Torque Teno Virus: Predictor of Infection After Solid Organ Transplant?

(See the Major Article Strassl et al, on pages 1191–9.)

Infection remains one of the greatest causes of morbidity and mortality after solid organ transplantation, due to suppression of the immune system from multiple factors, including immunosuppressive medications, leukopenia, malnutrition, diabetes mellitus, uremia, organ dysfunction, graft organ ischemia, and concomitant immunomodulatory viral infections. Optimal management of immunosuppressive medications after organ transplantation requires a fine balance between the risk of rejection due to inadequate immunosuppression and the risk of infection with excessive immunosuppression. At present, there is no single diagnostic test or algorithm that predicts that perfect amount of immunosuppression; this remains the elusive holy grail of transplant medicine. Such an assay would allow for “personalized medicine,” rather than the more global approaches currently in use.

Strassl et al, in this issue of The Journal of Infectious Diseases, evaluate whether quantification of torque teno virus (TTV) could be a predictive biomarker for risk of infection in kidney transplant recipients; this is an interim analysis of a larger clinical trial [1]. TTV is a highly prevalent and nonpathogenic human anellovirus first isolated in 1997. TTV is transmitted through saliva, blood transfusions, and possibly sexual transmission; transmission by organ transplantation would seem likely, although the impact is unclear, as most adults would have prior infection. Multiple subtypes of TTV have been described [2], and superinfection with new strains at the time of transplant might alter the impact of TTV, as has been seen with cytomegalovirus (CMV). Prior work suggests that low TTV levels are associated with graft rejection and higher TTV levels correlate with risk of infection. This work is the first publication of a prospectively designed trial to evaluate this.

The authors found that those with bacterial, viral, or fungal infection had higher levels of TTV compared to patients without infection (4.2 × 10⁸ [interquartile range {IQR}, 2.7 × 10⁷–1.9 × 10⁹] copies/mL vs 2.9 × 10⁷ [IQR, 1.0 × 10⁶–7.2 × 10⁸] copies/mL; P = .006). Differences in TTV load became evident almost 3 months before the infection (median, 77 [IQR, 19–98] days). Each log level increase of TTV increased the odds ratio for an infection by 23% (95% confidence interval, 1.04–1.45; P = .014). Risk factors for having infection and higher TTV levels included age >56 years for both donors and recipients, male sex, being CMV immunoglobulin G negative, and lymphopenia. Those undergoing retransplantation, often considered to be higher risk for infection, had lower infection rates and TTV levels. Common risk factors for infection did not have statistically different levels of infection and TTV, including obesity, prolonged years of dialysis before transplant, diabetes mellitus, other major comorbidities, use of thymoglobulin, donor-specific antibody, and ABO-incompatible transplants. Similarly, the types, doses, and trough levels of immunosuppressive medications used did not change between the infection and no infection cohorts. Combining multiple risk factors and a logistic regression model to test for an independent association of TTV with infectious disease resulted in odds ratios of 1.2–1.3, which may be statistically significant but perhaps of less clear clinical impact.

TTV levels >3.1 × 10⁹ copies/mL corresponded to a sensitivity of 90% to predict subsequent infections, with a specificity of 20%. The negative predictive value (for the prediction of an infection) was 92% and positive predictive value was 17%. Such sensitivity and negative predictive value could theoretically allow for transplant clinicians to focus on those at higher risk for infection, perhaps by prolonging their prophylaxis, increasing the intensity of monitoring, or changing the immunosuppression by decreasing the doses or altering the types of medications used. Interestingly, rates of rejection were similar in the infection (22%) and no infection (19%) groups, suggesting that TTV levels may not be useful for predicting the risk of rejection.

Prior publications have shown mixed results on TTV and transplantation. These authors previously demonstrated that kidney transplant patients who had donor-specific antibodies and a protocol biopsy with antibody-mediated rejection had 25% of the TTV levels measured in patients without antibody-mediated rejection (P = .003) [3]. Another group looked at both TTV and Epstein-Barr virus (EBV) levels in 98 lung transplant recipients and the risk of bacterial, fungal, CMV, and viral respiratory infections, along with acute rejection; they did not find any significant correlation, concluding that “quantification of TTV or EBV as biomarkers has limited potential for defining the net state of immune suppression” [4]. In a study of allogeneic hematopoietic stem cell transplant recipients, TTV levels fell after intense chemotherapy conditioning, and rose in the first 90 days after transplant, in parallel with the absolute lymphocyte count [5]. Levels were higher in those who developed severe (grades II–IV) acute graft-vs-host disease. Whether TTV would be prognostic of infection or graft-vs-host disease after allogeneic hematopoietic stem cell transplant, or would help with titration of immunosuppressive regimens, remains to be determined.

In this trial by Strassl et al, only 10% of the kidney transplant recipients received induction therapy with thymoglobulin [1], which is low compared with kidney transplantation done in the United States, where nearly 75% of kidney transplant recipients underwent immunosuppression induction with T-cell–depleting agents in 2016 [6]. Although this group did not see a change in TTV levels between those who got lymphodepleting therapy and those who did not, it is possible that lymphodepleting therapy and subsequent low lymphocyte levels may have a major impact on TTV levels, similar to that seen in allogeneic hematopoietic stem cell transplant recipients. Prophylaxis for CMV was shorter and monitoring for CMV and BK virus was less frequent than recommended in recent guidelines, which could impact the trial results, given the clinical interplay we commonly see with viral infections in the transplant setting [7, 8].

It will be interesting to see whether TTV levels could be used to reduce the risk of all infections after organ transplantation, or perhaps for management of infections that rely primarily on reduction of immunosuppression, such as BK nephropathy or EBV viremia. Determining the best threshold for reduction of immunosuppression will remain challenging, given the relatively wide range of TTV DNAemia results. As has been seen with CMV, BK, and EBV, determining the trend and plateau of viral DNAemia in the individual patient may be the most important factor, rather than defining universal thresholds. Given the lack of international standard for TTV, the thresholds from their trial might not be transferable to other centers; work will need to be done to show comparability of results among laboratories, similar to those done related to CMV DNAemia testing [8, 9].

The next study should be a randomized controlled interventional trial to evaluate the impact of TTV levels and changes in immunosuppressive regimens and antimicrobial prophylaxis. Numerous other assays reflect the “net state of immunosuppression”; whether TTV levels will be functional as a tool to manage the risk of infection and immunosuppression remains to be determined. Other tests have been used to evaluate the “net state of immunosuppression” [10], although none are currently widely used in clinical transplant medicine. Examples of such tests, evaluating the risk of infection and risk of rejection in organ transplant recipients, are shown in Table 1; there are additional tests in various states of development, and this table is not meant to be comprehensive. As no single test seems to be highly predictive, it is possible that a combination of tests may be more accurate. Given the risks with overlooking an opportunity to prevent or provide early interception of an infection or episode of rejection, it does not seem that we are ready as a transplant community to migrate to using any of these clinically. We need diagnostics with excellent predictive capacity that are easy to perform, with timely results that provide actionable information to the transplant clinician. Hopefully, this is the way of the future, allowing us to migrate toward a more personalized approach to transplant medicine.

Note

Potential conflicts of interest. C. N. K. has received personal fees from Qiagen and Oxford Immunotec, and grants from BIOHope. The author has submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References