Prognostic relevance of the FAB morphological criteria in chronic
lymphocytic leukemia: correlations with IgVHgene mutational status
and other prognostic markers*
J. SCHWARZ1, D. MIKULENKOVÁ1, K. ČERMÁKOVÁ1, V. POLANSKÁ1, K. MICHALOVÁ1, I. MARINOV1, V. CAMPR1, Š. RANSDORFOVÁ1,
J. MARKOVÁ1, L. PAVLIŠTOVÁ2, J. BŘEZINOVÁ1, J. SAJDOVÁ1, D. ŠPONEROVÁ1, Z. VOLKOVÁ1, K. BENEŠOVÁ1, J. ČERMÁK1, A. VÍTEK1,
1Institute of Hematology and Blood Transfusion, e-mail: firstname.lastname@example.org, CZ-128 Prague, Czech Republic;2Center of Oncocyto-
genetics, 1st Medical Faculty, Charles University, Prague, Czech Republic
Received October 6, 2005
Morphological examination is the routine first step in the diagnosis of hematological malignancies, including chronic
lymphocytic leukemia (CLL). Atypical cell morphology according to the FAB criteria is known to herald disease progres-
sion.Several years ago, it wasproposedthat FABmorphology at diagnosishadaconsiderable prognostic impact. However,
thisproposalhasnotbeenwidely adopted inpractice. Thuswequestioned theprognosticvalue ofthemorphological exami-
nation, which was performed retrospectively in 88 patients out of our 110 institutional registry patients (70 males and 40 fe-
males, median age 57 yrs) with CLL at diagnosis. We related the results to the more modern prognostic markers.
Atypical FAB morphology was shown to correlate with IgVHgene mutation status, trisomy of chromosome 12 and dele-
tionof17pdetected either by conventional G-bandingorby fluorescence insituhybridization (FISH)analysis. The correla-
tion of FAB morphology with CD38 antigen expression or with the histopathological pattern of bone marrow infiltration
was not significant. Overall survival (OS) data were available for 84 morphologically examined patients. The patients with
atypical morphology (64 patients) had a significantly shorter OS (103 months) than the 20 patients presenting with typical
CLLmorphology (237months;p=0.03). Only themutationstatusofIgVHgenescorrelated moreclosely withOS(p=0.002).
Of note, there was no leukemia-related death within “unmutated” cases who had typical FAB morphology (p=0.14), and
vice versa, the mutation statushad a significant prognostic impact within the morphologically atypical cases (p=0.01). Thus
FAB morphology and the mutation status may yield complementary prognostic information. OS was affected both by the
presence of cytogenetic aberrations (p=0.03) – most adversely by deletions of 17p and 11q, and by CD38 expression
We conclude that careful examination of peripheral blood smears according to FAB is a simple, cheap and valuable tool
inthe first-line assessmentofprognosisofCLL patients and shouldnotbeoverlooked even in3rdmilleniumwhenmoreso-
phisticated prognosticmarkersare athand. Thisoughttobeconfirmedinlarger prospective studieswithmultivariate analy-
sis of data.
Keywords:chroniclymphocytic leukemia, prognosticmarkers,morphology,FABclassification, IgVHgenemutationsta-
Managementof chronic lymphocyticleukemia(CLL) cur-
rently undergoes revolutionary changes. Efficient therapies
are developed and the therapeutic goal gradually shifts from
ies greatly, from an indolent disease to an aggressive one:
overall survival (OS) may be unaffected in some patients, or
it may be less than 3 years in others. An array of prognostic
markers is available for risk stratification of patients in order
to select candidates for aggressive treatment modalities,
NEOPLASMA, 53, 3, 2006219
*The study was supported by the grant IGA NM/7604-3 from the Czech
Ministry of Health. It was presented at the XI International Workshop on
CLL, New York, NY, September 15–17, 2005.
nostic markers, the mutational status of IgVHgenes (or its
“surrogate” – expression of the ZAP-70 antigen) and cyto-
genetic aberrations such as deletions of 17p and 11q have the
greatest impact. In addition, many other factors may predict
prognosis, e.g. bone marrow histopathology, the lymphocyte
doubling time, the extent of lymphocytosis, various bio-
chemical markers – see the recent reviews [1, 2].
Morphological examination of peripheral blood cells is
still the first-line contact of the physician with the patient’s
disease. The criteria elaborated in 1989 by the French-Amer-
ican-British (FAB) group  are still of value for diagnosing
CLL; they are based both on precise morphological evalua-
tion and on immunophenotypic analysis. These principles
were largelyadopted by the diagnostic criteriaof the Interna-
tional Workshop on CLL(IWCLL), and later by the National
Cancer Institute Working Group (NCI-WG) and more re-
cently by the WHO classifications [4–6]. The FAB criteria
 recognize 2 major categories of CLL: 1. the typical one
with quite monotonously looking mature-like CLL cells, 2.
atypical CLL, with varying percentages of larger cells
(“mixed cells”). One subclass of atypical CLL, CLL/PL, has
a “dimorphic” appearance with 2 subsets of cells – the larger
prolymphocytes with nucleoli and the smaller typical cells.
The other subclass comprises cells of variable size, from the
ing nucleoli. CATOVSKY has later called this subtype “pleo-
morphic” . The original article of the FAB group  has
noted that atypical morphology may be connected with dis-
ease progression. Other groups have shown that cell mor-
phology may be relevant to prognosis  or may correlate
with other prognostic markers, such as trisomy 12 [7–10].
However, the Belgic morphologically based concept of CLL
astwo (typicalor atypical)closelyrelatedentitieswith diver-
gent prognostic characteristics has never been widely
adopted into practice . Currently, the possible prognostic
relevance of cell morphology is shadowed by other markers,
such ascytogeneticaberrations, themutationalstatusof IgVH
genes, expression of CD38 or ZAP-70 antigens, or poten-
tially HLA-G expression [1, 2, 11–16].
Weattemptedtocheck how far relevantisinsistingon pre-
cise morphological examination of peripheral blood slides in
the beginning of the 3rd millenium with the numerous prog-
nostic markers at hand. Here we report the results of our ret-
CLL patients. To our surprise, even cell morphology could
efficiently predict their prognosis.
Patients and methods
Patients. As of June 30, 2005, altogether 110 CLLpatients
currently or recently followed-up at the Institute of Hematol-
ogy and Blood Transfusion, Prague, with diagnosis of CLL
according to the NCI-WG criteria , were included in our
70 males (64%) and 40 females, mean age 57 years (range
36–83), diagnosed 1982–2004, with a mediumfollow-up pe-
riod of 56 months. The mean WBC at presentation was 58.8
(7.1–569.0) x 109/l. Reliably documented clinical Rai stages
 were the following: Rai 0: n=46 patients; Rai 1: n=21;
Rai 2: n=7; Rai 3: n=2; Rai 4: n=7 patients. Treatment in the
majority of them was basically commenced according to the
NCI-WG recommendations  and was palliative in princi-
ple: either chlorambucil ± prednisone, the COP regimen
(cyclophosphamide + oncovin + prednisone) , or fludara-
bine monotherapy . A subset of patients, based on nega-
tive prognostic criteria (unmutated IgVHgenes and 11q or
17p deletions) and signs of progression, received more ag-
gressive therapy – FluCy (fludarabine-cyclophosphamide)
combination, some with rituximab in addition (the FCR pro-
tients an allograft. OS was evaluated in view of CLL-related
deaths (CLL progression, direct complications of CLL or
sequellae of antileukemic treatment). Patients who died of
diseases unrelated to CLL (n=8) were canceled the day prior
to death. All patients gave a written informed consent with
the investigational procedures.
Morphological evaluation. Altogether 88 patients (55
ripheral blood or bone marrow slides from the time of diag-
nosis for morphological evaluation. Preferentially peripheral
blood (83 cases) smears from untreated patients at diagnosis
were retrospectively read “blindly” by one of us (D.M.). In
the 5 remaining cases, bone marrow specimens were used. It
was acceptable, that in 25 patients (whose slides from the
date of diagnosis were either not available or were of poorer
quality), slides from later follow-up examinations were used
for morphological examination, provided that the patients
had not received any prior therapy. Cell morphology was
evaluated according to the FAB classification , on the ba-
sis of presence of atypicalcells (larger lymphocytes,prolym-
phocytoid cells and prolymphocytes) – see Figures 1 & 2.
The result was either typical or atypical (pleiomorphic or di-
Bone marrow histopathology. Results of a unilateral bone
marrow biopsy were available for 52 patients. The histo-
pathological pattern of bone marrow infiltration was evalu-
ated according to ROZMAN et al . Standard hema-
immunohistology (CD20, CD45, CD3) was performed for
discrimination of the lymphocyte subsets and for confirma-
tion of the B cell nature of the infiltration. The result was ei-
ther a diffuse or a non-diffuse (nodular, interstitial or mixed)
type of CLL infiltration .
Immunophenotypic studies. Diagnostic immunopheno-
typing was performed in all110 patients.Untilthe early90’s,
fluorescence microscopy was used. Since then, Becton-
Dickinson or Coulter Epics flow cytometry analyzers were
used. All patients included in the study had to conform with
the diagnostic criteria outlined by the IWCLL(CD5/CD19+,
220SCHWARZ, MIKULENKOVÁ, ČERMÁKOVÁ, POLANSKÁ et al.
weak surface Ig) . Cases diagnosed after 1995 had the
RoyalMarsden Score 4–5 [21, 22]. CD38 antigen expression
was analyzed in 73 patients, most of them in a CD19+ gate
using flow cytometry, and was deemed positive if >30% of
cells stained positive.
Cytogenetic and FISH analysis. Routine G-banding cyto-
genetics was performed in unstimulated bone marrow aspi-
rates. Since mid-90’s, standard interphase flurescence in situ
hybridization (FISH) analyses has been employed in 87 of
patients. FISH analyses were performed on bone marrow or
peripheral blood smear preparations (or both). For detection
of trisomy12, the centromeric CEP12 DNAprobe was used.
employed: LSI D13S319 for 13q14.3 band deletions, LSI
p53 for 17p13.1 and LSI ATM for 11q22.3 deletions. All
DNA probes were from Abbott-Vysis (Downers Grove, IL,
probe sets: Probe Set 1 LSI ATM/LSIp53 and Probe Set 2
of biallelic 13q14 deletions) has been used. Two hundred
nuclei were analyzed for each probe. For details, see ŠINDE-
LÁŘOVÁ et al .
87 patients, in some of them retrospectively, as the mutation
status is deemed to be stable in CLLpatients . Ficoll-Paq
isolated mononuclears were lyzed and RNA extracted using
the TriZol reagent (Invitrogen, KRD, Prague, Czech Repub-
lic). RNA was transcribed into cDNA using Superscript II
(Invitrogen), which was subjected to PCR amplificationwith
Ampli-Taq Gold polymerase (Applied Biosystems, Prague,
ation in the 7 families of IgVHgenes. The touch down meth-
odology using degenerated primers is described elsewhere
. RT-PCR products were purified and sequenced using
the Big Dye Terminator kit v. 3 and ABI Prism 310 Genetic
Analyzer (both from Applied Biosystems).The sequences of
IgVHgenes were compared to their germ-line configuration
using the BLAST program (www.ncbi.nlm.nih.gov/igblast/;
accessed September 5, 2005) and a cut-off value of 2% (i.e.
98% of homology) was set to discriminate between the mu-
tated and unmutated genes .
Statistical analyses. The correlation of various prognostic
parameters was analyzed in contingency tables using the
chi-square test. For analysis of quantitative data (WBC
counts) in 2 groups, medians were detected and a non-para-
metric two-tailed Mann-Whitney test was performed. Over-
allsurvival was analyzedusing the Kaplan-Mayer regression
method and the statistical significance was calculated using
the Mantel-Haenschel log-rank test. To reveal differences in
FISH analysis results, the log-rank test for trend was used.
All analyses were performed at the 95% confidence interval
and the p values were found using the GraphPad Prism ver-
sion 3.03 software (GraphPad Software, San Diego, CA,
FAB morphology and overall survival (OS). Typical FAB
morphology (see Fig. 1) was found in 25 (28.4%) and atypi-
cal morphology (see Fig. 2) in 63 (71.6%) of 88 patients.
Cases with atypical morphology had inferior OS (p=0.03),
the projected median OS was 103 and 237 months for pa-
tients with atypical and typical FAB morphology, respec-
tively (Fig. 3).
Correlation of FAB morphology and WBC counts. Initial
WBC counts were recorded in 82 patients. Interestingly, they
were higher in patients with typical morphology (Fig. 4). In
this group, median WBC was 45.0 x 109/l (range 13–382),
whereas this value was 29.0 (7–448) in patients with atypical
morphology (p=0.04). An arbitrary cut off value 30 (x 109/l)
of WBC discriminated the patients with respect to their OS
(Fig. 5): the projected median OS was 102 months and not
reached in patients with WBC >30 and WBC <30 (x 109/l),
FAB morphology and the mutational status of IgVHgenes.
Atypical morphology was found more often in patients with
unmutated rather than with mutated IgVHgenes: 36/44
(81.8%) and 23/38 (60.5%) patients, respectively (p=0.03;
Fig. 6). Inferior OS was characteristic of patients with
unmutated IgVHgenes (median 99 months), compared to pa-
tients with mutated IgVHgenes (237 months; p=0.002) – see
Figure 7. Interestingly, the analysis of 43 “unmutated” cases
with respect to morphology has revealed some trend to infe-
rior OS (p=0.14) in the subgroup of patients with atypical
reached, respectively. No CLL-related death was docu-
mented among the 7 “unmutated” cases with typical FAB
morphology. Vice versa, the analysis of the mutation status
within the morphologically atypical group (57 patients)
showed a significant difference (p=0.01) in OS: 99 months
and not reached for “unmutated” and “mutated” cases, re-
spectively (Fig. 9).
FAB morphology and cytogenetics. Two of the cytogenetic
aberrations, trisomy12(9patients)anddeletionof17p (7pa-
tients), were connected purely with atypical morphology
(Fig. 10). In the remaining patients, either with no aberration
detected, or with deletions of 13q or 11q, atypical morphol-
ogy was uniformly found in 68–71% of patients (17/25,
15/21 and 10/14 cases, respectively). There was an OS hier-
archy in patients according to the results of chromosomalab-
errations. Their rating was the following (from the most to
the least favorable): no cytogenetic aberrations, deletion of
13q, trisomy 12, deletion of 11q and 17p (Fig. 11). This hier-
archy was statistically significant when the log-rank test for
trends was employed (p=0.03).
FAB morphology and CD38 antigen expression. Atypical
morphology was found only slightly more frequently in
CD38+ patients than in CD38- cases (Fig. 12): 17/20 (85%)
any statistical significance (p=0.4). Patients expressing the
PROGNOSTIC RELEVANCE OF THE FAB CRITERIA221
222SCHWARZ, MIKULENKOVÁ, ČERMÁKOVÁ, POLANSKÁ et al.
Figure 4. WBC counts and FAB morphology. Medians are depicted.
Figure 5. OS according to WBC counts. An arbitrary cut-off of WBC
30x109/l was set.
Figure 6. The mutation status of IgVHgenes and FAB morphology.
Figure 1. “Typical” CLL morphology according to FAB . Small lym-
phocytes with monotoneous size and shape.
Figure 2. “Atypical” CLL morphology (“mixed cells”) according to
FAB . Small lymphocytes predominate, but somewhat larger lym-
phocytes and prolymphocytoid cells with more abundant basophilic cy-
Figure 3. OS according to FAB morphology.
Figure 7. OS according to the mutational status of IgVHgenes.
cell surface CD38 antigen had shorter OS than those lacking
it (median OS was 96 and 201 months, respectively; p=0.04;
FAB morphology and histopathology. Atypical morphol-
ogy was found in 4/5 (80%) with diffuse and in 31/44 (70%)
infiltration (p=0.7; Fig. 14). Some trend for worse OS was
seen in the only 4 patients with diffuse bone marrow
histopathology at diagnosis when compared with the 45 pa-
tients with a non-diffuse pattern of infiltration according to
ROZMAN et al  (p=0.12; Fig. 15).
PROGNOSTIC RELEVANCE OF THE FAB CRITERIA223
Figure 11. OS according to the cytogenetic aberrations.
Figure 12. CD38 antigen expression and FAB morphology.
Figure 13. OS according to CD38 antigen expression.
Figure 9. OS within the morphologically atypical FAB cases according
to the mutational status of IgVHgenes.
Figure 10. Cytogenetic aberrations and FAB morphology. The group
of 13q and 11q.
Figure 14. Histopathological pattern of bone marrow infiltration and
FAB morphology. Evaluated according to ROZMAN et al .
Figure 15. OS according to the histopathological pattern of bone mar-
row infiltration. Evaluated according to ROZMAN et al .
Morphological analysis(preferentiallyof peripheral blood
smears) is a first step of the routine diagnostic work-up of
chronic leukemias in every laboratory. It is quick, cheap and
However, in CLL, it is not generally accepted that cell
morphology as defined by the FAB criteria  might be of
prognostic value in CLL, despite that the Belgic group has
formulated a concept claiming that typical and atypical CLL
according to FAB are closely related but distinct diseases
with different outcomes . In 1999 and 2000, studies on
the prognostic impact of two important parameters, ie. the
mutation status of IgVHgenes and FISH detected chromo-
somal aberrations, were published [11–13] and somewhat
shadowed the possible prognostic potential of FAB defined
morphology. Therefore we tried to check whether FAB mor-
phology maybe of any value in the 3rd milleniumhaving the
newer prognostic markers at hand. Our data strongly support
the notion that FAB morphology is a reliable prognostic
marker. Notably, only a minority (28.4%) of our patients had
typical morphology. However, their OS was spectacular: the
first leukemia-associated death was documented 201 months
after the date of diagnosis. Interestingly, in the large Belgic
cohort of patients, 300 of 390 (76.9%) had typical morphol-
ogy. The reason of the considerable discrepancy between the
Belgic and our own results is not clear. There is some selec-
tion bias at our Institute, as patients having clinical problems
are preferentially seen. On the other hand, this selection is
IgVHgenes: 44 out of 82 (53.7%) evaluable patients in our
study had unmutated IgVHgenes, whereas only 45–51% of
seems,afterall,thatthediscrepancy of thepercentages of the
morphologically atypical cases in our and the Belgic study
[8, 10] were caused by the subjective nature of reading the
morphological slides. This is probably also the major reason
why morphological evaluation of CLLcells is not regarded a
reliable tool. However, it is interesting that despite these dif-
ies, prognostic relevance of the FAB classification could be
stilldemonstrated. We feel it is an advantage if a single expe-
rienced hematologistevaluates all the specimensover a short
period of time.This might have contributed to consistence of
our evaluation procedure performed on the basis of the FAB
Among the correlations of FAB morphology with other
markers, it was somewhat peculiar that higher WBC counts
However, this does not necessarily imply that patients with
higher WBC counts had a more progressive disease. It could
be the case that patients with typical CLLare diagnosed later
as their symptoms appear later. The other explanation might
be that the patients with atypical morphology could have a
more lymphomatoustype of the disease (with a higher tumor
distribution index, lower WBC counts and adverse progno-
sis) according to JAKSIC et al . On the other hand, our
study has confirmed that a WBC count >30 x 109/l is associ-
ated with poorer outcome.
Our results have revealed correlation of FAB morphology
and the mutation status of IgVHgenes. However, discrepant
cases (typical FAB-unmutated and atypical FAB-mutated)
were identified. Interestingly enough, it is shown here that
the mutation status of IgVHgenes could discriminate prog-
nostic subgroups within the morphologically atypical cohort
(Fig. 9) and vice versa, even within the unmutated cases, no
patient succumbed to CLL in case he/she had typical FAB
morphology (Fig. 8), although thelatterfinding did not reach
tistical power than FAB morphology in discriminating OS
probabilities in univariate analyses. However, the two char-
acteristics jointly bring additional prognostic information
and thus it seems that the two prognostic factors are
complemetary, if not independent (this remains to be proven
in future studies with multivariate analysis of data).
FAB morphology correlated with cytogenetic results. We
confirmed that there is a strong association between trisomy
12 and atypical morphology [7, 8, 10]. Above that, we dem-
onstrated the same was true for deletion 17p cases. In fact,
OS in our group with no cytogenetic aberration was superior
in comparison to deletion 13q cases. Otherwise the prognos-
tic hierarchy of the remaining aberrations was equal in the
The correlation of FAB morphology on the one hand and
CD38 expression or the histopathological pattern of bone
due to lower numbers of patients examined. However, CD38
expression stillhad impacton OS, thus suggesting thatCD38
expression and FAB morphology yield complementary (and
perhaps independent) prognostic information. It has been
shown byothersthatCD38expression andmutationstatusof
IgVHgenes are independent prognostic factors [2, 26].
In conclusion, our study shows that careful cytological
cording to the FAB classification may bring a prognostically
valuable information that is easily accessible at every hema-
tological laboratory within 10 minutes of analysis. The prog-
nostic impact of the morphological analysis (if we compare
gene mutation status or CD38 antigen expression. Our data
from a modestly-sizedsingle institution study should be con-
firmed in a prospective multiinstitutional study employing
divergent prognostic parameters (including e.g. ZAP-70) al-
lowing multivariate analysis. In that case, we would propose
central review of the blood smears.
224SCHWARZ, MIKULENKOVÁ, ČERMÁKOVÁ, POLANSKÁ et al.
FADERL SJ, KEATING MJ. Treatment of chronic lymphocytic
leukemia. Curr Hematol Rep 2005; 4: 31–38.
rations, and survival in chronic lymphocytic leukemia.
Blood 2002; 100: 1410–1416.
BENNETT JM, CATOVSKY D, DANIEL M-T, FLANDRIN G,
GALTONDAGetal. Proposalsfortheclassification ofchronic
(mature) B and T lymphoid leukaemias. J Clin Pathol 1989;
International WorkshoponChronicLymphocytic Leukemia.
Chronic lymphocytic leukemia: recommendations for diag-
nosis, staging, and response criteria. Ann Intern Med 1989;
CHESON BD, BENNETT JM, GREVER M, KAY N, KEATING MJ et
al. National Cancer Institute-sponsored Working Group
guidelines for chronic lymphocytic leukemia: revised guide-
lines for diagnosis and treatment. Blood 1996; 87:
MÜLLER-HERMELINK HK, MONTSERRAT E, CATOVSKY D,
HARRIS NL. Chronic lymphocytic leukemia / small cell lym-
phoma. In: Jaffe E, Harris N, Stein H, Vardiman J, editors.
WHO Classification of Tumours: Pathology & Genetics of
Tumours of Haematopoietic and Lymphoid Tissues. Lyon:
IARC Press, 2001: 127–130.
CATOVSKY D. Chronic lymphoid leukaemias. In: Hoffbrand
AV, Lewis SM, Tuddenham EGD, editors. Postgraduate
Haematology. 4. edn. Oxford: Butterworth-Heinemann,
CRIEL A, VERHOEF G, VLIETINCK R, MECUCCI C, BILLIET J et
al. Further characterization of morphologically defined typi-
cal and atypical CLL: a clinical, immunophenotypic, cyto-
genetic and prognostic study on 390 cases. Br J Haematol
1997; 97: 383–391.
OSCIER DG. Cytogenetic and molecular abnormalities in
CRIEL A, MICHAUX L, DE WOLF-PEETERS C. The concept of
typical and atypical chronic lymphocytic leukaemia. Leuk
Lymphoma 1999; 33: 33–45.
DÖHNER H, STILGENBAUER S, BENNER A, LEUPOLT E,
KRÖBERAetal. Genomicaberrations andsurvivalinchronic
lymphocytic leukemia. New Engl J Med 2000; 343:
DAMLERN,WASILT,FAIS F,GHIOTTOF,VALETTOAet al. Ig V
tic indicators in chronic lymphocytic leukemia. Blood 1999;
HAMBLIN TJ, DAVIS Z, GARDINER A, OSCIER DG, STEVENSON
FK. Unmutated Ig VHgenes are associated with a more ag-
gressive form of chronic lymphocytic leukemia. Blood
1999; 94: 1848–1854.
STILGENBAUER S, BULLINGER L, LICHTER P, DÖHNER H. Ge-
netics of chronic lymphocytic leukemia: genomic aberra-
tions and VHgene mutation status in pathogenesis and clini-
cal course. Leukemia 2002; 16: 993–1007.
CRESPO M, BOSCH F, VILLAMORN, BELLOSILLOB, COLOMER
D et al. ZAP-70 expression as a surrogate for immunoglobu-
lin-variable-region mutations in chronic lymphocytic leuke-
mia. N Engl J Med 2003; 348: 1764–1775.
NÜCKEL H, REBMANN V, DÜRIG J, DÜHRSEN U, GROSSE-
-WILDE H. HLA-G expression is associated with an unfavor-
able outcome and immunodeficiency in chronic lympho-
cytic leukemia. Blood 2005; 105: 1694–1698.
RAI KR, SAWITSKY A, CRONKITE EP, CHANANAAD, LEVY RN
et al. Clinical staging of chronic lymphocytic leukemia.
Blood 1975; 46: 219–234.
HALLEK M, EICHHORST BF. Chemotherapy combination
treatment regimens with fludarabine in chronic lymphocytic
leukemia. Hematol J 2004; 5 Suppl 1: S20–S30.
KEATING MJ, O’BRIEN S, ALBITAR M, LERNERS, PLUNKETTW
et al. Early results of a chemoimmunotherapy regimen of
fludarabine, cyclophosphamide, and rituximab as initial
therapy for chronic lymphocytic leukemia. J Clin Oncol
2005; 23: 4079–4088.
ROZMAN C, MONTSERRAT E, RODRÍGUEZ-FERNÁNDEZ JM,
AYATS R, VALLESPÍ T et al. Bone marrow histologic pattern –
the best single prognostic parameter in chronic lymphocytic
leukemia: a multivariate survival analysis of 329 cases.
Blood 1984; 64: 642–648.
MATUTES E, OWUSU-ANKOMAH K, MORILLA R, GARCIA
MARCO J, HOULIHAN A et al. The immunological profile of
B-cell disorders and proposal of a scoring system for the di-
agnosis of CLL. Leukemia 1994; 8: 1640–1645.
MOREAU EJ, MATUTES E, A’HERN RP, MORILLAAM, MORILLA
RM et al. Improvement of the chronic lymphocytic leukemia
scoringsystem withthemonoclonalantibody SN8(CD79b).
Am J Clin Pathol 1997; 108: 378–382.
ŠINDELÁŘOVÁ L, MICHALOVÁ K, ZEMANOVÁ Z, RANSDOR-
FOVÁ Š, BŘEZINOVÁ J et al. Incidence of chromosomal
anomalies detected with FISH and their clinical correlations
in B-chronic lymphocytic leukemia. Cancer Genet Cyto-
genet 2005; 160: 27–34.
PEKOVÁ S, MARKOVÁ J, PAJER P, DVOŘÁK M, CETKOVSKÝ P
et al. Touch-down reverse transcriptase-PCR detection of
RT-PCR quantitation of minimal residual disease in patients
with chronic lymphocytic leukemia. Molec Diagn 2005; 9:
JAKSIC O, VRHOVAC R, KUSEC R, KARDUM MM, PANDZIC-
-JAKSIC V et al. Clinical tumor cell distribution pattern is a
prognostically relevant parameter in patients with B-cell
chronic lymphocytic leukemia. Haematologica 2001; 86:
HAMBLIN TJ, ORCHARD JA, IBBOTSON RE, DAVIS Z, THOMAS
mutations are independent prognostic variables in chronic
lymphocytic leukemia, but CD38 expression may vary dur-
ing the course of the disease. Blood 2002; 99: 1023–1029.
PROGNOSTIC RELEVANCE OF THE FAB CRITERIA 225