Liver function tests in rheumatoid arthritis.

Interest in liver function in rheumatoid arthritis has arisen from investigations into the aetiology of the disease and from the observation that an attack of jaundice may induce a remission (Hench, 1940). Laboratory studies have, however, given conflicting results. Impaired glucose tolerance was found by Pemberton and Foster (1920) and by Andrews and Muether (1941) and Flynn and Irish (1946). Rawls, Weiss, and Collins (1937, 1939) in a detailed study of 100 cases came to the conclusion that liver function was definitely impaired. Using the dye azorubin S they found that 550% of the cases gave abnormal results. In many of these, however, the dye appeared in the bile much more rapidly than normal, a phenomenon which the authors ascribed to liver irritability. This interpretation has been criticized by Lyon and Wirts (1937), and exclusion of these cases reduces the abnormal results to 190. This test has never become widely used, because it requires duodenal intubation, though Rosenberg and Soskin (1940) found it at least as sensitive as the more familiar bromsulphalein test. Bromsulphalein (B.S.P.) was used by Robinson (1943) with negative results; he had, however, considerably modified the test as commonly used, and his B.S.P. " index " was based on blood levels found during 7.25 minutes following injection of 3.5 mg. per kg. Recent work on the fate of injected B.S.P. has shown that this time is too short for the demonstration of all but severe liver damage; the percentage disappearance rate from plasma does not fall off unless a larger dose of B.S.P. is used and time is allowed for the liver to become " saturated" with the dye (Ingelfinger, Bradley, Mendeloff, and Kramer, 1948 ; Mendeloff, Kramer, Ingelfinger, and Bradley, 1949). Movitt and Davis (1953) have also used B.S.P. in rheumatoid arthritis, and found a small degree of dye retention (up to 3 %) in seven out of 15 cases. Abnormal results have been obtained with the hippuric acid test by several workers (Rawls, Weiss, and Collins, 1939 ; Lovgren, 1953 ; Kersley, Mandel, and Jeffrey, 1953) ; this test is, however, open to the objection that impairment of renal function may also give low values. The occurrence of serum protein changes in rheumatoid arthritis is now well established, and they account for the positive flocculation tests which are found in a proportion of cases (Carter and Maclagan, 1946; Kersley et al., 1953 ; Poulsen, 1949), but they are not necessarily indicative of impairment of liver function.