PLENARY LECTURE

55. COMPUTERIZED IMAGE ANALYSIS

Kardum-Skelin I

Clinic of Internal Diseases KB "Merkur" Zagreb

Nowdays, computerized image analysis is not only a well established and highly developed methodology, but it is becoming widely used and more and more applied in various diagnostic fields in clinical laboratories.

Introduction of computers into image analysis has  enabled the following: processing, analysis, comparison, selection of certain image segments, and image storaging  with  possibility of creating data basis.

An automatic image analysor consists of microscope, video camera and computer with an adequate programme for the image analysis, which enables numeric objectivisation of the most subtle changes unapproachable to visual inspection, and, instead of subjective evaluation, the objective quantification of certain parametres  has turned up.

Numerous cellular parametres and  specifities  are acceptable for digital analysis: 1. morphometrical characteristics; 2. determination of DNA quantity - static DNA cytometry; 3. quantification of immunocytochemical staining; 4. determination of proliferational status (Ki-67, AgNOR); 5. karyometry etc.

Quantitative measurings, regardless to whether they are microphotometric, micromorphometric or stereologic, require the standardization of specimens: preparation, fixation and staining.

Cytologican specimens can be unstained and stained smears, cellular imprints or suspensions. Morphometry is a quantitative description of structural geometric characteristics  in all dimensions.

The morphometric parametres can be divided into four groups: 1. simple parametres (surface, diameter, range,  longest and shortest axis); 2.  shape factors (factors of symmetry and convexity); 3. two-phasal parametres (nucleocytoplasmatic relation and eccentricity), and 4. contextual parametres (surface of sheets within a cell or nucleus, number of elements for each sheet, shape of sheet and distance between sheets). Abnormal DNA contents, i.e. aneuploidy, is a malignant feature, and the determinaiton of DNA quantity has an important role in oncology for  safety improvement in the tumorous diagnosis as well as  prognosis.  

The cellular ploidity directly suggests  the proliferational status of a cell.

The determination of DNA quantity can also be performed by flow cytometre, but the slide computerized analysis is preferable  due to morphologic visualisation of the inspected cells (malignant and reference ones).

SYMPOSIUM NEW TECHNIQUES IN CLINICAL CYTOLOGY

56. LASER SCANNING CYTOMETRY:

NEW DIAGNOSTICAL TOOL IN ROUTINE CYTOLOGY

R.Bollmann,  J.Schmitz,  C.Vogel,  M.Bollmann

Laser-Scanning-Cytometry (LSC) is a newly developed technology to measure certain cytometric parametres like DNA-content and antigen expression or FISH. The method eliminates many of the drawbacks of flow-cytometry and image-analysis.

We are using the instrument in routine diagnostic pathology for ploidy analysis of solid tumors and cytological preparations. With the instrument we are also able to quantify immunological parametres,e.g. hormon receptors or MIB1. It is even possible to calculate which fraction of the tumor is more or less positive. Through multiparameter analysis immunotyping of lymphoma is also possible.

SYMPOSIUM NEW TECHNIQUES IN CLINICAL CYTOLOGY

57. COMPARISON OF MORPHOLOGY AND DNK ANALYSIS

IN ASPIRATED PROSTATE FRAGMENTS

Karmen Trutin Ostović,   Zorana Miletić,  Željka Znidarčić,

Goran Bedalov,  Renata Heinzl,  Tajana Štoos-Veić,

Gordana Kaić, Mladen Petrovečki

Aspirational prostate cytology has a very important role in the prostate lesion  diagnostics and in monitoring patients´ treatment. We wanted to examine if the DNK-content analysis by flow cytometry, as one of objective methods, can improve cytological (morphological) diagnostics.

From 1997 to 2000, we analysed, by flow cytometry, ploidy  and proliferation in 160 aspirated prostate fragments. The examination comprised patients  with the diagnoses of glandular hyperplasia, atypic glandular hyperplasia and carcinoma.  Aneuploidy is present in 29 (15%) patients: in 2(3,9%) patients out of 51  atypic glandular hyperplasia and in 26 (67,3%) patients out of 38  carcinoma. An increased proliferative activity was found in 79 (49,2%) patients: in 9 (20,5%) with glandular hyperplasia, in 37 (72,5%) with atypic glandular hyperplasia and in 29 (76,5%) with carcinoma. Aneuploidy correlates positively with carcinoma, and high proliferation with atypic glandular hyperplasia and carcinoma (Fisher´s test for probability p<0,001).

This brings us to the conclusion that, apart from morphological analysis, it is necessary to  make the DNK-contents analysis  since ploidy and cellular proliferation enable better proceding for monitoring patients with atypic glandular hyperplasia and carcinoma.

SYMPOSIUM NEW TECHNIQUES IN CLINICAL CYTOLOGY

58. DNA - IMAGE CYTOMETRY IN SUBTYPING

LARGE-CELLULAR LYMPHOMAS

BorovečkiA, Kardum-Skelin I, Šušterčić D, Ostojić S, Vrhovac R,

Radić-Krišto D, Jakšić B

Clinic of Internal Diseases KB "Merkur"

The malignancy is featured by  changes in DNA-contents - aneuploidy. Morphological characteristics of malignant cell nucleus, observed in standard May-Grunwald-Giemsa stained slides, suggest a change in DNA-contents by their hyperchromasia.

Since these changes are unapproachable to visual quantification, automatic image analysor is used in their analysis, which consists of microscope, video camera and computer with appropriate programme for the analysis of SFORM image that enables the numeric objectivisation in  staining intensity changes.

We have analysed 31 aspirated lymph node  fragments in patients suffering  large-cellular lymphomas - 6 patients with Burkitt and Burkitt-like lymphoma, 4 patients with FC III, 10 patients with ALCL (T or 0) and 11 patients with DLBCL. The slides were stained according to the Feulgen method.

The results revealed heterogenity in DNA quantity within the identical morphological subgroups of large-cellular lymphomas. The percentage of cells >30 in S+G2M phase or >30% cells with DNA>2N suggests a rapid disease relapse, shorter survival or correlation with complex chromosomal abnormalities.

The precise quantification of DNA-content is used not only in diagnosis, but also in the classification and prognosis of various diseases, malignant tumors at the first place.

SYMPOSIUM NEW TECHNIQUES IN CLINICAL CYTOLOGY

59. MORPHOMETRIC PARAMETRES OF LYMPHATIC

CELLS IN PATIENTS WITH

CHRONIC LYMPHOPROLIFERATIVE DISEASES

IN RELATION TO NORMAL LYMPHOCYTES

Borovečki A, Kardum-Skelin I, Šušterčić D, Vrhovac R,

Radić-Krišto D, Jakšić B

Clinic of  Internal Diseases KB "Merkur", Zagreb

Absolute lymphocytosis in peripheral blood, infiltration of bone-marrow, lymph nodes and/or spleen with small lymphocytes characterizes  chronic lymphocytic leukemia (CLL).

The research comprises 10 specimens of peripheral blood and bone-marrow in normal persons, 10 specimens of peripheral blood and bone-marrow in patients with CLL, and 4 aspirated lymph node fragments in the same patients. All together  1000 peripheral blood lymphocytes in normal persons and patients with CLL, 1000 bone-marrow lymphocytes in normal persons and patients with CLL and 400 lymphocytes of aspirated lymph node fragments in patients with CLL were analysed. The parametres of  whole cell were morphometrically analysed,  plus separately their nuclei: shape, surface, convexity, and factors of symmetry (factors of range, convexity and elongation). No statistically significant difference was  found not even in one of the mentioned parametres between normal lymphocytes and lymphocytes in patients with CLL. Similarly, there was no  lymphocytic  heterogenity in different compartments of the tumorous mass (in peripheral blood, bone-marrow and lymph nodes) observed, very possibly due to  a small number of specimens, but also due to the selection of typical CLL forms. The lymphocytic morphometry itself is not sufficient, as a separate criterium, for differentiation of normal from neoplastic lymphocytes in the CLL. Therefore, besides  pure appearance, some additional examinations are needed: absolute lymphocytosis (>5×109/L), bone-marrow infiltration, phenotype determination and lymphocytic clonality.

SYMPOSIUM NEW TECHNIQUES IN CLINICAL CYTOLOGY

60. SEROUS OVARIAN TUMORS -

AgNOR ANALYSIS IN CYTOLOGICAL SPECIMENS

Štemberger-Papić S, Verša-Ostojić D, Stanković T,

Vrdoljak-Mozetič D, SeiliBekafigo I

ABSTRACT: The aim of the work was to determine the number and size of particular AgNORs and Agnor sheets, as well as the  ratio of total AgNOR size  in relation to nucleus in cytological specimens of benign, boundary malignant and malignant serous ovarian tumors.

Cytological specimens of 15 patients were analysed with confirmed pathohistological diagnosis (5 benign, 5 boundary malignant, 5 malignant serous ovarian tumors). Cytological imprints were stained according to the  Papanicolaou method, then unstained and stained again by impregnation of silver for the AgNOR analysis. 100 nuclei were analysed for one specimen, with magnification of 1000 times, by  programme for digital procession of "SFORM" image. The obtianed results suggested  statistically siginificant difference in number and surface of particular AgNOR and AgNOR sheets among all the three analysed groups. Statistically significant difference was also observed in the size of total AgNOR, as well as in the ratio of total AgNOR  size in relation to nucleus.

The results suggested the possibility of AgNOR technique application as an additional method in differential cytological diagnosis of benign, boundary malignant and malignant serous ovarian tumors.

SYMPOSIUM NEW TECHNIQUES IN CLINICAL CYTOLOGY

61. AgNOR ANALYSIS IN CYTOLOGICAL SPECIMENS

OF BENIGN, BOUNDARY MALIGNANT

AND MALIGNANT OVARIAN TUMOR

Verša-Ostojić D, Štemberger-Papić S, Stanković T,

Seili-Bekafigo I, Vrdoljak-Mozetić D

Abstract

16 imprints of mucinous ovarian tumor have been elaborated in this work. Based on the pathohistological confirmation, they were classified in three categories: benign tumors (5),  tumors of boundary malignancy (5), malignant tumors (6).

The aim of the work was to examine the possibilities of differential diagnosis of benign tumors, tumors of boundary malignancy and malignant  mucinous ovarian tumors by usage of  Ag stained regions of nucleolar organizers (AgNOR). Cytological specimens were stained according to the Papnicolaou method, they were unistained and stained by silver impregnation. A hundred cells were chosen, and the analysis was performed at the 1000 time magnification by   the programme for "SFORM" image procession.

The obtained results suggested  statistically significant difference between all the three groups in the total surface of particular AgNOR   for each nucleus when using  the variance analysis.  Using the POST-HOC LSD test, the possibility of differing benign tumors and tumors of boundary malignacy is shown by measuring the total surface, minimal and maximal   separate AgNORs.

Statistically no significant difference was found in the distinction between boundary from malignant mucinous tumours, not even in one parametre.The results indicated the AgNOR technique demonstrated limited possibilities in differing mucinous ovarian tumors.

POSTER SECTION

62. DNA VALUES - CYTOMETRY IMAGE IN

DIAGNOSIS AND PROGNOSIS OF ALL IN ADULTS

Seili-Bekafigo I, Kardum-Skelin I, Šušterčić D, Jakšić B

KBC Rijeka - Internal Clinic

Lymphoblasts in the  bone-marrow of adult patients with acute lymphatic leukemia (ALL) were analysed by DNA-image cytometry at the time of rendering the diagnosis. The correlation between the DNA-parametres and FAS subtype of ALL was examined, as well as potential prognostic significance of those parametres regarding the survival duration.  The following parametres were determined: DNA-index (D1, cellular ratio in S+G2M phasis of cellular cycle, and cellular ratio with the content DNA>2n. The archival material of aspirated bone-marrow fragments in 49 adult patients with ALL, previously stained according to MGG,were in the analysis. The slides were first unstained, then again stained by the Feulgen method, and analysed by computer programme for the SFORM image analysis. The DNA-contents in  bone-marrow lymphoblasts of each patient are expressed in the form of DNY histogram. The main peak of blastic DNA-histogram is interpreted as a ploidy value of the majority of blast population. The results showed that, although DI mean is  the highest in ALL-L3 FAB subgroup, these mean values are within diploid frames, the same as in the rest 2 FAB ALL subgroups, and there is no significant difference in the DI values between the three FAB ALL subtypes. The ploidy, expressed as DI, also did not show  any significant correlation with the prognosis. In relation to the FAB subtype, the biggest % of cells in S+G2M phase was found in ALL-L3 (37,854%) which is in accordance with approachable literature data. A worse prognosis is predicted by the finding >30% cells in S+G2M cycle phase, as well as >30% cells with DNA>2n. The ploidy determination (DI)  itself  does not provide a useful prognistic information in the acute ALL:

The determinations of proliferational cellular fraction are expressed as cellular ratio in S+G2M cellular cycle phase, or as a cellular ratio with DNA>2n contents can contribute in the classification of patients into prognostic groups.