Wasting the Gift of Life?

Once a potential kidney donor has been identified, a simple but crucial question arises. Should the organ be accepted for transplant – or turned away? Factors such as donor and recipient characteristics, anatomic and immunologic information, and longevity matching considerations all influence this call and must be considered quickly. To help inform the decision, surgeons carry out procurement biopsies while the organ is being obtained; these occur in around half of all deceased-donor kidneys in the US (1). Deceased donor kidneys are a scarce and valuable resource – but, alarmingly, we find ourselves facing the highest kidney discard rate of all time, with one in five donated organs going to waste (2, 3). The findings of procurement biopsies play a telling role in these rates; they are listed as the main reason for discard in around 37 percent of kidneys that are procured but ultimately not used (4).

With the results of procurement biopsies contributing so heavily to the discard of potentially transplantable kidneys, it is vital to cast a critical eye over their efficacy at identifying organs that shouldn’t be transplanted. A number of analyses have already raised questions about the reproducibility and predictive value of procurement biopsies, so we conducted our own study to assess their reliability (4). Across 116 kidneys that had undergone multiple procurement biopsies, we found only a 64 percent agreement rate between different biopsies, suggesting low reliability and consistency when it comes to the information that they present. A similar agreement was found between procurement biopsies and gold-standard reperfusion biopsies performed after kidney implantation.

The problem with procurement

Why is there such poor agreement between procurement biopsies and reperfusion biopsies, as well as between sequential procurement biopsies? I think there are a number of factors that contribute to this inconsistency, starting with the environment in which they are conducted. There is great pressure to get these biopsies performed, processed, stained, and read in a relatively short space of time so that the surgeon can decide whether or not they want to keep the kidney. The biopsies tend to be frozen sections that are often read outside usual working hours (the “middle of the night phone call” is real!). Depending on who performs the biopsy and where it is completed, a hospital’s facilities can also have an impact. For example, smaller hospitals might not have a kidney biopsy needle and will therefore have to perform a wedge biopsy, which results in more subcapsular tissue and overestimation of scarring.

Other constraints with procurement biopsies arise because the frozen section procedure introduces a certain amount of artifact, making them difficult to interpret – an issue exacerbated by the fact that renal pathology is not a common skill. There are a relatively small number of renal pathologists available compared with, for example, surgical pathologists, so not everybody who reads the results of procurement biopsies has equal levels of experience. Additionally, given that they are often read in the middle of the night, it’s not infrequent for trainees with much less experience to take on the task, which naturally introduces the potential for error, even with supervision.

Variability also occurs because there isn’t enough standardization in the reporting of results amongst surgeons who carry out procurement biopsies. This is, in part, the result of a small procurement biopsy paper published around 25 years ago that investigated glomerulosclerosis. It showed that kidneys with glomerulosclerosis coverage over 20 percent were not suitable for transplant, whereas those with less than 20 percent glomerulosclerosis performed well (5). Although not true across the board, this threshold for kidney quality has taken hold and is used by many centers to decline an organ.

The issue is that there isn’t a standardized protocol to enforce this measure, so everyone uses their own version. Reports frequently vary from “less than/more than 20 percent glomerulosclerosis” to “mild/moderate/severe glomerulosclerosis.” It’s also not uncommon to see reports that simply read “good kidney” without any numbers or further explanation. This kind of “information” effectively blinds the surgeons and forces them to apply their own interpretation of what constitutes mild, moderate, or severe disease. Surgeons carrying out the transplant often don’t know the pathologist who completed the biopsy, where it was read, or whether it was a wedge or a core biopsy. This last unknown can have important effects because they are so different – wedge biopsies are more superficial and can show a lot more scarring and glomerulosclerosis than you’d expect in the rest of the organ, requiring the surgeon to adjust their interpretation accordingly.

Changing the rules

All of these issues contribute to the discard of a large number of kidneys that don’t necessarily need to go to waste. Our belief is that it is rare to find anything on procurement biopsy that should preclude a kidney’s use for transplantation. The primary concern for surgeons making the decision on whether to transplant or not is in deciding whether a deceased donor’s kidney is diseased or injured. For example, if a patient has died with an isolated creatinine level of 3, it can be hard to discern whether that value is indicative of chronic kidney disease or acute kidney injury. In these sorts of situations, we can apply procurement biopsy results in the right way – that is, by using them to rule in the transplant of a kidney once the surgeon concludes that it’s acutely injured, rather than chronically diseased. Unfortunately, in the US, many programs are using procurement biopsies the opposite way – applying the results to rule out the use of a kidney.

Part of the problem is that we have changed the way donors are described. Instead of paying close attention to the individual clinical characteristics of a patient, we place emphasis on a single score of clinical variables known as the kidney donor risk index (KDRI). The problem occurs when the kidneys of older donors are judged by their KDRI score and glomerulosclerosis percentage, which often leads to discard. In reality, 20 percent glomerulosclerosis in an older donor may be an age-appropriate amount of scarring, and the kidney would be fine for a recipient who is of a similar age. Essentially, this is double-counting the donor’s age against them.

The gold standard of biopsies in this scenario is the reperfusion biopsy, which is performed after anastomosis to learn more about the success of the transplant. Performed in the absence of any time pressure, these core needle biopsies are consistently read by experienced pathologists using paraffin-embedded tissue and multiple stains. When comparing the findings of 270 reperfusion biopsies with those from procurement biopsies on the same kidneys, we found a 64 percent agreement rate between the two (4).

This low agreement between the two types of biopsy highlights the need to rethink our use of procurement biopsies. Reperfusion biopsies read by renal pathologists were consistently associated with post-transplant outcomes, whereas the procurement biopsies read by on-call pathologists were not.

Operating change

We can – and must – do several things to rectify this growing issue. For one, I think we need to stop carrying out a biopsy when there isn’t a prior indication that calls for one. This will be tough because everyone has become so accustomed to having these results, but it would certainly bring down the high discard rates. Second, when we do carry out biopsies, they need to be standardized in terms of the process and the report. Pathologists can achieve this standardization by creating a set of criteria and developing a consistent approach to reporting results. I also think that pathologists should always have the clinical characteristics of the donor available to them. They should start by asking, “Is this an age-appropriate kidney?” and, “Is the amount of scarring we see on the biopsy consistent with the clinical history?” If so, then at least the surgeon can recognize that the findings are what healthcare professionals should expect and therefore not count it as a strike against the kidney.

We should also attach a degree of transparency to the person reading the procurement biopsy, perhaps by indicating their level of renal pathology expertise; although challenging to implement, I think it is a step in the right direction. Finally, when we perform a procurement biopsy, if we’re going to use a scoring guide such as the KDRI, I think we need to incorporate the biopsy findings into that score so that there isn’t an inadvertent double-counting of adverse factors.

From experience, my colleagues and I have found that kidneys are often turned away from other centers around the country due to the findings of a procurement biopsy. Some of those organs then get allocated to us in New York but, when we look at the biopsy, the quality is so poor that we can’t understand it. Our region frequently completes a second biopsy – and it often disagrees with the first. It’s possible that this is because of the haste with which the initial biopsy was completed. The median cold ischemia time in the US is currently around 16 to 17 hours, but is that truly the upper limit? Our center commonly transplants kidneys with over 30 hours of cold ischemia. Although fast action is key when it comes to kidney transplantation, this is still a reasonable amount of time and certainly enough to get things right. I urge pathologists conducting procurement biopsies to take time and care over the first one – better to spare a few additional moments than to risk discarding a perfectly acceptable kidney!

Limiting our reliance on procurement biopsy histology will result in greater organ utilization, which could drastically increase organ allocation efficiency and reduce the exceedingly high discard rates that we are currently facing. We all have the same goal: to ensure that every patient in need receives a healthy, usable kidney.