For those requiring a new kidney, receiving a transplant marks the end of a stressful period of securing a donor organ. However, it also signals the beginning of another nerve-racking chapter of facing the potential for organ rejection. Lorenzo Gallon, a nephrologist at Northwestern Medicine, has seen his patients grapple with this complex emotional toll. “They feel that they are receiving a precious organ; they’re lucky to have received it,” Gallon said. “The patient has no connection with the organ; there's no way for the patient to understand on a daily basis that their kidney transplant that is sometimes given by a loved one or family member is actually healthy.”

At any time after a transplant, the immune system may recognize the organ as foreign and attack. Rejection of the kidney, one of the mostly commonly transplanted organs worldwide, is a major driver in a cycle of organ loss, reinitiation of dialysis treatment, and growing transplant lists. The most accurate method for detecting kidney transplant rejection is a biopsy of the kidney tissue, an invasive procedure that is performed infrequently due to the risk of bleeding, infection, and damage to neighboring organs. Physicians rely on the blood biomarkers serum creatinine and blood urea nitrogen to obtain more regular updates on the kidney’s status, but these blood tests lack the specificity and sensitivity to reliably report rejection. The biomarkers can fluctuate due to other factors such as diet and medication, leading to false positive and negative results. 

“Serum creatinine is a measure of kidney filtration and function, but people can have a lot of injury and inflammation before we lose enough function for it to be detected on these routine clinical tests,” said Julie Ho, a transplant nephrologist at the University of Manitoba. Changes in the kidney’s filtration rate may also take days to weeks to show up in the blood as elevated serum creatinine. For these reasons, “even with normal blood biomarkers in the bloodstream, if you do a surveillance biopsy, in some of these organs, you have rejection,” Gallon said. 

Gallon and Ho are creating tools to better detect these instances of subclinical rejection, an early form of rejection that is apparent in the biopsy sample but not in the kidney function biomarkers in the blood. “[If] you have a red flag, you do a biopsy to confirm the rejection, then you can intervene and treat the rejection much earlier,” Gallon said. 

If rejection is caught in time, clinicians can adjust a patient’s anti-rejection medication regimen; they must do so judiciously to find the optimal balance of immunosuppression to combat rejection without increasing the risk for infection and disease. By developing an implantable temperature-sensing device and a simple urine test to detect kidney rejection earlier, the researchers aim to provide patients peace of mind, improve transplant outcomes, and preserve valuable donor organs. 

Turning up the temperature

Gallon recognized a need for an implantable device that could measure various biophysical parameters to enable continuous monitoring of a transplanted kidney. To bring his vision to life, he recruited the help of John Rogers, a biomedical engineer and materials scientist at Northwestern University. The researchers hypothesized that the inflammation that occurs during rejection may cause detectable changes in the organ’s temperature and blood flow rate.

Rogers’s team developed sensor technology that measures the electrical resistance of gold, which depends on temperature. The researchers can use this property to calculate the temperature of the kidney, and by incorporating a separate gold heating element, the rate of blood flow. Passing electric current through this element generates a small amount of local heat, which dissipates more rapidly the faster the blood is flowing through the organ. By measuring how quickly the increase in temperature decays, the researchers can determine the blood flow rate. 

The team then incorporated their temperature and blood flow sensors into a device that could be transplanted into a rat. They used ultrathin gold wires arranged in a serpentine configuration to create a flexible sensor that would not damage the fragile kidney tissue. They also designed the sensor to be small enough to slip under the capsule that surrounds the kidney, enabling strong thermal contact with the organ. The sensor is connected to an electronics module that is sutured to the abdominal wall. The module contains a Bluetooth system that transmits the data to a smartphone and a coin cell battery power supply that can be recharged through the skin using technology similar to a wireless charging pad. To prevent a foreign body response to the device that could affect the temperature measures, “we're pretty careful about the materials we select. Materials-level biocompatibility is very important, but you also want mechanical compatibility so you don’t create irritation,” Rogers said. “We didn't observe any effect like that, but it's a very important issue.” 

In a recent study in Science, the researchers transplanted kidneys into rats with different immune system genetics, creating an isogeneic model that leads to organ acceptance and an allogeneic model that results in organ rejection (1). They implanted the device into each type of rat and measured the organ temperature and blood flow rate in the absence of immunosuppressive drugs. They observed that the blood flow parameter did not display a distinct trend in the allogeneic model compared to the isogeneic model. While inflammation is expected to increase blood flow, damage to the blood vessels due to rejection may obstruct blood flow, canceling out this effect. Alternately, the uptick may be obscured by the high baseline blood flow rate of the kidney or occur nonuniformly throughout the organ, potentially outside of the sensor interface. 

The results for the temperature parameter, however, were consistent with the observation that inflammation generally increases temperature. Three days after the transplant, the team observed a spike of approximately 0.6°C in the allogeneic rats that was not present in the isogeneic rats, revealing a potential temperature signature of rejection. When they sacrificed the animals and analyzed the kidney tissue, they found histological signs of rejection. The researchers also measured the rats’ blood biomarkers at this timepoint and observed no significant difference between the allogeneic and isogeneic models. “Our device was able to detect an inflection point in the temperature... and that inflection point corresponded with the rejection of the kidney, even when the [serum] creatinine and the blood urea nitrogen were perfectly normal,” Gallon said.  

Our device was able to detect an inflection point in the temperature... and that inflection point corresponded with the rejection of the kidney, even when the [serum] creatinine and the blood urea nitrogen were perfectly normal. 
- Lorenzo Gallon, Northwestern Medicine

Next, the team administered an immunosuppressive drug to the rats and then discontinued it, simulating more realistic scenarios of patient noncompliance with or resistance to anti-rejection medication. They again detected an increase in temperature unique to the model of kidney rejection, which occurred after they stopped immunosuppression and was less dramatic (approximately 0.15 °C) compared to the unmedicated experiment. During the postmedication period, these rats also displayed a high-frequency temporal temperature cycle that deviated from the normal circadian rhythm of temperature. The researchers aren’t sure why this pattern emerged upon introduction of immunosuppressive drugs, but they found that the two temperature features coincided with histological rejection and preceded changes in blood biomarkers by two to three weeks. “We are actually able to diagnose subclinical rejections in this model by looking at temperature variations,” Gallon said. 

In light of these promising results, the team is now performing studies in pig models with the goal of translating their technology for use in humans. They envision a device that could be implanted at the end of a transplant surgery to continuously monitor the health of the kidney and provide early warning signs of rejection, guiding further testing and treatment. While there are other processes in the human body that could cause an increase in organ temperature, the researchers could distinguish this effect by placing one sensor on the transplanted kidney and another on the healthy kidney. “If it's something associated with an overall fever or something like that, then both temperature sensors would go up,” Rogers said. “But if the transplanted one's going up but the healthy one is not, that provides the asymmetry that would indicate something anomalous associated with the transplanted kidney and not with the overall body temperature.” 

The researchers are also exploring the potential to adapt the device for application to other organs. “As we learn more about the biophysical events that happen during rejection, the device can be designed to capture those,” Gallon said. “Temperature will probably [always] be there; then we can add different features based on the organ characteristics.” 

Go with the flow

Meanwhile, Ho’s team is conducting studies in human cohorts to evaluate biomarkers that may better detect subclinical kidney rejection. They focus on CXCL10, a signaling protein chemokine first identified as a potential marker of rejection in primate models by Stuart Knechtle, a transplant surgeon at Duke University (2). CXCL10 plays a role in rejection by recruiting T helper immune cells to areas of inflammation.  

The researchers measure CXCL10 in the urine, a biofluid that is noninvasive to collect and often used to evaluate kidney health. In observational studies of pediatric and adult kidney transplant patient populations, they compared urinary CXCL10 concentrations to biopsy results and found that the protein was elevated in cases of rejection (3,4). “There’s kind of a stepwise increase. With the increasing severity of rejection and inflammation, the higher the CXCL10 levels were,” Ho said.   

CXCL10 was elevated even when the biopsy was performed for routine transplant surveillance (while the serum creatinine level was normal) rather than indicated by a high serum creatinine concentration. This demonstrates that the urinary protein outperforms the blood biomarker and can successfully detect subclinical rejection. In fact, CXCL10 can rise up to four weeks in advance of an early clinical rejection episode that could be detected via serum creatinine. 

To quantify CXCL10’s ability to detect rejection, the team calculated the area under the curve (AUC), a statistical score that increases with diagnostic accuracy and has a maximum value of 1.0. In multiple patient cohorts, the researchers determined the AUC of CXCL10 to be approximately 0.74, compared to approximately 0.54 for serum creatinine. “It's not perfect; it's not 100 percent sensitive in unselected populations; it's not 100 percent specific — that's with many tests in medicine — but it is much better than our current gold standard for monitoring kidney transplants,” Ho said. 

The researchers found that CXCL10 appears to be elevated when inflammation occurs in kidney compartments associated with urine production, indicating that urinary tract infections and kidney infections such as polyomavirus could serve as confounding conditions. Alternately, the test may not pick up on inflammation that is restricted to the separate vascular compartment. However, isolated vascular inflammation is rare and a source of debate as to whether it truly constitutes rejection. 

Ho’s team is now conducting a large clinical trial at health centers throughout Canada and Australia to assess the potential of urinary CXCL10 testing to improve kidney transplant outcomes (5). The researchers measure CXCL10 concentration at fixed intervals for one year in patients who recently received a kidney transplant. Participants showing elevated levels of CXCL10 are randomized to either the intervention group, which undergoes a biopsy and appropriate treatment if rejection is confirmed, or the control group, which receives standard post-transplant monitoring. The researchers will then compare the groups’ outcome measures, including the incidence of rejection-associated de novo donor-specific antibody development, biopsy-proven rejection, and transplant failure, allowing them to evaluate whether early detection of rejection by CXCL10 leads to more successful kidney transplants (6).  

It's not perfect; it's not 100 percent sensitive in unselected populations; it's not 100 percent specific — that's with many tests in medicine — but it is much better than our current gold standard for monitoring kidney transplants. 
- Julie Ho, University of Manitoba

The trial investigators are currently measuring the CXCL10 protein with an antibody assay that uses equipment that is standard in research labs, but not clinical ones. To enable fast-turnaround, on-site testing in clinical transplant settings, the researchers are developing a bead-based antibody assay that relies on the Luminex instrument, which is widely available for antibody screening in transplant immunology labs. Ideally, the testing technology would one day be even more accessible. “Ultimately, the dream would be to have something like the pregnancy test, which is an antibody-based [test] on a paper strip. You just pee on it... and it would help with home and remote monitoring,” Ho said. “That’s going to take some development time.” 

Moving forward, Ho’s team is interested in integrating other biomarkers of kidney transplant rejection into their platform. “Are there things that you could add to a panel that are tracking different pathways? And by using multiple things at the same time, is it possible that you could improve the overall diagnostic performance than just using one marker?” Ho said. The team is now exploring the value of combining measures of CXCL10 and donor-derived cell-free DNA, which is shed upon injury to the kidney (7). “It’s something that has promise,” Ho said. “But getting a reproducible assay that can be implemented in labs might be one of the bigger barriers there.” 

Regardless of the directions her research takes, Ho ultimately hopes to enable a smoother post-transplant journey for the patients she treats. “Anything I'm doing is because I just want to make sure that when I'm looking after my patients, I'm doing my best for them,” she said. 

References

1. Madhvapathy, S.R. et al. Implantable bioelectronic systems for early detection of kidney transplant rejection. Science  381, 1105-1112 (2023).  
2. Hancock, W.W et al. Beneficial effects of targeting the CXCR3 chemokine receptor in non-human primate renal allograft recipients. Am J Transplant  4, 287 (2004). 
3. Blydt-Hansen, T.D. et al. Validity and utility of urinary CXCL10/Cr immune monitoring in pediatric kidney transplant recipients. Am J Transplant  21, 1545-1555 (2021). 
4. Hirt-Minkowski, P. et al. Detection of clinical and subclinical tubulo-interstitial inflammation by the urinary CXCL10 chemokine in a real-life setting. Am J Transplant  12, 1811-1823 (2012). 
5. Ho, J. et al. Multicentre randomised controlled trial protocol of urine CXCL10 monitoring strategy in kidney transplant recipients. BMJ Open  9, e024908 (2019). 
6. Zhang, R. Donor-specific antibodies in kidney transplant recipients. Clin J Am Soc Nephrol  13, 182-192 (2018). 
7. Bloom, R.D. et al. Cell-free DNA and active rejection in kidney allografts. J Am Soc Nephrol  28, 2221-2232 (2017).