- Author: Moro O Salifu, MD, MPH, FACP; Chief Editor: Vecihi Batuman, MD, FACP, FASN more...
Renal disease is a relatively common complication in patients with human immunodeficiency virus (HIV) disease.[1, 2] HIV nephropathy can result from direct kidney infection with HIV or from the adverse effects of antiretroviral drugs (see the image below).[3, 4, 5, 6] Further, patients with HIV disease are at risk for developing prerenal azotemia due to volume depletion resulting from salt wasting, poor nutrition, nausea, or vomiting.
HIV-associated nephropathy (HIVAN), formerly known as AIDS-associated nephropathy, is characterized by the following findings:
Nephrotic range proteinuria
Normal to large kidneys on ultrasound images
Focal segmental glomerulosclerosis (FSGS) on renal biopsy findings
Although FSGS is the predominant glomerular lesion in HIVAN, other reported glomerular lesions in patients with HIV include IgA nephropathy, cryoglobulinemia, amyloidosis, and a lupuslike immune complex glomerulopathy.
In the preantiretroviral therapy era, HIVAN was characterized by rapid progression to renal failure and end-stage renal disease (ESRD) leading to the need for dialysis. Highly active antiretroviral therapy (HAART) has changed the natural course of this disease, increasing the importance of prompt diagnosis and proper care. Baseline estimated glomerular filtration should be obtained and renal function monitored during HAART. Physicians must consider HIVAN in patients who develop proteinuria.
According to the US Renal Data System (USRDS), HIV-associated nephropathy (HIVAN) accounts for approximately 1% of new end-stage renal disease (ESRD) cases in the United States. HIVAN is observed in patients regardless of the route by which HIV was contracted.
HIVAN is observed predominantly among African Americans and is the third leading cause of ESRD among blacks aged 20-64 years.[9, 10] Most patients with HIVAN are young black men, and approximately 50% of patients with HIVAN are intravenous drug abusers.[11, 12] Overall, HIVAN is observed more often in men than in women, with a male-to-female ratio of 10:1. The mean age of persons with HIVAN is 33 years. HIVAN may occur in children.
Experiments using transgenic mice have provided perhaps the strongest evidence for a direct role by HIV type 1 (HIV-1) in the development of HIV-associated nephropathy (HIVAN). Researchers created transgenic mice by inserting HIV DNA constructs into the mice's genomes. The mice developed proteinuria and had a histologic picture similar to that observed in patients with HIVAN.
A genetic or environmental cofactor that has not yet been identified is required for patients to develop this disease. This unidentified factor might explain the predilection for HIVAN among black persons.
The cellular target in the development of HIVAN is probably the renal glomerular and tubular epithelium. Using in situ hybridization and polymerase chain reaction assays to detect HIV-1 DNA and messenger ribonucleic acid (mRNA), investigators have shown that renal glomerular and tubular epithelial cells are productively infected by HIV-1 in patients with HIVAN; this argues strongly for localized replication of HIV-1 in the kidney and for the existence of a renal viral reservoir.
Further, circularized viral DNA, a marker of recent nuclear import of full-length, reverse-transcribed RNA, has been detected in kidney biopsy samples from patients with HIVAN, suggesting active replication in renal tissue. However, the mechanisms of virus-induced renal injury remain undetermined.
Peculiar histopathologic features of HIVAN are the enhanced proliferation and the loss of differentiation markers of glomerular epithelial cells. In one study, HIV-1 infection was shown to kill renal tubular epithelial cells in vitro by triggering an apoptotic pathway involving caspase activation and Fas up-regulation, suggesting that apoptosis of nonlymphoid cells can be directly induced by HIV-1. The net and long-standing glomerular and tubular epithelial cell damage leads to proteinuria, glomerulosclerosis, and tubulointerstitial scarring.
The role of cytokines has not been established, and although their presence is not essential for the development of HIVAN, cytokines may modify the progression of infection or a patient's susceptibility to infection. The levels of cytokines are increased in renal biopsy samples from patients with HIVAN.
In one study, mesangial and tubular cell production of interleukin-6 and tumor necrosis factor–alpha was shown to be a potent stimulus for HIV-1 expression in HIV-1–infected monocytes. Viral replication in response to cytokines may play an important role in the pathogenesis of HIVAN.
The reason behind the increased predilection among black persons for the development of HIV-associated nephropathy is not clear. In general, black persons have a higher incidence of other renal diseases (eg, diabetic nephropathy, lupus); therefore, they may have an underlying genetic predisposition to severe renal disease, regardless of the etiology. The type of host response to the HIV infection itself may be what determines whether or not nephropathy develops in a specific individual.
Kopp et al studied genetic variants predisposing to idiopathic and HIV-1–associated focal segmental glomerulosclerosis (FSGS), and they concluded that genetic variation at the MYH9 locus substantially explains the increased burden of FSGS and hypertensive kidney disease among African Americans. They carried out admixture-mapping linkage-disequilibrium genome scanning on 190 African American individuals with FSGS and 222 controls and identified a chromosome-22 region centered on MYH9, a functional candidate gene expressed in kidney podocytes.
In one study, the rate of progression from the initial presentation to ESRD was 2.5 months in the pre-HAART (highly active antiretroviral therapy) era. With the introduction of HAART in 1996-1997, the traditional natural history of rapid progression of HIV-associated nephropathy (HIVAN) has been slowed significantly. HAART has been shown to retard the progression of renal disease in persons with HIVAN, and treatment with angiotensin-converting enzyme inhibitors may be beneficial.
Among patients with HIV infection and end-stage renal disease receiving hemodialysis, survival has improved significantly compared with the uniformly dismal outcomes in the 1980s.
Patients with HIV-associated nephropathy (HIVAN) typically present with a nephrotic syndrome consisting of nephrotic-range proteinuria (>3.5 g/d), azotemia, hypoalbuminemia, and hyperlipidemia. Edema is uncommon in HIVAN, yet many authors think that this is a characteristic of HIVAN. The salt-losing propensity and high oncotic pressure contributed by marked hypergammaglobulinemia in these patients have been suggested as possible explanations for this puzzling observation.
CD4+ T-cell count
The CD4+ T-cell count in patients with HIVAN is usually depressed below 200 cells/µL, but HIVAN has been reported in patients with higher CD4 counts. The prognosis for renal survival is worse in patients with AIDS, especially if their CD4 count is less than 50 cells/µL.
Ultrasound and CT scan
Patients with HIVAN are not typically hypertensive, even in the face of renal insufficiency, and their kidneys are usually normal to large in size and highly echogenic on ultrasonograms, as well as on CT scans. This may result from prominent interstitial expansion by cellular infiltrate and markedly dilated tubules containing voluminous casts.
Routine urinalysis may occasionally reveal findings of nonnephrotic proteinuria in patients being evaluated for other medical conditions. The urinalysis reveals microhematuria, leukocytes, hyaline casts, and oval fat bodies, but no cellular casts. Serum complement levels are normal.
Electrolyte abnormalities, such as hyponatremia and hyperkalemia, may be observed in patients with HIVAN and may reflect an increase in total body water (from nephrotic syndrome or syndrome of inappropriate secretion of antidiuretic hormone [SIADH]) or from hyporeninemic hypoaldosteronism, respectively.
SIADH may result from concomitant pulmonary infection or from persistent nausea from medications or gastrointestinal disease. Hyporeninemic hypoaldosteronism, a cause of type IV renal tubular acidosis manifesting as hyperkalemia with normal anion gap metabolic acidosis, is much more common when renal insufficiency is present.
The decision to obtain a biopsy sample is somewhat controversial in the general medical community. Even if a patient presents with the classic clinical features of HIV-associated neuropathy (HIVAN), clinical consideration is predictive of the biopsy diagnosis in only 55-60% of patients.
Therefore, to distinguish HIVAN from other forms of renal disease (eg, immune complex glomerulonephritis, immunoglobulin-A nephropathy), patients who are seropositive for HIV require a renal biopsy. The typical practice is to obtain a renal biopsy specimen if the patient's daily protein excretion is greater than 1 g.
Findings from light microscopy of kidney biopsy tissue are diagnostic in most cases. The most common histologic light microscopy finding is a collapsing form of focal segmental glomerulosclerosis. The glomerular capillary tuft is collapsed (see the first image below) and may be segmentally or globally sclerosed. Visceral epithelial cells are hypertrophied and form a characteristic pseudocrescent in the Bowman space. Tubulointerstitial scarring, atrophy, and marked dilatation of the tubules (microcystic dilatations) are usually present (see the second image below).
Immunofluorescent microscopy helps to identify positive staining for albumin and immunoglobulin G in epithelial cells and for immunoglobulin M, C3, and, occasionally, A in mesangial or sclerotic areas.
Electron microscopy reveals wrinkling of the basement membranes, epithelial cell proliferation, and focal foot process effacement. Tubuloreticular structures in the glomerular endothelial cells (consisting of ribonucleoprotein and membrane, the synthesis of which is stimulated by alpha interferon) is highly predictive of HIVAN (see the image below).
Antiretroviral Drugs and Renal Function
Most HIV medications are well tolerated, even in the presence of renal insufficiency . The (potential) toxicity of the nucleoside reverse transcriptase inhibitors (ie, zidovudine, didanosine, zalcitabine, stavudine, lamivudine, abacavir, emtricitabine) uniformly manifests as type-B lactic acidosis. However, didanosine may cause electrolyte abnormalities, such as hypokalemia, hyponatremia, hypermagnesemia, and hyperuricemia, and stavudine may cause hyperuricemia.
Tenofovir is a nucleotide reverse transcriptase inhibitor (NRTI) with known renal toxicity and hypophosphatemia, and therefore, dose adjustment is indicated when creatinine clearance is less than 50 mL/min. Except for nevirapine, which may cause lactic acidosis, the nonnucleoside reverse transcriptase inhibitors (ie, nevirapine, delavirdine, efavirenz, etravirine) have no reported significant renal toxicity.
As a class, the protease inhibitors (PIs) (ie, saquinavir, ritonavir, indinavir, nelfinavir, amprenavir, fosamprenavir, lopinavir, atazanavir, tipranavir, darunavir) may precipitate nephrolithiasis. A classic form of this is indinavir crystalluria, which occurs independently of renal function; however, the stones resolve after cessation of indinavir therapy. A study by Rockwood et al found that the rate of kidney stones was 7.3 per 1000 patient-years in patients receiving ritonavir-boosted atazanavir compared with 1.9 per 1000 patient years in patients receiving other commonly used antivirals.
Enfuvirtide (Fuzeon) is the first of a newer class of fusion inhibitors that targets the gp41 protein on the surface of HIV and stops the virus from entering cells. Enfuvirtide has no known renal effects for creatinine clearance of greater than 35 mL/min.
Maraviroc (Selzentry), approved in August 2007, is also a fusion inhibitor. It blocks the CCR5 coreceptor on CD4+ cells, preventing the virus from entering. Maraviroc does not require dose adjustment for creatinine clearance greater than 50 mL/min.
Raltegravir (Isentress) is the first of a newer class of integrase strand transfer inhibitors. It does not require dose adjustment in patients with abnormal renal function.
Dose adjustment should be made in patients receiving nucleoside reverse transcriptase inhibitors (NRTI) when the glomerular filtration rate falls below 50 mL/min. Patients receiving nonnucleoside reverse transcriptase inhibitors (NNRTIs) may also receive a dose adjustment when the glomerular filtration rate falls below 50 mL/min. No dose adjustment is required for patients taking protease inhibitors.
Some drugs used to treat opportunistic infections in HIV disease may also cause nephrotoxicity or electrolyte abnormalities (see the image below).
Although there are no guiding clinical studies, some experts recommend consideration of therapy in all patients with HIV-associated nephritis (HIVAN). Initiation or adjustment of antiretroviral therapy may be indicated. See Antiretroviral Therapy for HIV Infection for specific information, including recommendations for dosage adjustments for renal insufficiency.
Angiotensin-converting enzyme (ACE) inhibitors and corticosteroids have been studied for use in HIVAN. Some reports on pediatric populations suggest that cyclosporine can be effective in reducing proteinuria in persons with HIVAN. The usefulness of cyclosporine therapy for HIVAN warrants further study. Researchers are pursuing several promising therapeutic strategies. Patients who progress to end-stage renal disease (ESRD) require dialysis and consideration of renal transplantation in carefully selected cases.
Angiotensin-Converting Enzyme Inhibitors
In patients with advanced renal insufficiency, captopril was noted to improve renal survival for a mean length of 37-156 days. In a subsequent prospective follow-up of 44 patients, the median length of renal survival for patients who received fosinopril was 479.5 days, with only 1 patient developing ESRD. All untreated control subjects progressed to ESRD, with a median length of renal survival of 146.5 days.
The exact mechanism of action of ACE inhibitors in HIVAN is unknown, but it may be related to a hemodynamic effect, a reduction in the transglomerular passage of serum proteins, and an antiproliferative effect mediated in part by the inhibition of transforming growth factor beta. Use ACE inhibitors if patients do not have hyperkalemia.
A number of case reports have suggested that corticosteroids offer some short-term benefit in HIVAN. In one report, results from a pretreatment and posttreatment kidney biopsy suggested that an improvement in renal function was associated with a reduced number of lymphocytes and macrophages infiltrating the interstitium.
In another report, of 20 patients who were treated with prednisone, 60 mg/day for 2-11 weeks, followed by a slow taper, 8 patients required maintenance dialysis, 11 died from complications, and 7 were alive and no longer had ESRD after a follow-up of 44 weeks.
Experimental therapeutic strategies
Animal research shows promising results for retarding renal disease progression in HIVAN. In one study, the use of a cyclin-dependent kinase inhibitor decreased visceral epithelial cell proliferation in HIV-infected mice.
In another study, blocking nuclear factor kappa beta (a cell signaling pathway) in mice resulted in increased lifespan and kidney and lean body mass preservation. These benefits were associated with a reduction in the number of CD45+ cells infiltrating the kidneys, amelioration of the renal architecture, and a reduction in the level of circulating inflammatory cytokines. Further studies are needed to determine the role of these inhibitors on human HIVAN.
End-Stage Renal Disease
The care of patients with HIV-associated nephropathy (HIVAN) who progress to end-stage renal disease (ESRD) remains a clinical challenge. Physicians must anticipate progressive renal disease in patients with HIVAN and plan the placement of an arteriovenous fistula in a timely manner for future use in hemodialysis. In current practice, hemodialysis is the accepted modality of ESRD therapy in these patients.
A study by Ifudu et al showed that during a 68-month observation period, 17 (50%) of 34 patients with HIV infection and ESRD died, compared with 65 (50%) of 131 patients with ESRD alone. Mean survival was equivalent between patients with both HIV infection and ESRD and those with ESRD alone (47.4 mo and 50.2 mo, respectively).
Because of increased susceptibility to infections, peritoneal dialysis has not been widely advocated. Similarly, immunosuppression after kidney transplantation is thought to pose a substantive risk of opportunistic infections in patients with HIVAN. Consequently, kidney transplantation in these patients is performed with caution in compliant, stable patients with no prior opportunistic infections who have an undetectable viral load and a CD4+ T-cell count of more than 300 cells/µL.
Anecdotal reports drawn from small samples of this selected group of patients with HIVAN suggest no extra risk of opportunistic infections. Until larger studies are performed, however, transplantation in persons with HIVAN should be focused on stable patients.
In one study, 1- and 2-year actuarial patient survival was 85% and 82%, respectively, and graft survival was 75% and 71%, respectively. Plasma HIV-1 RNA remained undetectable, and CD4 counts remained in excess of 400 cells/µL with no evidence of AIDS for up to 2 years. These results were comparable to other high-risk populations receiving kidney transplantation.
In a prospective study of kidney transplantation in 150 HIV-infected patients, Stock et al reported 1- and 3-year graft survival rates of 90.4% and 73.7%, respectively. The rejection rate was higher than expected, however, with 1- and 3-year estimates of 31% and 41%, respectively. Living-donor transplants were protective. Before transplantation, all patients had CD4 counts of at least 200/µL and undetectable plasma HIV-1 RNA levels while being treated with a stable antiretroviral regimen, and HIV remained well controlled after transplantation.
In an analysis of the United Network for Organ Sharing (UNOS) database, Locke et al evaluated 39,501 patients undergoing renal transplantation between January 1, 2004, and June 30, 2006 and found no difference in patient survival among HIV-positive patients versus HIV-negative patients (95.4% vs 96.2%, respectively). However, death-censored 1-year graft survival was significantly lower among HIV-positive patients (87.9% vs 94.6%).
Locke et al concluded that with proper donor selection and transplant recipient management, including the avoidance of prolonged cold ischemic time, use of living donors, and determination of optimal immunosuppression dosing before transplant, long-term graft survival comparable to that in HIV-negative patients can be achieved.
Wyatt CM, Morgello S, Katz-Malamed R, et al. The spectrum of kidney disease in patients with AIDS in the era of antiretroviral therapy. Kidney Int. 2009 Feb. 75(4):428-34. [Medline].
Kaku M, Simpson DM. HIV neuropathy. Curr Opin HIV AIDS. 2014 Nov. 9 (6):521-6. [Medline].
Kalim S, Szczech LA, Wyatt CM. Acute kidney injury in HIV-infected patients. Semin Nephrol. 2008 Nov. 28(6):556-62. [Medline].
Atta MG, Deray G, Lucas GM. Antiretroviral nephrotoxicities. Semin Nephrol. 2008 Nov. 28(6):563-75. [Medline].
Calza L, Vanino E, Magistrelli E, Salvadori C, Cascavilla A, Colangeli V, et al. Prevalence of renal disease within an urban HIV-infected cohort in northern Italy. Clin Exp Nephrol. 2013 May 28. [Medline].
Fabian J, Naicker S, Goetsch S, Venter WD. The clinical and histological response of HIV-associated kidney disease to antiretroviral therapy in South Africans. Nephrol Dial Transplant. 2013 Feb 26. [Medline].
Foy MC, Estrella MM, Lucas GM, Tahir F, Fine DM, Moore RD, et al. Comparison of Risk Factors and Outcomes in HIV Immune Complex Kidney Disease and HIV-Associated Nephropathy. Clin J Am Soc Nephrol. 2013 May 16. [Medline].
Phair J, Palella F. Renal disease in HIV-infected individuals. Curr Opin HIV AIDS. 2011 Jul. 6(4):285-9. [Medline].
Schwartz EJ, Klotman PE. Pathogenesis of human immunodeficiency virus (HIV)-associated nephropathy. Semin Nephrol. 1998 Jul. 18(4):436-45. [Medline].
Winston JA, Burns GC, Klotman PE. Treatment of HIV-associated nephropathy. Semin Nephrol. 2000 May. 20(3):293-8. [Medline].
Carbone L, D'Agati V, Cheng JT, et al. Course and prognosis of human immunodeficiency virus-associated nephropathy. Am J Med. 1989 Oct. 87(4):389-95. [Medline].
Rao TK. Human immunodeficiency virus (HIV) associated nephropathy. Annu Rev Med. 1991. 42:391-401. [Medline].
Ray PE, Xu L, Rakusan T, et al. A 20-year history of childhood HIV-associated nephropathy. Pediatr Nephrol. 2004 Oct. 19(10):1075-92. [Medline].
O'Donnell MP, Chao CC, Gekker G, Modi KS, Kasiske BL, Keane WF. Renal cell cytokine production stimulates HIV-1 expression in chronically HIV-1-infected monocytes. Kidney Int. 1998 Mar. 53(3):593-7. [Medline].
Kopp JB, Smith MW, Nelson GW, et al. MYH9 is a major-effect risk gene for focal segmental glomerulosclerosis. Nat Genet. 2008 Oct. 40(10):1175-84. [Medline].
Locke JE, Montgomery RA, Warren DS, et al. Renal transplant in HIV-positive patients: long-term outcomes and risk factors for graft loss. Arch Surg. 2009 Jan. 144(1):83-6. [Medline].
Matavele Chissumba R, Silva-Barbosa SD, Augusto Â, Maueia C, Mabunda N, Gudo ES Jr, et al. CD4(+)CD25(High) Treg cells in HIV/HTLV co-infected patients with neuropathy: high expression of Alpha4 integrin and lower expression of Foxp3 transcription factor. BMC Immunol. 2015 Sep 2. 16:52. [Medline].
Chan KT, Papeta N, Martino J, et al. Accelerated development of collapsing glomerulopathy in mice congenic for the HIVAN1 locus. Kidney Int. 2009 Feb. 75(4):366-72. [Medline].
Pillay P, Wadley AL, Cherry CL, Karstaedt AS, Kamerman PR. Pharmacological treatment of painful HIV-associated sensory neuropathy. S Afr Med J. 2015 Sep 14. 105 (9):769-72. [Medline].
Ifudu O, Rao TK, Tan CC, et al. Zidovudine is beneficial in human immunodeficiency virus associated nephropathy. Am J Nephrol. 1995. 15(3):217-21. [Medline].
Rockwood N, Mandalia S, Bower M, Gazzard B, Nelson M. Ritonavir-boosted atazanavir exposure is associated with an increased rate of renal stones compared with efavirenz, ritonavir-boosted lopinavir and ritonavir-boosted darunavir. AIDS. 2011 Aug 24. 25(13):1671-3. [Medline].
Kimmel PL, Mishkin GJ, Umana WO. Captopril and renal survival in patients with human immunodeficiency virus nephropathy. Am J Kidney Dis. 1996 Aug. 28(2):202-8. [Medline].
Wei A, Burns GC, Williams BA. Long-term renal survival in HIV-associated nephropathy with angiotensin-converting enzyme inhibition. Kidney Int. 2003 Oct. 64(4):1462-71. [Medline].
Smith MC, Austen JL, Carey JT, et al. Prednisone improves renal function and proteinuria in human immunodeficiency virus-associated nephropathy. Am J Med. 1996 Jul. 101(1):41-8. [Medline].
Gherardi D, D'Agati V, Chu TH, et al. Reversal of collapsing glomerulopathy in mice with the cyclin-dependent kinase inhibitor CYC202. J Am Soc Nephrol. 2004 May. 15(5):1212-22. [Medline]. [Full Text].
Ifudu O, Salifu MO, Reydel C, et al. Survival of Patients with AIDS and End-Stage Renal Disease Receiving Hemodialysis. Dial Transplant. 2004. 33:176-182.
Sawinski D, Murphy B. End-stage renal disease and kidney transplant in HIV-infected patients. Semin Nephrol. 2008 Nov. 28(6):581-4. [Medline].
Abbott KC, Hypolite I, Welch PG, et al. Human immunodeficiency virus/acquired immunodeficiency syndrome-associated nephropathy at end-stage renal disease in the United States: patient characteristics and survival in the pre-highly active antiretroviral therapy era. J Nephrol. 2001 Sep-Oct. 14(5):377-83. [Medline].
Berggren R, Batuman V. HIV-associated renal disorders: recent insights into pathogenesis and treatment. Curr HIV/AIDS Rep. 2005 Aug. 2(3):109-15. [Medline].
Choi AI, Rodriguez RA. Renal manifestations of HIV. HIV InSite. Available at http://hivinsite.ucsf.edu/InSite.jsp?page=kb-04-01-10. Accessed: 2/3/09.
Conaldi PG, Bottelli A, Baj A, et al. Human immunodeficiency virus-1 tat induces hyperproliferation and dysregulation of renal glomerular epithelial cells. Am J Pathol. 2002 Jul. 161(1):53-61. [Medline]. [Full Text].
Dave MB, Shabih K, Blum S. Maintenance hemodialysis in patients with HIV-associated nephropathy. Clin Nephrol. 1998 Dec. 50(6):367-74. [Medline].
Dominguez S, Ghosn J, Peytavin G, et al. Efficacy and safety of tenofovir double-dose in treatment-experienced HIV-infected patients: the TENOPLUS study. J Med Virol. 2007 Feb. 79(2):105-10. [Medline].
Ifudu O, Mayers JD, Matthew JJ, et al. Uremia therapy in patients with end-stage renal disease and human immunodeficiency virus infection: has the outcome changed in the 1990s?. Am J Kidney Dis. 1997 Apr. 29(4):549-52. [Medline].
Marras D, Bruggeman LA, Gao F, et al. Replication and compartmentalization of HIV-1 in kidney epithelium of patients with HIV-associated nephropathy. Nat Med. 2002 May. 8(5):522-6. [Medline].
Rao TK. Human immunodeficiency virus infection and renal failure. Infect Dis Clin North Am. 2001 Sep. 15(3):833-50. [Medline].
Szczech LA, Gupta SK, Habash R, et al. The clinical epidemiology and course of the spectrum of renal diseases associated with HIV infection. Kidney Int. 2004 Sep. 66(3):1145-52. [Medline].
Tan HP, Kaczorowski DJ, Basu A, et al. Living-related donor renal transplantation in HIV+ recipients using alemtuzumab preconditioning and steroid-free tacrolimus monotherapy: a single center preliminary experience. Transplantation. 2004 Dec. 15;78(11):1683-8. [Medline].