Current antiretroviral therapy has allowed for an increased survival of patients infected by the human immunodeficiency virus (HIV).1 This increase has led to the occurrence of other associated complications that were less common 15 years ago.2 From the nephrological perspective, an increase has occurred in HIV patients with renal involvement, and their prognosis in renal replacement therapy has changed favorably3,4 as compared to the beginning of the 90s, when early studies reported high mortality and complication rates.5,6 There is currently no controversy about inclusion of these patients in a hemodialysis (HD) or peritoneal dialysis (PD) program when they need it, The situation of these patients on PD has progressed over these years, and the most determinant factor today is the HIV disease itself.7 The course over time and the characteristics of HIV-positive patients in our PD program from 1995 to date are reported.
METHODS
From November 1995 to November 2007, data were retrospectively collected from all HIV-infected patients who had started renal replacement therapy using peritoneal dialysis at our center. Demographic and laboratory data of all patients were reviewed. The survival curve was calculated using the Kaplan-Meier method. Patients who were switched to another replacement treatment modality (HD or kidney transplant) were considered lost for outcome purposes. Outcomes were compared using a Student¿s t test for paired data. Means and medians are given with their standard deviation and range respectively. A value of p < 0.05 was considered statistically significant.
RESULTS
Eight patients, seven males and one female, were included in this period. All patients were Caucasian. Mean patient age was 40,7 ± 5,3 years. The different underlying diseases are given in Table 1. Six patients had stage C3 HIV infection, one stage B3, and the other stage B2 HIV infection. Serological testing for HCV was positive in all patients, but only one pa- tient had data suggesting chronic liver disease. They all were former intravenous drug users.
Mean follow-up time on PD was 41.2 ± 32,1 months, with a median of 29,5 months (range: 12-103). Survival rates at one, two, and three years were 100%, 62.5%, and 50% respectively, as may be seen in Figure 1, with an overall mortality at study end of 62.5%. Five patients died, two from ischemic heart disease, a third patient from fulminant gangrene in his right leg caused by Pseudomonas aeruginosa, a fourth patient died at home for an unknown reason, and the final patient died from sclerosing peritonitis after six years on PD. A patient was switched to HD after peritonitis induced by Mycobacterium fortuitum, another patient received a kidney transplant after 9 years on dialysis, and a final patient continued on PD at study completion (table I). Table II compares the survival rates reported by studies involving HIV patients on peritoneal dialysis, showing an increased survival and a decrease in complications over the years.
Complications of the procedure
The infectious peritonitis (IP) rate and the number of hospitalizations were 0.36 IP/year and 0.69 admissions/year respectively. Staphylococcus epidermidis was the most common causative organism, but was only found in two patients. A patient experienced six IP episodes (two from Klebsiella pneumoniae, two from Staphylococcus epidermidis, one sterile peritonitis, and one from Pseudomonas fluorescens). Another patient experienced an IP by Mycobacterium fortuitum that required removal of his peritoneal catheter and switching to HD. In another case with a good outcome, the identified germ was Escherichia coli, and IP induced by Candida albicans occurred in a single case. Three patients suffered no peritonitis during this study period, one of them probably because of his short time on PD (12 months). A second patient continues on PD after 32 months, and the remaining patient received a kidney transplant after almost nine years on PD. Only 15.7% of hospital admissions were caused by an infectious peritonitis. The most common reason for admission was respiratory tract infection.
Course of HIV infection
All patients were on antiretroviral therapy. Although different schemes were used, all patients but one were on triple therapy. Among nucleoside reverse transcriptase inhibitors (NRTIs), the most commonly used drug was lamivudine (Epivir®) taken by 75% of patients. Thus, the most usual combination consisted of lamivudine plus stavudine (Zerit®), found in half the cases. These drugs were normally associated to a protease inhibitor (PI). Both stavudine and abacavir (Ziagen®) were taken by 62.5% of patients. The most commonly used PI was nelfinavir (Viracept®), received by 37.5% of patients. Non-nucleoside reverse transcriptase inhibitors (NNRTIs) were scarcely used, and only one patient received efavirenz (Sustiva®) in his treatment. From the clinical viewpoint, patients maintained a good viral response. None of the five patients with an undetectable viral load experienced any changes in viral load, that decreased to undetectable levels in the other three patients. CD4 levels increased in all patients but one. Mean CD4 percentage was 20.2% at baseline and 23.5% at one year, but this increase was not statistically significant. Another aspect assessed was weight gain (net weight), based on the assumedly greater malnutrition in PD due to protein loss in peritoneal fluid. No differences were found in our patients in serum albumin levels or nPCR, but a significant weight gain was seen in the first year. Mean patient weight was 60.6 ± 8 kg at baseline and 64.9 ± 10 kg at one year, p = 0.016 (fig. 2).
DISCUSSION
Our study confirms an increased survival of HIV patients on peritoneal dialysis. Survival rates at one, two, and three years were similar to another recent study3 and higher than the rates reported by Tebben and Kimmel.5,6 Early studies on HIV patients with end-stage renal disease (ESRD) and renal replacement therapy by Ortiz,8 Perinbasekar,9 and Feinfeild10 reported a poor outcome. New highly active antiretroviral therapy (HAART) has led, particularly since 1995, to a marked increase in life expectation1 which has allowed for survival rates in peritoneal dialysis longer than 5 years, and up to 9 years in our experience.3 However, mortality continues to be high in these patients despite more effective treatments.3,4 Survival of HIV-infected patients has improved,1 and does not change irrespective of the renal replacement therapy used.7 Rodríguez et al, in
their study on 100 HIV-infected patients, showed that those who started dialysis in the pre-HAART era survived 9.4 months, while patients starting dialysis in the HAART era survived 16.1 months. However, the corresponding figures for all HIV-positive patients not on dialysis were 16 and 81 months.11 Other studies suggested an inadequate antiretroviral treatment in patients on renal replacement therapy.12 However, an analysis of the different studies involving HIV-infected patients on PD showed an improved survival, particularly when the two pre-HAART studies were compared to studies conducted in the HAART era, from a mean survival of 17.9 months to longer than 40 months (table II).
From the therapeutic viewpoint, there are studies and guidelines with recommendations for antiretroviral therapy in ESRD and dialysis, but ART in PD is not clearly defined.13 Szczech reviewed this aspect in 89 HIV-positive patients, five of them on PD, and found that the most commonly used drug was lamivudine, in 36 patients, followed by stavudine in 33 patients and nelfinavir in 21 patients. These results were similar to those seen by our study group. These data are consistent with the widespread use of stavudine and nelfinavir during the final 90s and the beginning of the 2000 decade, the period mostly described in our study. These drugs are now used as first-line treatment because of their low potency (nelfinavir) or high associated toxicity (stavudine).14,15 Currently, and until studies providing new indications are available, it appears reasonable to
take the same measures as in hemodialysis, considering the special characteristics of these patients on PD.16,17
Complications of the procedure have decreased in parallel to the increase in survival. Thus, infectious peritonitis rates in HIV patients have decreased from 3.9 and 2.4 episodes/year in early studies to 1.4 episodes/year in the Khana study or 0.3 episodes/year in our study,4.6 though some studies have reported conflicting results. Thus, Tebben et al6 found in a group of 39 patients a higher number of peritonitis in HIV-infected patients, and Scholoth18 also noted in 9 similar patients a higher rate of infectious peritonitis, as did Khana.4 Other authors showed no differences between both groups.5,19 In our study, peritonitis rate was
not different from the rate seen in non-HIV-infected population in our unit.
Although opportunistic infections have been reported in HIV patients treated with PD, Gram-positive cocci are the most common cause of infectious peritonitis.6,20-22 Our results are very similar to those of the Kimmel group, with a case each of IP caused by Pseudomonas and Mycobacterium. Fungal involvement was seen in a single patient, and there were no cases of polymicrobial peritonitis. 23 It should be noted that 6 of the 10 episodes of peritonitis occurred in the same patient, and three patients experienced no IP while on peritoneal dialysis, in contrast to other studies where 100% of patients experienced some episode of IP.5 Course of peritonitis in our study was usually favorable, and removal of peritoneal catheter was only required in two cases. Hospitalization rate was also low as compared to previous studies.3-4
These results are possibly conditioned by advances in antiretroviral therapy, a decreased viral load, and an improved immune status, which are known good prognostic factors.24-26 In our study, all patients were on HAART, viral load was low or undetectable in most cases, and the CD4 percentage was adequate. Although higher serum albumin levels3 and an increased risk of peritonitis associated to them27 have been reported in PD, this does not appear to result in a lower survival in HIV patients.5 While no changes in serum albumin levels or nPCR were found in our patients, an increase in net weight was shown, at least during the first year. However, 62% of patients were hypertensive at PD start, and two patients died from coronary disease. Some drugs, such as lopinavir, have been associated to an increase in systolic blood pressure,28 other drugs such as fosamprenavir29 to a lipid-lowering effect, and still others such as ritonavir have been associated to both effects.30 NNRTIs and PIs usually have a poorer lipid profile and a higher risk of associated coronary disease.31-33 Association of risk factors may be determinant for a decreased overall survival in these patients.
On the other hand, increased prevalence of these patients raises therapeutic alternatives that were unthinkable not long ago. Thus, kidney transplant is a valid option if the following requirements are met: HAART, undetectable viral load, CD4 count above 200 cells/mm3, and absence of recent opportunistic infections.25,34,35 In our case, a patient meeting the above criteria received a kidney transplant with an excellent outcome. Long-term outcome in these kidney transplant patients is unknown, but short and mid-term course appears safe, or at least with a similar survival to other risk groups.36,37
In conclusion, while our study has the limitations inherent to a retrospective study on a small patient sample, it showed an improved survival and low rates of infectious peritonitis and hospitalization. Adequate treatment of HIV infection and other associated diseases conditions the current outcome of the procedure, which is similar to the non-HIV population in our study. Immune status does not appear to change in PD, and adequate nutrition is possible. PD is therefore a treatment of choice for these patients. Until greater experience and additional studies are available, individualized and multidisciplinary control appears reasonable in these patients.