Inclusion Criteria

- INCLUSION CRITERIA:

- Patients with podocyte diseases, MCD, FSGS (including idiopathic, secondary, and
hyperfiltration variants), and CG (including HIV-associated variant), who meet the
following criteria:

- An adequate renal biopsy, defined as having minimum 10 glomeruli for light microscopy
and minimum 3 glomeruli for electron microscopy, unless the podocyte disease is
diagnostic on fewer glomeruli. Patients who have non-diagnostic or technically
inadequate biopsies will be offered the opportunity to undergo a research biopsy to
determine eligibility. For some patients who have had a non-diagnostic or
technically inadequate biopsy, a repeat renal biopsy may be clinically indicated in
an effort to diagnose and treat a potentially serious kidney disease.

- Adults greater than or equal to 16 years of age. The rationale for excluding younger
children is that retinoids may carry greater toxicity in children, as there are
reports of premature epiphyseal closure, development of slender long bones, and
periostial thickening.

- Prior treatment with at least two immunosuppressive agents that have been shown to
induce remission in MCD and FSGS: glucocorticoids, cyclosporine, tacrolimus,
cyclophosphamide, chlorambucil, and mycophenolate mofetil. Therapy with each agent
must last at least 8 weeks. Exemptions will be made for those with contraindications
to these medications or those who cannot tolerate these medications. The rationale is
to recruit patients who have failed conventional therapy. All patients will be off
immunosuppression for at least 4 weeks before starting retiniods in order to avoid a
confounding effect.

- Three 24h urine collections with mean protein excretion greater than or equal to 3.5
g/d obtained within one month prior to enrolling in the study. These 24hr urine
collections will be collected after the patient has been on a stable dose of ACE
inhibitor or ARB for at least 4 weeks (the maximal antiproteinuric effect of these
medications occurs after 4 weeks). The rationale is that patients with
nephrotic-range proteinuria are at high risk for progressive renal disease,
justifying participation in a clinical trial with novel agents.

- Blood pressure of less than or equal to 130/80 on a stable dose of ACE inhibitor or
ARB for at least 1 month or greater than or equal to 75 percent of measurements
before the entry of the study. The rationale is that uncontrolled hypertension can
exacerbate proteinuria.

EXCLUSION CRITERIA:

- Pregnancy, breastfeeding, or unwillingness to use at least two contraceptive methods
(at least one of which must be primary, including tubal ligation, partner vasectomy,
oral contraceptives, implanted contraceptives, and intrauterine device). The
rationale is that retinoids are teratogenic and are excreted in breast milk.

- Abnormal liver function test, including AST, ALT, total bilirubin, or protime. The
rationale is that retinoids can be hepatotoxic. The only exception will be the
following: if the cause of abnormality of LFTs is felt to be due to a specific
hepatotoxic drug such as a statin and the levels are less than 2 times the upper
limit of the normal AND normalize upon holding the offending drug, patients may be
considered for study participation after consultation with hepatology service.

- Hypertriglyceridemia greater than 500 mg/dL despite statin/fibrate therapy. The
rationale is that retinoids can increase lipids, particularly triglyceride and this
can lead to pancreatitis.

- Any medical conditions requiring concurrent immunosuppression, as this pilot study is
designed to evaluate the effect of retinoids as a monotherapy.

- Any medical conditions requiring concurrent use of tetracycline, minocycline, or
doxycycline, due to enhanced risk of increased intracranial pressure.

- Hypersensitivity to retinoids.

- Presence of any unstable cardiovascular disease, uncontrolled diabetes with
hemoglobin A1c greater than 8 percent, chronic inflammatory or infectious conditions
except HIV-1 infection. Retinoids have been associated with chest pain of unclear
etiology, increased serum glucose, myelosuppression and increased risk of infection.
The etiology of infection is not clear but may be related to myelosuppression.

- Those with HIV-1 infection must not have any evidence for opportunistic infectious
complications and have CD4 count greater than or equal to 200. Both tretinoin and
isotretinoin have been safely studied in HIV-infected patients for other indications.

- GFR less than 25 ml/min/1.73m(2)estimated by 5-variable MDRD equation, as the
metabolities of retinoids are excreted in part in urine, and there is a concern for
increased toxicity. In participants less than 18 years of age, we will use Schwartz
equation.

- Untreated depression, as retinoids have been associated with depression, suicidal
ideation, and aggressive behavior. If patients manifest significant depressive
symptoms, they will be included in the study only if assessed and agreed by a
psychiatrist (either at NIH or other) and if they have regular follow-up visits with
a psychiatrist.

- Factors that increase the risk of renal biopsy. These include the following:
unwillingness to accept blood transfusion, bleeding diathesis, single kidney, small
kidneys (less than 9.5 cm).