HAIRY CELL LEUKEMIA: A N UNUSUAL

LYMPHOPROLIFERATIVE DISEASE
A Study of 24 Patients
M. GOLOMB,
HARVEY MD

A laboratory and clinical evaluation of 24 patients with hairy cell leukemia was
carried out over a 23-month period. Most patients had splenomegaly without
adenopathy or pancyotpenia. Nine of the patients had undergone splenectomy
prior to referral; their median WBC count was 6600/mm3.The median WBC
count for the 14 patients who had no prior therapy was 3550/mm3,and their
median platelet count was 80,500/mm3.Spleen weights ranged from 618 to 3780
g; there appeared to be no relationship between the size of the spleen and the
response in the blood counts after splenectomy. Four patients in whom the
majority of the WBC were hairy cells underwent splenectomy, which produced
no real change in their WBC count; however, there was improvement in the
platelet count in three. In contrast, the presence of leukopenia with a low
percentage of hairy cells predicted a beneficial response to splenectomy. The
study of surface immunoglobulins (SIg) in 16 patients demonstrated that
resynthesis had occurred in each case. Phagocytosis of zymosan was studied in
15 patients; in 8 of these, 25% or more of the hairy cells were capable of
phagocytosis; in 6 others, 0-9%; and in one, 13%. The resynthesis of SIg is a
feature usually associated with B-lymphocytes, but the phagocytosis of
zymosan is not. Thus, the existence of either a spectrum of functional capa-
bilities of hairy cells or several distinct subtypes is suggested by these data.
Platelet aggregation with epinephrine was abnormal in 7 of 14 patients studied
but there were no clinically significant bleeding problems. A chromosome
abnormality was present in 2 of the 19 patients from whom adequate samples
were obtained; the abnormality probably involved chromosome 12 in both
patients as well as absent Y and was associated with a rapidly progressive
clinical course. The presence of a predominant number of hairy cells with a
normal or increased peripheral blood WBC count or of a chromosomal ab-
normality suggests that splenectomy might not be beneficial as the initial ther-
apy and that chemotherapy should be considered.
Cancer 42:946-956, 1978.

H AIRY CELL LEUKEMIA (leukemic reticu- toplasmic projections. These projections can
loendotheliosis) is a distinct clinical- be very thin and look like hairs, or they may
pathological entity3,6,8,'5,24which is closely be rather thick and irregular and look like
associated with the lymphoproliferative dis- pseudopods. The diagnosis of the disease can
e a s e ~ .The
' ~ ~name
~ ~ is derived from the pres- be confirmed either by cytochemistry, in
ence in the peripheral blood of abnormal which staining for tartrate-resistant acid phos-
mononuclear cells which have exaggerated cy- phatase is e ~ t a b l i s h e dor
, ~from
~ the character-

Presented at the American Cancer Society- National Sprague Memorial Institute!: and the Thomas Moore
Cancer Institute National Conference o n the Lymphomas Fund.
and the Leukemias, September 29-October 1 , 1977, Address for reprints: Harvey M. Golomb, MD, Section
New York, New York. of Hematology/Oncology, University of Chicago Hospital
From the Department of Medicine, Section of and Clinics, 950 E. 59th St., Chicago, IL 60637.
HematologylOncology, University of Chicago Hospitals T h e author thanks .James Vardirnan and Daina
and Clinics, and T h e Franklin McLean Memorial Re- Variakojis for their collaboration, Janet Rowley and
search Institute (operated by T h e University of Chicago Valerie Lindgren for their assistance in karyotyping
for the U. S. Energy Research and Development Ad- cells, Donald Sweet for assisting in assessing platelet
ministration under Contract No. EY-76-C-02-0069), Chi- aggregation abnormalities, Deberah Simon, Eileen
cago, Illinois. Leatherman and Craig Rosner for the laboratory assist-
Supported in part by the Leukemia Research Founda- ance, and Elisabeth Land for editorial assistance.
tion, Hematology Research Foundation, Otho S. A. Accepted for publication February 3, 1978.

0008-543X-78-0800-0946-0115 0 American Cancer Society

946
No. 2 HAIRYCELLLEUKEMIA. Golomb 947

istic morphology seen in a bone core biopsy.30 formation have been previously documented.2o
Accurate diagnosis is important, since it is pre- Platelet aggregation studies were performed
sumably preferable not to use treatment or within one hour after collection by use of
advise only splenectomy, rather than to give a Chrono-Log Aggregometer (Chrono-Log
either radiotherapy or chemotherapy.*O Corp., Broomal, PA) according to the method
The clinical characteristics and therapy of of Mustard et aLZ9Peripheral blood and bone
hairy cell leukemia (HCL) has been studied, marrow samples were collected for chromo-
and laboratory investigations have been car- somal analysis, carried out by Valerie Lind-
ried OUt~2,4,10,13,14,16,1S-2Z,ZS,37-41 We now report gren and Janet Rowley, MD, of our institution.21
on a group of 24 patients with this unusual
lymphoproliferative disease who were exam- Controls
ined and evaluated during a 23-month period.
The history, physical findings, clinical course, Parallel laboratory studies for SIg and re-
and both routine and special laboratory ex- synthesis, phagocytosis, and E-rosette forma-
aminations will be detailed and compared with tion were made on the leukemic cells of 8
reports in the literature. A careful, compre- patients with chronic lymphocytic leukemia
hensive prospective evaluation allows defini- (CLL); 5 patients with malignant lymphoma,
tion of prognostic factors which have thera- poorly differentiated lymphocytic type (PDL);
peutic implication. 2 patients with acute myelogenous leukemia
(AML); and on the circulating mononuclear
MATERIALSAND METHODS cells of 3 normal subjects who served as con-
trols.
Clinical-Pathological
RESULTS
Twenty-four patients with a confirmed di-
agnosis of HCL were evaluated from May General Patient Characteristics
1975, through March 1977. In all but four of
these patients, the history was taken and the Of the 24 patients with HCL, 17 were men
physical examination was done by the author. and 7 were women. One man was black and
Routine hematology and chemistry laboratory one woman was of Mexican origin; the other
studies were performed in each case. Serum 22 were white. Since our institution has be-
protein electrophoresis and quantitative im- come a referral center for patients with HCL,
munoelectrophoresis were obtained when the referral diagnoses have been fairly spe-
possible. Bone marrow aspiration and bone cific; 18 of the 24 patients were referred for
core biopsy were performed when permitted evaluation of HCL, 1 for malignant lym-
by the patient. Follow-up was completed for phoma, 1 for hypersplenism, 1 for leukemia
all patients through June 1977. (type unknown), 2 for pancytopenia (etiology
unknown), and 1 with no specific diagnosis.
The history usually provided little specific
Cytological and Cytochemical information. Only 2 patients complained of
Smears of peripheral blood and bone mar- weight loss, fever, or night sweats. Four pa-
row aspirates were stained with Wright- tients had abdominal discomfort, 2 had minor
Giemsa stain. Each smear was examined, and bleeding problems, and 3 had some type of
the percentage of circulating hairy cells was infection. Only 1 patient had noted ade-
calculated. Peripheral smears and bone mar- nopathy.
row aspirates,if available, were tested for acid Physical examination showed that 21 pa-
p h o ~ p h a t a s e ,tartrate-resistant
~~ acid phos- tients were without adenopathy; of the others,
p h a t a ~ e , napthol
~ , ~ ~ AS-D-chloracetate ester- one had a node enlargement of 1-2 cm,
ate, and a-napthyl acetate e ~ t e r a s e peroxi-
,~ another 2-3 cm, and the third slighly more
dase,26and periodic acid-Schiff reactions. than 3 cm. Nine patients were postsplenec-
tomy at the time of first examination at our
Laboratory center. Of the remaining 15 patients, 4 had
no palpable spleen, 3 had a palpable spleen
The procedures for the isolation of hairy between 1 and 5 cm below the left costal mar-
cells and subsequent evaluations of the sur- gin (LCM), 5 between 6 and 10 cm, 2 at 15
face imrc ,noglobulins (SIg) and resynthesis cm, and 1 at 20 cm. Only 2 patients had a
of SIg, phagocytosis of zymosan and E-rosette liver span of more than 12 cm. A skin rash
948 August Supplement 1978
CANCER Vol. 42

TABLE
1 . Results of Initial Study of HCL, Patients at Presentation or Referral to the University of Chicago

Alpha-
napthyl Zymosan
PMN aretate Type phago-
and Mono- T.K.A.P. esterase E- PV of SIg cytosis Platelet
Case Age WBC/ bands cytes H.C. Plate- Splen- of H.C. of H.C. rosettes resyn- resyn- (% of aggrega- Karyo-
no. (yrs,) Sex mm3 (96) (%) (%j lets/mm' ectomy (%) ($6) (%) thesis thesis H.C.) tion type
~~~~~ ~~~ ~~

1 46 M 6.0 4 0 48 48,000 No 44 90 9 +§ MDK 33 NI Abn

2 52 F 8.4 3 3 70 149,000 No 82 61 13 +§ MDK 37 Ahn NI
3 40 F 2.9 29 0 9 192,000 Yes 100 20 20 +§ N.D. 41 Abn N1
4 75 F 5.9 22 7 5 2'L0,000 Yes 70 40 57 +§ N.E. 25 NI N.E.
5 58 M 5.5 9 0 58 84,000 No 6 10 25 +§ N.E. 34 N.D. NI
6 55 M 3.4 53 3 7 59,000 N" 70 95 13 + MA 13 Abn NI
7 80 M 9.4 40 6 35 167.000 NO 4 25 26 + N.D. 3 N.D. NI
8 66 M 7.2 41 10 7 128,000 No 50 70 29 +§ AMAi 9 NI NI
9 56 M 4.7 35 3 10 58,000 No 90 10 57 + N.D. 0 Ahn NI
10 82 M 13.5 12 2 73 66,000 Yes 38 N.D. 14 +§ MDK 0 N.D. NI
11 51 M 21.0 I3 1 59 123,000 Yes 55 95 5 +§ MD* 4 NI NI
12 60 M 4.7 52 10 5 278,000 Yes 6 40 48 + N.D. N.E. Ahn N1
13 52 M 1.3 10 2 5 115,000 No 100 N.D. N.D. N.D. N.D. N.D. Border- N.E.
line
14 28 M 1.0 13 0 27 77,000 No 15 95 N.D. N.D. N.D. N.D. Ahn NI
15 76 F 2.5 74 7 5 22,000 No +t N.E. N.D. N.D. N.D. N.D. N.D. NI
16 57 M 5.2 7 0 15 32,000 Yes 901. N.D. N.D. N.D. N.D. N.D. N.D. Abn
17 50 F 13.6 15 0 63 29 1,000 \Yes 75 46 3 + N.E. 0 N.E. N.E.
18 75 F 3.2 33 1 0 148,000 No N.E. N.E. 42 N.E. N.D. N.E. NI NI
19 50 M 4.7 34 3 22 77,000 No 92 5 29 + N.E. 35 NI NI
20 53 M 2.1 43 0 7 75,000 No 60 20 N.D. + N.E. 43 Abn N1
21 55 F 1.7 14 0 3 34,000 No N.D. N.D. N.D. N.D. N.D. N.D. N.D. N.D.
22 50 M 6.1 61 11 10 420,000 Yes +t N.D. N.D. N.D. N.D. N.D. N.D. NI
23 77 M 4.9 83 0 0 210,000 Yes N.E. N.D. N.D. N.D. N.D. N.D. N.D. N.D.
24 60 M 3.4 44 1 2 I 1 1,000 No +t N.E. 40 + N.D. 50 N.D. NI

* Light chain antisera not tested. N.D. = Not done; N.E. = Not evaluable; NI = Normal; Abn = Abnormal;
t Few cells available. PMN = Polymorphonuclear leucocytes; T.K.A.P. = Tartrate resistant acid
*Spleen cells. phosphataae; PV = Polyvalent goat anti-human antisera; H.C. = Hairy cells;
8 Trypsiniration technique. SIg = Surface immunoglobulin.
+ positive (PV resynthesis)

was present in 2 patients, but was not biopsied. ble 1). The percentage of hairy cells was low,
The results of the neurological examination except in 5 cases in which it was greater than
were negative in 23 patients; 1 had an un- 50%. A comparison of the postsplenectomy
associated abnormality. and no treatment patients is given in Table 2.
Routine laboratory tests showed a normal
Clinical and Laboratory Findings blood urea nitrogen in 12 of 14 patients; 2
had slightly increased values. Alkaline phos-
The most frequent finding was pancyto- phatase levels were normal in 14 of 16 patients
penia with a moderate decrease in hemoglo- tested, and the bilirubin was normal in 15
bin, white blood cell count, and platelets (Ta- tested. Total protein was measured in 15 pa-

TABLE
2. Comparison of Blood Counts for N o Treatment and Postsplenectomy Groups

Total No prior Post-

Type of blood count (24 patients) therapy (14)* splenectomy (9)T

Hematocrit (%)
Range 11-46 11-46 28-43
Median 37 34 39
White blood cells/mmg
Range 1300-2 1,000 1300-9400 2900-21,000
Median 4700 3550 6600
Platelets/mm3
Range 22,000-420,000 22,000- 167,000 66,000-420,000
Median 104,000 80,500 2 10,000

* Patient 2 had single-agent chemotherapy for 2 years and is excluded.

t At time of presentation to o u r institution.
No. 2 HAIRYCELLLEUKEMIA. Golomb 949

tients: 10 were normal, 2 had elevated values, tive for tartrate-resistant acid phosphatase
and 3 had depressed values. Serum protein (T.R.A.P.). In 2 patients (18 and 23), no hairy
electrophoresis was normal in 9 patients cells were seen in the peripheral blood. In one
tested. Immunoelectrophoresis (IEP) was ab- case, (21), the reaction was not performed.
normal in 3 of 7 patients tested, but was not In most cases, the proportion of hairy cells
monoclonal. Quantitative IEP was obtained which were acid-phosphatase-sensitive varied.
for 13 patients; the IgG was normal in 11 The number of acid-phosphatase-sensitive
and elevated in 2; the IgA was normal in 11 cells differed from patient to patient and, in
and elevated in 2; and IgM was normal in 8, the same patient, at different time intervals.
elevated in 2, and decreased in 3. The a-napthyl acetate esterase reaction was
positive, but was not intense in any case. There
Bone Marrow Findings was no correlation between the percentage of
Bone marrow aspiration was not attempted cells having esterase activity and the number
in 3 patients because permission was refused; of cells which showed tartrate-resistant acid
it was attempted, but unsuccessful in 7 pa- phosphatase activity or which ingested zymo-
tients. Aspirations were obtained, however, san (Table 1). Napthol AS-D-chloracetate es-
from the iliac crest in 10 patients, and from terase and peroxidase reactions were com-
the sternum in 4 patients. Three of the 4 pa- pletely negative in hairy cells. The Periodic
tients in whom sternal aspirations could be Acid-Schiff stain showed weakly positive
done had previously had “dry” iliac crest as- small granules which were scattered diffusely
pirations. The information obtained from the throughout most of the hairy cells.
aspirates was diagnostic of HCL in 12 of the
14 cases. Bone core biopsies, done on 20 pa- Clinical Course
tients, were diagnostic in every case.3oSeven- Prior to being seen at our institution, 9 pa-
teen of the 20 biopsies were available for eval- tients had had splenectomy and l patient (2)
uation of the pattern of involvement; in 5, had had chemotherapy (oral cyclophospha-
the pattern was patchy, and in 12 it was dif- mide 2 years previously). Fourteen had had
fuse. Cellularity was greater than 80% in 10 no treatment (Table 2). Fifteen were stable
patients, and more than 50% hairy cells were at the time of presentation, 4 had responded
found in 12 patients. Reticulin was increased to previous therapy ( i e . , splenectomy), and
in the 14 patients in whom it was evaluated. 5 had progressive disease and symptoms.
Of the 5 patients with progressive disease,
Cytochemistry 4 subsequently underwent splenectomy. Two
Cytochemical evaluation (Table 1) showed patients (1 and 2) had high percentages of
that, in 2 1 of 23 patients, hairy cells were posi- hairy cells initially; neither responded to the

3. Blood Counts and Spleen Size in Splenectomized Patients

TABLE

Interval
to post-
Presplenectomy Weight splenectomy Postsplenectomy
Case of spleen counts
no. Hct (%) WBC;/mm3 Plts/mm3 (P) (months) Hct (%) WBC/mm3 Plts/mm3
~

1 31 6,000 48,000 2,265 3 25 1 1,800 125,000

2 30 18,600 73,000 1.092 2 35 30,000 329,000
3 37 2,200 62,000 1,090 66 40 2,900 192,000
4 22 3,500 7 1,000 700 61 42 5,900 220,000
6 25 3,400 59,000 1,940 2 39 5,500 346,000
10 40 2,000 130,000 2.350 24 32 13,500 166,000
11 32 6,200 120,000 3,780 17 43 2 1,000 123,000
12 40 2,100 75,000 618 5 41 4,700 278,000
16 21 5,200 30,000 1,100 1 28 12,800 102,000
17 22 14,200 52,000 2,000 15 39 13,600 291,000
20 36 2,100 75.000 1,200 2 30 5,100 150,000
22 Leukopenia* 1,750 161 39 6,100 420,000
23 29 5,000 106,000 1,900 23 39 6,600 210,000

* Counts not available from 1963; “Leukopenia” stated o n Surgical Pathology Sheet.
950 August Supplement 1978
CANCER Vol. 42

operation. One of the two (1) died of progres- tested (Table 1); a clear monoclonal pattern
sive disease 6 months after splenectomy, and could be established in four patients (1, 2, 6,
within two days of the start of combination and lo), and a suggestive monoclonal pattern
chemotherapy which had lowered his WBC was seen in one other (11). Patient 14 had a
from 40,000 to 9,000. The second patient (2) GDK pattern initially, but the number of
was started on chlorambucil, 4 mg each day, cells available was inadequate for resynthesis
two months after splenectomy when her ini- studies.
tially elevated WBC continued to rise. The For 5 of the 16 patients whose cells were
third and fourth patients (6 and 20) had both evaluated for SIg resynthesis and phagocyto-
leukopenia and thrombocytopenia, which dis- sis of zymosan, the evaluation was performed
appeared after splenectomy; neither is receiv- two (17) or three times ( 1 , 2 , 6 , 11). Polyvalent
ing therapy now. Pre- and postoperative blood resynthesis was present in the majority of the
cell counts and spleen weights for 13 splenec- hairy cells from these 5 patients in repeated
tomized patients are given in Table 3. determinations, both when culturing alone
A patient (14) who did not undergo splenec- and trypsinization with 48-hour culturing
tomy, but who had progressive disease and a were used. No more than 20% of nonhairy
history of 3 weeks of acute illness with fever mononuclear cells were positive for PV SIg
and dyspnea, was felt to be a poor surgical after 48 hours. The predominant SIg was not
risk for splenectomy to correct his leuko- consistently the same; in each case, however,
penia; he was started on combination chemo- IgM was present at some time on the majority
therapy. Open lung biopsy several weeks later of the hairy cells. IgD was also found on cells
revealed the presence of-Mycobacteriumknnsasii; of 4 of the 5 patients.
the patient died secondary to dissemination Results of studies on the hairy cells from
of this complicating infection. the spleens of four patients are shown in Table
Eighteen of the 24 patients are currently 4. SIg initially and resynthesis values in pa-
alive. Besides the 2 patients whose deaths are tients with other lymphoproliferative diseases
described above, 3 patients (15, 16, and 21) and in controls are shown in Table 5.
had leukopenia and died of infection. Patient
15 died of pneumonia as the diagnosis of HCL Phagocytosis of Zymosan
was being established, patient 16 died of pneu-
monia 3 months after splenectomy, and pa- The presence or absence of the capacity to
tient 21 died of pseudomonas sepsis several phagocytose zymosan resulted in segregation
days prior to an electively scheduled splenec- of the hairy-cell patients into two groups. In
tomy. The sixth (patient 10) died during the 8 of 15 evaluable patients, zymosan phagocy-
leukemic phase of the disease, two years after tosis was present in 25% or more of the hairy
splenectomy, with a white blood cell count of cells (Fig. 1). Of the 6 patients who had phag-
200,000/n11113. ocytosis by less than 10% of their hairy cells,
one (11) was studied three times; successive
Surface Ig and its Resynthesis values of 3 , 4,and 2% were found. Cases 8
and 9, with values of 9 and 0%, respectively,
I n 19patients(l-14, 16, 17, 19,20,and24), were studied twice. Phagocytosis was observed
surface immunoglobulin was demonstrated in 5% and 0% of the hairy cells, respectively,
initially in 90- 100% of the hairy cells. Resyn- the second time. In patient 6, only 13% of
thesis of PV SIg occurred in all 16 patients the hairy cells from the peripheral blood, but

TABLE
4. Functional Markers on Cell Suspensions from Spleens

Zymosan
Case PV initial PV resynthesis Type of SIg E-Rosettes phagocytosis
no. (% of H.C.) (% of H.C.) resynthesis (%) (% of H.C.)

1 100 100 M 2 12
2 100 100 MDK* 3 N.D.
6 N.D. 100 AGK* 17 38
20 100 90 GK* 6 50

* 1gA not available.

H.C. = Hairy cells.
No. 2 HAIRYCELLLEUKEMIA. Golomb 95 1

TABLE
5. Controls: Functional Data on Normals and Patients with the Leukemic Form of Other Diseasfs

Zymosan
No. of %TO ‘SIgT48 Type of SIg E-Rosettes phagocytosis
Diagnosis patients (W) (5%) resynthesis (%) (%)

AML 2 0,s 0,1 None 02 0

CLL 7 60- 100 66- 100 4-MK 1-21 0- 1
2-MA
1-A
1 2 0 None 60 0
PDL 2 0,6 4,10 None 4,33 0
1 77 7 None 10 0
2 100 100 MK 1,10 0
Normal 3 16-20 9- 16 N.D. 53-54 PMN only

38% of his spleen hairy cells, phagocytosed myelogenous leukemia did not phagocytose
zymosan. zymosan (Table 5).
Zymosan phagocytosis in hairy cells, as eval-
uated in 1 micron sections, usually involved E-Rosettes
only one and never more than two zymosan
granules. This was in sharp contrast to the The percentage of E-rosettes in 7 HCL pa-
much larger number of granules phagocy- tients who had a high percentage of hairy cells
tosed in polymorphonuclear leukocytes and (Table 1: 1, 2, 5, 7, 10, 11, 17) ranged from
monocytes. Malignant cells from patients with 3 to 26%, with a median of 13%. The values
other lymphoproliferative diseases and acute for 8 patients in whom hairy cells comprised
less than 10% of the peripheral blood cells
ranged from 13 to 5796, with a median of
41%. Two patients in the latter group had
values of 57%, similar to those in our controls
(Table 5).

Platelet Aggregation Studies

In 14 patients, the platelet counts were suf-
ficiently high to allow aggregation studies to
be performed. Seven patients demonstrated a
marked reduction of platelet aggregation with
epinephrine (Table 1). N o abnormalities were
seen in secondary release and aggregation
with ADP or collagen. None of the patients
with reduced epinephrine aggregation had a
history of hemorrhage or a bleeding diathesis.
Platelets from six laboratory investieitors
were used as controls.

Chromosome Analysis
Cell samples for chromosome analysis were
obtained from 22 of the 24 patients. Chromo-
some samples from 19 patients were obtained
from peripheral blood cultures grown 24 and
48 hours without phytohemagglutinin, or
FIG. I . Zymosan ingestion by a hairy cell from patient 1 from bone samples. T~~ male pa-
shown by transmission electron microscopy. A ribosome-
,amellae complex can he identified in this cell (arrow) tients had similar, consistent abnormalities:
(X7000). Inset: Higher magnification of the complex the karyotypeof one (16) was462 x, +12; that
(~35,000). of the second (1) was 46, X, + C marker. In
952 CANCER
August Supplement 1978 Vol. 42

the latter case, the distal long arm of the C instead of our success of 13 of 20 (65%). Thus,
marker most closely resembled chromosome I would suggest one attempt at an iliac crest
No. 12 from band q 14 to q terminal, but aspirate initially. If the tap is “dry,” a bone
the short arm and proximal long arm were core biopsy should be obtained and suitable
of undetermined origin. Both karyotypes touch preparations made so that cytochemical
lacked the Y chromosome. A third patient (9) stains can be performed; then an aspirate
is listed as normal, but had, in one sample, frbm the sternum should be attempted.
a single abnormal cell with an extra No. 3 The appearance of the bone core biopsy
and an extra No. 12 (48, XY, +3, +12) and, specimens is quite characteristic for HCL.6,30
in a later sample, a second cell of poor mor- As Naeim and Smith described it, “The pat-
phology which also could have been trisomic tern is characterized by uniform, elongated,
for No. 12. fusiform, vesicular nuclei interspersed among
an increased delicate reticulin fibrillar net-
DISCUSSION work.”30 Naeim and Smith found a diffuse
pattern in all 14 of their biopsy specimens;
The clinical characteristics of the 24 patients Burke et al., however, reported that 7 of their
discussed in this report are similar to those 14 patients had focal involvement.6 Five of our
W,e’ ~
described in earlier s t ~ d i e s . ~ , ~ , ~ ~a
have 17 patients had a patchy or focal involvement.
slightly higher percentage (29%) of women Reticulin was increased in the samples from
than previously reported by Katayama and all our patients, as suggested by Naeim and
FinkelZ4(23%), by Burke et aL6 (19%), or by Smith.30
Catovsky’O (21%). The nonspecific nature of The percentages of hairy cells positive for
the presenting complaints as well as the lack tartrate-resistant acid phosphatase in the pe-
of symptoms are consistent with the previous ripheral blood could be scored for 18 patients
findings. Spenomegaly was present in 73% (Table 1). The percentages can be grouped
and extended more than 6 cm below the left as follows: hairy cells between 1 and 14%, 3
costal margin in 8 of the 11 patients with a patients (17%); between 15 and 20%, 1 (6%);
palpable spleen. In a small percentage of pa- between 30 and SO%, 4 (22%); and between
tients in each review, the spleen was not palp- 60 and loo%, 10 (55%). Our percentages of
able.6,’0~24
Hepatomegaly in a small percentage 17, 6, 22, and 55 can be compared with those
of patients as well as rare adenopathy are char- for 31 cases scored by Catovsky’O in which
acteristic of this disease. he reported the percentages in each group
Although pancytopenia is the characteristic to be 16, 23, 29, and 32, respectively. The
finding in the blood counts, 3 patients (10, presence of alpha-napthyl acetate esterase in
11, and 17) had elevated white blood cell the hairy cells of 15 evaluable patients differs
counts at the time of presentation to our from the observations of Utsinger et al. ,41 but
center. One of these patients (10) had been supports those of Rozenszajn et ~ l .who, ~ sug-
~
leukopenic two years earlier when he was ad- gested that this reaction, as well as evidence
mitted to a community hospital where a sple- of phagocytosis of latex particles, indicates
nectomy was done. Patient 11 had a normal that hairy cells share a common origin with
white blood cell count at presentation in 1973 monocytes or histiocytes.
and underwent splenectomy in late 1974. Pa- Regarding the clinical management of pa-
tient 17 had an elevated white blood cell count tients with HCL, Catovsky states that “splenec-
at the initial visit to her local physician. The tomy appears to be the best single measure
percentage of white blood cells scored as hairy for improving well-being and life expectancy
cells was greater than 50% in 5 patients and in HCL.”’O The 14 patients in our series who
between 20 and 50% in 4 patients. One pa- had no prior therapy upon presentation to
tient had 15% hairy cells, and 12 had 10% our institution had a median WBC count of
or less. Two patients (18 and 2 3 ) had only 3550/mm3 and a platelet count of 80,500/mm3
very rare circulating hairy cells, but the bone (Table 2), whereas those who were post-
core biopsies led to the diagnosis of HCL. splenectomy had a median WBC count of
The bone marrow aspirate has been re- 6600/mm3 and a platelet count of 210,000/
ported to be “dry” in one-half6 or morelo of mm3. Three of the patients who had had no
the patients. If the sternal aspirates had not prior therapy were recommended for sple-
been attempted after a “dry” iliac aspiration, nectomy-patient 1 within three months of
we would have had 10 out of 20 “dry” taps diagnosis because of decreasing platelets in
No. 2 CELLLEUKEMIA. Golomh
HAIRY 953

association with an increasing transfusion re- (PMN’s) was always less than 5% in contrast
quirement, patient 6 within fifteen months of to that of patients 11 and 17 who remained
diagnosis because of increasing abdominal ‘‘leukemic’’also, but who had more than 10%
girth which was interfering with his work as a PMN’s. Thus, splenectomy is of little benefit
farmer, and pat.ient 20 within four months o f in the “leukemic” patient unless anemia or
diagnosis because of symptomatic pancyto- thrombocytopenia is the major problem. In
penia. Patient 2 had received oral cyclophos- addition, a low percentage of PMN’s in the
phamide 2 years prior to presentation and leukemic patient postsplenectomy could indi-
underwent splenectomy nine months later be- cate susceptibility to recurrent infections due
cause of symptomatic splenic infarction which to granulocytopenia.
was documented at splenectomy. T h e absolute number of PMN’s might also
In the 13 patients who were splenectomized be an indicator for splenectomy to prevent
(Table 3 ) , spleen weights ranged from 618 fatal sepsis. Patients 14 and 2 1 had a low WBC
to 3780 g. Eleven of the 13 patients had at count and a low percentage of PMN’s; both
least a doubling of their platelet count fol- died of infection prior to undergoing splenec-
lowing splenectomy; one patient (10) had an tomy. An early elective splenectomy might be
increment of only 25%, whereas patient 11 the treatment of choice in patients with PMN’s
showed little change. T h e postsplenectomy of less than 500/mm3.
counts available for patients 10 and 11 were Of the 7 patients who were leukopenic
obtained at 24 and 17 months, respectively; at the time of splenectomy, 6 had an improve-
it is possible that the increment immediately ment in blood counts, and all are still alive,
postsplenectomy was greater. T h e WBC 3 for more than 5 years, and 1 patient for
showed a less dramatic increase in all cases. 14 years. All 6 patients had, and still have,
Catovsky has reported that, although most pa- less than 10% hairy cells in the peripheral
tients will benefit initially, only one-third will blood. Thus, patients with leukopenia and
obtain a long-term benefit from splenectomy with a low percentage of hairy cells might
alone.l0 Our series of patients will require a have a long-term beneficial response to
longer follow-up time before such a conclu- splenectomy. This group could comprise the
sion can be drawn. one-third of HCL patients who, Catovsky
In most series, results of splenectomy in states, will obtain a long-term benefit from
patients who were pancytopenic have been splenectomy alone.1°
reported,6”0 but little attention has been Our laboratory results support the sugges-
given to patients in the “leukemic” state (k., tion by both Fu et and Utsinger et d 4 1
normal or increased WBC with the majority that HCL consists of a proliferation of Ig-
of WBC consisting of hairy cells). In 4 (1,2, 11, bearing cells which are capable of phago-
and 17) of the 13 patients who underwent cytosis. T h e resynthesis of surface immuno-
splenectomy, 48% or more of their white globulin in all 16 of the patients whom we
blood cells were hairy cells. Two had normal tested supports the view of these authors
WBC counts and two had increased WBC that the neoplastic cells possess features
counts. The only change in the WBC count usually associated with B-lymphocytes. The
postsplenectomy was a rise in patients 1 and 2. discrepancy in SIg resynthesis patterns be-
The sole benefit due to splenectomy in these tween the hairy cells in the peripheral blood
4 patients was an improvement in the plate- and spleen (patients 1 and 6 in Tables 1
let count in 3 cases and improvement in the and 4) remains to be explained; a similar
hematocrit in 2. Of these 4 patients, 2 (11 lack of consistency has been observed by
and 17) have had little difficulty following Catovsky.lo Golde et al.lS who reported on a
splenectomy, even though they remain in the HCL patient with IgM macroglobulinemia,
leukemic phase about 2 years later. Patients regarded HCL as a neoplasm of B-lympho-
1 and 2 had a progressive increase in their cytes. Catovsky recently demonstrated the
WAC and percentage of hairy cells. Patient 1 presence of surface immunoglobulin in 14 of
succumbed to hairy cell infiltration of the 15 caseslO; in 11 of these IgD and/or IgM
lung, but patient 2 has responded to oral were the predominant surface immunoglo-
administration of chlorambucil. Both patients bulins. In our HCL patients, resynthesis of
had episodes of fever and infection within IgM and usually IgD was detectable (Table 1).
several months of splenectomy. Their per- This finding was similar to that seen in 6 of
centage of polymorphonuclear leukocytes our CLL patients and in 2 PDL patients (Table
954 CANCER
August Supplement 1978 Vol. 42

5); it is in keeping with the observations was trisomic for chromosome No. 12 (46, X,
of others for B-lymphoproliferative dis- +12). The second patient (1) had a similar
order~.',~~ karyotype (46, X, marker), also missing the Y
We have shown that the capacity to phago- chromosome, with the addition of a C-sized
cytose differs among HCL patients. In 8 of marker, the long arm of which most closely
our patients, phagocytosis of zymosan by 25% resembled chromosome No. 12 from band q
or more of their hairy cells was demonstrated, 14 to q terminal. The short arm and proximal
whereas 6 others had zymosan particles in long arm had a dull fluorescence of undeter-
only 0-9% of their hairy cells. Four of these mined origin. In one of these cells there
latter patients had many hairy cells in the pe- was a deletion of chromosome No. 3 (q12).
ripheral blood, and the low percentage of Both patients had normal lymphocytes con-
phagocytosis was not difficult to assess. The taining brightly fluorescent Y chromosomes.
absence of phagocytosis of staphylococci in a Although the hairy cells of both patients had
single patient in three s t ~ d i e s ~of, carbonyl
~~,~~ a normal chromosome count of 46, they were
iron particles or bacteria in 2 patients, of pseudodiploid, since each lacked a Y and since
latex particles in 1 patient,45and of Candida one was + 1 2 and the other +marker. A
in 4 patientsg led the respective investigators third patient (9) had an abnormal cell with
to conclude that hairy cells do not have a an extra No. 3 and an extra No. 12 (48,XY,
phagocytic capacity. Other investigators have + 3 , + 12) in one sample and a second cell with
found phagocytosis of staphylococci in a poor morphology, in a later sample, which
single case,25 of both latex particles and may also have been trisomic for No. 12.
staphylococci in 2 cases," and of latex particles Sakurai and Sandberg found that only 4% of
in 3 cases.34 control subjects between the ages of 50 and 59
This phagocytic activity does not exclude a years have a missing Y cell line in their bone
lymphoid origin for the hairy cell, since marrow, and that none of the subjects younger
phagocytic activity by lymphocytes has been than 50 years have such a cell line.36 Thus,
suggested46;however, this feature may indi- it seems reasonable to view the lack of the
cate some relationship between hairy cells Y chromosome as being a part of the leu-
and monocytes. Whether this means that both kemic karyotype since both men (1 and 16)
cells are derived from a precursor cell with are middle-aged (46 and 57 years old).
phagocytic potential, or whether the phago- The 46,X,+12 karyotype could be a non-
cytic activity in the hairy cell represents a random chromosomal abnormality specifi-
function acquired through a process of cally associated with HCL; however, more
dedifferentiation by a neoplastic cell, is cases need to be evaluated before such a con-
unclear. clusion is reached. The determination of the
In addition to doing studies of SIg regrowth histogenesis of HCL on the basis of the pres-
and zymosan phagocytosis, we tested platelet ence of a specific chroposome abnormality is
aggregation in 1 4 of the 24 patients. Our re- a theoretical possibility, but also requires fur-
sults confirm the observations of Levine and ther data. At this time, the No. 12 chromo-
Katayama that the platelets may be qualita- some abnormality cannot be related to the
tively abnormal in patients with HCL.26The abnormal No. 14 observed in other lympho-
latter observed a total lack of aggregation proliferative d i ~ e a s e s , ' ~ to ~ , ~819
, ~the ~ ,trans-
~~
with epinephrine in 6 patients. However, location present in the cells of a patient with
our 7 patients with abnormal platelet func- acute monocytic l e ~ k e m i a ,or ~ to the non-
tion demonstrated some aggregation with random abnormalities seen in acute nonlym-
epinephrine. There were no clinically sig- phocytic leukemia.33 The two patients with
nificant bleeding problems associated with the similar, consistent chromosome abnormalities
abnormal platelet function. (1 and 16) had a brief clinical course and
In spite of the low chromosome specimen rapidly progressive disease. Possibly the pres-
yields, several conclusions can be drawn from ence of a chromosome abnormality in these
this study. Most of the patients in this sample 2 patients signaled the evolution of a clone
were karyotypically normal. In two patients, of hairy cells with a proliferative advantage
however, consistent chromosomal abnormali- that resulted in this rapidly fatal course.
ties were evident in cells from unstimulated In the light of two recent reports on ag-
peripheral blood samples. One of these pa- gressive chemotherapy which provoked a
tients (16) was lacking the Y chromosome and dramatic clinical response in patients with
No. 2 HAIRYCELLLEUKEMIA. Golomb 955

active HCL,’o,12 patients with abnormal for aggressive chemotherapy, especially if fur-
chromosome patterns and/or patients with the ther data support the observation that an ab-
leukemic form of the disease at presentation normal karyotype or the “leukemic” state
could perhaps be considered as candidates are poor prognostic signs.

RE:FERENCES
1. Aisenberg, A . C., and Wilkes, B.: Lymphosarcoma 16. Fu, S. M., Winchester, R. J., Rai, K. R., and Kunkel,
cell leukemia: T h e contribution of cell surface study H. G.: Hairy cell leukemia: Proliferation of a cell with
of diagnosis. Blood 48:707,715, 1976. phagocytic and B-lymphocyte properties. Scand. J .
2. Boldt, D. H., Speckart, S. F., MacDermott, R. P.,
Nash, G. S., and Valeski, J. E.: Leukemic reticulo- 17. Fukuhara, S., Shirakawa, S., and Uchino, A.:
endotheliosis: “Hairy cell leukemia,” functional and Specific marker chromosome 14 in malignant lympho-
structural features of the abnormal cell in a patient with mas. Nature 259:210-23 1, 1976.
profound leukocytosis. Blood 49:745-757, 1977. 18. Golde, D. W., Saxon, A ,, and Stevens, R. H.:
3. Bouroncle, B. A,, Wiseman, B. K., and Doan, C . A.: Macroglobulinaemia and hairy-cell leukemia. N . E n d . J .
Leukemic reticuloendotheliosis. Blood 13:609-630, M e d . 296:92-93, 1977.
1958. 19. Golomb, H. M., Braylan, R., and Polliack, A.:
4. Braylan, R. C . , Jaffee, E. S., Triche, T. J., Nanba, “Hairy” cell leukaemia (leukaemic reticuloendotheliosis):
K., Fowlers, B. J., Metzger, H., Frank, M. M., Dolati, 4 scanning electron microscopic study of eight cases.
M. S., Yee, C. L., Green, I., and Berard, C. W.: Str-uc- Br. J . Haematol. 29:455-460, 1975.
tural and functional properties of the “hairy” cells of 20. Golomh, H. M., Vardiman, J., Sweet, D. L., Jr.,
leukemic reticuloendotheliosis. Canrer 41 :210-227, 1978. Simon, D., and Variakojis, D.: Hairy cell leukemia: Evi-
5. Brynes, R. K., Golonib, H. M., Desser, R. K., dence for the existence of a spectrum of functional
Recant, W., Reese, C., and Rowley, J. D.: Acute mono- capabilities. Br. J . Haematol. 38: 161- 170, 1978.
cytic leukemia: Cytological, cytochemical, ultrastruc- 21. Golomb, H. M., Lindgren, V., and Rowley, J. D.:
tural, and cytogenetic observations. Am. J . Clzn. Pathol. Chromosome abnormalities in patients with hairy
65:471-482, 1976. cell leukemia. Cancrr 41:1374-1380, 1978.
6. Burke, J. S., Byrne, G. E., Jr., and Rappaport, H.: 22. Haak, H. I,., d e Man, J. C. H . , Hijmans, W.,
Hairy cell leukemia (leukemic reticuloendotheliosis). I. Knapp, W., and Speck, B.: Further evidence for the
A clinical pathologic study of 21 patients. Canrrr lymphocytic nature of leukaemic reticuloendotheliosis
1399-1410, 1974. (hairy cell leukaemia). Br. ,/.Haemulo/. 27:31-38, 1974.
7. Caspersson, T., Zech, L., Johansson, C., and Modest, 23. Kaplow, I,. S.: Simplified myeloperoxidase stain
E. S.: Identification of human chromosomes by DNA- using benzidine dihydrochloride. Blood 26:215-2 19,
binding fluorescent agents. Chromosuma 30:2 15-227, 1965.
1970. 24. Katayama, I . , and Finkel, H. E.: Leukemic reti-
8. Catovsky, D., Pettit, J. E., Galton, D. A. G., Spiers, culoendotheliosis. A clinico-pathologic study with re-
A. S. D., and Harrison, C. V. Leukaemic reticuloendo- view of the literature. A m . J . ’Wed. 57:115-126, 1974.
theliosis (“hairy” cell leukaemia): A distinct clinico- 25. King, G. W., Hurtubise, P. E . , Sagone, A. L., Jr.,
pathological entity. Rr. J . Huemato/. 26:9-27, 1974. Lo Buglio, A. F., and Metz, E. N.: Leukemic reticulo-
9. Catovsky, D., Pettit, J. E., Galetto, J., Dkas, A,, and endotheliosis: A study of the origin of the malignant
Galton, D. A. G.: T h e R-lymphocyte nature of the hairy cell. ,4m.J. Med. 59:411-416, 1975.
cell of leukaemic reticuloendotheliosis. Rr. J . Harmritol. 26. Levine, P. H., and Katayama, I.: ‘ r h e platelet
26:29-37, 1974. in leukemic reticuloendotheliosis. Cancer 36: 1353- 1358,
10. Catovsky, D.: Hairy-cell leukemia and prolympho- 1975.
cytic leukaemia. Clin. Haematol. 6:245-268, 1977.
27. Li, C. Y., Yam, 1,. T., and Lam, K. W.: Acid phos-
11. Daniel, M. T., and Flandrin, G.: Fine structure
phatase isoenzyme in human leukocytes in normal
of abnormal cells in hairy cell (tricholeucocytic) leu-
and pathologic conditions. J . Hzstochem. Cytochem. 18:
kemia, with special reference to their in vitro phagocytic
capacity. Lab. Invest. 30:l-8, 1974.
473-481, 1970.
12. Davis, T. F., Waterbury, L., Abeloff, M., and 28. Manolov, G., and Manolova, Y.: Marker band in
Burke, P. J.: Leukemic reticuloendotheliosis: Report of a one chromosome 14 from Burkitt lymphoma. Nrcturr
case with prolonged remission following intensive chemo- 237:33-34, 1972.
therapy. Arch. Intrrn. ‘Wed. 136:620-622, 1976. 29. Mustard, J. F., Hegardt, R . , Rowsell, H. C., and
13. Dehusscher, L., Bernheim, J. L., Collard-Range, MacMillan, R. L.: Effect of adenosine nucleotides in
E. E., Govaerts, A,, Hooghe, R., Lejeune, F. J.. Zeicher, platelet aggregation and clotting times. J . Lab. Clin.
M., and Strychmans, P. A,: Hairy cell leukemia: Func- M r d . 64:548-559, 1964.
tional, immunologic, kinetic and ultrastructural char- 30. Naeim, F., and Smith, G. S.: Leukemic reticulo-
acterization. Blood 46:495-507, 1975. endotheliosis. Cancrr 34:1813- 1821, 1974.
14. Deegan, M. J., Cossman, J., Chosney, B. T., and 31. Preud’Homme, J. L., Brouet, J. C., Clauvel, J . P.,
Schnitzer, D.: Hairy cell leukemia. Cancer 38: 1952- 1961, and Seligmann, M.: Surface IgD in immunoprolifera-
1976. tive disorders. Scand. ,I .
Immunol. 3:853-858, 1974.
15. Flandrin, G., Daniel, M. T., Fourcase, M., and 32. Rowley, J . D.: A new consistent chromosomal
Chelloul, N.: LeucCmie A “Tricholeucocyte” (hairy cell abnormality in chronic myelogenous leukemia identified
leukemia), etude clinique et cytologique d e 55 observa- by quinacrine fluorescence and Giemsa staining. Na-
tions. Nouvelle Revue Francoire d’H6rnatologie 13:609-640, ture 243:290-293, 1973.
1973. 33. Kowley, J. D., and Potter, D.: Chromosomal
956 August Supplement 1978
CANCER Vol. 42

handing patterns in acute nonlyrnphocytic leukemia. 40. Sweet, D. L., Jr., Golomb, H. M., and Ultmann,
Blood 47:705-721, 1976. J. E.: Chronic lymphocytic leukaemia and its relation-
34. Rozenszajn, L. A,, Gutman, A., Radnay, J., Ben ship to other lymphoproliferative disorders. Clin.
David, E., and Shoham, D.: A study of the nature of Haematol. 6:141-157, 1977.
“hairy cells,” with emphasis on enzymatic markers. Am. J . 41. Utsinger, P. D., Yount, W. J., Fuller, C. R., Logue,
Clzn. Pathol. 66:432-441, 1976. M. J., and Orringer, E. P.: Hairy cell leukemia: B-
35. Rubin, A. D., Douglas, S. D., Chassin, L. N., Glade, lymphocyte and phagocytic properties. Blood 49: 19-27,
P. R., and Dameshek, W.: Chronic reticulolymphocytic 1977.
leukemia. Reclassification of “leukemic reticuloendo- 42. Wurster-Hill, D. H., McIntyre, 0. R., Cornwell,
theliosis” through functional characterization of the G. G., and Maurer, L. H.: Marker chromosome 14 in
circulating mononuclear cells. Am. j. Med. 47:149- multiple myeloma and plasma cell leukaemia. Lancet
162, 1969. 2:1031, 1973.
36. Sakurai, M., and Sandberg, A. A,: T h e chromo- 43. Yam, L. T., Li, C. Y . , and Lam, K. W.: Tartrate-
somes and causation of human leukemia. XVIII. T h e resistant acid phosphatase isoenzyme in the reticular
missing Y in acute myeloblastic leukemia. Cancer 38: cells of leukemic reticuloendotheliosis. N . Engl. J . Med.
762-769, 1976. 284:357-360, 1971.
37. Scheinberg, M., Brenner, A. I., Sullivan, A. L., 44. Zech, L., Haglund, V., Nilsson, K., and Klein, G.:
Cathcart, E. S., and Katayama, I.: T h e heterogeneity of Characteristic chromosomal abnormalities in biopsies and
leukemic reticuloendotheliosis, “hairy cell leukaernia.” lymphoid cells from patients with Burkitt and non-
Cancer 37:1302-1307, 1976. Burkitt lymphomas. Int. J . Cancer 17:47-56, 1976.
38. Seshadri, R. A,, Brown, E. J., and Zipursky, A,: 45. Zidar, B. L., Smith, W. I., Winkelstein, A,, Shad-
Leukemic reticuloendotheliosis: A failure of monocyte duck, R. K., Zeigler, Z., Rabin, B. S., Whiteside, T.,
production. N . Eng.1. Med. 295:181-184, 1976. and Breitfeld, V.: Hairy cell leukemia: A case with B-
39. Stein, H., Kaiserling, E.: Surface immunoglo- lymphocyte origin. Proc. Am. Soc. Hematol. 145a, 1975.
bulins and lymphocyte-specific surface antigens on 46. Zucker-Franklin, D.: T h e percentage of monocytes
leukemic reticuloendotheliosis cells. Clzn. Exp. Immunol. among “mononuclear” cell fractions obtained from nor-
18:63-71, 1974. mal human b1ood.j. Immunol. 112:234-240, 1974.