WO1990010225A1

WO1990010225A1 - Method for the determination of lipid-bound sialic acid content of blood

Info

Publication number: WO1990010225A1
Application number: PCT/HU1989/000022
Authority: WO
Inventors: Béla SCHUMANN; Tamás PULAY
Original assignee: Reanal Finomvegyszergyar Rt
Current assignee: Reanal Finomvegyszergyar Rt
Priority date: 1989-02-24
Filing date: 1989-05-22
Publication date: 1990-09-07
Anticipated expiration: 1991-08-24
Also published as: HU204131B; AU3697189A; KR920700401A; CN1045185A; JPH02242160A; ES2013572A6; HUT53227A

Abstract

The invention relates to a method for the determination of lipid-bound sialic acid content of blood, primarily of human blood, in which serum is diluted with water, then extracted with a water-immiscible organic solvent or solvent mixture to remove neutral lipids, a lipoprotein fraction containing lipid-bound sialic acid is precipitated from the aqueous phase, the precipitated lipoprotein fraction is redissolved, a mineral acid, a copper(II) salt and resorcine are added to the solution, and the lipid-bound sialic acid content of the serum is calculated on the basis of the colour reaction produced with resorcine. According to the invention a) dextrane with a molecular weight of 10,000 to 20,000 is applied as precipitating agent in a solution with a concentration of 15-35 mg/100 ml and precipitation is performed at a pH of 5.8 to 7.5, or b) polyethylene glycol with a molecular weight of 4,000 to 10,000 is applied as precipitating agent in a solution with a concentration of 15-55 g/100 ml and precipitation is performed at a pH of 9 to 11, or c) a water-soluble manganese(II) salt is applied as precipitating agent in a solution corresponding to a Mn2+ ion concentration of 8.7-13 g/100 ml and precipitation is performed at a pH of 5.5 to 7.5.

Description

METHOD FOR THE DETERMINATION OF LIPID-BOUND SIALIC ACID

CONTENT OF BLOOD

The invention relates to a new method for the determination of lipid-bound sialic acid content of blood, primarily of human blood.

One of the major problems of therapeutics is the early recognition of malignant diseases (in the following: cancer), since this is a decisive factor of their curability. In this respect the developments in biochemical methods are very promising. One of the. major lines of cancer research is to find endogeneous substances the concentration of which in human body fluids (blood, urine, spinal liquor, etc.) changes significantly and characteristically in cancerous processes. These substances are termed as tumor markers (in the following: TM). More than 50 substances possessing TM properties have been discovered till now, but still there is no "cancer test" as taken in common sense, i.e. no laboratory diagnostic method exists which could give a definite and completely reliable "yes" or "no" answer in the detection of cancer. One of the reasons of this insufficiency is that for a significant part of TM substances the biological basis of the determination, particularly the correlation between tumorous state and the amount of the TM substance, has not been elucidated or the results obtained are ambiguous. An other reason is that tumor markers are substances of complicated structure, their isolation and determination is very difficult and in the majority of cases cannot be solved with the existing routine laboratory implements and staff, consequently the measurements are rather expensive and usually do not provide appropriately reliable results.

Sialic acid (N-acetylneuraminic acid) is a component of membrane-forming glycolipids and of some glycoproteins and blood group substances. When examining the changes in glycoprotein and glycolipid composition of cell membrane it has been observed that in tumorous cells and plasm membranes thereof the ganglioside levels increase, thus the examination of the changes in ganglioside level can be applied as a TM test (G. Yogoaswaren: "Cell Surface Glycoproteins in Malignant Transformation", Advances in

Cancer Research, Vol. 38, Academic Press, Inc. 1983, IBSN 0-12-006638-6). The methods known so far for the analytical determination of gangliosides are, however, very sophisticated, time and labour consuming, thus they cannot be applied in practical clinical laboratories as a routine methodo

Serum sialic acid level or the concentration of lipid-bound sialic acid (LSA, one of the sialic acid fractions) is in good correlation with ganglioside levels. This component can be detected by simpler analytical methods.

Several research groups have examined and verified the applicability of LSA as a tumor marker on real clinical cases. Katopodis et al. [Cancer Research 42, 5270-5275

(1982)] determined the serum LSA level in samples originating from 850 patients, 670 of them suffering from clinically verified cancer and 180 of them being tumor-free. The authors observed that the LSA levels measured in the

cancerous group were significantly different from those measured in the tumor-free group. In 97 % of the samples taken from cancerous patients the LSA level exceeded

20 mg/100 ml (the average value was 26.3 mg/100 ml), whereas an average LSA level of 17.6 mg/100 ml was observed in the tumor-free group consisting of healthy persons and patients suffering from non-tumorous benign diseases. The authors regarded LSA levels exceeding 20 mg/100 ml as pathologic. The authors also observed that the determination of LSA level can be applied to follow up the course of the disease, i.e. to indicate healing or worsening.

R.S. Shamberger [J. Clin. Chem. Biochem. 22, 647-651 (198427 examined serum sialic acid levels on 134 tumor-free persons (healthy persons and patients suffering from benign diseases) and on 160 cancerous patients. In 94 % of the cancerous patients the sialic acid concentration of the serum was pathologically high. Within this group the sialic acid levels of the 63 metastatic patients were significantly higher than those observed on patients with no metastasis, indicating that a correlation exists between the stage and mass of tumor and sialic acid concentration of the serum. On 10 cancerous patients the author made a comparison between the commonly known carcinoembryonal antigen (CEA) test and sialic acid determination with respect to their diagnostic usability. Pathologically high CEA levels were observed only on 4 of the patients, whereas all the 10 patients had an increased sialic acid level. The author has also observed that some inflammatory disorders (arthritis, psoriasis, ulcer, etc.) are also accompanied with an increase in sialic acid level.

A.M. Dnistrian et al. [Clin. Chem. 27/10, 1737¬

-1739 (1981); Cancer 50, 1815 (1982)] examined LSA as a tumor marker in mammary cancers. The authors observed a significant increase in LSA level compared to that measured in the healthy control group. The authors examined the changes of LSA level on patients treated with different methods and observed a correlation between the progress of the disease and the changes in LSA level. The authors also compared the LSA and CEA levels of serum samples obtained from 125 patients with clinically verified cancer. Pathologically high LSA levels were found with 78 % of the patients, whereas the CEA level was pathologically high only with 35 % of the same patients.

Prom the above publications it appears that the determination of LSA level is well applicable to indicate the existence of cancer, to estimate the mass and stage of the tumor and to monitor the progress of the disease.

Although numerous methods have been elaborated for the determination of the LSA content of human blood, only the method disclosed in USP 4,342,567 is a relatively simple one enabling routine application in series under average clinical laboratory conditions. According to this method serum is diluted with distilled water, the diluted serum is extracted with a mixture of chloroform and methanol to remove the lipid-soluble fraction, thereafter the

aqueous-methanolic phase is treated with phosphotungstic acid, the separated precipitate is dissolved in water, hydrochloric acid, a copper(II) salt and resorcine are added to the aqueous solution, and the sialic acid content of the serum is determined photometrically on the basis of the resulting colour reaction.

We tried to repeat the method described above under exactly keeping the prescriptions of the specification.

Serum samples originating from three persons were analysed in 30 parallels, each, and the analyses were repeated on four consecutive days. We found an extremely high deviation (25-30 %) between the results of the 30 measurements performed with identical serum samples, which is unacceptable for clinico-chemical purposes. The reproducibility between different days was insufficient as well, the deviation of the average values exceeded 20 %. A further disadvantage of the method is that the slope of the absorbance vs. concentration curve is critically low; consequently, the change in absorbance related to unit change in LSA concentration can be measured only with a great error with photometers having an accuracy of 0.0005 AU in average, routinely applied in clinical laboratories. Tests performed with blanks (i.e.

with samples comprising distilled water instead of serum) have shown that, particularly for samples with an LSA concentration below 20 mg/100 ml, the useful signal/noise ratio is below or at about 3:1 which is the accepted

critical limit of measurability. Presumably these conditions have lead to the fact that the method of USP 4,342,567 has not been put into practice.

When examining the individual steps of the method disclosed in USP 4,342,567 we have found that precipitation of the sialic acid-containing fraction has a decisive role in the accuracy of the measurement. Phosphotungstic acid, utilized as precipitating agent according to USP

4,342,567, does not enable the selective precipitation of the sialic acid-containing fraction. Beside the aimed fraction other disturbing protein fractions precipitate as well and get denaturated. This is reflected by the fact that even according to the cited reference the precipitated protein fraction cannot be completely redissolved. The amount and quality of the disturbing protein fractions cannot be regulated, thus they distort the results in an uncontrollable manner. As a principal deficiency, USP

4,342,567 does not disclose which fraction(s) should be separated from the multicomponent proteins present in human blood, highly varying in structure and composition. Thus, the result of precipitation cannot be checked with an independent method.

We have elaborated first an ultracentrifugal method as an independent method to determine the characteristics of the protein fraction to be examined. The ultracentrifugal separation applied by us was performed under the following conditions:

temperature: 20°C

preparative tube volume: 11.5 ml

gradients: 8.5 ml of a 0.9 % by weight aqueous sodium chloride solution as upper layer and 3 ml of serum

(d = 1.3 g/ml) as lower layer

speed: 50,000 r.p.m. rotation period: 120 minutes

K' factor: 68.7

The basis of ultracentrifugal separation is that the individual protein and lipoprotein fractions float with different velocities depending on the density of their hydrated forms and separate from one another as well-defined individual bands. The fraction for LSA determination separates as the fourth individual band from the meniscus. This band could be removed quantitatively from the preparative tube via a syringe and could be subjected to further

examinations. Thus the amount of the fraction for LSA determination could be measured very exactly by ultracentrifugation. The physico-chemical characteristics of the separated fraction were as follows:

density: 1.125 - 1.210 g/ml

molecular weight: (3o86 - 1.86) × 10⁵ daltons

lipid/protein ratio: 52/48

Ultracentrifugal method, which is too expensive and lengthy for being applied in serial analysis but is very accurate, has been utilized as a reference to check the reliability of chemical precipitation.

Our further work has been directed to elaborate a method for the selective precipitation of the LSA-containing fraction. It has been found that the LSA-containing fraction can be precipitated selectively and quantitatively a) with a 15-35 g/100 ml solution of dextrane sulfate (molecular weight: 10,000 to 20,000) at a pH of 5.8 to 7.5 , or

b) with a 15-55 g/100 ml solution of polyethylene glycol (molecular weight: 4,000 to 10,000) at a pH of 9 to 11, or

c) with a solution of a water-soluble manganese(II) salt corresponding to a Mn ²⁺ ion content of 8.7-13 g/100 ml at a pH of 5.5 to 7.5,

and the sialic acid content of the precipitated fraction can be determined reproducibly and with very high accuracy by the colour reaction disclosed in USP 4,342,567. The accuracy of measurement can be increased when an aqueous buffer solution with a pH of 7 to 9 is applied to redissolve the precipitate.

Based on the above, the invention relates to a method for the determination of lipid-bound sialic acid content of blood, primarily of human blood, in which serum is diluted with water, then extracted with a water-immiscible organic solvent or solvent mixture to remove neutral lipids, a lipoprotein fraction containing lipid-bound sialic acid is precipitated from the aqueous phase, the precipitated lipoprotein fraction is redissolved, a mineral acid, a copper(II) salt and resorcine are added to the solution, and the lipidbound sialic acid content of the serum is calculated on the basis of the colour reaction produced with resorcine. According to the invention

a) dextrane with a molecular weight of 10,000 to 20,000 is applied as precipitating agent in a solution with a concentration of 15-35 mg/100 ml and precipitation is performed at a pH of 5.8 to 7.5, or

b) polyethylene glycol with a molecular weight of

4,000 to 10,000 is applied as precipitating agent in a solution with a concentration of 15-55 g/100 ml and precipitation is performed at a pH of 9 to 11, or

c) a water-soluble manganese(II) salt is applied as precipitating agent in a solution corresponding to a Mn²⁺ ion concentration of 8.7-13 g/100 ml and precipitation is performed at a pH of 5.5 to 7.5.

According to the invention the precipitated lipoprotein fraction is redissolved preferably with an aqueous buffer solution of pH 7 to 9.

It is particularly preferred to apply a water-soluble manganese(II) salt, such as manganese(II) sulfate, manganese(II) nitrate, manganese(II) chloride, manganase(II) bromide or manganese(II) iodide, as precipitating agent, of which

manganese(II) chloride is particularly preferred. This salt is applied in a solution with a concentration of 20-30 g/100 ml, preferably 22-27 g/100 ml, and precipitation is performed preferably at a pH of 6.0±0.1.

A 10 mmole/litre aqueous tris(hydroxymethyl) aminomethane buffer solution (pH = 8.6) proved to be particularly preferable for redissolving the precipitate.

It has been found that when serum is stored at -20°C its LSA content does not change for 6 months. It has also been found that the colour density of the coloured complex formed with resorcine (which density is proportional to the LSA concentration) does not change at room temperature within 6 hours. This fact is essential with respect of serial analysis.

The method of the invention can be applied primarily in human diagnostics for a quick cancer screening to be performed in series and to monitor the healing process of cancerous diseases evaluating thereby the therapy applied. However, the method can also be applied in pharmacological examinations for cancer research, such as for evaluating the effects of prospected antitumor agents.

The invention is elucidated in detail by the aid of the following non-limiting Example.

Example

Determination of the LSA content of human blood

Venal blood is used for the determination. After blood sampling the sample is centrifuged for 15 minutes at 4000 r.p.m., and 0.6 ml of the sample is frozen. The lipidbound sialic acid concentration of the sample stored at -20°C remains unchanged for 6 months.

Prior to starting the analysis the test tubes are thoroughly rinsed with methanol, dried, and precooled in a refrigerator. 0.2 ml of the serum sample and 0.2 ml of bidistilled water are filled into each of the test tubes, the mixture is stirred with a Vortex stirrer for 15 seconds and then placed into ice water bath (0°C). Two parallel samples, each, are prepared from each of the sera to be examined and from the controls.

3 ml, each, of a pre-cooled 2:1 v/v mixture of chloroform and methanol are filled into all of the test tubes, and the contents of the test tubes are stirred with a Vortex stirrer for 30 seconds. Then the test tubes are placed again into the ice water bath. 0.5 ml, each, of bidistilled water pre-cooled to +4°C is added to the individual samples, the contents of the test tubes are stirred with a Vortex stirrer for 15 seconds, and the mixtures are allowed to stand at room temperature for 5 minutes. Thereafter the samples are centrifuged for 10 minutes at 3000 r.p.m. 1 ml, each, of the supernatant is filled cautiously into numbered centrifuge tubes, 0.1 ml each of manganese(II) chloride solution is added, the mixtures are stirred with a Vortex stirrer and then allowed to stand for 5 minutes. Manganese(II) chloride solution is prepared as follows: 25 g of anhydrous manganese(II) chloride are dissolved in 80 ml of bidistilled water, the pH of the solution is adjusted to 6.0±0.1 with 0.1 n aqueous sodium hydroxide solution or 0.1 n aqueous hydrochloric acid, and the volume of the resulting solution is adjusted to 100 ml with bidistilled water.

In a further test series dextrane sulfate solution is applied as precipitating agent. 20 mg of dextrane sulfate with a molecular weight of 15,000 are dissolved in 100 ml of a 0.9 % by weight aqueous sodium chloride solution. 1 ml of the resulting solution is added to 1 ml of the supernatant, the mixture is stirred with a Vortex stirrer, and then allowed to stand at room temperature for 5 minutes.

In a still further test series polyethylene glycol solution is applied as precipitating agent. 20 mg of polyethylene glycol with a molecular weight of 6,000 are dissolved in 80 ml of a 0.2 mole/litre glycine buffer, the pH of the solution is adjusted to 10.0±0.1 with 0.1 n aqueous hydrochloric acid or 0.1 n aqueous sodium hydroxide solution, and the volume of the solution is adjusted to 100 ml with bidistilled water. 1 ml of the resulting solution is added to 1 ml of the supernatant, the mixture is stirred with a Vortex stirrer, and then allowed to stand at room temperature for 5 minutes.

At the end of the standing period the mixtures are centrifuged for 15 minutes at 4000 r.p.m. and the supernatant is separated completely from the precipitate by decanting (the supernatant is discarded).

1 ml, each, of a tris buffer solution is introduced into the individual centrifuge tubes, redissolving thereby the precipitate. Tris buffer solution is prepared by dissolving 10 mmoles of tris(hydroxymethyl) aminomethane in one litre of bidistilled water and adjusting the pH of the resulting solution to 8.6.

Standard and blank samples are prepared for the photometric measurements as follows:

50 mg of analytically pure sialic acid are dissolved in 100 ml of bidistilled water, and the resulting solution is utilized as stock solution to prepare two further dilutions (25 mg/100 ml and 12.5 mg/100 ml in concentration) with bidistilled water. The standard solutions are stored at +4°C. Samples of 0.2 ml, each, of the standard solutions with three different concentrations are filled into centrifuge tubes, and the volume of the solution is adjusted to 1 ml with bidistilled water in each of the centrifuge tubes. 1 ml of bidistilled water is applied as blank.

1 ml, each, of resorcine reagent solution is added to all of the samples (standard, blank, control, test serum). The resorcine reagent solution is prepared as follows: 2 g of analytically pure resorcine are dissolved in 100 ml of bidistilled water. 2.49 g of anhydrous copper(II) sulfate are dissolved in 100 ml of bidistilled water. Thereafter 10 ml of the resorcine solution are admixed with 0.25 ml of the copper(II) sulfate solution and 9.75 ml of bidistilled water, and 100 ml of analytically pure concentrated aqueous hydrochloric acid are added to the mixture.

The test tubes are placed into a 1C0°C water bath for exactly 10 minutes. Thereafter all of the test tubes are immediately immersed into a 0°C ice water bath and kept there for 10 minutes. 2 ml, each, of a 85:15 v/v mixture of butyl acetate and n-butanol are introduced into the individual test tubes and the contents of the test tubes are stirred with a Vortex stirrer for 5 minutes. The mixtures are allowed to stand with occasional stirring, the samples are centrifuged for 10 minutes at 2500 r.p.m., and then the blue supernatants are filled from each centrifuge tube into a measuring cell. The density of the blue colour does not change within 6 hours.

Photometric measurement is performed at a wavelength of 580 nm. The zero point of the photometer is adjusted for bidistilled water. Thereafter, depending on the type of the photometer used (single-beam or double-beam), the measurement is performed either against the blank, or the extinction of the blank is subtracted from the measured extinction of the sample obtaining a corrected value:

E(corr) = E(measured) - E(blank)

LSA content is calculated by one of the following two methods:

1) Prom the results obtained with standard sialic acid samples of known sialic acid concentrations extinction vs. concentration curves are constructed either graphically or by calculation. Due to the dilution steps, the original concentration should be recalculated according to the following table:

Original concentra- Measured concentra- tion (mg/100 ml) tion (μg/200 ml) Extinction 50 100 0.508

25 50 0.252

12.5 25 0.127

Prom the calibration curve constructed as described above the sialic acid content (A) of the supernatant sample, 1 ml in volume, can be determined. The curve gives the sialic acid content in units of μg of sialic acid/1 ml of supernatant. The sialic acid content of the blood serum can be calculated from the formula LSA (mg/100 ml) = A × 0.72,

where A is the sialic acid content of the supernatant in units of μg/ml and 0.72 is a factor which comprises the degree of dilution and the conversion of μg to mg.

2) According to the second method of calculation reference serum samples with known LSA concentrations are applied. Prom several parallel measurements a factor F is calculated according to the formula

F = C(ref)/E(ref),

where C(ref) is the LSA content of the reference serum in mg/100 ml and E(ref) is the extinction of the reference serum. The LSA content of the sample to be tested can be calculated from the formula

LSA (mg/100 ml) = E(m) × F,

where E(m) is the measured extinction of the sample to be tested.

The reliability of the measurement was checked with the ultracentrifugation method described above. The L_SA contents of serum samples originating from 15 different persons were measured by ultracentrifugation (UC), precipitation with manganese(II) chloride (MnCl₂), precipitation with dextrane sulfate (DS), precipitation with polyethylene glycol (PEG), and with the method described in USP

4,342,567 (known method). The results are listed in Table I. Table I

Number of LSA content mg/100 ml

the sample UC MnCl ₂ DS PEG Known method 1 14.9 15.1 14o7 15.3 12.5

2 19.0 18.8 19.1 19.5 21.8

3 16.9 16.6 16.5 17.1 19.1

4 21.0 22.0 23.1 20.1 26.5

5 19.5 19.0 18.4 21.0 15.1

6 30.8 31.3 32.1 29.1 24.2

7 25.6 25.0 24.2 27.1 18.2

8 26.9 27.1 25.6 28.0 20.2

9 14.9 14.3 16.0 15.5 17.9

10 23.5 21.6 22.4 23.7 18.7

11 20.3 19.6 18.7 21.6 24.4

12 16.1 16.4 18.0 17.3 12.8

13 18.3 18.3 16.7 17.1 23.5

14 28.5 29.3 31.2 26.9 21.8

15 44.3 42.8 41.2 46.3 28.7 The data listed in Table I indicate that the

results obtained in the ultracentrifugation method used as reference, which provides absolutely precise data, are in a very good correlation with the results obtained according to the method of the invention. The best results are obtained when applying manganese(II) chloride as precipitating agent; for this method the correlation coefficient is 0.995. To the contrary, the results obtained with the known method differ significantly from those obtained in the reference measurement.

To check the reproducibility of the method of the invention within the same laboratory, the LSA level of Hyland normal reference serum was determined on 30 different days. The statistical evaluation of the results has shown that the reproducibility of the different examination series is better than +10 %, i.e. it is acceptable.

To check the reproducibility between different laboratories, several hundreds of serum samples were subjected to analysis in two different laboratories. All essential features of the analysis (personal, types of equipment, etc.) were different. Comparing the results it has been found that, when applying identical reference sera, the reproducibility between different laboratories does not exceed the precision limit of +10 %.

Based on the experiences of several hundred LSA determinations we have found that a well-trained laboratory assistant can perform at least 80-100 determinations within a workday in completely manual manner.

The applicability of the method according to the invention was checked in real clinical cases. The results were classified according to the following grading:

normal level: below 18 mg/100 ml,

borderline of normal and pathological level; 18-20 mg/100 ml, pathological level: above 20 mg/100 ml. 1) General applications:

LSA determinations performed in 115 healthy persons or patients suffering from benign diseases have shown levels below 20 mg/100 ml in 95 % of the cases. With five patients actually treated for pneumonia the LSA levels fell within the range of 20-25 mg/100 ml, in full harmony with the observations described in the literature.

LSA levels of 94 patients suffering from various malignant diseases (pulmonary cancer, colon cancer, stomach cancer, testicular cancer, mammary cancer, laryngeal cancer, etc.) were examined in parallel with the examination of other tumor marker substances (neopterin, polyamine, etc.).

In one group of the patients (30 persons) the measurements were performed in the clinically verified active period of the disease. Apart from one patient suffering from

advanced hepatic cancer, pathologically high LSA levels were measured for each persons.

In the other group of the patients (64 persons) the tumors had been removed surgically years before. In this group the goal of the examination was to decide whether the patient is still tumor-free or to obtain an indication on the possible recurrences. With 20 % of the patients (13 persons) both the LSA levels and the values obtained with other TM substances uniformly revealed a malignant diseas or an activation stage. For these patients a diagnosis with the qualification "high probability of active tumorous process" was given. For 25 % of the patients (16 persons) the LSA levels and the values obtained with other TM substances indicated an ambiguous picture; some of the values were normal, whereas others were pathologically high. These patients were qualified as "doubtful cases", and more frequent checks and the use of other diagnostic tools were recommended to their doctors, simultaneously drawing their attention to the possibility of other non-malignant diseases, such as inflammations. In 55 % of the patients (35 persons) both the LSA levels and the other TM values were within the normal range, thus they were tumor-free with high probability.

2) Examination of a well-defined tumor type:

Of the malignant gynaecological tumors ovarian cancer was examined. Complete and detailed case records (histological examinations, surgery, type and time of therapeutic interventions, etc.) of the patients were placed at our disposal, which enabled us to make a realistic judgement on the tumor marker performance of LSA.

LSA levels of 142 patients suffering from histologically verified ovarian cancer were determined before operation, and then, associated with the different therapeutic interventions, 5-7 tests in average were performed. Of the laboratory parameters examined together with LSA level determinations the CA-125 monoclonal antigen values were compared to the LSA levels, since CA-125 is now the internationally accepted most sensitive tumor marker used to indicate and follow up epithelial ovarian cancer. The results are summarized below.

a) With 95 % of the patients, pathologically increased LSA levels were measured before surgery, whereas the ratio of pathological CA-125 values was below 95 % in the same stage,

b) The LSA values returned to the normal one within 45 days in average after the successful and complete removal of the tumor; this is practically identical to the tenfold half life-time of CA-125 (48-50 days).

c) LSA monitored the progress of the disease practically with the same sensitivity as CA-125. LSA indicated propagation with a 100 % security. LSA also indicated the improvement and recovery processes occurring upon the therapy, but the rate and extent of the decrease in LSA level were lower than those of CA-125.

d) LSA indicated recurrences in 80 % of the cases before the appearance of clinical symptoms.

e) In all of the clinically verified recurrences the LSA level was pathologically high.

f) In non-epithelial tumors (germinal tumors, etc.) which amount to about 20 % of the ovarian cancer cases, CA-125 cannot be applied for reliable indication. It is essential that LSA proved to be a tumor marker of unchanged sensitivity in the examination of such cancers as well.

g) In order to assess its value as a differentiating agent between benign and malignant tumors, the LSA levels of 25 patients suffering from myoma and 12 patients suffer ing from benign ovarian cyst were examined. Pathologically high LSA levels could not be observed in any of these cases.

The main advantages resulting from the use of the method of the invention are as follows:

1) Laboratory aspects:

a) The measurement does not require special training and can be performed by any person having ordinary skill in laboratory work.

b) The measurement can be performed in laboratories equipped with tools and instruments on the average level, and does not require any specific additional installation.

c) When following the analytical prescriptions precisely, reliable measurements can be performed in different laboratories.

d) On the basis of its labour and time consumption the method can be applied as a serial analysis for mass examinations.

e) The costs of chemicals and tools are much

(about ten times) lower than those appearing in connection with the recently applied tumor marker examinations (CEA, CA-125, APP, etc.).

f) On the basis of the method of the invention test kits can be assembled very easily, by which time and labour can be saved and the reliability of the examination between different laboratories can be increased.

2) Clinical aspects:

The lipid-bound sialic acid content of human blood can be applied as information in the therapy of malignant diseases particularly in the following fields:

a) Screenings, early recognition, support for other diagnoses:

Depending on the type and stage of the malignant process, the LSA value indicates the possible existence of the disease with a probability of 70-93 %. Since the method is relatively easy to perform and is inexpensive, it can be applied for mass screenings. Its introduction as a general screening method does not require particular additional financial resources and investments. Of course, it should be stressed that LSA determination alone does not solve cancer recognition problems, nevertheless it is a useful element of a well-organized diagnostic system.

In screening programs LSA can be applied to detect the probable existence of disorders, and thus the other more complicated, consequently more expensive, specific diagnostic tests which can be performed on a much narrower scale can be concentrated on the doubtful cases. The introduction of such a multistep screening system would lead to a much more effective recognition of the disorders than the recent practice.

b) Monitoring the status of patients suffering from malignant tumors, checking the effects of therapeutic interventions:

Prom the aspects of the chances of recovery, the continuous monitoring of the status of tumorous patients is at least as important as early recognition. LSA determination, coupled with other laboratory and diagnostic methods , could be used as a valuable element of an obj ective cancer monitoring system. In this form of application the change of LSA level with time is an important information in differentiating tumor-free and progression stages and in checking the effects of various therapeutic interventions.

Claims

What we claim is:

1. A method for the determination of lipid-bound sialic acid content of blood, in which serum is diluted with water, then extracted with a water-immiscible organic solvent or solvent mixture to remove neutral lipids, a Hpoprotein fraction containing lipid-bound sialic acid is precipitated from the aqueous phase, the precipitated Hpoprotein fraction is redissolved, a mineral acid, a copper(II) salt and

resorcine are added to the solution, and the lipid-bound sialic acid content of the serum is calculated on the basis of the colour reaction produced with resorcine, characterised in that

b) polyethylene glycol with a molecular weight of 4,000 to 10,000 is applied as precipitating agent in a solution with a concentration of 15-55 g/100 ml and precipitation is performed at a pH of 9 to 11, or

c) a water-soluble manganese(II) salt is applied as precipitating agent in a solution corresponding to a Mn ²⁺ ion concentration of 8.7-13 g/100 ml and precipitation is performed at a pH of 5.5 to 7.5.

2. A method as claimed in claim 1, characterised in that a 20-30 g/100 ml solution of manganese(II) chloride is applied as precipitating agent.

3. A method as claimed in claim 1 or 2, characterised in that the precipitated lipoprotein fraction is redissolved with an aqueous buffer solution of pH 7 to 9.

4. A method as claimed in claim 2 or 3, characterised in that a 22-27 g/100 ml solution of manganese(II) chloride is applied as precipitating agent and precipitation is performed at a pH of 6.0±0.1.

5. A method as claimed in any of claims 1 to 4, characterised in that a 10 mmole/litre aqueous solution of tris(hydroxymethyl) aminomethane is applied as aqueous

buffer to redissolve the precipitated lipoprotein fraction.