metabolitesin urine [Abstract]. Clin Chem 1987;33:972. Fibronectin Is Unsuitable as a Tumor Marker in Pleural Effusions. Patrick Deipuech, Gerard Deech, and ...
cannabinol-9-carboxylic acid. J Anal Toxicol 1988;12:136-40. 2. Sutheimer CA, Yarborough R, Hepler BR, et al. Detection and confinnation of urinary cannabinoids, J Anal Toxicol 1985;9:15660. 3. Kanter SL, Hollister LE, Williams M. Extraction of glucuronide metabolites of9-tetrahydrocannabinol by diethyl ether. J Chroma-
togr 1982;234:255-60. 4. Kaistha KK, Tardrus R Semi-quantitative
thin-layer massscreening detection of 11-nor-9-tetrahydmcannabinol-9-carboxylic acid in human urine. J Chromatogr 1982;237:528-33. 5. Vinson JA, Lopatofsky DJ. A semi-automated extraction and spotting system for drug analysis by TLC. I. Procedure for analysis ofthe major met.aboliteof9-tetrahydrocannabinol in urine. J Anal
Toxicol 1985;9:6-9. 6. Dupont Technical Note: Use of the Prep A extraction cartridge for confirmation of 9-THC by gas chromatography. Wilmington, DE: Dupont Clinical and Instruments Division, 1983. 7. Abercrombie ML, Jewel JS. Evaluation of EMIT and RIA high volume test procedures for THC metabolites in urine utilizing GC/ MS confirmation. J Anal Toxicol 1986;10:178-80. 8. Kogan MJ, Newman E, Willson WJ. Detection of marijuana metabolite 11-nor-i9-tetrahydrocannabinol-9-carboxylic acid in human urine by bonded phase adsorption and thin-layer chromatography. J Chromatogr 1983;306:441-3.
9. Nakamura GR, Stoll WJ, Masters RG, et al. Analysis of urine for 11-nor-9-tetrahydrocannabinol-9-carboxylic acid using Sep-Pak cartridges for sample clean-up. J Anal Chem 1985;57:1494-6. 10. Touchstone JC, Dobbins MF. Practice of thin-layer chromatography. New York: John Wiley and Sons, 1983;90-5. 11. Fenimore DC, Davis CM. High performance thin-layer chromatography. J Anal Chem 1981;53:252A-66A. 12. Michaud JD, Jones DW. Thin-layer chromatography for broadspectrum drug detection. Am Lab 1980;12:104-7. 13. Harkey MR, Stolowitz ML. Solid phase extraction techniques for biological specimens. 1n Baselt RC, ed. Advances in analytical toxicology. Vol I. Foster City, CA: Biomedical Publications,
1984;255-70. 14. Foltz RL, McGinnis KM, Chinn DM. Quantitative measurements of 9-tetrahydrocannabino1 and two major metabolites in physiological specimens using capillary column gas chromatography negative ion chemical ionization mass spectrometry. Biomed Mass Spectrom1983;10:316-23. 15. Baker TS, Harry JV, RussellJW, et al. Rapid methodfor the GC/MS confirmation of 11-nor-9-carboxy-9-tetrahydrocannabinol in urine. J Anal Toxicol 1984;8:255-9. 16. King DL, Gabor MJ, Martel PA, O’Donnell CM. Evaluation of a rapid TLC technique for the detectionof cannabinoid metabolitesin urine [Abstract]. Clin Chem 1987;33:972.
CLIN. CHEM. 35/1, 166-168 (1989)
Fibronectin Is Unsuitable as a Tumor Marker in Pleural Effusions Patrick Deipuech, Gerard Deech, and Fabienne Fructus
We studied 75 patients with nonmalignant pleural effusions (50 with pneumopathy, 16 with pulmonary tuberculosis, and nine with congestive heart failure) and 33 patients with malignant pleural effusions. We selected 105 mg/L as the most suitable cutoff concentration of fibronectin for distinguishing between the two groups. We found high concentrations of fibronectin in 21 of the 33 patients with malignant pleural fluid but also in 37 of the 75 patients with nonmalignant pleural fluid. Evidently, measuring fibronectin in pleural fluid will not help in differentiating nonmalignant from malignant pleural fluids (diagnostic accuracy: 55%). AdditIonal Keyphrases: cutoff value heart and lung disease
.
cytology
cancer
The differential diagnosis of pleural effusions is still a clinicalproblem. Cytological investigation, despite its high specificity, has often been unreliable, owing to the high proportion
of false-negative results (1). To our knowledge,
few data (2) have been published about fibronectiri, a multifunctional glycoprotein, in pleural fluid. Fibronectin accumulates locally at sites of injury and forms a provisional extracellular matrix in various inflammatory reactions involving macrophages, including delayed hypersensitivity responses (3) and healing wounds (4). Fibronectin has also been implicated as an important nonspecific opsonin. Here
Laboratoire de Chimie Clinique, H#{244}pital de la Durance, F-84025 Avignon Cedex, France. Received September 20, 1988; accepted October 31, 1988. 166
CLINICAL CHEMISTRY, Vol. 35, No. 1, 1989
we report
our
evaluation
fibronectin in distinguishing pleural effusions.
of the diagnostic accuracy of nonmalignant and malignant
Patients and Methods We studied prospectively 108 consecutive hospitalized patients (67 men, 41 women), ages 25-82 y (mean, 60 y). Their diagnoses were based on tissue biopsy, cytological examination, and response to therapy. We divided the patients into two groups. The first group comprised 24 patients with lung cancers, one with renal cancer, two with breast cancers, and one with coloniccancer; five more patients had non-Hodgkin’s lymphoma with positive lymph node biopsy. The lung-cancer patients were classified into four categories, depending on the predominant histological pattern (including homogeneous and heterogeneous tumors): adenocarcinoma (nine patients), squamous carcinoma (eight), large-cell carcinoma (four), and small-cell carcinoma (three). Three of the lungcancer patients died within a month after hospitalization. Group 2 consisted of 75 patients with nonmalignant conditions, all of whom were negative for cancer cells by cytological examination and pleural biopsy. Sixteen had tuberculosis; the 59 other patients included single cases of rheumatoid arthritis, renal disease, supraclaviculitis, intestinal obstruction, bone fracture, septicemia, polycystic kidney and liver, uremic pleuritis, alcoholic liver cirrhosis, perforated uterus with generalized peritonitis; nine cases of congestive heart failure; 14 cases of parapneumonia; 11 casesof bacterial endocarditis with pneumonia; and 15 cases of bacterial infection.
The pleural-effusion samples were freshly collected in polystyrene test tubes and examined cytologically within 2 h of aspiration. The sediment obtained by centrifugation of 10- to 20-mL samples at 500 x g for 10 mm(5) was smeared onto microscope slides, stained with the Gram and ZiehiNeelsen stains, and examined. The corresponding supernates were used for biochemical analysis: total protein measurement by the sulfosalicylic acid method (6) and fibronectin measurement by an immunoturbiclimetric method (7), in the Rotochem Ha (Roche Kontron) centrifugal analyzer. Reagents for fibronectin determination were obmined from Behring Diagnostics (Marburg, F.R.G.). Fibronectin antiserum (code no. OUND) was diluted fivefold with accelerator I (code no. OSVX) in isotonic sodium chloride (1:2). The reference curve was determined from a geometric dilution series of Protein Standard Plasma (code no. OTJID). Samples were diluted fourfold in isotonic NaC1. Kinetic turbidimetric measurement of the antigen-antibody reaction was recorded at 340 nm according to the principle of the fixed-time method (10 and 600 s). The 16 patients with advanced pulmonary tuberculosis (category ifi of the American Thoracic Society) were positive for acid-fast bacilli in sputum before admission to the study; positive cultures were also confirmed six weeks later. The data for the patients with nonmalignant pleural effusions and with low protein concentrations (30 gIL were found in pleural effusions from 14 of 33 patients with malignant pleural effusions (Table 1) and in 33 of 75 patients with nonmalignant pleural effusions (sensitivity: 42%, specificity: 56%, predictivity of a positive test: 25%, diagnostic accuracy: 42%). The results of the cytological tests were often also negative (22 of our 33 patients with cancer) (Tables 1, 2). In our study, pleural fluid from the 16 patients with active advanced pulmonary tuberculosis was a serous exudate with high fibronectin content (Table 3). Epstein et al. (10) also showed that tuberculous pleural effusions are generally serous exudates. Such fluids are thought to result from rupture of a subpleural caseous focus in the lung into the pleural space (11, 12). Most probably, they accumulate as a
Table 1. Diagnostic Value of Test Results Difterentiating Nonmalignant from Malignant Pleural Eftusions Nonmalignant
pleural effusions (n
=
Malignant pleural = 33)
effuslons (n
75)
Positivecytological examination Pleuraltotal protein >30 g/L
0/75
11/33(33%)
33/75 (44%)
14/33 (42%)
Pleural fibronectin >105 mg/L
37/75 (49%)
21/33 (64%)
Table 2. Fibronectin Concentrations and Cytological Test Results in Malignant Pleural Eftusions from 33 Patients DIagnosIs
Small-celllungcaa Squamouslungca Small-celllungca Squamouslungca Lungadenoca Lungadenoca Lungadenoca Squamouslungca Large-celllungca Lungadenoca Lungadenoca Squamouslungca Lungadenoca Lungadenoca Breastca Lung adenoca Small-cell lung ca Renalca
Breastca Non-Hodgkin’s lymphoma
Large-celllungca Non-Hodgkin’siymphoma
Flbronectln, mg/L
Cytology
474
+
374 335
+ + + + +
325 301 300 253
242 195 193 184
+ -
+
155
-
139
+
127 123 117 115 110 110 108 107
87
Non-Hodgkin’s lymphoma Squamous lung ca Lung adenoca
73 71 70
Squamouslungca Squamous lung ca Non-Hodgkin’s lymphoma
69 66
Colonic ca Squamous lung ca Large-cell lung ca Large-cell lungca Non-Hodgkin’s lymphoma ca: carcinoma.
result
62 44 39 33
-
+
-
-
+ -
-
29
-
26
-
CLINICALCHEMISTRY, Vol. 35, No. 1, 1989
167
Table 3. Fibronectin Concentrations Effuslons
accuracy: 56%). The histological pattern of lung cancer was not a variable related to flbronectin concentration (Table 2).
in Pleural Fibronectin concn, mg/L
No. subjects
Diagnosis
9 11
Congestiveheart failure Bacterialendocarditis
Mean
14 15
Parapneumonia Bacterial infection
16 10 33
Pulmonary
with pneumonia
tuberculosis
Other (see text)
Cancer
SD
139 150
65 82
141 151 143 j538
76 64 88 113
a Not
significantly(P >0.05) different from mean value for nonmalignant pleuraleffusions.
result of a hypersensitivity reaction to tuberculoprotein (12). If these patients’ fluids are omitted from consideration, the diagnostic accuracy of fibronectin concentration is higher (63% vs 55%), but still not sufficient to differentiate malignant from nonmalignant fluid. The mean fibronectin concentration in malignant pleural fluid was not significantly (P >0.05) different from the mean value for sterile pleural effusions (Table 3). Table 2 gives the fibronectin concentrations and the results of the cytological examination of every patient with malignant pleural effusions. Histological heterogeneity of lung cancers is well-documented. In our study, 30% of the tumors were homogeneous, 20% showed minor or subtype heterogeneity (e.g., mixtures of papillary and acinar adenocarcinoma, or mixtures of oat-cell and intermediate smallcell carcinoma), and 50% showed major heterogeneity, with all combinations of the four major patterns of small-cell, large-cell, squamous-cell, and adenocarcinoma being observed. Eighteen of 25 patients with lung carcinoma had fibronectin concentration >105 mgIL (Table 2) (diagnostic
We conclude that flbronectin marker in pleural effusions.
is unsuitable
as a tumor
References 1. Tomb J. A cytological study on serous fluid in cancer. Lab Med J 1974;27:51-8. 2. Klockars M, Pettersson T, Vartio T, Riska H, Vaheri A. Fibrenectin in exudative pleural effusions. J Clin Pathol 1982;35:723-7. 3. Clark RA, Dvorak HF, Colvin RB. Fibronectin in delayed-type hypersensitivity skin reaction association with vessel permeability and endothelial cell activation. J Imniunol 1981;126:787-93. 4. Fujikawa LS, Forster CS, Harnst TJ, Lanigan JM, Colvin RB. Fibronectin in healing rabbit corneal wounds. Lab Invest 1981;45:120-9. 5. Berge T, Hellsten S. Cytological diagnosis of cancer cells in pleural and ascitic fluids: comparison of results obtained by pathologist and cytologist. Acta Cytol 1966;10:138-40. 6. Meulemans 0. Determination of total protein in spinal fluid with sulphosalicylic acid and trichloroacetic acid. Clin Chim Acts 1960;5:757-61. 7. Delpuech P, Grangeon F, Desch G, Hassanaly F. Dosage de la fibronectine plasmatique en inununoturbidim#{233}trie:adaptation sur analyseur centrifuge (Rotochem ila). Comparaison avec use tech-
nique immunonephelem#{233}trique (BNA Behring). Rev Fr Lab 1987;161:9-14. 8. Galen RB,GambinoSR.Beyondnormality-the predictive value and efficiencyof medical diagnosis. New York: John Wiley and Sons, 1975.
9. Delpuech P, Desch G, Grangeon F, Hassanaly F, Sautecoeur M. Fibronectine plasmatique: influences de l’ageet du sexe. Ann Biol Clin 1988;46:313-8. 10. Epstein DM, Kline LR, Albeda SM, Miller WT. Tuberculous pleural effusions. Chest 1987;91:106-9. 11. Light RW. Tuberculous pleural effusions. In: Light RW, ed. Pleural diseases. Philadelphia: Lea & Febiger, 1983:119-25. 12. Berger HW, Mejia E. Tuberculous pleurisy. Chest 1973;63:8892.
CLIN. CHEM. 35/1, 168-172 (1989)
Salivary Progesterone Excellently Reflects Free and Total Progesterone in Plasma during Pregnancy P. M. M. Meulenberg1 and J. A. Hofman2 To see if saliva is a valid substitute for plasma in assay of
progesterone even when concentrations of hormone and binding proteins are fluctuating, we determined the concentrations of total and free progesterone in plasma and salivary
progesterone in specimens from 36 women volunteers during the course of pregnancy and six weeks postpartum, using a highly specific RIA for total progesterone after extraction and chromatographic purification of the steroid. The free fraction in plasma was determined via equilibrium dialysis,
‘Department
of Medicine,
Division
for Endocrinology
and Repro-
duction, St. Radboud Hospital, Nijmegen, The Netherlands. (Address correspondence to this author at Tolhuis 76-20, 6537 RV Nijmegen, The Netherlands.) 2lmmuno Technology Service, Nieuweweg 172, Wychen, The Netherlands. Received August 18, 1988; accepted October 31, 1988. 168
CLINICAL CHEMISTRY, Vol. 35, No. 1, 1989
followed by the same AlA analysis for progesterone in the dialysate. Despite the dramatic increases in concentrations of total progesterone and binding proteins in plasma during pregnancy, we found highly significant correlations between total and free progesterone in plasma and salivary progesterone in the group as a whole as well as individuals (P
