Bone Marrow Transplantation, (1999) 23, 335–346  1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt Immune reconstitution following allogeneic peripheral stem cell transplants

S Shenoy1, T Mohanakumar2, G Todd3, W Westhoff2, K Dunnigan4, DR Adkins3, RA Brown3 and JF DiPersio3

Departments of 1Pediatric -Oncology (Division), 3Bone Marrow Transplantation and Stem Cell Biology (Division), 4Biostatistics, and 2Surgery and Pathology, St Louis Children’s Hospital, Washington University School of Medicine, St Louis, MO, USA

Summary: Keywords: immune reconstitution; PBSCT; cytokines; function; GVHD Growth factor-mobilized peripheral blood stem cells (PBSCs) engraft rapidly in myeloablated recipients compared to conventional BM, but this procedure also mobilizes mature and which Allogeneic bone marrow transplantation (BMT) is curative can impact immune reconstitution and GVHD. Hence, therapy for many malignant hematologic and genetic dis- we serially evaluated immune reconstitution and cyto- orders.1–3 Growth factor-mobilized allogeneic PBSCT, by kine expression in PBSCT recipients in the first year. providing more CD34+ stem cells compared to bone mar- Engraftment of neutrophils and monocytes stabilized row, results in an advantage for recipients with earlier neu- early but NK cells, B cells and CD4؉ T cell numbers trophil and recovery thus decreasing transfusion were significantly (P Ͻ 0.05) low with persistently requirements and duration of hospital stay.4–12 The pro- reversed CD4:CD8 ratios. NK function remained low longed intense supportive care when awaiting engraftment throughout the first year. The quantitative decrease in that is necessary following conventional BMT as well as -CD4؉ T cells resulted in significantly decreased prolifer- the morbidity and logistics of harvesting marrow from nor ation in response to mitogens and alloHLA antigens. mal donors under general anesthesia can be avoided by Yet, a qualitative analysis of T cell function measured using cytokine-mobilized allogeneic PBSC. -by Ca؉؉ influx after T cell activation with antiCD3 as The lymphocyte and composition of cytokine well as T-dependent polyclonal Ig secretion by mitogen- mobilized peripheral blood products as well as their cellular stimulated B cells was preserved even early post trans- function is different from bone marrow and resting periph- plant. TNF␣ mRNA was detected in almost all recipi- eral blood.13–16 As analyzed at our institution, compared to ents in the first year. IL-10 mRNA was detected in 77%, resting peripheral blood, monocytes are increased four-fold IL-2 in 22% and IFN␥ in 44% of recipients in the first in the peripheral blood of normal allogeneic donors mobil- 6 months. Only 30% expressed IL-10 in the second 6 ized with G-CSF.17 CD19+ B cells increase three-fold and months post transplant while expression of IL-2 and CD3+ T cells are increased two-fold. This increase is IFN␥ was detected in 38% and 46% respectively. reflected as proportionate increases in CD4+ and CD8+ T Thirty-seven percent of PBSCT recipients developed cell subsets thus maintaining a normal CD4:CD8 ratio.17,18 grades II–IV acute GVHD but 72% went on to develop In addition to increases in the absolute numbers of B and chronic extensive GVHD at a median of 120 days. Sixty- T cells, NK cells increase 150% from baseline after G-CSF two percent developed CMV viremia and 5.4% mobilization. Other recent studies have confirmed the developed overt CMV disease in the first year post observations that growth factor promotes the mobilization PBSCT. Lymphocyte engraftment is quantitatively of lymphocytes and NK cells in normal allogeneic donors. delayed but CD4 functions are preserved while NK Korbling et al described a 16-fold increase in CD3+ cells, numbers and function are compromised post PBSCT. a 13-fold increase in CD4+ T cells (increased CD45RO+ IL-10 expression decreases after the first 6 months post memory cells as well as ␥␦ cells has been described by transplant while TNF␣ is continually expressed. The Mills et al) and 27.4-fold increase in CD8+ T cells after balance between quantitative lymphocyte reconstitution peripheral blood growth factor mobilization compared to and qualitative lymphocyte functions as well as changes bone marrow harvests. An 11-fold increase in CD19+ B in lymphokine patterns may influence infection and cells and a 19.4-fold increase in CD56+ NK cells was also GVHD and thus the clinical outcome post PBSCT. noted.19 Furthermore, autologous stem cell transplantation results in significantly more rapid recovery of monocytes, NK cells and naive CD4+ T lymphocytes, which then trans- lates into early immune function recovery compared to Correspondence: Dr S Shenoy, St Louis Children’s Hospital, Room 1044, delayed reconstitution especially of naive CD4+ T lympho- Washington University School of Medicine, Division of Pediatric Hema- 20,21 tology-Oncology, One Children’s Place, St Louis, MO 63110, USA cytes in conventional BMT. Therefore, it is conceivable Received 21 January 1998; accepted 28 September 1998 that the infusion of stem cell products with such altered Immune reconstitution following PBSCT S Shenoy et al 336 numbers of lymphocytes, monocytes and NK cells in an procedures were performed once or twice to collect a allogeneic setting could affect post transplant immune mandatory threshold of donor CD34+ PBSCs (mean 9.96; reconstitution as well as the extent and severity of GVHD. range 1.5–32.4 ϫ 106 CD34+/LinϪ/kg).26 Unmanipulated Conversely, immune reconstitution which is crucial follow- allogeneic PBSCs were infused into allogeneic transplant ing transplantation is closely linked to the development recipients on day 0. Twenty-six recipients were conditioned and treatment of GVHD and influences infectious with total body radiation (TBI) 175 cGy × 6 doses, etopo- complications.22–25 side 30 mg/kg × 1 dose and cyclophosphamide 50 mg/kg We evaluated immune reconstitution in the first year in × 2, and four recipients with cyclophosphamide 60 mg/kg recipients following allogeneic PBSCT from HLA-matched × 2 and busulphan 1 mg/kg × 16 doses (due to inability to family donors for various resistant hematologic malig- receive TBI). Cyclosporine 3 mg/kg by continuous intra- nancies. Measurement of neutrophils, monocytes, B cell, venous infusion was commenced on day Ϫ1 in all recipi- NK cell and T cell subset engraftment and function were ents prior to PBSC infusion. Cyclosporine levels were performed serially for 1 year post transplant. Additionally, maintained at 250–350 ng/ml for 100 days, and gradually constitutive post-transplant lymphokine mRNA expression tapered or continued based on recipient GVHD status. Ster- patterns were determined in a subset of patients using oids (0.5 mg/kg methylprednisolone) were commenced on RT-PCR. day +7 post transplant, increased to 1 mg/kg on day +14 and tapered between days 28 and 56 in the absence of GVHD. No recipient received immunoglobulin infusions in Materials and methods the post-transplant period. White cell engraftment (ANC Ͼ500/␮l) occurred at a Recipient population demographics median of 9 days, (Ͼ20 000/␮l) at a median of 10 days. Eleven of 30 (37%) patients developed grades II–IV These transplants and studies were performed after infor- acute GVHD graded as described by Glucksberg et al;27 med consent was obtained under protocols approved by the 12% developed grade III–IV disease. Skin was the com- Human Studies Committee and Institutional Review Board monest site of significant involvement. Twenty-one of 30 of Washington University. Recipients with resistant hema- (72%) developed chronic GVHD at a median of 120 days tologic malignancies who received PBSCT from HLA- graded as per Shulman et al28 and included extensive skin, matched sibling donors were the subjects of this study. Ͼ gut and liver involvement which necessitated continuation Stable donor chimerism ( 95% donor cells) was confirmed of immunosuppressive therapy. Eight of 10 (80%) recipi- by serial VNTR analyses on bone marrow aspirates ents evaluated at 1 year post transplant continued to receive between days 50 and 120. Recipients with graft rejection, prophylaxis/treatment for GVHD with CsA and steroids. recurrent malignancy or severe post-transplant compli- Isolated cGVHD was uncommon. The actuarial risk of cations and organ failure requiring intensive care and life- chronic GVHD in survivors at day +180 was 92%. The support were excluded. Thirty patients (20 males and 10 development of acute or chronic GVHD did not correlate females; mean age 39 years; range 14–64 years) were with the numbers of CD34+, CD4+, CD8+, CD3+ or mono- evaluated at different time points post PBSCT. Indications nuclear cells/kg.29 A high incidence of clinical extensive for PBSCT as well as recipient demographics are summar- chronic GVHD (relative risk 2.37 at 2 years) has also been ized in Table 1. All recipients were heavily pretreated with reported in allogeneic PBSCT recipients by Storek et al.30 multiple cycles of standard chemotherapy with or without radiation using established protocols for their disease. No recipient underwent splenectomy prior to transplant. Since CMV infection recipients were treated with PBSCT for six hematologic CMV viremia and infection post PBSCT has been reported diseases, combination chemotherapy varied according to from our institution by DiPersio et al.31 Twenty-two of 30 the lymphoma or treated. Frontline standard ther- (74%) of PBSCT recipients were at risk for CMV infection apy was used in all these diseases and transplants were (seropositive donor 18% and/or recipient 56%). Nineteen undertaken for poor response and high risk of relapse. of 30 (62%) developed CMV viremia as measured by shell All donors were mobilized with 10 ␮g/kg/day of G-CSF vial cultures at a median of 34 days, 25% had second epi- for 4 days prior to leukapheresis. Twenty liter leukapheresis sodes and 8.3% had third episodes and 5.4% developed CMV disease. Table 1 Indications for PBSCT in patient population studied for immune reconstitution Cell preparation Indications for PBSCT No. of recipients PBMCs were separated from heparinized blood on Ficoll– Hypaque density gradients (Pharmacia Biotech, Uppsala, Acute myeloid leukemia 13 Acute lymphatic leukemia 1 Sweden). Mononuclear cells were washed twice with phos- Chronic myeloid leukemia (accelerated phase) 4 phate-buffered saline (PBS) (Sigma, St Louis, MO, USA) Refractory chronic lymphatic leukemia 2 and suspended in RPMI-1640 (Sigma) supplemented with Myelodysplastic syndrome (RAEB-t) 2 2mml-glutamine, 1 mm sodium pyruvate, 25 mm Hepes, Non-Hodgkin’s lymphoma 8 100 U/ml penicillin and 100 ␮g/ml streptomycin and 15% Total 30 fetal bovine serum (Biocell Laboratories, Rancho Doming- uez, CA, USA). Immune reconstitution following PBSCT S Shenoy et al 337 Normal control cells were obtained from donors prior to a 120 mobilization with growth factors and were used for all con- trol assays described. These cells were evaluated as normal n = 15 controls since immune reconstitution post PBSCT was a 100 reflection of the function of these cells following *P < 0.05 engraftment in PBSCT recipients whose intrinsic cells had 80 been effectively ablated. Since recipients were only evalu- PHA ated if they were a donor chimera and immunologic status Anti-CD3 was a donor-derived function, recipient immune functions 60 prior to transplant were not evaluated. It is assumed that n = 8 high-dose chemotherapy in recipients prior to transplant 40 ** n = 8 n = 18 would suppress immune functions and immune reconsti- n = 9

Proliferation (mean SI) ** tution with donor cells was necessary to normalize these ** ** functions again. 20

0 Lymphocyte subset analysis (immunofluorescence studies) Control 1–3 4–6 7–9 10–12 Time in months post PBSCT The total number of lymphocytes, neutrophils and mono- cytes were calculated from peripheral blood using a Coulter b 160 STKS Automated Hematology Analyzer. Lymphocyte sub- sets were estimated on peripheral blood samples. For lym- 6 phocyte subsets, four tubes containing 1 ϫ 10 nucleated Control + + + + 140 WBC were stained for CD3 , CD4 and CD8 , CD19 and Rec 1 CD56/16+ cells using the following antibodies: (1) IgG1- Rec 2 FITC/IgG1-PE/CD45-PE cy5 Imm (Immunotech, Marse- Rec 3 120 illes, France) No. 1673, (2) CD4-FITC/CD8-PE/CD3-PE Cy5 Imm No. 1650, (3) CD3-FITC/CD19-PE/CD45-PE Cy5 Imm No. 1671, (4) CD3-FITC Imm No. 1281, CD16- 100 PE Imm No. 1328, CD56-PE Imm No. 1324, and CD45- PE Cy5 Imm No. 1575. The cells were analyzed on a Coul- ter Elite flow cytometer (Miami, FL, USA) equipped with 80 an argon laser emitting a 488 nm beam. The lymphocyte population gated was identified by forward and side scatter characteristics. The percentages of lymphocyte subsets 60 obtained were multiplied by the number of absolute lym- Proliferation to PHA (SI) phocytes to determine subset numbers in the peripheral blood of premobilization normal donor controls and 40 allogeneic PBSCT recipients 1, 3, 6 and 12 months post transplant.

20 Proliferation assays 0 Normal control and patient PBMCs were suspended in cul- 1–3 4–6 7–9 10–12 ture medium at 1 ϫ 106 cells/ml, plated in triplicate in flat- bottomed 96-well microtiter trays at 105 cells per well and Months post PBSCT ° Mean background control c.p.m.: 926 cultured at 37 C with optimum concentrations of mitogens Mean background test c.p.m.: 234 (PHA 10 ␮g/ml or anti-CD3 150 ng/ml) for 72 h or a mix- ture of 2 ϫ 104 irradiated (50 Gy) allogeneic lymphoblas- Figure 1 (a) Proliferation of PBLs in response to PHA and anti-CD3 serially in the first year post PBSCT. Numbers tested in each group are toid cell lines (LCLs) (allogeneic MHC stimulators) for 5 indicated by n. Significance is indicated by *. (b) Proliferation to PHA days. PHA was obtained from Sigma and anti-CD3 from shown serially in three representative recipients at different time points Orthobiotech (Raritan, NJ, USA). 3H-thymidine (ICN, post PBSCT and compared to controls. Background c.p.m. are indicated Costa Mesa, CA, USA) at 1 ␮Ci/well was added during the at the bottom of figure. final 24 h of culture. Cells were harvested on to glass fiber filters and counted on a microbeta counter (Wallac, Gai- In vitro immunoglobulin (Ig) production by B cells thersburg, MD, USA). Results are expressed as a stimu- lation index (SI): the ratio of mean test c.p.m. to mean Twenty PBSCT recipients between 1 and 3, 4 and 6, 7 and background c.p.m. of unstimulated cells. The number of 9 and 10 and 12 months were evaluated for Ig secretion, recipients tested at each time point is indicated in Figures five at each time point. Five normal controls were simul- 1 and 2. taneously evaluated. One hundred thousand recipient Immune reconstitution following PBSCT S Shenoy et al 338 35 normal control PBMCs stimulated for 12 h with 10 ␮g/ml of PHA served as positive controls. Six unstimulated volun- 30 *P < 0.05 teer donor PBMCs as well as sterile water substituted for 25 ␤ n = 15 template cDNA were also used as controls. -actin was used as positive control for PCR reactions. Twenty micro- 20 liters of each PCR product were subjected to electro- 15 phoresis on a 2% agarose gel (Midwest Scientific, Valley Park, MO, USA) with ethidium bromide staining and blot- n = 18 10 n = 8 n = 5 ted on to nylon membranes (Genescreen Plus, NEN n = 5 5 Research Products, Boston, MA, USA). After pre-hybridiz- * 32 Mean SI to allogeneic LCLs ation for 2 h at 42°C, the blots were probed with P-lab- 0 eled specific oligonucleotide probes. Radiographs were Control 1–3 4–6 7–9 10–12 generated by exposure to X-ray film (Kodak X-OMAT Time in months post PBSCT XAR-5; Sigma) at Ϫ70°C for 48 h. Figure 2 Proliferation of recipient PBLs to allogeneic HLA antigens evaluated serially in the first year post transplant. n is the number of Intracellular calcium measurements patients. Five million normal control or recipient PBMCs were acti- vated from their normal peripheral resting state by incubat- PBMCs were cultured in 0.1 ml of medium containing 15% ing overnight in 5.0 ml culture medium with 10 ␮g/ml FCS with 10 ␮g/ml of pokeweed mitogen (PWM) in 96- PHA. Two million cells were used to measure the Ca++ well culture plates. Fifty microliters of supernatant was har- influx in each experiment. Cells were suspended in Hank’s vested after 6 days of culture at 37°C and Ig secretion was balanced salt solution (HBSS) without Ca2+ and Mg2+ con- determined using a sandwich ELISA technique.32 Super- taining 0.5% BSA (Sigma) and incubated for 30 min at natant was harvested at 6 days since standard protocols 37°C with 1 ␮m fura-2-acetomethyl ester (Molecular indicate that maximum polyclonal Ig secretion occurs at 6– Probes, Eugene, OR, USA). Intracellular calcium measure- 7 days following T-dependent B cell stimulation with ments were conducted using the method described by Nazi- PWM.33 Control PBLs were subjected to similar stimu- ruddin et al.35 Cells were washed twice with HBSS, resus- lation with PWM and harvested supernatant was also tested pended in 1.0 ml HBSS with 0.5% BSA and incubated for for total Ig secretion. Briefly, ELISA plates were coated 10 min with 5 ␮g anti-CD3 (Orthobiotech, Raritan, NJ, overnight with 0.1 ml/well of optimally diluted affinity pur- USA) or 5 ␮g isotype-matched negative control (MOPC) ified antibody to human Igs (Kirkegaard & Perry Labora- on ice. Fluorescence changes in HBSS containing Ca2+ 2+ tories, Gaithersburg, MD, USA). After blocking with 3% (CaCl2 140 mg/l) and Mg (MgCl2 97.7 mg/l) were moni- BSA (bovine serum albumin), appropriately diluted patient tored with a F-2000 Hitachi Fluorescence Spectrophoto- and control samples as well as the standard Ig dilutions meter (Danbury, CT, USA), using 340- and 380-nm exci- were incubated for 1 h at room temperature following the tation and 510-nm emission wavelengths. Goat anti-mouse addition of optimally diluted peroxidase-conjugated rabbit Ig as a cross-linking agent was added at 60 s from the start affinity purified anti-human Ig (Cappel, Durham, NC, of fluorescence measurement. Sixty seconds later, cells USA). Plates were washed with PBS-Tween buffer. were lysed with 1% Triton followed by calcium chelation Microwell peroxidase substrate (Kpl) was added to the with 0.1 m EDTA to end the reaction. Changes in intracellu- plates and optical density read out at 410 nm. Test and lar Ca2+i concentrations in the nm range were calculated control sample Ig concentrations were determined using from individual flux patterns as described by Grynkiewicz curves generated from standards. Irradiated control cells et al.36 Thirteen normal volunteers served as controls. produced Ͻ1 ng/ml of Ig. Group one consisted of 11 patients, 1–6 months post PBSCT and group two had six patients, 7–12 months post transplant as described in Figure 3. Cytokine mRNA expression by RT-PCR and Southern analysis NK-mediated cytolysis Constitutive mRNA expression in PBMCs was determined in two groups, nine recipients 0–6 months and 13 recipients K562 cells (NK cell targets) were labeled for 60 min with 7–12 months post PBSCT, using the methods of Sundare- 250 ␮Ci of 51Cr (ICN). After washing three times, 103 san et al.34 RNA was extracted from 5 to 10 ϫ 106 unstimu- K562 cell targets were plated in triplicate in 96-well plates lated PBMCs. Five micrograms of RNA were reverse tran- and incubated with 2.5 ϫ 104 PBMCs (E:T ratio 25:1) for scribed and amplification of cytokine-specific cDNA 6 h prior to harvest. 51Cr release was measured using a fragments was accomplished by 30 cycles of PCR, adding microbeta system (Wallac). Spontaneous lysis was calcu- 0.5 ␮l of cDNA dNTP mixture, DMSO, 10 × buffer lated from targets incubated in medium and maximum lysis (Geneamp PCR Buffer, Cetus, Norwalk, CT, USA), Taq from targets incubated with 1% Triton. Percent-specific DNA Polymerase (Promega, Madison, WI, USA), and the lysis was calculated using the following formula: experi- 5Ј- and 3Ј-specific primers for IL-2, IL-4, IL-10, TNF␣, mental release Ϫ spontaneous release divided by the IFN␥ and ␤-actin (control). Annealing temperature for ␤- maximum release Ϫ spontaneous release × 100%. Eight actin was 58°C and 54°C for other cytokines. mRNA from normal controls, 16 recipients between 1 and 3 months, Immune reconstitution following PBSCT S Shenoy et al 339 200 to rise. This increase accounted for the increase in total 180 lymphocyte numbers in the presence of persistently low numbers of CD4+ T cells. CD4+ T cells remained signifi- 160 Ͻ P = 0.24 cantly (P 0.05) lower than control levels even at 12 )

M 140 months. The CD4/CD8 T cell ratio was inverted at 1 month

(n and stayed inverted through the first 12 months of this ++ 120 analysis. CD19+ B cells gradually increased through the 100 P = 0.07 first year but did not reach control levels even at 12 months and remained significantly low. CD56+/16+ NK cell num- 80 bers were higher (1.8 ϫ 105/ml) (Table 2) at 1 month 60 (although less than control levels) but subsequently con-

Intracellular Ca tinued to decrease such that their numbers were signifi- 40 cantly less than 30% of pre-mobilization donor levels 20 (controls) at 12 months post transplant (0.7 vs 2.7 ϫ 105/ml) (P Ͻ 0.05). 0 Control 1–6 7–12 n = 11 n = 6 T cell proliferation remains impaired in the first year Time in months post PBSCT post PBSCT Figure 3 Intracellular calcium levels measured following stimulation T cell function was evaluated by proliferation in response to with anti-CD3 in PBSCT recipients. Calcium concentrations in control PHA, anti-CD3, and allogeneic MHC antigen stimulators. cells are on the left. Recipients were evaluated at 0–3 months, 4–6 months, 7– 9 months and 10–12 months post PBSCT and compared eight between 4 and 6 months, nine between 7 and to normal donor controls. Figure 1 shows proliferation of 9 months and eight between 10 and 12 months were PBMCs to PHA and anti-CD3 compared to proliferation in control PBLs. Proliferation in response to PHA and anti- evaluated for NK function. CD3 was significantly (P Ͻ 0.05) less than in controls at all time points tested. Similarly, the ability of recipient Statistical analysis PBMCs to proliferate in response to mixed allogeneic radi- Proliferation and NK assays were subjected to the one-way ated LCLs was 4–10 times lower than in control cells (Figure 2). The mean SI in normal controls was 13.26. In ANOVA test with Tukey’s simultaneous confidence inter- the study population, evaluated at 1–12 months post vals. Pair-wise differences in stimulation indices (SIs) between controls and recipients at different time points PBSCT, SIs ranged between 1.71 and 4.07. This difference in the stimulation index was significant (P Ͻ 0.05) at 1–3 were evaluated for each assay in addition to differences in months and 7–9 months. Sample numbers were too small control and recipient NK-mediated lysis. Significant obser- vations (P Ͻ 0.05) are indicated by asterisks on each figure. to comment upon in the other two groups evaluated at 4– 6 months and 10–12 months although the stimulation indi- Lymphocyte subset numbers as well as immunoglobulin ces were similar. Overall, proliferation of engrafted T cells assays and intracellular calcium concentrations were com- pared with control values and analyzed for significance to mitogens and allogeneic MHC was impaired at all time points in the first post transplant year, with no demonstrable using the t-test. recovery of function during the interval tested. Serial evalu- ation of proliferation to PHA in three index PBSCT recipi- ents to demonstrate the lack of recovery of proliferative Results function is shown in Figure 1a. Engraftment of CD4+ T cells, NK and B cells remains low through the first year Intracellular calcium measurements following recipient T cell activation with anti-CD3 is comparable to control Lymphocyte engraftment was measured by immunofluo- cells rescence studies on peripheral blood at 1, 3, 6 and 12 months and is presented in Table 2. Premobilization mean Since T cell proliferation assays were markedly depressed, lymphocyte and lymphocyte subset numbers in normal engrafted cells were evaluated for their ability to respond donors are considered as control values. CD3+ cells to stimulation with anti-CD3 with subsequent increase in approached normal levels between 9 and 12 months prim- intracellular calcium levels. Calcium concentrations calcu- arily due to an increase in CD8+ T cell numbers. Monocyte lated by fluorescence changes after 12 h activation with numbers remained stable post transplant although an initial PHA and cross-linking with anti-CD3 are shown in Figure rise in was noted. The general trend was for 3 and compared with levels measured in similar numbers all subsets to gradually increase with the exception of NK of normal cells (2 ϫ 106). The intracellular calcium concen- cells. However, except for CD8+ T cells which increased tration was calculated from baseline and peak levels of above mean control values by 6 months, other lymphocyte response following cross-linking. Mean and standard devi- subset numbers never reached normal range. CD8+ T cells ations were calculated from each group and significant dif- increased above control levels by 3 months and continued ferences evaluated. Intracellular calcium levels tended to Immune reconstitution following PBSCT S Shenoy et al 340 Table 2 WBC, neutrophils, monocyte and lymphocyte subset numbers in peripheral blood of normal controls (donors prior to mobilization) and PBSCT recipients measured serially in the first year

Lymphocyte subsets Donor control Peripheral blood levels post PBSCT mean × 105/ml (s.d.)

1 month 3 months 6 months 12 months

Total lymphocytes 21.2 (5.3) 5.7 (5.2) 12.9 (10.6) 12.1 (6.3) 20.9 (15.6) CD3+ 15.0 (3.7) 4.5 (3.8)a 9.1 (9.7) 9.8 (6.0) 15.6 (12.3) CD3+4+8Ϫ 9.5 (3.1) 1.5 (1.5)a 2.3 (1.9)a 2.8 (1.5)a 4.8 (2.5)a CD3+4Ϫ8+ 5.3 (2.2) 2.7 (2.2)a 6.2 (7.6) 6.5 (5.0) 10.3 (9.4) CD19+ 8.7 (5.5) 0.3 (0.4)a 0.9 (2.9)a 0.9 (1.0)a 3.3 (3.3)a CD16+/56+ 2.7 (2.4) 1.8 (1.6) 1.5 (1.3)a 1.0 (0.9)a 0.7 (0.7)a Monocyte 5.7 (2.0) 5.3 (3.9) 5.7 (4.0) 4.2 (2.3) 6.6 (2.3) Neutrophils 44.1 (15.&0 88.6 (50.9) 49.1 (29.8) 40.3 (23.6) 45.0 (15.1) WBC 70.1 (18.5) 101 (54.9) 68.9 (31.1) 57.7 (26.9) 74.4 (22.2)

Standard deviations are indicated in parenthesis. aLevels significantly lower (P Ͻ 0.05) at different times post PBSCT compared to control donor premobilization levels.

decrease when evaluated at 6–12 months post PBSCT but measured between 9 and 12 months post PBSCT but the the decrease was not statistically significant (P = 0.07). difference was not significant (P = 0.07). Thus activated engrafted cells were capable of responding to stimulation by increasing intracellular calcium concen- trations and showed no major changes from baseline NK cell-mediated lysis of target cells is impaired control levels. significantly in the first 12 months post PBSCT The ability of engrafted NK cells to lyse susceptible targets Immunoglobulin production by B cells following in vitro (K562) was measured as an assay of NK cell function. NK stimulation recovers early following PBSCT assays are shown as percent lysis of K562 targets by PBMCs at E:T ratios of 25:1 (Figure 5). Mean control lysis The functional ability of engrafted B cells to secrete Ig on was 26.72%. Mean NK-mediated lysis was 4.82% at 3 stimulation with PWM was measured by harvesting culture months, 7.1% at 6 months, 12.48% at 9 months and 7.08% ␮ supernatants of PBMCs incubated for 6 days with 10 g/ml at 12 months which was significantly (P Ͻ 0.05) less than of PWM at 1–3, 4–6, 7–9 and 10–12 months post transplant control levels at all time points as indicated in Figure 5. (Figure 4). No difference was noticed in total Ig production Although minimum improvement was seen after 3 months, in the recipients studied (range 20.6–193.3 ng/ml) when NK function continued to remain significantly low through- compared with six controls (31.5–90.6 ng/ml). Immuno- out the first year. Three recipients are shown serially fol- globulin levels tended to be higher than control levels when lowed in Figure 5a to demonstrate the lack of recovery of NK cell function. 250 Constitutive cytokine mRNA expression patterns by P = 0.07 peripheral blood mononuclear cells changed during the 200 first year Since different cytokines are associated with acute and chronic GVHD, resting recipient PBMCs were analyzed for 150 ␥ ␣ P = 0.36 P = 0.42 expression of IL-2, IL-4, IL-10, IFN and TNF mRNA by RT-PCR analysis (Figure 6). Nine recipients were evalu- P = 0.24 ated 1–6 months post transplant. Of these, seven (77%) 100 strongly expressed IL-10 and three expressed IL-4. Three of the nine demonstrated weak expression of IL-2 and four

Immunoglobulin ng/ml others showed a weak message for IFN␥. TNF␣ was 50 strongly expressed by all recipients. Of the 13 recipients tested 7–12 months post transplant, six strongly expressed IFN␥ and five expressed IL-2. Only four recipients (30%) 0 from this group expressed IL-10 and three expressed IL-4. Control 0–3 4–6 7–9 10–12 All recipients but one continued to express TNF␣ (Table Time in months post PBSCT 3). When recipient PBLs were stimulated for 12 h with Figure 4 Serial measurement in PBSCT recipients of total immunoglob- PHA, all cytokines were strongly upregulated and easily ulin levels in ng/ml following in vitro stimulation with PWM compared detectable in both groups supporting appropriate responses with normal controls. Number of samples tested at each time point (n) = 5. to stimulation. Positive controls were normal PBLs stimu- Immune reconstitution following PBSCT S Shenoy et al 341 a 40 PHA C

35 IL-10 0.35 kb *P < 0.05 30

25 IL-4 0.45 kb K562 targets 20 * * TNF-α 0.47 kb 15 * * % specific lysis 10

5 IL-2 0.48 kb

0 Control 1–3 4–6 7–9 10–12 IFN-γ 0.54 kb n = 8 n = 16 n = 8 n = 9 n = 8 Time in months post PBSCT

β -actin 0.66 kb b 30 ab c Control Rec 1 Figure 6 Representative radiographs showing the presence or absence Rec 2 of cytokines in PBSCT recipients. (a) C: unstimulated control cells; PHA: control cells stimulated for 12 h with 10 ␮g/ml of PHA; (b) 0–6 months 25 Rec 3 and (c) 7–12 months post PBSCT. ␤-actin is present in all samples.

Table 3 Constitutive cytokine expression in engrafted PBLs detected by RT-PCR 20

Cytokine 1–6 months 7–12 months Resting post transplant post transplant PBMC PBSCT PBSCT control 15 (n = 9) (n = 13) (n = 6)

IL-2 2 (22)a 5 (38) 1 (17) IL-4 3 (33) 3 (23) 0 IL-10 7 (77) 4 (30) 0 10 TNF␣ 9 (100) 12 (92) 0 IFN␥ 4 (44) 6 (46) 0 % specific lysis (E:T 25:1) PHA stimulated normal cells were positive controls. Sterile water substi- tuted for template in PCR reactions were negative controls. 5 n = number of recipients tested. aNumber positive (% positive).

Discussion 0 1–3 4–6 7–9 10–12 The use of cytokine-mobilized peripheral blood as a source of stem cells in the autologous transplant setting has been superior to BMT in achieving early engraftment of neutro- –5 phils and platelets thus decreasing the duration of hospi- Time in months post PBSCT talization and utilization in the post-trans- Figure 5 (a) NK-mediated lysis evaluated serially post PBSCT. Lysis plant period. Such transplants which were predominantly mediated by control cells is indicated on the left. The E:T ratio is 25:1. undertaken as autologous rescue following high-dose (b) Serial evaluation of NK-mediated lysis in three PBSCT recipients com- pared with control lysis by normal PBMCs. chemotherapy proved to be safe and effective. In light of the ease, efficacy and cost saving associated with PBSCT, growth factors (G-CSF and GM-CSF) have now been used to mobilize stem cells from peripheral blood of normal allo- lated for 12 h with PHA (10 ␮g/ml) following which all geneic transplant donors.19,37–39 cytokines tested were strongly expressed. Five of six nor- Following conventional allogeneic BMT, 40–85% of mal unstimulated control PBMC samples failed to show the recipients develop significant acute GVHD depending on presence of any of the cytokines, one was positive for the the level of HLA identity. The development of GVHD is presence of IL-2. All were positive for ␤-actin which was dependent on the immune and cytokine profiles in the used as a positive control for all assays (Figure 6). recipient.27,39 Immune reconstitution is also an important Immune reconstitution following PBSCT S Shenoy et al 342 component of a successful allogeneic transplant, since viral, different time points tested and hence did not seem to be fungal and bacterial infections contribute significantly to a function of the number of engrafted cells. This is in con- morbidity and mortality in recipients. Immune reconsti- trast to conventional BMT recipients who recover NK cell tution has been carefully evaluated in recipients of allo- numbers as well as function as early as 1–3 months post geneic bone marrow with and without T depletion.41–45 transplant.47,52 This is of clinical concern since NK cells are Quantitatively, cellular components of the immune system cytolytic to virus infected cells and show cytolytic activity return to normal levels in 4–6 months following conven- against leukemic cell targets.53–57 As a result, the risk of tional BMT with the exception of CD4+ T cell and B cell viral infections and disease relapse may vary in allogeneic engraftment which is often delayed.42,43 Naive CD45RA PBSCT recipients compared to those patients undergoing cell numbers are depressed following BMT for over a year, BMT. with early recovery of memory CD45RO+ cells suggesting The pattern of engraftment of WBCs reflected increased engraftment of mature cells from the donor/recipient rather numbers of granulocytes in the first few months post than lymphocytes arising from maturing stem cells.20,21 PBSCT. The number of monocytes was stable throughout Qualitatively, NK and LAK functions have been reported the first year thus maintaining the same lymphocyte to to normalize very early in the first 2–3 months.45–47 T cell monocyte ratio and unlikely influenced changes in proliferation as well as immunoglobulin production remains lymphocyte functions. impaired usually until the second half of the first year post We found that the pattern of engraftment of lymphocyte transplant. Yamagami et al48 further characterized T cell subsets was similar to that reported in traditional BMT immune defects following conventional BMT and found recipients.45 There was persistent reversal of the CD4:CD8 that 40% of recipients had blunted calcium flux responses ratio. CD8+ T cells increased to above baseline levels at 3 after CD3 stimulation and was able to attribute this pre- months whereas CD4+ T cell numbers remained low (4.8 dominantly to decreased numbers of responding cells. This at 12 months vs 9.5 ϫ 105/ml in normal premobilization immune deficient picture is associated with a reversed donor controls). The higher numbers of CD8+ T cells CD4:CD8 T cell ratio in the first year. However, immune resulted in mean lymphocyte counts approaching normal functions return to baseline levels at about 12 months in range at 10–12 months post PBSCT. In spite of the mobiliz- the absence of ongoing GVHD.45 Not surprisingly, mature ation and infusion of mature lymphocyte subsets with a T lymphocyte depletion of allogeneic bone marrow further normal CD4:CD8 ratio,17 CD8 T cells seemed to preferen- compromises T cell engraftment and function and is asso- tially engraft in the first year post transplant, similar to that ciated with additional delay in immune recovery.41,45 described in recipients following autologous PBSCT.15 It There is limited knowledge on immune functions when is unclear whether preferential engraftment of CD8+ T cells peripheral blood cells are used as a stem cell source in is favored by the post-transplant milieu in recipients. Tal- an allogeneic setting following cytokine mobilization. As madge et al15 when evaluating recipients of autologous previously mentioned, the mobilized product varies vastly stem cell transplants found early NK cell engraftment but from harvested bone marrow as well as normal resting per- increased levels of ‘irradiation resistant suppressor cell’ ipheral blood in numbers and function. Mills et al,16 in activity as well as increased numbers of TCR ␣␤ CD4Ϫ, addition to reporting increased numbers of T (bearing the CD8Ϫ natural suppressor cells post autologous PBSCT. We ␣␤ and ␥␦ receptor) cells, NK cells and CD45RO+ memory failed to identify cells with this ‘suppressor’ phenotype in cells found that cytokine-mobilized T and B cell products our recipient population but could demonstrate a lag in had significantly depressed proliferation and that NK- mitogenesis similar to the above study. It is possible that mediated lysis was not increased. G-CSF-mobilized mono- the administration of steroids in our recipient population as cytes have been shown to impair and inhibit T cell function GVHD prophylaxis resulted in depleting these cells. Per- in PBSC products and this phenomenon has been attributed sistently low T and B lymphocyte numbers compared to to T cell apoptosis achieved by cell to cell contact as well normal controls in the first year is apparent from the analy- as the elaboration of IL-10 and TNF␣ by monocytes.49,50 sis described by Beelen et al58 although in contrast to con- Furthermore, G-CSF-mobilized monocytes have been ventional BMT, CD3+CD4+CD45RA cells seem to engraft shown to inhibit the induction of co-stimulatory molecules better along with CD3+CD4+CD45RO cells following auto- like the CD28-responsive complex thus compromising logous PBSCT.20,58–60 CD4+ T cell function.51 Since little is known about immune We found that functional T cell recovery when evaluated function reconstitution following allogeneic PBSCT with by response to standard mitogens (PHA and anti-CD3) as the infusion of 10–50 times more lymphocytes and four- well as allogeneic HLA antigens presented by mixed lym- fold more monocytes compared to bone marrow, we evalu- phoblastoid cell stimulators in primary proliferation assays ated lymphoid engraftment, immune reconstitution and was significantly lower than in normal controls. There was cytokine expression in allogeneic PBSCT recipients during no evidence of recovery in the first year reflecting the per- the first 12 months following transplantation. sistently low numbers of CD4+ T cells. Defective endogen- In our study, NK cell engraftment and function were ous IL-2 production by engrafted lymphocytes has also deficient throughout the first year post allogeneic PBSCT. been described as a mechanism that downregulates prolifer- NK cell numbers were comparatively higher in the first ation to allogeneic antigens in BMT recipients especially month (although numbers never reached control levels) but in the first 6 months post transplant.61,62 Other studies sug- subsequently decreased. Functional NK activity remained gest that the addition of exogenous IL-2 or anti-CD3 significantly depressed and showed no evidence of recovery restores this deficient proliferative response.63 We were in the period assessed. NK function did not vary at the unable to demonstrate a proliferative response to exogenous Immune reconstitution following PBSCT S Shenoy et al 343 anti-CD3 or induce a significant PHA response throughout Recipient PBMCs constitutively expressed predomi- the first year in PBSCT recipients. Keever et al45 described nantly IL-10 and TNF␣ in the first 6 months post allogeneic brisk proliferation early in the first few months post trans- PBSCT. This pattern changed to markedly decreased plant and associated this with the infusion of mature T cells expression of IL-10 in the second half of the first year with from the donor. In spite of the transfer of significantly continued expression of TNF␣. TNF␣ expression is likely higher numbers of functionally competent peripheral T cells related to the significantly increased numbers of donor we were unable to observe such initial enhanced T cell monocytes infused although peripheral blood levels did not responses. Antigen presentation by /monocytes reflect these monocyte numbers (Table 2). Mobilized donor does not seem to be the basis for these immune defects as cells tend to express more Th2 cytokines (S Shenoy, manu- bypassing antigen presentation still results in persistently script in preparation), similar to murine PBMCs mobilized low T cell proliferation. A possible mechanism could be by G-CSF in vivo).67 The presence of these cells may polar- inhibition of T cell responses by the increased monocytes ize recipients towards Th2 cytokines in the early post-trans- present in the infused product50 or inhibitory cytokines pro- plant period. On the other hand, Th2 may predominate sec- duced by engrafted cells. The presence of acute and/or ondary to a paucity of Th1 or due to signal transduction chronic GVHD also resulted in the continued use of deficiencies in such cells.68 Th1 cytokines IL-2 and IFN␥ immune suppression in symptomatic recipients. Both these are associated with the development of GVHD69 and propa- factors, as well as the presence of CMV infection, have ␣ + gated by inflammatory cytokines like TNF . Th2 cytokines been associated with delayed CD4 T and B cell have been associated with protection from GVHD engraftment and quantitative differences can influence especially in animal BMT models.70 Consistent with the 64 proliferative responses. observed patterns of PBMC cytokine expression and a late Conventional BMT as well as T cell-depleted BMT decrease in IL-10 expression, a large percent of recipients recipients have demonstrated an inability to respond to anti- developed chronic GVHD.29 IL-10 can also inhibit TNF␣ 41 CD3 stimulation with increased intracellular calcium. The synthesis by monocytes.71 The impact of persistent quantitative decrease in response following BMT has been expression of the inflammatory cytokine TNF␣ on acute related to low numbers of specific T cell subsets and low + 48 and chronic GVHD and the influence of other cytokines numbers of responding cells, especially CD4 T cells. B needs further evaluation post PBSCT. cells have been shown to respond as well as controls.65 Our + In keeping with the profoundly compromised T cell and patient population had significantly low B cell and CD4 NK cell function, 62% of PBSCT recipients developed T cell numbers. Yet, analysis of calcium responses follow- CMV viremia in the first year. Some recipients developed ing CD3 cross-linking in PBSCT patients revealed normal multiple viremic episodes and 5.4% developed CMV dis- intracellular calcium flux indicating that TCR-mediated sig- ease. Immune reconstitution and cytokine profiles in nalling was intact. Nonspecific responses to anti-CD3 addition to other variables like conditioning and GVHD through Fc receptor cross-linking by non-T cells was ruled prophylaxis and treatment are likely to influence the devel- out by simultaneous analysis using MOPC antibody (data opment of such infectious complications and their signifi- not shown) which showed no increase in intracellular cal- cance needs to be further evaluated with fairly large cium. Our results suggest that even though lymphocyte 2+ randomized trials. numbers are low, when evaluated by intracellular Ca con- 46 centration changes post stimulation, engrafted lymphocytes In contrast to our findings, Ottinger et al described were qualitatively normal. early recovery of NK cells and function as well as normal B cell numbers (CD19+ cells) remained low (less than engraftment of lymphocyte subsets (normal CD4:CD8 50% of control levels) through the first year, yet total ratios) by 6 months post PBSCT. T and B cells demon- immunoglobulin synthesis was comparable at all time strated normal proliferation to PHA and PWM, respect- points measured to those levels in control B cells following ively. Adequate proliferative responses to candida and teta- stimulation with PWM at least at 6 days post stimulation. nus were also present by 4–6 months. It is unclear whether Since conventional protocols recommend this standard per- this difference in immune recovery is related solely to dif- iod of incubation with PWM, stimulation was not continued ferences in the incidence of GVHD and concurrent immune beyond this time period. Hence it is unclear whether a suppression. Myeloablation protocols can influence the longer period of stimulation would have demonstrated a development of GVHD and may influence immune recon- difference in Ig synthesis between normal controls and stitution.72 It is unlikely that this difference is age or disease PBSCT recipients. Higher Ig levels than controls were related. Our finding of delayed immune reconstitution as noticed at one year. Previous studies have demonstrated well as changes in cytokine profiles matched clinical that following conventional BMT, high IgM responses are GVHD and CMV infection patterns in our recipients. present in the first year especially with the development of PBSCT provides an important source of stem cells that chronic GVHD.66 It is possible that a similar IgM increase has advantages for both donor and recipient. However, may be present following PBSCT. This was not separately immune reconstitution was quantitatively delayed in some measured in our assay system. Immunoglobulin secretion areas following allogeneic PBSCT but was qualitatively in response to PWM stimulation is CD4+ T cell dependent unimpaired in the group of patients studied. This may pro- and again demonstrates the functional competence of these vide a clinical difference in GVHD patterns and suscepti- cells in spite of low numbers. This method of analysis does bility to infectious pathogens compared to conventional not evaluate specific antibody-mediated B cell responses BMT. These issues should be explored in patients receiving which may be deficient post PBSCT. allogeneic PBSCT as their stem cell source. Immune reconstitution following PBSCT S Shenoy et al 344 Acknowledgements CSF mobilization on lymphocyte subsets, monocytes, NK cells, RBCs, platelets and CD34+/LinϪ progenitors in normal We would like to thank Ms Billie Glasscock for her secretarial allogeneic PBSC donors. Blood 1996; 88: 679a. assistance in preparing this manuscript. This work was supported 18 Hassan HT, Stockschlader M, Schleimer B et al. Comparison by NIH DK32253(TM), ACS-IRG(SS) and NIHKO8 of the content and subpopoulations of CD3 and CD34 positive CA72587(SS). cells in bone marrow harvests and G-CSF-mobilized periph- eral blood leukapheresis products from healthy adult donors. Transpl Immunol 1996; 4: 319–323. 19 Korbling M, Huh YO, Durett A et al. Allogeneic blood stem References cell transplantation: peripheralization and yield of donor- derived primitive hematopoietic progenitor cells (CD34+ Thy- 1 Thomas ED, Storb R, Clift RA et al. Bone marrow transplan- 1dim) and lymphoid subsets, and possible predictors of tation. New Engl J Med 1975; 292: 832–843. engraftment and graft-versus-host disease. Blood 1995; 86: 2 Storb R, Thomas ED, Weiden PL et al. Aplastic anemia 2842–2848. treated by allogeneic bone marrow transplantation: a report on 20 Talmadge JE, Reed E, Ino K et al. Rapid immunologic recon- 49 new cases from Seattle. Blood 1976; 48: 817–841. stitution following transplantation with mobilized peripheral 3 Thomas ED, Buckner CD, Banaji M et al. One hundred blood stem cells as compared to bone marrow. Bone Marrow patients with acute leukemia treated by chemotherapy, total Transplant 1997; 19: 161–172. body irradiation, and allogeneic marrow transplantation. Blood 21 Storek J, Witherspoon RP, Storb R. T cell reconstitution after 1977; 49: 511–533. bone marrow transplantation into adult patients does not 4 Henon PR, Liang H, Beck-Wirth G et al. Comparison of resemble T cell development in early life. Bone Marrow haemopoietic and immune recovery after autologous bone Transplant 1995; 16: 413–425. marrow or blood stem cell transplants. Bone Marrow Trans- 22 Atkinson K, Storb R, Prentice RL et al. Analysis of late infec- plant 1997; 9: 285–291. tions in 89 long term survivors of bone marrow transplan- 5 Kiesel S, Pezzutto A, Korbling M et al. Autologous peripheral tation. Blood 1979; 53: 720–731. blood stem cell transplantation: analysis of autografted cells 23 Meyers J. Infection in bone marrow transplant recipients. Am and lymphocyte recovery. Transpl Proc 1989; 21: 3084–3088. J Med 1986; 81: 27. 6 Appelbaum FR. Hemopoietic reconstitution following auto- 24 Paulin T, Ringden O, Nilsson B et al. Variables predicting logous bone marrow and peripheral blood mononuclear cell bacterial and fungal infections after allogeneic marrow infusion. Exp Hematol 1979; 7: 7–11. engraftment. Transplantation 1987; 43: 393. 7 Bensinger WI, Weaver CH, Appelbaum FR et al. Transplan- 25 Wingard JR. Infections in allogeneic bone marrow transplant tation of allogeneic peripheral blood stem cells mobilized by recipients. Semin Oncol 1993; 20: 80–87. recombinant human colony stimulating factor. 26 Brown RA, Adkins D, Goodnough LT et al. Factors which Blood 1995; 85: 1655–1658. influence the collection and engraftment of allogeneic periph- 8 Korbling M, Przepiorka D, Huh YO et al. Allogeneic blood eral blood stem cells. J Clin Oncol 1997; 15: 3067–3074. stem cell transplantation for refractory leukemia and lym- 27 Glucksberg H, Storb R, Fefer A et al. Clinical manifestations phoma: potential advantage of blood over marrow graft. Blood of graft-versus-host disease in human recipients of marrow 1995; 85: 1659–1665. from HLA-matched sibling donors. Transplantation 1974; 18: 9 Schmitz N, Dreger P, Suttorp M et al. Primary transplantation 295–304. of allogeneic peripheral blood stem cells mobilized by filgras- 28 Shulman H, Sullivan K, Weiden P. Chronic graft-versus-host tim (granulocyte colony stimulating factor). Blood 1995; 85: syndrome in man. Am J Med 1980; 69: 204–217. 1666–1672. 29 Brown RA, Adkins D, Goodnough LT et al. Graft-vs-host dis- 10 Russel N, Gratwohl A, Schmitz N. The place of blood stem ease and transplant related mortality after allogeneic periph- cells in allogeneic transplantation. Br J Haematol 1996; 93: eral blood stem cell transplantation for resistant hematologic 747–753. malignancy. J Clin Oncol (in press). 11 Goldman J. Peripheral blood stem cells for allografting. Blood 30 Storek J, Gooley T, Siadak M et al. Allogeneic peripheral 1995; 85: 1413–1415. blood stem cell transplantation may be associated with a high 12 Kessinger A, Armitage J, Bierman P et al. Clinical outcome risk of chronic graft-versus-host disease. Blood 1997; 90: of peripheral blood stem cell support. Med Oncol 1994; 11: 4705–4709. 43–46. 31 DiPersio J, Martin B, Abboud CN et al. Increased rates of 13 Tjonnfjord GE, Steen R, Evensen SA et al. Characterization CMV viremia in recipients of peripheral blood stem cells of CD34+ peripheral blood cells from healthy adults mobilized (PBSC) allogeneic transplants: rationale for both prophylactic by recombinant human granulocyte colony-stimulating factor. therapy and short course single daily dosing of gancyclovir Blood 1994; 84: 2795–2801. for 21 days. Blood 1996; 88: 424a. 14 Weaver CH, Longin K, Buckner CD, Bensinger W. Lympho- 32 Barry TS, Jones DM, Richter CB, Haynes BF. Successful cyte content in peripheral blood mononuclear cells collected engraftment of human postnatal thymus in severe combined after administration of recombinant granulocyte colony stimu- deficient (SCID) mice: differential engraftment of thymic lating factor. Bone Marrow Transplant 1994; 13: 411. components with irradiation versus anti-asialo GM-1 immuno- 15 Talmadge JE, Reed EC, Kessinger A et al. Immunologic attri- suppressive regimens. J Exp Med 1991; 173: 167. butes of cytokine mobilized peripheral blood stem cells and 33 Current Protocols in Immunology 1994; vol. 1, 7.11.1–7.12.6. recovery following transplantation. Bone Marrow Transplant 34 Sundaresan RS, Alevy YG, Steward N et al. Cytokine gene 1996; 17: 101–109. transcripts for tumor necrosis factor-alpha, interleukin-2, and 16 Mills KC, Gross TG, Varney ML et al. Immunologic pheno- interferon-gamma in human pulmonary allografts. J Heart type and function in human bone marrow, blood stem cells Lung Transpl 1995; 14: 512–518. and umbilical cord blood. Bone Marrow Transplant 1996; 18: 35 Naziruddin B, Duffy B, Tucker J, Mohanakumar T. Evidence 53–61. for cross-regulation of Fc-gamma-RIIIB (CD16) receptor- 17 Todd G, Hang JS, Brown R, DiPersio JF. The effect of G- mediated signaling by Fc-gamma-RII (CD32) expressed on Immune reconstitution following PBSCT S Shenoy et al 345 polymorphonuclear neutrophils. J Immunol 1992; 149: DNA fragmentation and apoptosis in allogeneic target cells. 3702–3709. Cell 1994; 76: 977–987. 36 Grynkiewicz G, Poenie M, Tsien RY. A new generation of 54 Kagi D, Ledermann B, Burki K et al. Cytotoxicity mediated Ca2+ indicators with greatly improved fluorescein properties. by T cells and natural killer cells is greatly impaired in per- J Biol Chem 1985; 260: 3440–3450. forin deficient mice. Nature 1994; 369: 31–37. 37 Rosenfeld C, Collins R, Pineiro L et al. Allogeneic blood cell 55 Arase H, Arase N, Saito T. Fas-mediated cytotoxicity by transplantation without posttransplant colony-stimulating fac- freshly isolated natural killer cells. J Exp Med 1995; 181: tors in patients with hematopoietic neoplasm: a phase II study. 1235–1238. J Clin Oncol 1996; 14: 1314–1319. 56 Henkart PA. Mechanism of lymphocyte mediated cytotoxicity. 38 Bacigalupo A, VanLint MT, Valbonesi M et al. Thiotepa Ann Rev Immunol 1988; 3: 31–58. cyclophosphamide followed by granulocyte colony-stimulat- 57 Rouvier E, Luciani MF, Golstein P. Fas involvement in ing factor mobilized allogeneic peripheral blood stem cells in Ca(2+)-independent T cell mediated cytotoxicity. J Exp Med adults with advanced leukemia. Blood 1996; 88: 353–357. 1993; 177: 195–200. 39 Bensinger WI, Clift R, Martin P et al. Allogeneic peripheral 58 Beelen DW, Ottinger HD, Elmaagacli A et al. Transplantation blood stem cell transplantation in patients with advanced hem- of filgrastim-mobilized peripheral blood stem cells from HLA- atologic malignancies: a retrospective comparison with mar- identical sibling or alternative family donors in patients with row transplantation. Blood 1996; 88: 2794–2800. matologic malignancies: a prospective comparison on clinical 40 Vogelsang GB, Wagner JE. Graft-versus-host disease. outcome, immune reconstitution, and hematopoietic chimer- Hematol/Oncol Clin 1990; 4: 625–639. ism. Blood 1997; 90: 4725–4735. 41 Soiffer RJ, Bosserman L, Murray C et al. Reconstitution of 59 Charbonnier A, Sainty D, Faucher C et al. Immune reconsti- T cell function after CD6-depleted allogeneic bone marrow tution after blood cell transplantation. Hematol Cell Ther transplantation. Blood 1990; 75: 2076–2084. 1997; 39: 261–264. 42 Ueda M, Harada M, Shiobara S et al. T lymphocyte reconsti- 60 Autran B, Malphettes M, Dhedin N et al. Studies of T cell tution in long-term survivors after allogeneic and autologous reconstitution after hematopoietic stem cell transplant. Hema- marrow transplantation. Transplantation 1984; 37: 552–556. tol Cell Ther 1997; 39: 252–256. 43 Witherspoon RP, Lum LG, Storb R. Immunologic reconsti- 61 Cayeux S, Meuer S, Pezzutto A et al. T cell ontogeny after tution after human marrow grafting. Semin Hematol 1984; 21: autologous bone marrow transplantation: failure to synthesize interleukin-2 (IL-2) and lack of CD2- and CD3-mediated pro- 2–10. + + 44 Roberts MM, To LB, Gillis D et al. Immune reconstitution liferation by both CD4 and CD8 cells even in the presence following peripheral blood stem cell transplantation, autolog- of exogenous IL-2. Blood 1989; 74: 2270–2277. ous bone marrow transplantation and allogeneic bone marrow 62 Azogui O, Gluckman E, Fradelizi D. Inhibition of IL-2 pro- transplantation. Bone Marrow Transplant 1993; 12: 469–475. duction after human allogeneic bone marrow transplantation. 45 Keever CA, Small TN, Flomenberg N et al. Immune reconsti- J Immunol 1983; 131: 1205–1208. 63 Welte K, Ciobanu M, Moore MAS et al. Defective interleu- tution following bone marrow transplantation: comparison of kin-2 production in patients after bone marrow transplantation recipients of T cell depleted marrow with recipients of con- and in vitro restoration of defective T lymphocyte proliferation ventional marrow grafts. Blood 1989; 73: 1340–1350. by highly purified interleukin-2. Blood 1984; 64: 380–385. 46 Ottinger HD, Beelen DW, Scheulen B et al. Improved immune 64 Kook H, Goldman F, Padley D et al. Reconstruction of the reconstitution after allotransplantation of peripheral blood immune system after unrelated or partially matched T cell stem cells instead of bone marrow. Blood 1996; 88: 2775– depleted bone marrow transplantation in children: immuno- 2779. phenotypic anlaysis and factors affecting the speed of recov- 47 Dokhelar MC, Wiels J, Lipinski M et al. Natural killer cell ery. Blood 1996; 88: 1089–1097. activity in human bone marrow recipients. Early reappearance 65 Jin NR, Lum LG, Buren EV et al. Signal transduction by B of peripheral natural killer activity in graft-versus-host dis- and T cells early after bone marrow transplantation: B cell ease. Transplantation 1981; 31: 61–65. calcium flux responses are intact whereas lack of CD4 cells 48 Yamagami M, McFadden PW, Koethe SM et al. Failure of T accounts for impaired T cell responses. Bone Marrow Trans- cell receptor-anti-CD3 monoclonal antibody interaction in T plant 1995; 16: 103–109. cells from marrow recipients to induce increases in intracellu- 66 Saxon A, McIntyre RE, Stevens RH, Gale RP. Lymphocyte lar ionized calcium. J Clin Invest 1990; 86: 1347–1351. dysfunction in chronic graft-versus-host disease. Blood 1981; 49 Ino K, Singh RK, Talmadge JE. Monocytes from mobilized 58: 746–751. stem cells inhibit T cell function. J Leukocyte Biol 1997; 61: 67 Pan L, Delmonte J, Jalonen CK, Ferrara JLM. Pretreatment 583–591. of donor mice with granulocyte colony-stimulating factor pol- 50 Mielcarek M, Martin PJ, Torok-Storb B. Suppression of arizes donor T lymphocytes toward Type-2 cytokine pro- + alloantigen-induced T cell proliferation by CD14 cells duction and reduces severity of experimental graft-versus-host derived from granulocyte colony-stimulating factor-mobilized disease. Blood 1995; 86: 4422–4429. peripheral blood mononuclear cells. Blood 1997; 89: 1629– 68 Schneider LC, Antin JH, Weinstein H et al. Lymphokine pro- 1634. file in bone marrow transplant recipients. Blood 1991; 78: 51 Tanaka J, Mielcarek M, Torok-Storb B. Impaired induction of 3076–3080. the CD28-responsive complex in granulocyte colony-stimulat- 69 Dickinson AM, Sviland L, Hamilton PJ et al. Cytokine ing factor mobilized CD4 T cells. Blood 1998; 91: 347–352. involvement in predicting clinical graft-versus-host disease in 52 Swine S, Dutton R, Schwab R, Yamamoto J. Xenogenic allogeneic bone marrow transplant recipients. Bone Marrow human anti-mouse T cell responses are due to the activity of Transplant 1994; 13: 65–70. the same functional T cell subsets responsible for allospecific 70 Rus V, Svetic A, Nguyen P et al. Kinetics of Th1 and Th2 and major histocompatibility complex-restricted responses. J cytokine production during the early course of acute and Exp Med 1983; 2: 720–729. chronic murine graft-versus-host disease. J Immunol 1995; 53 Heusel JW, Wesselschmidt RL, Shresta S et al. Cytotoxic 155: 2396–2406. lymphocytes require granzyme B for the rapid induction of 71 De Waal Malefyt R, Abrams J, Bennett B et al. Interleukin Immune reconstitution following PBSCT S Shenoy et al 346 10 (IL-10) inhibits cytokine synthesis by human monocytes: of cytokines and secondary messages after T cell depleted and an autoregulatory role of IL-10 produced by monocytes. J Exp non-T cell depleted bone marrow transplantation. Transplan- Med 1991; 174: 1209–1220. tation 1996; 62: 947–953. 72 Schwaighofer H, Kernan NA, O’Reilly RJ et al. Serum levels