Diagnostic Specificity

A limited number of laboratory tests will be reviewed in this section, in order to examine their specificity in relation to a specific disease. Various tests have been proclaimed as being virtually pathognomonic, at least if a certain value is exceeded. Some tests have considerable diagnostic value; with others, early enthusiasm has been tempered with experience.

Many diagnostic tests for cancer, based on analysis of blood or urine, have been suggested and described, but apparently none has proved sufficiently reliable. In 1 test, a small drop of serum is floated on the surface of a mixture of phenol and glycerin adjusted to pH 8. Normal serum is said to float as a discrete drop, whereas serum from cancer patients spreads out and dissolves in approximately 2 min.146 The chemical basis for this test is unknown. Claims for a high degree of specificity were not confirmed by others.160

Another suggested diagnostic test for cancer proposed the measurement of the optical density at 500 mu of serum diluted with an equal volume of saline solution.134 In general, serum from cancer patients exhibited lower optical densities than that from normal persons. Factors contributing to the optical density of serum at 500 mu include bilirubin, hemoglobin, and iron-sidero-phyllin, but the cause for the differences in optical density was not postulated by the authors. A subsequent evaluation of this test was published, revealing a high incidence of false-positive and false-negative results.137

A proposed biochemical test for pregnancy, presumably based on the reaction of free estrone with ra-dinitrobenzene or 2,4-dinitrophenylhydrazine to produce colored derivatives, was reported by the original author to have unusually high specificity.141 These observations were essentially confirmed in a quantitative spectrophotometric modification of the original method.123 Later reports have not confirmed these early findings.59 In 1 study it was concluded, in fact, that the amount of estrone in pregnancy urine is too small to be detected by the original method.50

It is difficult to reconcile original findings on such tests, as reported above, with subsequent inability to reproduce the results. Various factors must be considered, such as unconscious bias (for or against), failure to duplicate experimental conditions exactly, and failure to include a sufficient range of pathologic material other than the specific condition under investigation.

Proposed diagnostic tests with no plausible scientific basis should be regarded with skepticism until a rational explanation is forthcoming. A more recent example is the p-toluenesulfonic acid test for acute systemic lupus erythematosus.100 This test consisted of mixing 0.1 ml. of serum with 2 ml. of a 12 per cent solution of p-toluenesulfonic acid in glacial acetic acid and observing for a clot or precipitate after 20 min. Numerous subsequent studies have demonstrated that the test is nonspecific and will not distinguish

May 1962

between lupus and rheumatoid arthritis. The positive reactions seem to correlate with increased serum beta or gamma globulins and with other nonspecific compounds, such as C-reactive protein and serum mucopro-tein.11'25•52•129,171

A specific enzyme, leucine aminopeptidase, has been described as being consistently elevated in the serum of patients with carcinoma of the pancreas.73,145 Subsequent studies have revealed that this is not invariably so, and that the enzymatic activity is elevated in patients with a wide spectrum of hepato-biliary-pancreatic diseases, including cirrhosis, viral hepatitis, infectious mononucleosis, common duct stone, metastases to the liver, acute cholecystitis, and acute pancreatitis.6'27'93 Although nonspecific, the test still has limited value, in that a normal finding is presumptive evidence against the presence of carcinoma of the pancreas.

An elevated level of serum acid phosphatase is often regarded as pathognomonic for metastatic carcinoma of the prostate gland when the value exceeds 5 Bodansky units or 10 King-Armstrong units.32 Ozar and co-workers128 have evaluated various factois that might affect the reliability of this determination: hemolysis, endocrine therapy, prostatic manipulation, stability of substrates, and effect of anticoagulants. In 17 of a group of 20 men, the serum acid phosphatase rose to abnormal levels within 1 hr. after prostatic massage, but returned to normal within 24 hr.92 Elevated values are encountered in the newborn111 and in females receiving androgen therapy for metastatic carcinoma of the breast. Platelets may release the enzyme into the serum during the clotting process.197 Some discrepancies may be related to the presence of several acid phosphatases in serum and to the lack of substrate specificity. The subject has been reviewed recently by Woodard.187 A case has been reported with an elevation in both total and tartrate-inhibited serum acid phosphatase, although no demonstrable carcinoma of the prostate gland was found on autopsy.165

Estimation of serum amylase is a valuable aid in the diagnosis of acute pancreatitis.

Values of more than 1000 Somogyi units are believed by some authors to be pathognomonic.9 In the absence of any pathologic process in the pancreas, however, values ranging from 1300 to 3200 have been reported in cases of peritonitis, mesenteric adenitis, and staphylococcal pneumonia.20 Highly elevated values are also found in patients with biliary lithiasis.19 In 1 series, 67 patients were explored surgically within 6 weeks of an attack of clinical acute pancreatitis. Fifty of these patients had values for serum amylase that exceeded 1000 Somogyi units, and, of these 50, there were 43 with biliary lithiasis, but with a grossly normal pancreas.3 Serum amylase is also greatly elevated in mumps, although serum lipase is not.31 It is recognized, of course, that the serum amylase is greatly elevated in acute pancreatitis,1*116 but this does not make the test specific for this disease.

The excretion of porphobilinogen in the urine is believed to occur rather specifically in acute idiopathic porphyria.179 A routine survey of 1000 specimens of urine revealed no false-positive reactions.82 In another study, however, false-positive reactions were observed to occur in approximately 50 per cent of patients with epilepsy.117 The material manifesting the positive test was demonstrated not to be porphobilinogen.

In a recent study, the Congo red test was performed on 20 patients with a histologic diagnosis of amyloidosis.71 Only 9 of these had more than 60 per cent disappearance of Congo red from the blood in 1 hr., and of these only 5 had more than 90 per cent disappearance. Again, it is recognized that a positive Congo red test is strong presumptive evidence for the presence of amyloidosis. Few false-positive tests occur, but false-negative tests may be expected in more than 50 per cent of patients with amyloidosis.

The estimation of serum protein-bound iodine is regarded as a measure of circulating thyroxine and is an excellent index of thyroid activity. Occasionally, normal values have been found in cretins.154 Presumably the protein-bound iodine may be present as a biologically inactive iodinated protein. It is significant that these patients had very low serum levels of butanol-extractable io

Was this article helpful?

0 0
Arthritis Relief and Prevention

Arthritis Relief and Prevention

This report may be oh so welcome especially if theres no doctor in the house Take Charge of Your Arthritis Now in less than 5-Minutes the time it takes to make an appointment with your healthcare provider Could you use some help understanding arthritis Maybe a little gentle, bedside manner in your battle for joint pain relief would be great Well, even if you are not sure if arthritis is the issue with you or your friend or loved one.

Get My Free Ebook


  • Myrtle
    What is the specificity for clinical chemistry?
    2 years ago

Post a comment