greater than 1.23 indicates increased risk of transplant rejection, and an IR less than 1.23 indicates a decreased risk of transplant rejection. For posttransplant samples, an IR greater than 1.1 indicates increased risk of transplant rejection, and an IR less than 1.1 indicates a decreased risk of transplant rejection.
ASSESSMENT OF TECHNOLOGY
ImmuKnow®
Cadillo-Chávez et al. (2006) evaluated records of all individuals in a limited sample size of 64 kidney transplant recipients at a single institution and reviewed data for associations among ATP levels and immunosuppression type, doses, and levels; creatinine levels; white blood cell (WBC) count; tissue typing; performed antibodies; as well as ATP levels on infection and rejection and any changes in the ATP levels. There was no correlation found between ATP levels and immunosuppression type, doses, or levels; creatinine levels; WBC counts; HLA; and panel-reactive antibody. However, individuals with moderate or high pretransplant ATP levels had more rejection episodes, while those with ATP levels in the low immune response had more infections. Although acute rejections occurred mostly above 300 ng/mL, this was not significant. When pretransplant ATP values were compared with posttransplant ATP values at the second week, a decrease in ATP values did not correlate with infection.
Gupta et al. (2008) studied 125 adult heart transplant recipients, the majority of whom underwent ImmuKnow® assay testing for more than 1 year posttransplantation. There was no apparent correlation between ATP level and rejection. For individuals who developed an infection, the ATP levels did not differ from those individuals who did not develop an infection. During this study, a unique incident occurred where there was a significant correlation between ATP levels and WBC count but not between ATP level and absolute lymphocyte count, suggesting that nonlymphocytes also may influence the ATP response.
A study by Bhorade et al. (2008) found that the ImmuKnow® assay had high sensitivity but poor specificity for infection in 57 lung transplant recipients. Because ImmuKnow® assay levels were lower in infected lung transplant recipients compared with noninfected recipients and increased with the treatment of these infections, it was unclear if the ImmuKnow® assay reflected those who were over-immunosuppressed and at risk of infection of the bone marrow due to infectious agents. The authors concluded that further study of this assay is needed to determine its role in optimizing immunosuppression in a given transplant recipient.
Reinsmoen et al. (2008) reported on a study of 126 kidney transplant recipients that focused on whether pretransplant immune parameters (i.e., human leukocyte antigen [HLA] mismatch, HLA-specific antibodies, CD4+ cellular response as measured by ATP synthesis, and interferon-gamma precursor frequencies to donor or third-party cells) were associated with posttransplant early acute rejection, unstable creatinine courses, and poor graft outcome. The immune parameters were not used to guide a tailored immunosuppressive therapy. Graft survival and incidence of infection were not reported in this study.
Russano et al. (2009) found no correlation between ATP levels as determined by the ImmuKnow® assay and outcomes in 83 pediatric cardiac transplant recipients. ImmuKnow® assays were performed for which there was no difference in ATP levels between patients with or without acute rejection or with or without infection. Further, the manufacturer's reported target response zone with ATP levels 225 ng/mL or less for infection or 525 ng/mL or greater for rejection were not predictive of infection or rejection. Consideration is also given to the fact that pediatric risk for infection and rejection may correspond to different ATP levels.
In another study, Cabrera et al. (2009) assessed the ability of ImmuKnow® assay to differentiate between acute cellular rejection and recurrent hepatitis C in 42 individuals with hepatitis C–related end-stage liver disease as the indication for liver transplant. The ImmuKnow® assay was performed prior to liver biopsy, and the biopsy samples were interpreted by histopathologists blinded to the results of the ImmuKnow® assay. In individuals whose liver biopsy showed evidence of cellular rejection, the ATP value was very high, and in individuals with recurrent hepatitis C, the ATP value was very low. A statistically significant difference of P<0.001 was noted.The data showed the potential utility of the ImmuKnow® assay in distinguishing hepatitis C from acute cellular rejection and the assay's potential usefulness as a marker for the progression of hepatitis C.
Husain et al. (2009) reported on the correlation between ImmuKnow® assay values and various infectious episodes such as cytomegalovirus disease, viral infection, and bacterial pneumonia, fungal disease, and tracheobronchitis in a prospective cohort of 175 lung transplant recipients. The study findings demonstrated lower ImmunKnow® assay values that independently correlated with increased risk of infection. In addition, the study results also suggested that ImmuKnow® assay may be potentially useful in identifying those with fungal colonization. Further testing is needed in order to validate the study results.
Kobashigawa et al. (2010) studied the effect of ImmuKnow® assay values for 296 heart transplant recipients. The authors concluded that ImmuKnow® assay values may be able to predict the risk of infection in heart transplant recipients, but the association between higher scores and rejection was inconclusive because the average assay value was not significantly different during rejection as compared to the steady state of nonrejection. A larger sample size of rejection episodes and a randomized trial in which immunosuppression could be prospectively guided by the assay values are needed.
Gesundheit et al. (2010) studied immune monitoring in 40 individuals in post–allogeneic stem cell transplantation. These immunosuppressed individuals are susceptible to opportunistic infections, and uncontrolled function of the graft can result in graft-versus-host disease. ImmuKnow® assay values were compared in clinically immunocompromised versus immunocompetent individuals after allogeneic stem cell transplantation. The researchers also compared the reconstitution of WBC count to the ImmuKnow® assay values. The individuals' clinical course correlated with the stratification of immune response established by the ImmuKnow® assay values for the solid organ transplantation, which often differed from their WBC count. However, due to the small number of individuals studied, and the fact that many of the individuals died, relapsed, or were lost to follow-up, this study was not able to establish a clear cause-effect relationship. The authors concluded that ImmuKnow® assay should be evaluated prospectively in clinical trials.
Manga et al. (2010) reported on a study of 16 individuals with hematologic malignancies (multiple myeloma, B- or T-cell lymphoma, and acute myeloid leukemia) undergoing hematopoietic stem cell mobilization prior to autologous transplantation in which the immune cell function assay was used to measure ATP activity. The ATP activity was found to be significantly higher than that measured in immunocompetent individuals. The authors concluded that larger studies are needed to assess the significance of ATP activity in individuals with hematologic malignancies.
Torio et al. (2011) grouped 227 samples from 116 kidney transplant recipients by clinical course. The first group included individuals that were stable, and had no infectious sequelae or acute rejection episode 1 month before or after the immune cell assay. Those who demonstrated an infection, defined as a fever plus at least one positive culture or positive polymerase chain reaction, were included in the second group. The last group included those individuals who demonstrated rejection qualified by a biopsy-proven rejection. Healthy blood donors served as controls. Mean ATP levels in the stable group and in the control group were higher than in the infection group. No difference was observed between the rejection group and the stable group or the control group.
Huskey et al. (2011), in a large retrospective analysis, evaluated ImmuKnow® assay values and compared them with subsequent events of infection and acute rejection over a 5-year period of 583 kidney transplant recipients. The authors concluded that ImmuKnow® assay values at a single point did not identify individuals at risk for future development of clinically significant events. Further studies are needed to define the role of ImmuKnow® assay in immune monitoring of kidney transplant recipients.
Two retrospective studies of kidney transplant recipients found statistically significant correlations between ATP production and WBC count. In a study of 39 patients at a single center in Japan, Nishikawa et al. (2014) reported correlation coefficients (R2) of 0.573 (P=0.03) and 0.510 (P=0.02) for associations between WBC and neutrophil counts, respectively. In this study, ATP levels in five patients who developed viral infections in the early posttransplantation period (<50 days) were within normal limits. Methodologic limitations prevented any conclusion about the association of ATP levels with infections in eight patients in the late posttransplantation period (>120 days).
In a study of 306 patients at a single US center, Sageshima et al. (2014) reported correlation coefficients (R2) of 0.264 (P<0.001) for the association between ATP production and WBC. In this study, mean (standard error [SE]) ATP production in patients with biopsy-proven rejection (389 [56] ng/mL) and borderline/clinical rejection (254 [41] mg/mL) were not statistically higher compared with ATP production in patients without rejection (not reported). Mean (SE) ATP production in patients with opportunistic (349 [48] ng/mL) and other (345 [27] ng/mL) infections were not statistically lower compared with ATP production in patients without infection (not reported).
In a prospective cohort study of 55 patients followed for 3 years, Libri et al. (2014) observed that ATP production was often lower in patients with acute rejection compared with patients without acute rejection, and was often greater in patients with infection compared with patients without infection. Using labelled cutoffs for ImmuKnow®, area under the summary receiver operating characteristic (ROC) curve (AUC) was 0.44 (95% confidence interval [CI], 0.18–0.71) for acute rejection and 0.37 (95% CI, 0.22–0.53) for viral or major respiratory tract infections. In a prospective study of 67 patients undergoing kidney transplant, patients with low preoperative ATP production had statistically fewer rejection episodes than those with high preoperative ATP production (P<0.001). The cutoff used for this analysis was 300 ng/mL.
To optimize ImmuKnow® performance, Quaglia et al. (2014) and Wang et al. (2014) both proposed assessing change in ATP production over time, rather than single values. In a retrospective study of 118 patients, Quaglia et al. reported AUC of 0.632 (95% CI, 0.483–0.781) for infection risk using a cutoff of −30 ng/mL for the decrease in ATP production from month 1 to month 3.29 In a prospective study of 140 patients, Wang et al. reported AUC of 0.929 for risk of acute rejection using a cutoff of 172.55 ng/mL for the increase in ATP production from “right before” the rejection episode to the occurrence of rejection.
Natsuda et al. (2014) assessed ATP production in 28 patients coinfected with human immunodeficiency virus (HIV) and hepatatis C virus (HCV). These patients were all receiving antiretroviral therapy with undetectable viral load in most, and were classified Child-Pugh class A. Results were compared with those of 24 HCV-infected liver transplant recipients and 11 healthy volunteers. Median ATP levels in the HIV/HCV coinfected group (259 ng/ML [range, 30–613]) were statistically higher compared with the HCV monoinfected group (33 ng/mL [range, 6–320]; Mann–Whitney U test, P<0.001) and significantly lower compared with healthy volunteers (446 ng/mL [range, 309–565]; Mann–Whitney U test, P=0.001). In HIV/HCV-coinfected patients, ATP production was significantly correlated with CD4+ cell count (Spearman rank correlation, P=0.03) but not with CD4/CD8 ratio (Spearman rank correlation coefficient, P=0.76). The clinical significance of these findings, for either HCV monoinfected liver transplant recipients or HIV/HCV coinfected patients, is unclear.
Liu et al. (2014) compared ATP production in 22 patients with lupus nephritis (LN) and severe infection requiring hospitalization, 74 patients with lupus nephritis and no infection, and 28 healthy controls. Mean ATP production was significantly lower in patients with lupus nephritis and severe infection compared with noninfected patients and healthy controls (P<0.01 for both comparisons). Mean ATP production in noninfected LN patients did not differ statistically from that in healthy controls. Using a cutoff of 300 ng/mL, sensitivity and specificity for severe infection were both 77%. Strength of the correlation between ATP production and severe infection (r=−0.040, P<0.001) was less than that between C-reactive protein and severe infection (r=0.962, P<0.001).
Pleximmune™
The FDA documents contain a summary of clinical information from a validation study with 122 specimens from 87 transplant patients. In 33 pretransplantation patients, Pleximmune™ had a 57% sensitivity and an 89% specificity at identifying rejection. In 64 posttransplantation patients, Pleximmune™ had an 84% sensitivity and an 80% specificity at identifying rejection. No other details were provided regarding how the study was validated or what standard was used as a comparator. Limitations from this study include its small sample size, that the results of the study are not published in a peer-reviewed journal and therefore did not undergo the peer-review process, and the lack overall study details. The investigators did identify that the performance of the test may be affected by gender- and ethnicity-related effects. This potential gender bias could not be further explored due to the small sample size.
Ashokkumar et al. (2017) performed a preclinical validation of allospecific CD154+TcM to predict acute cellular rejection after liver or intestine transplantation in children. The study included cryopreserved samples from 214 children (<21 years old): 127 (120 analyzable samples) were included in the training set, and 87 (72 analyzable samples) were in the validation set. The IR was calculated as a ratio comparing the reaction of the donor-induced CD154+TcM response to that of reference PBL. If the donor-induced response was greater than the reference-induced response, the ratio would be above 1. An IR of greater than 1.1, the cutoff in posttransplant samples had an 84% sensitivity (95% CI, 60–96), an 80% specificity (95% CI, 65–90), a positive predictive value of 64% (95% CI, 43–81), a negative predictive value of 92% (95% CI, 78–98), and an area under the summary ROC curve of 0.878 in the test set and 0.791 in the validation set. An IR of greater than 1.23, the cutoff for pretransplant samples had a 57% sensitivity (95% CI, 30–81), an 89% specificity (95% CI, 65–98), a positive predictive value of 80% (95% CI, 44–96), a negative predictive value of 74% (95% CI, 51–89), and an area under the summary ROC curve of 0.82 in the training set and 0.842 in the validation set. The limitations of this study include statistically significant differences between the organ type used in the training and validation sets (P<0.007) along with wide confidence intervals for the specificity and sensitivity results.
PROFESSIONAL SOCIETIES
The American Society of Transplantation (AST) does not include the use of the immune cell function assay in their publication of recommendations for screening, monitoring, and reporting of infectious complications in the evaluation of recipients of organ transplantation (AST 2006). The 2019 American Society of Transplantation Infectious Diseases and Community of Practice discusses that immune assays such as ImmuKnow® have “undergone preliminary evaluation as adjunct markers of post-transplant lymphoproliferative disorders risk when combined with viral load testing in pediatric thoracic transplant recipients but require further testing.” In addition, the 2010 guidelines for the care of heart transplant recipients by the International Society for Heart and Lung Transplantation do not include ImmuKnow®. Further, the International Cytomegalovirus (CMV) Consensus Group of the Transplantation Society states that there are no clinical studies that demonstrate that management decisions based on immunologic monitoring affect patient outcomes, and that routine immunologic monitoring is not recommended.
SUMMARY
Published studies to date have primarily been small single-center retrospective studies. Studies indicate that ATP levels vary among transplant recipients who have evidence of acute infection or transplant rejection, compared with clinically stable patients. Sensitivity and specificity of immune cell function assay have varied in studies reporting these parameters. Based on results from two 2012 systematic reviews of observational studies of ImmuKnow® in adult transplant recipients, estimates of sensitivity range from 52% to 88% for infection and 34% to 75% for rejection. Estimates of specificity, on the other hand, range from 66% to 79% for infection and 72% to 84% for rejection. Given the significant heterogeneity observed across studies, performance characteristics of ImmuKnow® have not been conclusively demonstrated. Further, it remains unclear whether different types of organ transplants or different immunosuppressive regimens affect CD4+ T cells’ response to phytohemagglutinin stimulation variably or whether cutoff values require adjustment for various clinical scenarios. Prospective trials are needed to better define the predictive ability of ImmuKnow® compared with current methods of assessing immune status. Other studies evaluate immune cell function for conditions other than solid organ transplantation not supported in the FDA 510(k) clearance for immune cell function assay.
The clinical utility of ImmuKnow® to impact net health outcome in comparison with current methods of care for solid organ transplant recipients has not been evaluated using prospective trials with multiple time point measurements of ATP production. Thus, it is unknown how current methods of assessing immune status in solid organ transplant recipients, e.g., immunosuppressant drug-level monitoring or empiric use of anti-infective agents, might be changed by use of ImmuKnow®.
The clinical utility of Pleximmune™ to impact net health outcomes in comparison with current methods of care for solid organ transplant recipients has not been examined. Evidence for the validity of the test has only been studied in FDA documents in a small population and in a separate preclinical validation study with wide confidence intervals for specificity and sensitivity. There are no studies of the clinical validity of Pleximmune™ in hematopoietic cell transplant. Thus, the clinical validity of Pleximmune™ is uncertain.
