JN 2000; Vol.13 N°4: 267-270

Research protocol

JNEPHROL 2000; 13: 267-270

Diet or dialysis in the elderly? The DODE Study: a prospective randomized multicenter trial

Rosario Maiorca¹, Giuliano Brunori¹, Battista Fabio Viola¹, Roberto Zubani¹, Giovanni Cancarini¹, Giovanni Parrinello², Adriano De Carli² - ¹Institute and Division of Nephrology, University School of Medicine and Spedali Civili, Brescia - Italy ²Section of Biometry, University of Brescia, Brescia - Italy

ABSTRACT: There are no solid data on the real advantage of an early start of dialysis, as suggested by the DOQI guidelines. Uremic patients frequently have a poor nutritional status. However, we cannot distinguish between the detrimental effect on nutrition of too low a residual renal function or too long a period of low protein-diet, per se. However, it appears that a very-low-protein diet (VLPD) supplemented with essential amino acids and keto-analogs of amino acids, and with an adequate quantity of calories, can prevent hypoalbuminemia at the start of dialysis and can slow the progression of chronic renal failure.
EDTA and USRDS data suggest that most patients starting dialysis nowadays are elderly, who also have the highest incidence of morbidity and mortality. Moreover, hospitalization rate becomes higher after the start of dialysis compared to the pre-dialysis period.
Can an aminoacid-supplemented VLPD, prolonged beyond the GFR limits suggested by DOQI, offer elderly patients better survival and better quality of life than dialysis? The answer can only come from a prospective, randomized trial, in elderly patients, starting at the GFR values suggested by the NKF-DOQI for starting dialysis, comparing outcomes with a vegetarian VLPD supplemented with a mixture of keto-analogs of amino acids and essential amino acids, and with dialysis.

Key words: Supplemented very-low-protein diet, Dialysis, Elderly, Mortality, Morbidity

Background and Rationale of the DODE

An adequate dose of dialysis is needed, in order to improve patient outcome, in both peritoneal dialysis (PD) and extracorporeal dialysis (HD). The debate, however, has renewed interest in when to start dialysis. The NKF-DOQI group suggested using the amount of solute clearance for regular dialysis as a value for
starting dialysis treatment, i.e. creatinine clearance 9-14 ml/min (1).
Elderly patients on dialysis have a low survival rate. According to the Italian Lombardy Registry, patients 70-75 years old survive 2.9/3.6 years (male/female) after the start of dialysis, those 75-80 years old 2.7/2.7 years, and those 80-85 years old 2.4/1.4 years (2). This low survival is largely due to the combined effects of aging and comorbidities, on which an early start of dialysis has no or very limited positive effects.
Moreover, the hemodialytic treatment might have more negative effects in the elderly: intermittent hemodynamic stress, continuous fluctuation of electrolytes, metabolites, acid-base equilibrium, and the bioincompatibility of the membrane. In PD patients, the negative effects are also related to episodes of peritonitis and to daily loss of protein and a higher frequency of malnutrition.
An early start of dialysis should improve survival and well-being, and reduce morbidity. But a rapid decline of residual renal function (RRF) is observed after the start of dialysis (3-5). Maintenance of RRF as long as possible is very important, because the kidneys do some things a dialysis membrane cannot do: tubular and endocrine function, larger clearances of middle molecules, some of them potentially toxic such as parathyroid hormone, granulocyte inhibitory protein, and substances inducing anorexia.
In a prospective study on CAPD patients, for every
increase of 1 ml/min of GFR, the relative risk (RR) of death decreased 50% (6). In the CanUsa study, using the time-dependent Cox multivariate analysis, a 5% reduction in RR of death was observed for each 5L/week of residual renal GFR at baseline (7). These results suggest that, at least in elderly patients, the fast decline of GFR observed after the start of dialysis could be another cause of the low survival rate.
In our center, 42 elderly patients started a very-low-protein diet (VLPD: 0.3 g of protein/kg body weight), supplemented with a mixture of keto acids and amino acids when their creatinine clearance reached values usually considered as a parameter for starting dialysis. The survival of this group of patients was compared with that of 71 patients directly put onto dialysis with the same creatinine clearance at which we started the VLPD. Subsequently, 24 patients in the diet group started dialysis when their mean GFR was 3.9 ml/min, i.e. lower than the 7 ml/min usually considered adequate in PD patients and much lower than 9-14 ml/min suggested by the NKF-DOQI group.
In this retrospective, non-randomized study, survival was better in the patients treated with diet than patients with dialysis (8). This seems to bear out that RRF does more than a dialytic membrane and must be preserved as long as possible.

Objectives

The primary hypotheses to be tested are whether, at least in the elderly:

* the supplemented VLPD does not increase the risk of morbidity and mortality,

* a prolonged period of VLPD can postpone the need for dialysis in elderly patients, and dialysis can start at a RRF lower than usually suggested.

The secondary hypothesis to be tested is:

* that the VLPD does not induce malnutrition.

Study design

The DODE Study is a prospective randomized controlled trial comparing 200 elderly patients, divided into two groups when their RRF (expressed as the mean of creatinine and urea clearance) is 5-7 ml/min/1.73 m²: one being started on dialysis, the second one kept on supplemented VLPD. Patients will be enrolled in at least 50 centers and the randomization ratio will be 1:1, using Zelen's method (9). Patients willing to participate will be randomly allocated to dialysis or to diet, after approval of the Ethical Committee of each center participating in the study. The enrollment period will be 18 months. The planned duration of follow-up is 48 months. Each patient should terminate the study three years after enrollment.
The VLPD is a diet with a calorie intake of 35 kcal/kg BW/day and a protein intake of 0.3 g protein/kg BW supplemented with 100 mg/kg BW of a mixture of keto-analogs of amino acids and essential amino acids (ALFA-KAPPA^®, Farma Biagini). Patients randomized to the dialysis group (and those on the VLPD diet group and subsequently shifted to dialysis) will eat a diet with 35 kcal/kg/day and 1.2 g of protein/kg/day.

Inclusion and exclusion criteria

All new ESRD patients of both sexes, eligible for the trial, who are willing to co-operate, who have given their informed consent and fulfil the following criteria, may be admitted. §
The inclusion and exclusion criteria are reported in Table I.

TABLE I

Inclusion criteria Exclusion criteria

1) males or females >= 70 years of age 1) diabetes

2) RRF 5-7 ml/min/1.73 m² 2) severe cerebral disease

3) left ventricular ejection fraction >40% 3) proteinuria >3 g/24 h

4) signed informed written consent 4) severe hepatic failure

5) active malignancy

6) COPD requiring supplementary oxygen

7) AIDS, or HIV-positive

Allocation of treatment

Patients will be randomly allocated to one of the two treatments groups using dialysis or diet as established by the protocol. Patients will be treated with bicarbonate HD or hemodiafiltration or PD, according to their clinical conditions and choice. Randomization will be centralized by the Study Secretariat.

Trials procedures

Patients considered eligible on the basis of the inclusion/exclusion criteria will be given information about the study, in order to enable them to give their informed consent, and start randomization.
The baseline visit will include a record of the patient's data (co-morbid conditions, renal disease, age, sex, race), medical history, laboratory parameters, RRF, protein nitrogen appearance (PNA), anthropometry (BSF, TSF, SSF), subjective nutritional global assessment (SGA), according to Desky (10).
During the follow-up period, blood samples will be drawn every month, and tested for the following: prealbumin, albumin, hematocrit, cholesterol, triglycerides, bicarbonate, calcium, phosphate, RRF, Kt/V, PNA. PTHi, ferritin, iron, transferrin will be determined every four months. At the same interval, patients will be evaluated for anthropometry, SGA, peroneal nerve conduction velocity.
Residual renal function will be measured every month, and as long as the urine output is more than 200 ml/day in patients on dialysis. In HD patients, the target Kt/V is 1.2, as calculated by the Daugirdas 2^nd generation formula (11); in PD patients, the Kt/V must be at least 2. Body water volume (V) is calculated by Watson's formula (12).

Medication and hospital treatments of any patient will be collected.

All the data will be sent to the coordinating group in Brescia, Italy.

For the diet group, dialysis must be started when one of the following events appears:

1) severe fluid overload, in non-responders to diuretics;

2) severe hypertension (>170/95 mmHg) in spite of pharmacological therapy;

3) hyperkaliemia;

4) signs of malnutrition:

a) serum albumin < 30 g/L,

b) serum triglycerides less than 20% of minimum
reference value,

c) total serum cholesterol less than 20% of minimum
reference value,

d) body weight loss more than 8% of body weight at
the start of diet (in patients without edema);

5) signs of uremic syndrome:

a) persistent nausea,

b) vomiting,

c) pericarditis,

d) severe neuropathy (peroneal nerve conduction
velocity less than 35 m/sec).

Statistical analysis

Calculation of the sample size

According to Blackwelder's approach, a sample of 83 patients is needed in order to evaluate with 80% power and *=10% whether the occurrence of the primary endpoint at one year (one tailed-test) of diet patients is at least equivalent to that of dialysis patients (13). Fifteen percent of patients can be expected to be lost to follow-up. Accordingly the required sample size has been adjusted to 100 patients per group (under the usual assumption of independence between survival and censoring due to lost to follow-up). Each center is expected to enroll at least four patients.

Efficacy analysis

The analysis will take into consideration - when available from patients' charts - the previous management (particularly as regards compliance with conservative therapy) and the rate of progression of renal failure in the previous year. In the diet-group patients, the comparison will be made with the entire follow-up period.
A minimum of three months will be required for the efficacy analysis for the diet group.
Two types of analysis will be done: an "intention-to-treat analysis" (patients starting on VLPD and subsequently transferred to dialysis will be dealt with by diet group intention-to treat analysis) and a "per protocol analysis" (patients will be analyzed in relation of the treatment really received).
Univariate analysis of survival from entry into the study will be done using the Kaplan-Meier method, and the log-rank test will be used to compare survival between the two groups (14).
Mortality will also be analyzed using the Cox proportional hazards model (15) adjusted for age, sex, baseline co-morbidity and type of treatment (diet or dialysis), SGA, serum albumin, Kt/V, GFR. Death due to all causes will be considered as an event, but patients will be censored at the time they reach the end of follow-up. The cause of death will be confirmed by death certificate.
Separate subpopulation analyses will be done: the multivariate analysis will focus on the assessment of treatment effect in relation to age (intervals of five years) (2).

The trial will also focus on the following:

- patient morbidity: number of hospital admissions and duration of stay, related to:

1 - co-morbidity,

2 - therapy (diet or dialysis),

3 - new events;

- RRF, Kt/V urea, weekly creatinine clearance (for PD patients);

- nutritional status assessed by SGA, nPNA, transferrin, prealbumin, albumin, triglycerides, cholesterol;

- other laboratory tests, such as PTHi, calcium, phosphorus, hemoglobin, total leucocyte count.

Although this is a multicenter trial, the regressor will not include indicators for evaluating center effect and the treatment center interaction, in view the small numbers of patients per center.

Ethical considerations

The trial will be carried out in accordance with the Declaration of Helsinki and subsequent modification (16). Before it starts, patients must be informed about the aim of the study, the possible benefits to them, the risks and inconvenience involved, and their right to withdraw at any time without sanction. Written consent must be obtained before randomization (17).

Reprint requests to: Prof. Rosario Maiorca Chair and Division of Nephrology University and Spedali Civili Piazza Ospedale, 1 25125 Brescia Italy maiorca@master.cci.unibs.it

References (when available, each reference has been linked to PubMed)

1. NKF-DOQI Clinical practice guidelines for peritoneal dialysis Adequacy. Am J Kidney Dis; 1997; 30 (suppl 2): S70-3.

2. Malberti F, Conte F, Limido A, Marcelli D, Spotti D, Lonati F, Locatelli F. Ten years' experience of renal replacement treatment in the elderly. Ger Nephrol Urol 1997; 7: 1-10.

3. Rottembourg J. Residual renal function and recovery of renal function in patients treated by CAPD. Kidney Int 1993; 43 (suppl 40): S106-10.

4. Cancarini GC, Brunori G, Camerini C, Brasa S, Manili L, Maiorca R. Renal function recovery and maintenance of residual diuresis in PD and hemodialysis. Perit Dial Bull 1986; 6: 77-9.

5. Lysaght MJ, Vonesh EF, Gotch F, Ibels L, Keen M, Lindholm B, Nolph KD, Pollock CA, Prowant B, Farrell PC. The influence of dialysis treatment modality on the decline of remaining renal function. Trans ASAIO 1991; 37: 598-604.

6. Maiorca R, Brunori G, Zubani R, Cancarini GC, Manili L, Camerini C, Movilli E, Pola A, d'Avolio G, Gelatti U. Predictive value of dialysis adequacy and nutritional indices for mortality and morbidity in CAPD and HD patients. A longitudinal study. Nephrol Dial Transplant 1995; 10: 2295-305.

7. Bargman JM, Thorpe KE, Churchill DN. The importance of residual renal function for survival in patients on peritoneal dialysis (Abstract). 30^th Annual Meeting of the American Society of Nephrology, San Antonio, Texas, November 2-5, 1997.

8. Maiorca R, Brunori G, Gregorini G, Pola A, Viola BF, Zani R, Zubani R. Very low protein diet in elderly ESRD patients prolongs residual renal function, improves survival and dialysis can be started late (Abstract). 31^st Annual Meeting of the American Society of Nephrology, Philadelphia, Pennsylvania, October 25-28, 1998.

9. Zelen M. A new design for randomized clinical trials. N Engl J Med 1979; 300: 1242-5.

10. Detsky AS, McLaughlin JR, Baker J et al. What is subjective global assessment of nutritional status? J Parent Ent Nutr 1987; 11: 8-13.

11. Daugirdas JT. Second generation logarithmic estimates of single-pool variable volume Kt/V: an analysis of error. J Am Soc Nephrol 1993; 4: 1205-13.

12. Watson PE, Watson ID, Batt RD, Phil D. Total body water volumes for adult males and females estimated from simple anthropometric measurements. Am J Clin Nutr 1980; 33: 27-39.

13. Blackwalder WC. Proving the null hypothesis in clinical trials. Controlled Clinical Trials 1982; 3:345-53.

14. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Statist Assoc 1958; 53: 457-81.

15. Cox DR. Regression models and life tables (with discussion). J Royal Statist Soc 1972; B34: 197-220.

16. World Medical Association. Declaration of Helsinki: Recommendations guiding physicians in biomedical research involving human subjects. IV^th adaptation 1989.

17. Lebacqz K, Levine RJ. Respect for persons and informed consent to participate in research. Clin Res 1977; 25: 101-7.

Received: October 30, 1999 Revised: April 14, 2000 Accepted: June 08, 2000

Back to Home Page