Search

Molecular remissions were observed in some relapsed post allogeneic transplant chronic myeloid leukemia (CML) patients who received donor peripheral blood lymphocyte infusions (DLI). This was indirect evidence of immune-mediated tumor cell killing, and the process was termed graft-versus-leukemia (GVL). Mechanisms were hypothesized to be direct T cell mediated cell lysis or augmented programmed cell death. These observations formed the basis for exploring the role of DLI first in relapsed allogeneic transplant patients with other hematologic malignancies, then as prophylaxis for patients at high risk for relapse and then as adjunctive therapy for transplant strategies utilizing non-myeloablative conditioning regimens. Multiple clinical trials are underway to define dose, schedule, clinical and molecular response and laboratory assays to distinguish lymphocyte subsets mediating GVL and those responsible for graft-versus-host disease (GVHD) toxicity. This review outlines some of the donor/patient and therapy variables involved and some principles of mononuclear cell collection for cellular immunotherapy.

Guest Editor's Introduction: By 1977, the NCI/IBM 2990 and the Amico Celltrifuge, a conceptually similar instrument that had an identical centrifuge bowl, were being used worldwide. However, a major technical problem remained in that the blood path was still not completely disposable. To overcome this drawback, IBM 2977 was developed. The centrifuge bowl was changed from an axial flow separator to a circumferential flow geometry. The IBM 2997 single-stage channel instrument was first introduced and followed by the dual-stage channel instrument to obtain platelet concentrate with minimum lymphocyte contamination. This paper describes the basic design, operation conditions, and collection performance of the single-stage channel IBM 2977. This paper was printed in Blood, vol 54, page 254–268 (1979) and reprinted here with permission. A single-stage disposable channel and seal that provides for leukocyte and granulocyte collection by continuous-flow cell separation (CFCS) has been designed by the IBM Corporation. This paper describes (1) the separation characteristics of whole blood as it responds to varying gravitational (G) forces and flow rates through the channel; (2) the mechanism by which the buffy coat accumulates and is extracted; (3) the efficiency of extraction; (4) those donor and procedural variables that contribute to the final yield; (5) posttransfusion increment response in patients; and (6) the functional integrity of the cells collected.

Platelet engraftment, the time course and magnitude of platelet recovery (PR) post-transplant, is imprecisely defined but is most often reported as the time to transfusion (tx) independence and/or a platelet count ⩾20,000/μl. While correlations between engraftment time for granulocytes (PMN) and the dose of CD34-positive cells per kilogram are established, such associations have not been established for platelet engraftment. The objective of this study was to quantify subpopulations of CD34-positive cells in peripheral blood stem cell (PBSC) collections of normal, colony-stimulating factor-granulocyte) (G-CSF) primed donors that might represent megakaryocyte (MK) precursors, and to determine whether there is a statistical association between the dose transfused and the time course of the recovery. Based on previously published data of the sequential expression of CD34, HLA-DR, and CD61, among others, during MK maturation, a combination of corresponding antibodies for the detection of various antigen coexpressions by flow cytometry fluorescence-activated cell sorting [FACS] was chosen. CD34-positive cells were further subdivided into CD34++ (bright) and + (dim). Ploidy of density-gradient separated cells was examined in subsequent donor samples by FACS. For the entire group of patients, there was no strong correlation between any of the studied subpopulations and time to PR. Only in a selected groups of patients whose platelet counts showed a sustained increase during the first 6 days after engraftment, there was a weak correlation between the time to PR and the quantity of CD34+/+CD61+ (r = −0.57) and CD34++HLA-DR-CD61+ (r = −0.62) cells infused. The magnitude of platelet production in these pt., a product of the peripheral blood platelet count and the patient's blood volume, was correlated with the time to PR (r = −0.73). We conclude from this study that subpopulations within CD34+ cells are making some contribution to PR in allogeneic peripheral blood stem cell transplantation, but the correlations are not sufficiently strong because there are probably too many unpredictable and unknown variables in the allogeneic setting that influence the pattern of engraftment. J. Clin. Apheresis 13:7–15, 1998. © 1998 Wiley-Liss, Inc.

Fifteen patients with prolonged neutropenia (a median of 23 days with granulocyte [PMN] < or = 500/microliters) and established fungal infections that had not responded to adequate antifungal therapy were transfused with PMN concentrates collected from 35 cytokine-primed granulocyte colony-stimulating factor (GCSF) donors. Patients received a median of six transfusions. Leukocytosis and granulocytosis were observed within 24 hours of the first GCSF injection, which yielded concentrates averaging 55 x 10(9) white blood cells and 41 x 10(9) PMN. Data analysis suggested that response might be related to the duration of neutropenia and known infection, as patients given PMN tx earlier in the infectious course tended to have a better response. No significant toxicity was observed in donors.

In order to determine the minimum dose of buffy coat cells necessary to achieve hematopoietic rescue following supralethal irradiation, mongrel dogs under general anesthesia were subjected to leukacytapheresis using three different techniques of cell separation. The buffy coats were frozen with dimethylsulfoxide and stored at -196 degrees C until transfused. Sixteen dogs were irradiated with 800 rads and were supported with antibiotics and transfusions of irradiated homologous blood. They were transfused with the frozen and thawed buffy coat cells, and, if they survived, they were followed for 100 days, sacrificed, and their tissues studied. The mean yield of mononuclear cells during leukocytapheresis ranged from 4.1 +/- 2.0 X 10(9) (mean +/- SD) to 6.0 +/- 4.0 X 10(9) for the three leukacytapheresis methods; one technique was not as satisfactory as the other two. Six of the 16 dogs fully recovered with evidence of marrow rescue; however, only one had a dose of mononuclear cells less than 11.1 X 10(9). These data indicate that seven to 17 leukacytapheresis procedures would be required to reconstitute a 70 kilogram patient. These preliminary findings suggest that, because the yields of transplantable cells with current technology are not adequate, the transplantation potential of buffy coat cells exposed to mobilizing agents should be evaluated.

A high dose combination chemotherapy regimen (CBV) consisting of cyclophosphamide (1.5 gm/m2 day 1 to day 4); BCNU (300 mg/m2 day 1) and etoposide (100 mg/m2 every 12 hours for 6 doses), followed by bone marrow transplant from human leukocyte antigen (HLA) identical sibling donors, was evaluated in 29 patients in whom acute leukemia was in relapse or remission. Engraftment of donor cell type occurred in all but one of 21 patients, in whom marker differences between donor and recipient were established. Two of 11 patients transplanted during relapse of the disease, lived beyond 1 year after bone marrow transplantation. One patient died free of leukemia, 41 months after transplantation of meningitis. Two of seven patients transplanted during the second remission of the disease, are alive and free of leukemia at 42+, and 8+ months. All patients transplanted during the third or fourth remission of the disease have died from either a further relapse, or transplant related causes. The low incidence of organ toxicity with CBV allows for further dose escalation of its drug components.

Sixty-two adult patients with acute myeloid leukemia were treated with a chemotherapy program of Adriamycin, Cytosine Arabinoside (Ara-C), vincristine, and prednisone (Ad-OAP). Immunotherapy with BCG by scarification was administered before and/or during remission induction therapy. Maintenance chemoimmunotherapy consisted of Ara-C, vincristine, prednisone, and BCG. Patients still in remission after 12 months received late intensification chemotherapy for three courses before discontinuing chemotherapy. Forty-four (71%) patients achieved a complete remission (CR). The CR rate for patients less than 50 years of age ws 84% (32/38). Age was strongly predictive for response and survival. An additional 14 patients with acute lymphoid leukemia were treated with the same program; ten patients responded, all have relapsed and died. Nine patients with AML remain alive 4 1/2-6+ years from diagnosis.

Large numbers (2.9 +/- 1.2 X 10(9)) of mononuclear cells can be obtained from incidental samples collected during routine plateletapheresis. We conducted studies comparing characteristics and functions of mononuclear cells derived from venous blood samples and from routine plateletapheresis in the same normal donors. Cell viability was similar in both samples (96 +/- 1% plateletapheresis vs 97 +/- 2% venous blood). Higher concentration of monocytes were observed in the plateletapheresis samples (32.3 +/- 6%) than in the venous blood (14.3 +/- 4%). The procedure of plateletapheresis does not seem to alter lymphocyte or monocyte function. Thus, the functional integrity of these cell populations was demonstrated in terms of natural killer cell activity, blastogenic response to mitogens, local graft-versus-host reactions, monocyte-mediated antibody-dependent cellular cytotoxicity against human red cells, monocyte-mediated tumor cell cytotoxicity, latex phagocytosis, and monocyte-dependent lymphocyte blastogenesis. We conclude that monocytes and lymphocytes obtained during routine plateletapheresis are functionally intact.

A total of 201 leukaphereses have been performed on 73 donors to collect granulocytes for transfusion to infected leukopenic recipients. The pretreatment of the donor with etiocholanolone and the addition of hydroxyethyl starch to the extracorporeal circulation significantly increased the total number of granulocytes collected. Adverse side effects were minimal. The majority of donors underwent the procedure on 13 occasions over a period of eight weeks. Follow-up donors two or more years later has revealed no long-term adverse side effects from leukapheresis or from the addition of etiocholanolone and hydroxyethyl starch.

Patient response to plasma exchange is variable. Clinical experience indicates that the outcome depends on the size of the exchange relative to the patient's plasma volume, the amount of material available for removal in both the intravascular and extravascular pools, and the mobility of the material between the pools. Post-exchange, the patient's course also depends on the mobility of the material between the pools as well as its production and catabolic rates. The work reported here is an attempt to construct a mathematical model that describes events occurring intra- and post-plasma exchange. An attempt to model the kinetics of plasma exchange promotes an understanding of the requirements of volume removal in individual exchange procedures for optimal results, and establishes the interval between exchanges required to sustain the response to each individual exchange procedure.

This article compares the outcome of 14 patients with primary refractory acute leukemia who underwent bone marrow transplantation from human leukocyte antigen (HLA)-identical donors with that of 18 age-matched control patients who received chemotherapy. Complete clearing of leukemia was seen in all 14 transplanted patients. Five of the transplanted patients are alive 98 to 1790 days posttransplant, and four are free of leukemia. Nine patients have died, eight with severe graft-versus-host disease associated with interstitial pneumonia or systemic infections and one with relapse from chemotherapy-associated infections. Engraftment was seen in all patients. Severe graft-versus-host disease (grades III and IV) was seen in ten patients and resolved in three patients following high-dose corticosteroid treatment. Three of the 18 control patients are alive, none of them in complete remission. It appears that the combination of piperazinedione and total-body irradiation followed by allogeneic transplant is effective induction treatment for primary refractory acute leukemia and will be considered in the future as first salvage treatment for patients failing induction treatment.

The effect of oral methylprednisolone and the sedimenting agent, hydroxyethyl starch, on granulocyte recovery, morphology, and function was studied in a volunteer donor programme. Using the IBM 2997, 10 litres of whole blood were processed, with an average procedure time of 2.4 hours and a collection volume of 300 ml. Donors not receiving methylprednisolone (n = 80) had a mean total granulocyte count of 3.5 X 10(9)/litre (range 1.6-5.3 X 10(9)/litre) and mean granulocyte yields were 1 X 10(10) (range 0.2-3.0 X 10(10)). Those receiving 48 mg oral methylprednisolone 6-8 hours before the procedure (n = 320) had a mean granulocyte count of 6.3 X 10(9)/litre (range 3.2-11.4 X 10(9)/litre) and significantly superior mean granulocyte yields of 2.0 X 10(10) (0.3-6.5 X 10(10)) (P less than 0.05). For both groups the mean packed cell volume of 0.08 litre/litre (range 0.02-0.17) and platelet contamination 1.9 X 10(11) (range 0.3-5.0 X 10(11)). In all these procedures, hydroxyethyl starch was added to the blood entering the centrifuge channel. In none of the procedures were any untoward symptoms experienced by the donors. Light microscopy and ultrastructural studies showed no difference between control granulocytes and those collected following the addition of hydroxyethyl starch or after oral methylprednisolone. Similarly, granulocyte function measured with a random migration, chemotaxis, phagocytosis, and intracellular killing was not significantly different between control cells and those exposed to the sedimenting agent or the adrenocorticosteroids (P greater than 0.10). It is concluded that donor premedication with methylprednisolone significantly enhances granulocyte yields in the presence of hydroxyethyl starch and neither agent has any demonstrable effect on granulocyte morphology or function.

Mononuclear cell separation and collection by continuous-flow centrifugation relates to centrifugal acceleration, total blood processed, peripheral blood concentration, and probable mobilization from extravascular sites during the procedure. At low-G forces, 70-80% of mononuclear cells are cleared. As G forces are increased, 95-100% of mononuclear cells can be cleared, but excessive platelet depletion occurs. An average of three billion cells were harvested from 2-3 liters of blood, rising to nine billion when 6 liters were processed. The total blood processed is defined by citrate infusion rates and ionized calcium changes. Large quantities of mononuclear cells were collected from normal donors without any significant depletion. If the objective of harvesting mononuclear cells is depletion, intensive schedules may be needed.

Eighty-five patients with acute myeloblastic leukemia (AML) presenting with hyperleukocytosis (HL) were analyzed to assess morbidity and mortality in early induction. Patients who failed to achieve remission were older and more often had pulmonary leukostasis (62% vs 23%, p = .01) and hepatomegaly (54% vs 31%, p = .06) at presentation. Thirty-seven patients (44%) did not achieve complete remission (CR); 17 (54%) died early in induction therapy, 11 directly as a result of pulmonary hemorrhage with respiratory failure, while 5 had both pulmonary hemorrhage with respiratory failure and CNS hemorrhage. Early death patients were older and more often had pulmonary leukostasis (88% vs 29%, p less than .0001), hepatomegaly (71% vs 34%, p = .01), hyperbilirubinemia (60% vs 16%, p = .01) and hypofibrinogenemia (47% vs 12%, p less than .01) at presentation. Primarily for technical reasons, preinduction leukapheresis was not employed as often in this very-high-risk group as in other patients (56% vs 82%, respectively). Thus, sufficient heterogeneity exists in patients presenting with HL to define a subset of patients at particularly high risk for early mortality. Preinduction leukapheresis applied in a prospective controlled fashion should be evaluated to assess if such treatment may decrease early mortality in this group.

Eighty-seven patients from 18 institutions with a confirmed diagnosis of chronic myelogenous leukemia were registered on a Southwest Oncology Group protocol for multiagent induction and single-agent maintenance chemotherapy, with randomization to an immunotherapy arm. Elective surgical splenectomy was performed for 42 patients at the completion of 3 months of induction therapy. Final analysis of the study revealed statistically significant survival advantages were correlated with age, splenectomy, the absence of hepatic leukemic infiltrate at the time of splenectomy, and race.

The natural history and response to treatment of 46 patients with acute myelogenous leukemia and a pretreatment leukocyte count of 100,000/microliters or higher were reviewed to identify the clinical features and response characteristics to the treatment of this group of patients. While the response rate of 52% was similar to that of patients with lower leukocyte counts, remission durations were shorter and related inversely to the height of the initial leukocyte count and to the number of treatment courses necessary to achieve a complete remission. A high incidence of hemorrhagic deaths was observed during the first 8 days of treatment. These hemorrhages occurred at a time when the leukocyte count was falling secondary to chemotherapy and the platelet count was still greater than 15,000/microliters. Pretreatment coagulation disorders and poor performance status were factors associated with this fatal complication. Antimetabolites to rapidly arrest leukemic cell proliferation and leukapheresis to avoid further leukostatic plug formation may be useful immediate measures to reduce the incidence of these fatal hemorrhages until the underlying pathogenic mechanisms have been elucidated.

This study chronicles leukocyte- and erythrocyte-depleted platelet concentrate collection by a dual stage channel in which three variables: 1) donor peripheral blood platelet concentration, 2) total blood processed, and 3) collection volume were statistically correlated with platelet yield as determined by a multiple regression analysis of single variables. Platelet concentration in the final yield was related to donor precount and collection rate, and could be varied as indicated for individual applications. Total blood processed was established by procedure time, which in turn was defined by citrate-induced calcium changes in the donor. Reduction in peripheral blood platelet concentration averaged 24% for a mean platelet yield of 3.8 X 10(11). An average of 40% of transfused platelet concentrates were recovered in recipient peripheral blood 1 hr posttransfusion and were hemostatically effective, as determined by correction of bleeding time. Platelet yields and patient response were sustained during current collection procedures, verifying the principles described during the investigative period.

Splenectomies have been performed on 58 patients with chronic myelogenous leukemia (CML) during the last 16 years. For the 27 patients operated upon during the first 12 years, the operative mortality was 26%. Four patients also had to be re-explored for bleeding and three required drainage of subphrenic abscesses. There has been no operative mortality in the 31 patients operated using a standardized procedure during the last four years but two had to be re-explored for bleeding and one required drainage of a subphrenic abscess. The operative risks for leukemic patients are infection, perhaps related to granulocyte abnormalities and hemorrhage in patients with thrombocytopenia or qualitative platelet abnormalities. When the splenectomy in patients with CML is timed with the patient's chemotherapy cycle, the use of local antibiotics and platelet concentrates should permit a safe operation. Survival rates were not comprised even when the patients had their splenectomy during CML-blast crisis.