LETTERS 78 1

pako (4) on in vitro PGE2-induced contractions of the rat stomach strip. We have personally observed a non- selective antagonism of chloroquine (5) on several agon- ist-induced contractions, including the effect of PG in the guinea pig isolated ileum. In vivo chloroquine is also able to antagonize PG actions. It suppresses cardiac rhythm disturbances induced by PG in the rat (6). It is able to promote closure of patent ductus arteriosus in premature infants, a condition known to persist when high levels of PG are achieved (2).

On the basis of these results, Manku and Horro- bin (2) suggested that binding sites for chloroquine in- cluding those to nucleic acids might be binding sites for PG. Actually they have shown that PG can block the binding of chloroquine to DNA. But more important is the fact that chloroquine might interfere with PG mem- brane binding sites which trigger cyclic-AMP produc- tion. The results of Teitz and Chrisman (7) cited by Ful- kerson et a1 (1) are easily explained by this hypothesis. The intraarticular injection of PG in the rabbit knee in- duces a synovitis that is not prevented by systemic salic- ylate, since it is not due to stimulation of PG synthetase but is abolished by systemic chloroquine because of its PG antagonist effect.

We propose a similar interpretation of the results of Fulkerson et a1 (1). If chloroquine is truly a PG an- tagonist, it is not surprising that the in vitro effects of PGE on cartilage slices are abolished by this drug. However, in our interpretation the main effect of chloroquine must take place at the membrane binding sites for PG. An additional action on DNA-dependent RNA synthesis is still possible, but it would not be nec- essary for obtaining the inhibition of the stimulation of hexosamine cartilage depletion by PGE.

J.P. FAMAEY, MD J. FONTAINE, PhD Laboratory of Pharmacology Rheumatology Unit School of Medicine and Institute of Pharmacy University of Brussels Belgium

REFERENCES

1. Fulkerson JP, Landenbauer-Bellis IM, Chrisman OD: In vitro hexosamine depletion of intact articular cartilage by E-prostaglandins: prevention by chloroquine. Arthritis Rheum 22:1117-1121, 1979

2. Manku MS, Horrobin DF: Chloroquine, quinine, procaine, quinidine and clomipramine are prostaglandin agonists and antagonists. Prostaglandins 12:789-80 1, 1976

3. Horrobin DF, Manku MS, Karmazyn M, Ally AI, Morgan RO, Karmali RA: Quinacrine is a protaglandin antagonist. Biochem Biophys Res Commun 76:1188-1193, 1977

4. Okpako DT: Interactions of aminoquinolines and mepa- crine with prostaglandin E2 in tissues of rats and guinea pigs. Gen Pharmacol 9:25-28, 1978

5. Famaey JP, Fontaine J, Reuse J: An analysis of the inhib- itory effects and of possible prostaglandins antagonism of chloroquine in the guinea pig isolated ileum. J Pharm Pharmacol29:761-762, 1977

6. Swift A, Karmazyn M, Horrobin DF, Manku MS, Karmali RA, Morgan RO, Ally AI: Low prostaglandin concentra- tions cause cardiac rhythm disturbances: effect reversed by low levels of copper or chloroquine. Prostaglandins 15:651-657, 1978

7. Teitz CC, Chrisman OD: The effect of salicylate and chloroquine on prostaglandin-induced articular damage in rabbit knees. Clin Orthop 108:264-274, 1975

Renal disease in rheumatoid arthritis To the Editor:

The association of renal disease with rheumatoid arthritis (RA) has been debated for many years. A re- cent review (1) and case report (2) have suggested that a membranous glomerulonephritis (GN) occasionally oc- curs in the course of the joint disease. The nature of the relationship has been difficult to clarify, since agents used in the treatment of RA have themselves been im- plicated in the production of proteinuric renal disease. In fact, Samuels ( 1 ) hypothesized that the GN might be pathogenetically linked to the rheumatoid disease and that gold or penicillamine treatment could produce the nephrotic syndrome in only appropriately predisposed patients. Most patients studied who developed protein- uria in the course of rheumatoid disease had also taken some potentially nephrotoxic therapy before the detec- tion of proteinuria. Those with well documented RA had long-standing joint disease before the renal disease became manifest.

Case report. We have recently seen a 40-year-old man who came to the New York Veterans Administra- tion Medical Center complaining of pain and swelling in his right ankle and right temporomandibular joint. He had experienced a symmetrical but migratory pain in his shoulders and hands and morning stiffness of 1- hour duration for 1 year before admission. Physical ex- amination revealed moderate synovitis of the meta- carpophalangeal and proximal interphalangeal joints bilaterally and reduced range of motion in both shoul- ders. His laboratory data consisted of a latex fixation ti- ter of 1:2560 and sheep cell agglutination at 1:320. His hematocrit (HCT) was 47.3% and white blood cell count

782 LETTERS

Figure 1 . Electron micrograph of the first renal biopsy Only a few electron-dense deposits (ar- rows) are present in subepithelial sites of the glomerular basement membrane (b) of a peripheral capillary loop (X 8000)

Figure 2. Electron micrograph of the second renal biopsy. There are more numerous electron- dense deposits than previously. Deposits are separated by projections of the glomerular base- ment membrane (b) in a peripheral capillary loop (X 4800).

LETTERS 783

Figure 3. Glomerulus from the second renal biopsy stained with fluo- rescein-conjugated antiserum to IgG. Diffuse uniform finely granular deposits are present in glomerular capillary walls (X 800).

(WBC) was 6,300/ml. The antinuclear antibody was weakly positive with a diffuse pattern, while the ENA, HbAg, LE prep and cryoglobulins gave negative results. The ESR was 71 mm/hour total; serum hemolytic com- plement was 178; the BUN was 16 mg/dl; serum choles- terol, 299 mg/dl; serum albumin-to-globulin, 4.1/3.1; creatinine 0.8 mg/dl; 24-hour urine protein, 6 gm/liter. The urine sediment contained 2 to 4 white blood cells per high power field.

A diagnosis of rheumatoid arthritis was made. The patient began taking enteric-coated aspirin, 16 tab- lets per day. He noted lessening of his morning stiffness and joint pains over the next few months. A renal biopsy showed focal segmental proliferative glomerulo- nephritis in I of 10 glomeruli, while others were normal by light microscopy. Ultrastructurally, there were scat- tered electron-dense deposits restricted to subepithelial sites (Figure 1). Immunofluorescence microscopy re-, vealed mild, diffuse, finely granular staining for IgG and C3; there was focal staining for IgM but none for IgA. He was seen monthly over the next year and con- tinued to take both salicylates and a nonsteroidal anti- inflammatory agent (ibuprofen).

His physical examination continued to show mild-to-moderate synovitis of the ankles and hands. No erosive changes were noted on x-ray. He subsequently developed peripheral and periorbital edema. Labora- tory results during this period revealed a serum albu- min-to-globulin ratio of 3.7 to 3.0, serum cholesterol of 430 mg/dl, ESR of 96 mm/hr, and a 24-hour urine pro- tein of 6 gm/liter. The HCT was 41%, and the WBC

6,30O/ml. He was given a thiazide diuretic which re- duced his peripheral edema. Over the next several months, the arthritis of his ankles became more severe. He was hospitalized for additional antiinflammatory therapy, bed rest, and intraarticular steroid injections. After 2 weeks he improved and was discharged. Despite diuretic therapy, his peripheral edema recurred. A sec- ond renal biopsy 1 year after the first contained 18 glomeruli, all showing normal cellularity and more cap- illary wall thickening than in the previous biopsy. Elec- tron microscopy revealed more numerous deposits than in the previous biopsy in a subepithelial location, and more glomerular basement membrane thickening (Fig- ure 2). There was now strong staining for all Igs and C3 (Figure 3).

Discussion. The course of our patient demon- strates that progressive glomerular disease can occur with rheumatoid arthritis in the absence of any therapy known to produce the nephrotic syndrome. It was not associated with either long-standing or particularly ag- gressive joint disease but was clearly progressive mor- phologically with increasing immune complex deposi- tion.

Given the occurrence of such an entity in several other reported patients who received no potential neph- rotoxic therapy (3), it seems germane to ask if those pa- tients who developed the nephrotic syndrome during penicillamine or gold therapy for RA did so as a result of their disease or their therapy. Perhaps patients whose renal disease did not resolve after cessation of therapy were, in fact, suffering from rheumatoid membranous glomerulonephritis, an unusual complication of rheu- matoid arthritis.

ROBERT FRIEDMAN, MD GLORIA R. GALLO, MD JOEL N. BUXBAUM, MD New York University Medical Center New York VA Medical Center New York City

REFERENCES

1. Samuels B, Lee JC, Engelman EP, Hopper S Jr: Mem- branous nephropathy in patients with rheumatoid arthritis: relationship to gold therapy. Medicine 57:3 19-327, 1978

2. Davis JA, Cohen AH, Weisbart R, Paulus HE: Glomerulo- nephritis in rheumatoid arthritis. Arthritis Rheum

3. Row PG, Cameron JS, Turner DR, Evans DJ, White RHR, Ogg CS, Chantler C, Brown CB: Membranous ne- phropathy: long-term follow-up and association with neo- plasia. Quart J Med (new series) 44:207-239, 1975

22: 101 8- 1023, 1979