Introduction: Choices in Therapy for the Diabetic Uremic Patient

Suicide (withdrawal)

Diabetic end-stage renal disease (ESRD) patients are managed differently depending on age, severity of comorbid disorders, available local resources and patient preference. Coping with a seemingly unending series of comorbid complications throughout the course of ESRD therapy, diabetic patients opt to discontinue treatment, meaning passive suicide, more frequently than do nondiabetic patients.(1) Withdrawal from ESRD therapy is understandable for blind, hemiparetic, bed-restricted limb amputees for whom life quality is unsatisfactory. By contrast, with a strong social support system providing attention to the total patient, satisfactory life quality unforeseen at the time of ESRD evaluation may be attained.(2) Indeed, survival of diabetic ESRD patients - whether treated with a kidney transplant or by peritoneal dialysis or hemodialysis - has improved progressively over the past decade (Fig. 1).

Figure 1. Improving Prognosis: Diabetic ESRD Patients.

Whether treated by dialysis (combined peritoneal and hemodialysis) or a kidney transplant, survival of diabetic ESRD patients has improved continuously over the past decade. A continuing rise in absolute number and proportion of total diabetic ESRD patients is shown in the statistics provided by the United States Renal Data System 1998 Report.

Diabetes mellitus in 1998 is the leading cause of ESRD in all industrialized (well-fed) nations. Both the incidence and prevalence of ESRD in diabetic patients has risen yearly over the past decade. Most recently, as listed in the 1998 report of the United States Renal Data System (USRDS), in 1996, of 283,932 U.S. patients receiving either dialytic therapy or a kidney transplant, 92,211 had diabetes,(3) a prevalence rate of 32.4%. Even more instructive is the fact that during 1996, 30,393 (43%) of 72,000 new (incident) cases of ESRD were attributed to diabetes.

Adequate comprehensive medical care is only infrequently given to individuals with diabetic nephropathy. Whether assessed in Europe or the US, the usual level of support for diabetic patients with symptomatic kidney disease is incomplete. In Germany, for example, Pommer et al. termed "preterminal care in diabetic patients with ESRD" deficient in amount and quality(4) with ineffective attention to control of hypertension, hyperlipidemia or ophthalmologic intervention.(5) Similarly, in a US study, Ifudu et al. found that because of excessive comorbidity: "Few elderly, diabetic hemodialysis patients conduct any substantive portion of their lives outside their homes."(6) Dissatisfaction with life quality during renal replacement therapy in the United Kingdom prompted a high rate of withdrawal therapy due to comorbid medical problems (89%), accounting for 17% of all deaths.(7)

Hemodialysis

Maintenance hemodialysis is the only renal replacement regimen that will be employed for the large majority - over 80% - of diabetic persons who develop ESRD in the US, Europe or Japan. Although enthusiasts for peritoneal dialysis for diabetic ESRD patients previously imputed survival superior to that attained by hemodialysis, careful registry analyses do not support such advantage. According to the USRDS, approximately 12% of diabetic persons with ESRD will be treated by peritoneal dialysis while the remaining 8% will receive a kidney transplant.

Establishment of a durable vascular access to the circulation is a precondition for performance of maintenance hemodialysis. The most reliable variety of vascular access - an internal arteriovenous fistula in the wrist - is often more difficult in a diabetic than in a nondiabetic person because of advanced systemic atherosclerosis. Because of peripheral vascular calcification and/or atherosclerosis, creation of an access for hemodialysis in many diabetic ESRD patients necessitates resort to synthetic (Dacron) prosthetic vascular grafts.

A typical hemodialysis regimen consists of three weekly treatments lasting 4 to 5 hours each, during which extracorporeal blood flow is sustained at 300 to 500 ml/min. Anemia in hemodialysis patients responds to recombinant erythropoietin, while metabolic bone disease is minimized by treatment with synthetic vitamin D which binds dietary phosphate to calcium supplements within the gut. Motivated patients trained to perform self-hemodialysis at home gain the longest survival and best rehabilitation afforded by any dialytic therapy for diabetic ESRD. As conducted at an outpatient facility, hemodialysis-treated diabetic patients fare less well, receiving significantly less dialysis than nondiabetic patients due, in part, to dialysis treatments that have been interrupted or discontinued by hypotension and reduced access blood flow.(8)

Rehabilitation of diabetic patients during maintenance hemodialysis ranges from poor to abysmal. Lowder et al., in 1986, reported that, of 232 diabetics on maintenance hemodialysis, only seven were employed while 64.9 percent were unable to complete their routine daily activities without assistance.(9) Approximately 45% of diabetic patients begun on maintenance hemodialysis die within two years of their first dialysis. Death on hemodialysis most often results from heart disease though cerebrovascular disease and infection also contribute to excess mortality.

Peritoneal Dialysis

Peritoneal dialysis is successfully employed as a first choice regimen for diabetic ESRD patients.(10) Key advantages of continuous ambulatory peritoneal dialysis (CAPD) or machine sustained continuous cyclic peritoneal dialysis (CCPD) over maintenance hemodialysis include: freedom from a machine, performance at home, rapid training, minimal cardiovascular stress and avoidance of heparin.(11) A typical CAPD prescription consists of four to five, two to three liter dialysate exchanges, seven days a week. Anemia and metabolic bone disease are managed as in hemodialysis with erythropoietin and synthetic vitamin D and oral calcium phosphate binders. CAPD can be a difficult to sustain, enervating regimen that delivers suboptimal solute removal.

No matter how sorted and analyzed, survival of diabetic ESRD patients treated by CAPD/CCPD is inferior to that attained by hemodialysis (Figs. 2-7).

Figure 2. ESRD Death Rate by Modality: PD vs. Hemodialysis.

In each age group, peritoneal dialysis (CAPD/CCPD) in diabetic ESRD patients has inferior survival compared with hemodialysis. United States Renal Data System 1998 Report.

Figure 3. ESRD Death Rate in Diabetes: PD vs. Hemodialysis.

In men, peritoneal dialysis (CAPD/CCPD) in diabetic ESRD patients has inferior survival compared with hemodialysis. United States Renal Data System 1998 Report.

Figure 4. ESRD Death Rate in Diabetes: PD vs. Hemodialysis.

In women, peritoneal dialysis (CAPD/CCPD) in diabetic ESRD patients has inferior survival compared with hemodialysis. United States Renal Data System 1998 Report.

Figure 5. ESRD Death Rate in Diabetes: PD vs. Hemodialysis.

In blacks, peritoneal dialysis (CAPD/CCPD) in diabetic ESRD patients has inferior survival compared with hemodialysis. United States Renal Data System 1998 Report.

Figure 6. ESRD Death Rate in Diabetes: PD vs. Hemodialysis.

In whites, peritoneal dialysis (CAPD/CCPD) in diabetic ESRD patients has inferior survival compared with hemodialysis. United States Renal Data System 1998 Report.

Figure 7. ESRD Death Rate in Diabetes: PD vs. Hemodialysis.

Deaths attributed to myocardial infarction occur at a higher rate in diabetic ESRD patients treated with peritoneal dialysis (CAPD/CCPD) compared with hemodialysis. United States Renal Data System 1998 Report.

Rubin et al. experienced difficulties in utilizing CAPD in a largely black diabetic population;(12) only 34% of their patients continued CAPD after two years, and, at three years, only 18% remained on CAPD. Peritoneal dialysis patients in the US have a 14% greater risk of hospitalization than patients on hemodialysis.(13) Restraint is mandatory when attempting to compare mortality and comorbidity in hemodialysis versus peritoneal dialysis patients because of selection bias in assigning subjects to one or the other regimen. Indeed, there have been no prospective controlled studies in which allocation to either peritoneal dialysis or hemodialysis was randomized. Illustrating this limitation is the CANUSA study comparing survival and other variables in CAPD patients in the United States and Canada.(14) Interpretation of all results in this international comparison is an unexplained higher relative risk of death (1.93) in the US, reflecting a two-fold greater acceptance rate for ESRD including purportedly sicker (greater comorbidity) patients.

Extracted data subsets(3) from the USRDS Report for 1998 show that diabetic patients including both genders, black and white races and all age subgroups have a higher death risk on CAPD than on hemodialysis (Figs. 2-6). Opposed to this finding, Fenton et al. extracted data from the Canadian Organ Replacement Register for 11,970 patients, who began ESRD therapy between 1990 and 1994, concluded that peritoneal dialysis "is not associated with increased mortality rates relative to hemodialysis."(15) The decision to treat a diabetic patient with CAPD, therefore, must be individual-specific.

Kidney Transplantation

A functioning kidney transplant provides the uremic diabetic patient better survival with superior rehabilitation than does either CAPD or maintenance hemodialysis. As for comparisons of hemodialysis versus CAPD, however, there have been no prospective controlled studies of dialysis versus kidney transplantation in diabetic patients whose therapy was assigned randomly. Grouping all dialysis-treated diabetic patients in comparison with all diabetic renal transplant recipients shows that - whatever the explanation - survival is markedly superior in recipients of a kidney transplant.

Sutherland et al.(16) correctly predicted that following a renal transplant, patient survival at one and two years would be equivalent in diabetic and nondiabetic recipients, (17) though graft survival, in some large series, remains marginally lower in diabetic persons. A single center retrospective review of all kidney transplants performed between 1987 and 1993 found no significant difference in actuarial 5-year patient or kidney graft survival or mean serum creatinine concentration between diabetic and nondiabetic recipients overall or when analyzed by donor source.(18) Fewer than four in 100 diabetic ESRD patients treated by either peritoneal or hemodialysis will live for a decade while cadaver donor and living donor kidney allograft recipients fare far better. Furthermore, not reflected in crude survival statistics, following a renal transplant, is the incomparably superior rehabilitation. It is mainly because of the enhanced life quality permitted by a kidney transplant that this therapeutic option is preferred for all newly evaluated diabetic persons with ESRD under the age of 60. More than half of diabetic kidney transplant recipients in most series live for at least three years. Many survivors return to occupational, school and home responsibilities.

Pancreas Plus Kidney Transplantation

Although still viewed by some as investigational (19) and, even when successful, applicable to no more than 9% of uremic diabetic patients who have Type 1 diabetes, pancreatic transplantation is growing in acceptability and technical success.(20) Survival one year post-renal transplant can be remarkably good (84%) as evidenced by a series of 995 diabetic kidney recipients who also received a pancreas transplant renal allograft.(21) As reported in 1998 by the International Pancreas Transplant Registry, world-wide results in simultaneous kidney-pancreas transplants show that, after one year, 94% of recipients were alive, one-year kidney graft survival was 89% while one-year pancreas survival was 82%.(22) In 1998, consensus of clinical nephrologists is that the ESRD patient with Type 1 diabetes should consider a simultaneous kidney and pancreas transplant as, at least, a temporary cure of inexorable disease.(23) Anecdotal (unpublished) reports of pancreas transplantation in Type 2 diabetes are sparse and do not permit judgment as to its value - if any - in C-peptide negative diabetes other than Type 1. Not surprisingly, steroid treated Type 1 diabetic pancreas recipients may develop Type 2 diabetes.(24)

Patient Survival with ESRD

By current practice, younger patients with fewer complications are assigned to renal transplantation while residual older, sicker patients are treated by dialysis. Combined kidney/pancreas transplants are restricted to Type 1 diabetic patients younger than age 50 who have minimal cardiac disease.

Survival on CAPD and maintenance hemodialysis are lower in the US than in Europe. An explanation for diabetic dialysis patients' better survival in Europe is not evident, though the growing application of American practices of dialyzer reuse and shortened treatment hours have been incriminated as promoting fatal underdialysis.(25) A rebuttal to the allegation that US dialysis mortality is higher than that in Europe contends that because the US has approximately twice the incidence of ESRD acceptance than Europe, older and sicker patients in the US obviously will experience greater mortality.(26)

Rehabilitation

Unless patient cohorts (hemodialysis versus CAPD/CCPD versus renal transplantation versus combined kidney and pancreas transplantation) have equivalent comorbidity when treatment is initiated, conclusions about the effect of modality on outcome are erroneous. To quantify the course of diabetic patients over the course of ESRD treatment, we periodically inventory the type and severity of common comorbid problems. Numerical ranking of this inventory constitutes a comorbid index (Table 1).

To obtain a numerical Co-Morbidity Index for an individual patient, rate each variable from 0 to 3 (0 = absent, 1 = mild - of minor import to patient's life, 2 = moderate, 3 = severe). By proportional hazard analysis, the relative significance of each variable can be isolated from the other 12.

Karnofsky scoring(27) assesses patient well-being.(28) Applying the Karnofsky score to 2,481 dialysis patients, irrespective of location or type of dialysis, Gutman et al. in the 1970s found that diabetic patients achieved very poor rehabilitation; only 23% of diabetic patients (versus 60% of nondiabetic patients) were capable of physical activity beyond caring for themselves.(29) Confirming and extending this conclusion a decade later, Lowder et al reported abysmal rehabilitation in diabetic hemodialysis patients.(9) More recent confirmation of this point was afforded by Ifudu et al. who documented pervasive failed rehabilitation in multicenter studies of diabetic, nondiabetic(30) and elderly(6) inner-city hemodialysis patients.

Conclusions

An inescapable conclusion of studies to date is that maintenance hemodialysis as performed in the US does not permit return to life's responsibilities for the large majority of diabetic individuals with ESRD. The same statement is correct for peritoneal dialysis. However, while enthusiasts for CAPD/CCPD expound the concept that peritoneal dialysis affords unique or special advantage for the diabetic ESRD patient, with relatively rare exception, long-term survival is only attained with hemodialysis or the distinctly preferable option of kidney transplantation. No matter which choice in therapy is selected for the diabetic ESRD patient, the crucial aspect of management is careful and continuing attention to extrarenal comorbid conditions that ultimately threaten life.