We examined the stem cell compartment of patients with acquired aplastic anemia (AA) using the long-term cutture- initiating cell assay (LTC-IC), in parallel with ...
From bloodjournal.hematologylibrary.org by guest on July 12, 2011. For personal use only.
1996 88: 1983-1991
A severe and consistent deficit in marrow and circulating primitive hematopoietic cells (long-term culture-initiating cells) in acquired aplastic anemia JP Maciejewski, C Selleri, T Sato, S Anderson and NS Young
Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.
From bloodjournal.hematologylibrary.org by guest on July 12, 2011. For personal use only.
A Severe and Consistent Deficit in Marrow and Circulating Primitive Hematopoietic Cells (Long-Term Culture-Initiating Cells) in Acquired Aplastic Anemia By Jaroslaw
P. Maciejewski, Carmine Selleri, Tadatsugu Sato, Stacie Anderson, and Neal S. Young
We examined the stem cell compartment of patients with acquired aplastic anemia (AA) using the long-termcuttureinitiating cellassay (LTC-IC), in parallel with measurements of CD34+ cells and mature hematopoietic progenitors. Secondary colonies from cells surviving 5 weeks of long-term bone marrow culture(LTBMC) were determined for the peripheral blood (PB) of 68AA patients and13 normal controls and for BM of 49 AA patients and 14 controls; because of low cell numbers, formal limiting dilution analysis could only be performed in 10 patients. The relationship of cell input in LTBMC and the outputof secondary colonies was linear, allowing quantification of LTC-IC number from bulk cultures. Secondary colony formation was markedly abnormal in severe AA. In contrast t o 7.8 colony-forming cells (CFC)/105 mononuclear cells in normal BM and 0.14 CFC/105 normal PB mononuclear cells, patients with severe disease showed 0.024 CFC/105 in BM and 0.0068 CFC/105 in PB. Under limiting dilution conditions, patients’ cells also showed markedly lower colony-forming ability. In contrast t o 4.3 1 colonies/normal LTC-IC, we obtained only 1.27 0.09 and 2.0 & 0.35 colonies from BM of acute and recovered cases, respectively. These values were used t o extrapolate LTC-IC numbers from secondary colony formation in suspension cul-
tures. In PB, calculated LTC-IC were decreased 7.4-fold in new and relapsed severe AA and 2.8-fold in recovered AA. In BM, LTC-IC were decreased 10-fold in new and relapsed AA and sixfoldin recovered cases. Compared with measurements obtained on presentation, LTC-IC were lowerin posttreatment samples from patients who had failed t o recover after intensive immunosuppression and relatively higher in cases at relapse. In recovered patients, LTC-IC number increased but remained below the normal range in 20 of 25. In patients studied serially for 3 t o 12 months after treatment, LTC-IC numbers remained stable but low. LTC-IC number correlated with concurrentlydetermined CD34+ cell number and primaryhematopoietic colony formation.These results indicate that stem cell numbers, as quantitated by the LTC-IC assay, are markedly diminished in number in all severe AA. Additionally, the functionof the stem cellor the stem cell compartment in AA is also abnormal, as inferred fromthelow clonogenic potential in secondary colony assays. Early hematologic improvement in some patients occurs without increasing numbers ofLTC-IC, and a minority of recovered cases show apparent repopulation of the LTCIC compartment years after treatment. 0 1996 by The American Societyof Hematology.
E
markers; kinetically, both LTC-IC and repopulating cells are quiescent, as determined by their insensitivity to 5-fluoro~ r a c i lLTC-IC .~ would appear to be either identical to stem cells or a surrogate assay for the stem cell compartment. We have now applied this assay to measure hematopoiesis in patients with AA.
*
*
ARLY OBSERVERS OF aplastic anemia (AA) inferred “anhematopoiesis” from the empty bonemarrow (BM).’ With more sophisticated tests, hematopoietic failure was evident from functional assays of lineage-committed and multipotential colony-forming cells (CFC; summarized inYoung’) and low numbers of phenotypically defined CD34’ cells.’,4 Absence of stem cells, which could not be directly measured in humans, was extrapolated from these data; AA has been called a stem cell disease: the basis for replacement treatment by BM transplantation. Recovery after immunosuppressive treatment: without stem cell replacement, suggested the possibility that stem cells might not be absent in all AA. Stem cells might be spared due to their quiescence or infrequency, and marrow failure may then develop as a result of toxic or immune effects on more mature cells. In the mouse, cycle-active drugs can produce BM hypocellularity yet spare stem cells (reviewed in Young5). Failure to recover with immunosuppression could reflect the severity of stem cell deficiency rather than the absence of an autoimmune mechanism.6Clonogenic hematologic diseases arising after immunosuppressive treatment may also relate to stem cell number, reflecting either stressed hematopoiesis from a limited number of stem cells or the consequences of rapid proliferation within a primitive cell compartment. The long-term culture-intitiating cell (LTC-IC) is defined as a primitive hematopoietic cell capable of producing clonogenic progenitor cells after 5 weeks of long-term BM culture (LTBMC).7Murine LTC-IC have about the same frequency as stem cells rigorously definedby their ability to effect long-term repopulation of an animaL8LTC-IC copurify with repopulating cells and share with stem cells rare phenotypic
Blood, Vol 88,No 6 (September 15). 1996: pp 1983-1991
MATERIALS AND METHODS Patients. BM and peripheral blood (PB) samples from 50 healthy volunteers and 90 patients with AA were obtained between October1994 and June 1996 and analyzed.Thediagnosis of AA wasestablished by BM biopsy and PBcountsaccording to the criteria of the International Studyof Aplastic Anemia and Agranulocytosis’’; severity was classified by the criteria by Camitta et al.” In the acute AA group, BM samples were analyzed from 41 patients with severe disease and for 8 with moderate disease. Of the severe cases, 29 had not previously been treated with immunosuppressive therapy, 8 were refractory to therapy and continued to fulfill the severity criteria, and 3 had relapsed after a period of improvement. We also analyzed 41 PB and 40 BM samples from patients who had
From the Hematology Branch, National Heart, Lung and Blood Institute, Bethesda, MD. Submitted November 3, 1995: accepted May 8, 1996. Address reprint requests to Jaroslaw P. Maciejewski, MD, Hematology Branch, National Heart, Lung and Blood Institute, Bldg 10, Room 7C103, Bethesda, MD 20892-1652. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. section 1734 solely to indicate this fact. 0 1996 by The American Society of Hematology. 0006-4971/96/8806-0$3.00/0
1983
From bloodjournal.hematologylibrary.org by guest on July 12, 2011. For personal use only. 1984
recovered hematopoietic function after immunosuppressive therapy, with recoverydefinedassubstantialimprovement in at least two lineages: decreased transfusion requirementsor transfusion-independence and/or elevation of the absolute neutrophil counts to greater than S X 107/L.12 The numbers of CD34' cells, CFC, and LTC-IC determinationsdifferbecausenot all assayscould be performed in every patient. For some patients, sequential measurements were performed at 3, 6 , and 12 months after treatment. BM cell prepararion. BM was obtained by aspiration from the posterioriliaccrestintosyringescontaining media supplemented I :10 with heparin (O'Neill and Feldman, St Louis, MO). informed consent was obtained according to a protocol approved by the Institutional Review Board of the National Heart, Lung, and Blood Institute. MononuclearBMand PB cells(BMMNCand PB MNC) were isolated by density gradient centrifugation using lymphocyte separationmedium(Organon,Durham, NC). Afterwashing in Hank's balanced salt solution (HBSS; Life Technologies, Gaithersburg, MD), cells were resuspended in Iscove'smodified Dulbecco's medium (IMDM; Life Technologies) supplemented with 20% fetal calf serum (FCS; Life Technologies). Flow cvtometric analysis. Phycoerythrin (PE)-conjugated HPCA-Imonoclonalantibody(MoAb)directedagainst human CD34 antigen (Becton Dickinson, Mountain View. CA) was used for quantitation of CD34' cells in PB and BM according to a modification of theproceduredescribed by Sutherlandetal." Briefly, 100 p L of PB was stained for 30 minutes with appropriate MoAb or isotypic control, lysed, and fixed using Q-prep reagent (Coulter, Hialeah, FL). Samples were analyzed using an Epics ELITE flow cytometer(Coulter).Dependingonthecellularity of thesample. variable numbers of cells were analyzed until 100 to 1,000 positive events were collected in a gate set according to the isotypic control. The frequency of CD34' cells was then calculated by dividing the number of CD34' cells by the total number of cells analyzed. The frequency of CD34' cells in BM was measured after isolation of BM MNC. BM cells were incubated with the appropriate directly conjugated MoAb and analyzed in a similar fashion as described for PB. The frequency of CD34' cells was calculated and expressed for BM per IO' MNCandfor PB per 10' MNC or per milliliter o f blood. Separation of CD34' cells. After washing with phosphate-buf2% human fered saline (PBS; Life Technologies) supplemented with albumin, cells were applied to an affinity column containing biotincoated beads and the CD34' cell fraction was eluted with PBS. An aliquot of eluted cells was stained with PE-conjugated anti-CD34 HPCA-2 MoAb (Becton Dickinson) to assess purity; usually, 70% to 90% of separated cells were CD34'. For higher purity preparations,cellswerefurtherfractionated.Column-purifiedcellswere stained with fluorescein isothiocyanate (F1TC)-labeled anti-CD34MoAb,washedwithPBS,andsorted by flow cytometry(Epics ELITE; Coulter); purity of 97% to 99% was obtained by this combined method. Hematopoietic cell culture. The numbers of hematopoietic CFC in BM and PB ofAA patients were measured in methylcellulose cultures under standard conditions. Freshly isolated BM cells were plated in methylcellulose (Stem Cell Technologies, Vancouver,British Columbia, Canada) in the presence of S0 ng/mL interleukin-3 (IL-3; Amgen, Thousand Oaks, CA), 20 ng/mL granulocyte-macrophagecolony-stimulatingfactor(GM-CSF), 50 ng/mLstemcell factor (SCF), and 2 U/mL erythropoietin (EPO; all Amgen). Total BM cells from normal volunteers were plated at a density of l X ID' andcellsfromAApatientsat 3 X IO5 cells in I mLculture medium in 35-mm dishes. CD34' cells were plated at a density of I X 1 0 ' cells/O.S mL methylcellulosein 48-well I l-mm plates.
MACIEJEWSKI ET A1
Colonies on duplicate plates were counted and the average number of CFC per IOc cells was then calculated. LTBMC for the determination of LTC-IC numbers was performed according to a modification of described methods.'"" BM MNC ( I O X IOb) were used to initiate stromal culture. Allogeneic stroma was grown to confluency. Culture medium consisted of stem cell media mol/L hydro(StemCellTechnologies)supplemented with 1 X cortisone 2 1 -hemisuccinate (Sigma, St Louis, MO) and was replaced weekly. After 3 weeks of culture, stromal cells were briefly treated with trypsin (LifeTechnologies),washed,and placed in 48-well plates. After reestablishment of a confluent cellular layer, the plates were irradiated (IS Cy of 2.50 kV x-rays) and used for the measurement of LTC-IC content in MNC fractions derived from the BM and PB of patients and healthy volunteers. To assure consistency i n all experiments, stroma from only five normal donors was used, and these preparations showed comparable feeder function as measured by the number of LTC-IC grown from the BM and PB of controls. At leasttwo(if cell numberswerelow)but more oftenthree or more cell concentrations were applied to preestablished, irradiated stomal feeder layers and cultured 33°C at for 5 weeks. Media changes wereperformedweekly.After 5 weeks, the adherentcells were harvested by treatment with trypsin, washed, and replated in duplicate in methylcellulose to estimate the numbersof cells able to form secondary colonies. When possible, to convert the numbers of secondary CFC cells after S weeks to numbers of LTC-IC, limiting dilution experiments were performed to determine the numbers of colonies derlved from a single LTC-IC.'q~'"-'7 Decreasing dilutions of BM and PB MNC or purified CD34' cells were cultured in 96-well flat-bottom plates containing preformed irradiated allogeneic stroma. At each dilution, 12 or 24 wellswereseeded.After 5 weeks of culture under the previously described conditions. cells from each well were treated with trypsin and transferred into methylcellulose supplemented with growth factors. After an additional 2 weeks, wells containing colonies were scored. The frequency of LTC-IC and the mean number of colonies per positive well (per LTC-IC; clonogenic capacity of individual LTC-IC) were calculated by determining the cell dilution that resulted in 537% negative wells, equivalent to single hit kinetics ( I LTC-IClwell) according to the Poisson distribution. The clonogenic capacity of a single LTC-IC also was calculated by dividing the numbers of colonies derived frombulk cultures by the frequency of LTC-IC. Because limiting dilution experiments were not possible for all patients, for cultures from all AA cases the number of clonogeniccellswasconvertedtoanabsolutenumber of LTC-IC by dividing the numbers of colonies bulk in cultures by the mean clonogenic capacity of LTC-IC measured in the limiting dilution experimentsandexpressedasLTC-ICper IO5 MNC.Forestimationof LTC-IC per milliliter of PB, the following equation was used: LTCIC/mL = White Blood Cells (WBC)/mL X 100/% of MNC X LTCIC/lOh. Stcrristical merlwd,i. Differences between groups were evaluated using the nonparametric Krusal-Wallis test or, if cells contained low numbers, by the Fisher one-tail and Iwo-tail tests. Correlation between variables was assessed using linear regression analysis.
RESULTS
Clonogenic capacity of LTC-IC ,from AA putierzts. T h e number o f LTC-IC ordinarilyiscalculated based on the number of secondary colonies formed and the clonogenic capacity of a single LTC-IC. LTC-IC can be formally measured by limiting dilution analysis.in which varying concentrations of cells are tested to define the concentration that achieves single hit kinetics by Poisson distribution. Using
From bloodjournal.hematologylibrary.org by guest on July 12, 2011. For personal use only. 1985
LTC-IC IN APLASTICANEMIA
i!m:m E2m: 8 40
1
2
3
4
3
1.5
0
4.5
1
l 0 0.5
2
1
.
-0 Fig 1. Limiting dilution analysis. Normal BM cells were serially diluted and placedon irradiated allogeneic stroma. The proportion of wells containing secondary colonies was assessed, and the absolute frequency of LTC-ICwas determinedat the dilution achieving a single cell per well (37% negative wells). The numberof secondary colonies f standard error of the meanlLTC-IC was1.27 f 0.09 for severe AA (N = 6),2.0 f 0.35 for moderate and recoveredAA IN = 51, and 4.3 f 1.0 for normal controls (N = 5). (BM LTC-IC are shown; similar results were obtained for PB LTC-IC.)
10
2.5 4 5
10
0
2.5
5
7.5
10
2.55~~~
0
200
75
15
30
0 0250.5
1
2
cell number x103
180
Fig 3. Relationship between the day 1 cell input of LTBMCand weeks of culture. the number of secondary colonies measured 5after (A through G) Individual AA patients’ BM and PB; (H) normal volunteers.
160 140
2 120 U >
L
{ 100 8
% 80 60 40
-
20
i 100
v
500
1000 5000 cell number
10000
Fig 2. Analysis of secondary clonogenic capacityafter LTBMC of CD34+tells obtainedfrom BM ofnormal volunteers andtwo patients with AA.
this technique, the ability of a single LTC-IC to give rise to multiple secondary colonies has been determined for normal individuals and is relatively constant. LTC-IC measurement can then be simplified by determination of the number of total secondary colonies generated in LTBMC, if the relationship between the number of initial cells (input) and the number of secondary colonies (output) is linear. We determined by limiting dilution analysis that, from normal donors under our culture conditions, approximately four secondary colonies were generated by a single LTCCIC (Fig 1). Limiting dilution analysis was applied to those AA patient samples that contained sufficient cells to perform the required multiple dilutions. The clonogenic potential of LTC-IC from new AA patients and from recovered cases was lower than normal (Fig 2). For other samples from AA patients, total cell numbers were too low to permit limiting dilution analysis. Instead, we approximated LTC-IC number using bulk cultures of patients’ PB and BM, enumeration of secondary 5-week CFC from a single concentration of input cells, and division of this number by the estimated clono-
12.5
From bloodjournal.hematologylibrary.org by guest on July 12, 2011. For personal use only. MACIEJEWSKI ET AL
1986 Table 1. LTC-IC,CFC, and CD34' Cells in AA BM CD34+
CFC
2,892 2 206 N = 50
Normal control
Severe AA Moderate AA Recovered AA
211 t 124 N = 22 52
230 t- 50* N = 41 854 t 418* N=8 888 t- 129*S N = 40
1.42*
Table 2. LTC-IC Number in Severe AA LTC-IC
1.95 t 0.5 N = 14
N = 15 27 -c l l t
0.19 t- 0.03* N = 17 0.29 t 0.13*
N=4 32 t- 9*S N = 17
N-7 0.33 ? 0.08** N = 25
All values expressed per lo5 MNC. LTC-IC numbers were derived from bulk cultureexperimentsandcalculatedusingfactorsdetermined in limiting dilution analysis: 4 for the clonogenic capacity of for Severe AA, and 2,3 for or recovered
AA. Statistical analysis was performed using the Kruskal-Wallis test. * P < ,131. normal control versus severe AA. moderate AA. and recovered AA. t P i.01, moderate AA versus severe AA. P < .05, recovered AA versus severe AA.
*
7000-
LTC-IC/105
On presentation N-8 Refractory to therapy
BM
PE
0.129 t 0.19
0.046 t- 0.018
N=8 0.11 t 0.2 N=6 0.69 t 0.4
Relapsed
N=3
0.07 2 0.04 N=9 0.006 i 0.002 N = 2
Values represent mean t standard error of the mean.
genic capacity. To ensure that the number of CFC measured was alinear function of cell input, culturesafter serialdilution were performed for 'pecimens containing sufficient numbers of cells. Not only was the output linear(Fig 31, but for mostcultures, duetothe lownumbers of secondary colonies formed, single hit kinetics were apparently achieved. In general, there was good agreement in patient specimens among (1) limiting dilution analysis to determine LTC-IC frequency; (2) secondary CFC output in bulk
0
6ooo0
8
0 Qx)
00 0
00
v 0
0
00
& b o oooo 1000
O
t
b 0 00 v
Fig 4. CD34+, primary CFC. and secondaryCFC measured in BM of AA patients and normal controls. Each dot represents a patient sample studied. Severe AA ( S U I includea patients on presentation with their disease, cases refractory to immunosuppressivetherapy,and patients who had relapsed after a period of recovery.Among the recovered AA (rAA) cases shown are (0) patients more than 2 years from treatment with antithymocyte globulin and (0) patients less than 2 yean from treatment with antithymocyte globulin. N, normal; mAA. moderate AA.
From bloodjournal.hematologylibrary.org by guest on July 12, 2011. For personal use only. 1987
LTC-IC IN APLASTICANEMIA
1
300-
250
-
U 200-
S
m
a
0
m
0
5 150-3
0" U
0
0
0 0
"8" 0
0
0
100- -b
8
0
00
0 0
8
0
0
0
0.001
Fig 5. CD34+cells, primary CFC, and secondary CFC measuredin PB of AA patients and normal controls. See legend to Fig 4.
LTBMC, divided by the estimated frequency of LTC-IC; and ( 3 ) the number of colonies observed in individual wells under single hit kinetic conditions. (This concordance suggested that we had not overestimated the clonogenic capacity of AA LTC-IC by performance of limiting dilution analysis on an unrepresentative sample ofBM specimens; inany event, our measured value of 1.27 differed little from the theoretical minimum of 1 .O). Frequency of CD34+ cells, CFC, and LTC-IC in AA BM. MNC from BM were assayed by flow cytometry for CD34+ cells and by colony culture methods for primary CFC and after LTBMC for LTC-IC (Fig 3 and Table l). As previously reported byus and other^,^.^ CD34+ cell
numbers were markedly decreased in almost all patients with acutesevere AA. In recovered patients, the mean CD34' cell number was higher, but the majority still showed very low values as compared with normal. Deficiency of primary hematopoieticcolony formation by aplastic BM, which is repeatedly documented in the literature, was observed in this series as well. Secondary colony formation after LTBMC, a measure of LTC-IC, was extremely low in all patients with acute AA. Average numbers of LTC-IC were relatively higher in patients with moderate disease or after hematologic recovery, but these values were still below the normal range in almost all cases (Table 2). Only about 20% of recovered AA cases
From bloodjournal.hematologylibrary.org by guest on July 12, 2011. For personal use only. MACIEJEWSKI ET AL
1988
Table 3. LTC-IC, CFC, and CD34' Cells in AA PB ~
LTC-IC
CD34
Per lo5 M N C
Normal control
Per rnL
323 2 30 N = 13 27 f lo* N = 18 f 38* 88903 N=9 64 t 12** N = 38
Severe AA Moderate AA Recovered AA
Per
5,867 f 602 551 i 133' f 250*t
1,567
?
260**
0.34
?
N
=
io5
0.04 13 0.055 ? 0.02' N = 19 0.11 i 0.05* N=8 0.14 f 0.04*t 1.88 N = 41
-
Per r n L
5.24 f 0.5 0.7 f 0.31* 1.7 i 0.81* f 0.23*t
See caption for Table 1. Statistical analysis by Kruskal-Wallis test. * P < .01, normal control versus severe AA, moderate AA, and recovered AA. t P < .05, recovered AA versus severe AA. P .01, recovered AA versus severe AA.
*
showed secondary colony formation within the normal range (mean ? 2 standard deviations [SDI; Fig 4). Frequency of CD34' cells, CFC, and LTC-IC AA PB. Because of the possibility of contamination of hypocellular BM specimens with variable amounts of PB, we also made similar measurements in blood from patients and normal controls (Fig 5 and Table 3). Low numbers of CD34+ cells and secondary CFC were measured in AA. As WBC counts are depressed in AA, using PB MNC as a denominator might lead to an overestimate of CD34' or LTC-IC numbers; by adjustment for the WBC number andthe proportion of MNC, we derived a concentration of CD34' and LTC-IC per unit volume of blood (Table 3). These values were similar. Although the total WBC count in patients was lower than normal, this factor was counterbalanced by the higher proportion of MNC in neutropenic individuals. Patients with AA do not have lymphocytosis, so the reduced numbers of CD34' cells and LTC-IC likely represents a frank deficit in hematopoietic cells, whether expressed as proportional to either MNC or per volume of blood. We compared CD34+ cell numbers and LTC-IC numbers obtained in BM and PB in patients and controls (Table 4). As anticipated for normals, both CD34+ cells and LTC-IC were more abundant by almost a factor of 10 in BM than in PB. A similar ratio was observed for CD34' cells in aplastic BM compared with PB, suggesting that dilution of the hypocellular BM specimens by PB had not occurred. The reduc-
Table 4. CD34+ and LTC-IC PB v BM
Normal control Severe AA Moderate AA Recovered AA
BM CD34'IPB CD34'
BM LTC-ICIPB LTC-IC
9.2 8.7 9.7 14.5
6.8 3.5 2.1 2.4
Calculated ratios are based on 14 BM and 13 PB samples for normals; 17 BM and 18 PB samples for severe AA patients; 7 BM and8 PB samplesfor moderateA A ; and 25 BM and38 PB samples in recovered patients. The ratios are derived from observed CD34' cells or LTC-IC determined/105 mononuclear cells of BM or PB.
tion in LTC-IC, although severe in both PB and BM, was modestly more pronounced for marrow. Frequency of CFC and LTC-IC in the CD34' compartment. To determine the proportion of CFC and stem cells within the hematopoietic cell compartment, we directly measured colony formation from purified CD34' cells in a few patients in whom adequate amounts of marrow was obtained (Table 5). For other cases in which samples were insufficiently cellular for sorting by fluorescent microflow cytometry, we calculated representation by division of the number of CFC or LTC-IC by CD34+ cell number (Table 5). For four patients with severe disease, LTC-IC determined in purified CD34' cells were measured at 1/653, 1/714, 1/833, and l / 1,000; these values were in the same range if slightly higher than the frequency of 1/1,250 when LTC-IC were determined by calculation. Using both methods, LTC-IC frequency in the CD34' compartment in AA was normal. In contrast, primary colony formation when directly measuredfrom sorted CD34' cells was consistently lower thannormalin severe AA, ranging from 1/71 to 1/250, compared with control frequencies greater than 1/25. Calculated frequencies from the current series also showed a marked defeciency in progenitors in severe disease, with improvement towards normal in patients with moderate AA or who hadhematologically recovered with treatment (Table 5 ) . Clinical correlations. We attempted to relate a number of clinical parameters to LTC-IC number. There was no relationship between LTC-IC number and the duration of disease before treatment (Table 6). For all patients (N = 48). LTC-IC did not correlate with absolute neutrophil count at the time of marrow or blood sampling; the correlation coefficient between these two variables (using BM LTC-IC) was .34. However, for all AA cases, LTC-IC were significantly lower in patients with more severe degrees of neutropenia (absolute neutrophil count
