We present a system for approaching kidney transplants with late deterioration of function. In the past, renal transplant deterioration was approached as a problem of “chronic rejection,” a poorly defined term that lumped together many conditions. The concept of chronic rejection implied an assumption that deterioration of a kidney occurred because its past injuries had programmed inexorable loss of function. New evidence challenges this assumption. In most cases, deterioration occurs because of a new or recent source of injury. It cannot be assumed to be programmed by early damage but is rather a consequence of new active disease, often rejection but also others. Thus we now approach kidney transplant deterioration differently. If a graft is losing
function, a cause is operating in the present, and the aim is to identify the distinct disease entities or stresses that are causing that deterioration. Some causes will be treatable, while those that are not should be diagnosed and studied so that they may be treatable in the future. To this end, we need to approach late deterioration (loss of function or proteinuria) of a renal transplant by identifying distinct entities or states or components of the problem. We should (a) assess the state of the parenchyma and arteries, including fibrosis and atrophy, which reflects previous nephron injury and loss; (b) establish whether there is active rejection (alloimmune injury), and whether it is T-cell-mediated or antibody-mediated, and if so, why the immunosuppressive strategy has failed; (c) identify transplant glomerulopathy, defined by double contours, which reflects anti HLA antibody in at least some cases; (d) search for specific diseases: recurrent disease, de novo disease, viral and bacterial infections, drug-induced damage, and obstructive uropathy; and (e) identify and treat progression factors such as hypertension and proteinuria which can affect the outcome of any of the above pathologies. With this approach many causes of both early and late transplant deterioration become potentially treatable, providing a step toward achieving the goal of permanent survival of renal transplants as a cure for end-stage renal disease (ESRD).

Both early and late kidney graft survival has improved considerably over recent years, due to many advances in clinical care, with immunosuppression being the most important. The old cliché that control of rejection changes short-term outcomes but not long-term outcomes is not correct and should be discarded. It is reasonable to assume that many of the kidney transplants functioning today will serve their new owners for their life expectancy. In a sense, death with kidney function is the desired endpoint for everyone, renal transplant or not. Premature death is a problem in transplants that must be addressed (1) and is outside the scope of this discussion. Our focus is on those kidneys that are deteriorating in the months and years after the transplant. Admittedly, it is artificial to separate death with function from renal insufficiency: renal insufficiency increases mortality from other conditions, and other morbidities aggravate renal insufficiency. Nevertheless our focus will be on kidney function and survival.

The main cause of late kidney deterioration is still rejection, defined as graft injury due to an ongoing alloimmune response, early or late. Often this reflects a failure of the immunosuppressive prescription as a result of medical decisions or patient noncompliance, leaving patients on ineffective immunosuppression. Breakthrough rejection also occurs because physicians want to minimize exposure to risks of immunosuppressive drugs (ISDs), so called drug minimization. The range of other renal diseases is increasing, and ESRD in a transplant, like ESRD in native kidneys, is certain to have some cases that cannot be classified or diagnosed. But many who follow transplant patients believe that there is a system problem: inadequate follow-up and response to change. The main message of this chapter is that long-term renal transplant care demands intelligent and knowledgeable follow-up and prompt response to change.

There are currently about 110,000 to 120,000 kidney transplants functioning in the United States and Canada, of which about 7% fail each year. Death with a functioning graft accounts for about 40% to 50% of late graft failure, but the remainder are lost because the kidney function deteriorates. About 4,200 renal transplants fail per year and return to dialysis, making transplant deterioration the fourth largest cause of ESRD and entry into dialysis in the United States. Although far behind some causes of ESRD (e.g., diabetic nephropathy causes about 40,000 patients to enter ESRD programs per year), transplant deterioration is nonetheless an important renal disease for the clinician to manage.

Data in national databases on the causes of failure are of limited value because the classification of failure is not standardized and many kidneys are not biopsied. Our estimates from multiple sources are provided in Table 25.1. The causes differ in the early versus late period, and with live donors versus cadaver donors. Although outright rejection still destroys some kidneys, death with a functioning graft is the largest single cause of transplant failure, accounting for approximately half of all losses. Of those kidneys that fail without death of the patient, the commonest phenotype is the general category of non-specific deterioration, in the absence of a specific kidney disease (2,3). This category was previously called chronic rejection or chronic allograft nephropathy (CAN): progressive deterioration with scarring and atrophy. Nonspecific deterioration causes about 30% of all kidney failures. The remainder includes recurrent glomerulonephritis (3), other recurrent and de novo renal diseases (approximately 10%), and multifactorial, unclassified, and undiagnosed causes (about 10%), e.g., development of ESRD in the context of a severe medical illness.

The concept of chronic rejection arose when researchers realized that rejection sometimes followed a slow time course in animal models (4), and that one of the first human kidney transplants followed a slow progressive course over months (5). The latter case showed extensive obliterative arterial lesions and may have initiated the belief that all deteriorating kidneys have obliterative arterial lesions. Indeed, nephrectomy specimens usually do have very advanced obliterative lesions, but they have often been removed late after their failure, and the arterial lesions may have been late developments in a failed kidney, not causes of failure. Obliterative arterial lesions are seldom observed in biopsies of kidneys that are beginning to show loss of function.

The term chronic rejection was in regular use by the mid 1960s to denote a slow late rejection process (6), with proteinuria and nephrotic syndrome as a manifestation (7). Still later, the problem of transplant coronary artery disease appeared in heart transplants (8). Jeannet et al (9) described some cases in which alloantibodies were present in association with graft deterioration, and Pierce et al (8) described donor-specific immunoglobulin G (IgG) antibody in relationship to chronic rejection of human renal allografts.

Thus by the mid 1970s the main elements in a vague definition of chronic rejection had been described, elements that would persist for about 2 decades: slow deterioration, obliterative arterial changes, glomerulopathy, and a possible role for alloantibody. In addition, a number of investigators began to appreciate that early acute rejection was associated with an increased risk of late graft loss of unspecified pathology. This introduced a new concept that late deterioration was somehow an effect of early rejection. Rat models helped drive some of the concepts in this area, despite the fact that these models mainly develop focal segmental glomerulosclerosis (FSGS), a lesion very distinct from human kidney deterioration of transplant glomerulopathy (TGP). Simply put, the relevance of rodent models of “chronic rejection” or “chronic allograft nephropathy” to human problems is not known and cannot be assumed.

From the mid 1990s many investigators concluded that the term chronic rejection was encrusted with different and even contradictory meanings that could not be shed, and that the term must be avoided. To some it meant a specific vascular lesion, to others all late graft loss, to yet others a glomerular disease with proteinuria. Many people used the term “chronic rejection” to mean that it may have something to do with rejection but without the need for any rigorous definition. For decades this was good enough: our focus was early rejection and improvement in early outcomes.

Some attempts at a rigorous definition simply underlined the problem. Once when attempting to construct a clinical trial, I defined chronic rejection as having four elements: dysfunction, steady progression (deteriorating at an arbitrary rate), proteinuria, and specific arterial lesions. We then went back to the renal transplant population and found that cases meeting the rigorously defined disease did not exist. Thus our belief system was not compatible with the behavior of the transplant population.

The multiple meanings of chronic rejection in the literature create a circle of confusion that could easily mean that two people using the term were describing fundamentally different processes. Part of the problem lies in the word chronic, which can mean late, persistent, slow, irreversible, or can imply certain pathologic features (fibrosis, atrophy). Strangely, chronic inflammation refers to the presence of lymphocytes and mononuclear cells. Thus acute rejection is chronic inflammation. But some used the term chronic rejection for cases that had no inflammation at all, only scarring and atrophy.

As we now focus on improving late outcomes, we must develop precise definitions of rejection and identify relatively homogeneous conditions. Rejection can occur early or late and can be rapid or slow in its course and can be antibody-mediated or T-cell-mediated. Our goal above all is to say what we mean in a way that everyone understands the terms.

Slowly investigators began to realize that the course of late deterioration was not relentless or predictable but actually erratic (10). Both the level of glomerular filtration rate (GFR) and the rate of previous loss of GFR correlate with the probability of kidney graft survival. A 30% chronic decline in inverse serum creatinine is a correlate of late renal allograft failure (11). As many non-immunologic population factors correlate with an increased risk of late graft deterioration, late graft loss can reflect nonimmunological influences (12), such as obstruction, infection, drug toxicity, and recurrent disease.