Selectivity of compounds isolated from the leaves of Nerium indicum Mill. on various human cancer cell lines.

The leaves of Nerium indicum Mill. have been utilized traditionally to cure cancer. By Bioassay (BST) guided isolation method, six compounds were isolated from the CHCl3 extract of the leaves. Selectivity of these compounds (in 0.6-12,500 ng/ml) was tested on various human cancer (MCF7, EVSA-T, T47D, H226, IGROV, A498, WIDR, M19, HeLa) and normal (Vero) cells in vitro. Doxorubicin and cysplatin were used as positive controls. The result indicated that NiO2D (5alpha-oleandrin) possessed the best cytotoxic effect on HeLa cells (IC50, 8.38 x10(-6) mM) and NiO2C (16, 17-dehidrodeasetil-5alpha-oleandrin) on A498 cells (IC50, 1.43 x 10(-6) mM). Those two compounds were not cytotoxic to normal cell.

Increased levels of soluble CD27 in the cerebrospinal fluid are not diagnostic for leptomeningeal involvement by lymphoid malignancies.

Soluble CD27 (sCD27) reportedly is a sensitive and specific marker for leptomeningeal involvement (LI) of CD27-expressing lymphoproliferations such as B-cell non-Hodgkin's lymphoma (B-NHL) or chronic B-lymphocytic leukemia (B-CLL). Because morphological analysis of cerebrospinal fluid (CSF) in patients suspected of LI is false negative in one-third of patients, a diagnostic marker for LI by B-NHL or B-CLL would be very valuable. sCD27 was determined in the first CSF sample from each of 102 unselected patients submitted for (immuno)morphologic detection of malignant cells. The patients were considered to have LI if either (immuno)morphologic analyses showed tumor cells or if neuroradiological evaluation showed typical abnormalities consistent with LI. Patients were suspected of having LI if CSF samples revealed atypical lymphocytes and/or if clinical symptoms and signs suggestive of LI were present, but clinical follow-up was shorter than 3 months because of deterioration of the patient. LI was considered absent if (immuno)morphologic analyses of CSF samples were negative without evidence for LI during 3 months of clinical follow-up. In patients with chronic lymphoproliferative disorders [mainly B-non-Hodgkin's lymphoma (NHL)], sCD27 concentrations were significantly higher in the CSF samples of 16 patients with confirmed or suspected LI than in those of 46 patients without LI. However, sCD27 was also increased in a variety of other predominantly inflammatory neurological disorders including herpes simplex and zoster infections. The positive predictive value of sCD27 determination for LI was only 54%, but the negative predictive value was 92%. Normal sCD27 concentrations in CSF samples of patients with chronic lymphoproliferation makes LI unlikely, but the determination of CSF sCD27 is not sufficiently specific to serve as a reliable tumor marker.

Circumventing tolerance to a human MDM2-derived tumor antigen by TCR gene transfer.

We identified a tumor-associated cytotoxic T lymphocyte (CTL) epitope derived from the widely expressed human MDM2 oncoprotein and were able to bypass self-tolerance to this tumor antigen in HLA-A*0201 (A2.1) transgenic mice and by generating A2.1-negative, allo-A2.1-restricted human T lymphocytes. A broad range of malignant, as opposed to nontransformed cells, were killed by high-avidity transgenic mouse and allogeneic human CTLs specific for the A2.1-presented MDM2 epitope. Whereas the self-A2.1-restricted human T cell repertoire gave rise only to low-avidity CTLs unable to recognize the natural MDM2 peptide, human A2.1+ T lymphocytes were turned into efficient MDM2-specific CTLs upon expression of wild-type and partially humanized high-affinity T cell antigen receptor (TCR) genes derived from the transgenic mice. These results demonstrate that TCR gene transfer can be used to circumvent self-tolerance of autologous T lymphocytes to universal tumor antigens and thus provide the basis for a TCR gene transfer-based broad-spectrum immunotherapy of malignant disease.

Tetramer-based quantification of cytomegalovirus (CMV)-specific CD8+ T lymphocytes in T-cell-depleted stem cell grafts and after transplantation may identify patients at risk for progressive CMV infection.

Recovery of cytomegalovirus (CMV)-specific T-cell-mediated immunity after allogeneic hematopoietic stem cell transplantation (SCT) is critical for protection against CMV disease. The study used fluorochrome-conjugated tetrameric complexes of HLA-A2 molecules loaded with the immunodominant NLVPMVATV (NLV) peptide derived from the CMV protein pp65 to quantify A2-NLV-specific CD8+ T cells in partially T-cell-depleted grafts administered to 27 HLA-A*0201+ patients and to monitor recovery of these T cells during the first 12 months after SCT. None of the 9 CMV-seronegative patients became infected with CMV, whereas 14 of 18 CMV-seropositive patients developed CMV antigenemia after SCT. CMV-seropositive recipients of grafts from CMV-seronegative donors required more preemptive treatment with ganciclovir (GCV) than those of grafts from CMV-seropositive donors (3 [1-6] versus 1 [0-3] courses, respectively; P =.009). The number of A2-NLV-specific CD8+ T cells in the grafts correlated inversely with the number of preemptive GCV courses administered (r = -0.61; P =.01). None of the 9 CMV-seronegative patients mounted a CMV-specific immune response as measured by monitoring A2-NLV-specific CD8+ T cells after SCT. Thirteen of 14 CMV-seropositive patients without CMV disease recovered these T cells. In spite of preemptive GCV treatment, CMV disease developed in 4 patients, who all failed to recover A2-NLV-specific CD8+ T cells after SCT (P =.002). Thus, enumeration of HLA-restricted, CMV-specific CD8+ T cells in the grafts and monitoring of these T cells after SCT may constitute a rapid and sensitive tool to identify SCT recipients at risk for developing CMV disease.

Interleukin 12 induces activation of fibrinolysis and coagulation in humans.

Interleukin 12 (IL-12) has potential efficacy in malignant, infectious and allergic diseases. Its side-effects include activation of coagulation and fibrinolysis, as documented in chimpanzees. We assessed the coagulative and fibrinolytic response in 18 patients with renal cell carcinoma after subcutaneous injection of 0.5 microg/kg recombinant human IL-12. IL-12 induced a fibrinolytic response in 17 patients (94%): plasmin-alpha2-anti-plasmin complexes (PAPc) increased from 11.8 +/- 6.6 nmol/l (mean +/- SD) to a maximum of 18.8 +/- 7.4 nmol/l at 72 h. Baseline levels of tissue plasminogen activator (tPA) and plasminogen-activator inhibitor-I (PAI) were elevated in eight and 14 patients respectively. tPA increased from 12.6 +/- 5.2 ng/ml to a maximum of 19.0 +/- 6.7 ng/ml at 72 h. PAI decreased from 111 +/- 69 ng/ml to a minimum of 65 +/- 53 ng/ml at 8 h, thereafter remaining below baseline. Elevation of PAPc correlated with elevation of tPA and reduction of PAI. A coagulative response occurred in nine patients (50%): thrombin-anti-thrombin III complexes increased from 29 +/- 53 ng/ml to a maximum of 460 +/- 322 ng/ml at 12 h. Patients with and without a coagulative response had similar levels of recombinant human IL-12, interferon-gamma or tumour necrosis factor-alpha. We conclude that IL-12 can activate both fibrinolysis and coagulation in a significant proportion of patients with cancer. The time-frame and sequence of these activation processes differ from those known for other cytokines.

A phage display selected fab fragment with MHC class I-restricted specificity for MAGE-A1 allows for retargeting of primary human T lymphocytes.

The clinical benefit of adoptive transfer of MHC-restricted cytotoxic T lymphocytes(CTL) for the treatment of cancer is hampered by the low success rate to generate antitumor CTLs. To bypass the need for tumor-specific CTL, we developed a strategy that allows for grafting of human T lymphocytes with MHC-restricted antigen specificity using in vitro selected human Fab fragments fused to the Fc(epsilon)RI-gamma signaling molecule. Retroviral introduction of a Fab-based chimeric receptor specific for MAGE-A1/HLA-A1 into primary human T lymphocytes resulted in binding of relevant peptide/MHC complexes. Transduced T lymphocytes responded to native MAGE-A1/HLA-A1POS target cells by specific cytokine production and cytolysis. Therefore, peptide/MHC-specific Fab fragments represent new alternatives to TCR to confer human T lymphocytes with tumor specificity, which provides a promising rationale for developing immunogene therapies.

Grafting primary human T lymphocytes with cancer-specific chimeric single chain and two chain TCR.

Primary human activated T lymphocytes were genetically grafted with chimeric T cell receptors (TCR). Three domain single chain (sc-) TCR as well as two chain (tc-) TCR gene constructs were derived from the melanoma-specific cytotoxic human T cell (CTL) clone 82/30, and linked to the CD3-zeta signaling element. Chimeric TCR alpha and beta receptor genes were structurally designed to prevent pairing with endogenous TCR alpha and beta chains in order to prevent the generation of unpredictable immune specificities. After transduction of polyclonally activated human peripheral blood lymphocytes with retroviral vectors harboring the chimeric receptor genes, genetically engineered cells specifically recognized and responded to MAGE-A1POS/HLA-A1POS cells. Importantly, each type of transduced T lymphocytes that bound specifically to peptide/MHC complexes also showed specific antitumor reactivity as well as lymphokine production. Genetically engineered primary human T lymphocytes expressing chimeric sc- or tc-TCR therefore hold promise for disease-specific therapies.

An entirely humanized CD3 zeta-chain signaling receptor that directs peripheral blood t cells to specific lysis of carcinoembryonic antigen-positive tumor cells.

Recombinant T-cell receptors with antibody-like specificity for tumor-associated antigens are successfully used to direct the cytolytic activity of T cells toward tumor cells. Clinical application, however, needs to comply with the low immunogenicity of the recombinant receptor, efficient gene transfer into peripheral blood T cells, and enrichment of receptor-grafted cells. Here, we address these issues and describe an entirely humanized immune receptor for use in adoptive immunotherapy of colorectal carcinoma. The receptor consists of a single-chain antibody (scFv) binding domain specific for carcinoembryonic antigen (CEA), the IgG hinge and CH2/CH3 (Fc) joining region, and the transmembrane and intracellular CD3 zeta signaling chain. To express the receptor in peripheral blood T cells, both GALV envelope and MuLV 4070A pseudotyped retrovirus turned out to be equally efficient, with transduction efficiencies of about 5% to 40%, depending on the lymphocyte donor. Furthermore, receptor-grafted T cells could be 2- to 6-fold enriched by magnetic activated cell sorting, utilizing an antibody directed to the extracellular IgG domain of the receptor. Upon co-culture with CEA(+) tumor cells, receptor-grafted T cells are specifically and efficiently activated to cytolysis and IFN-gamma secretion, demonstrating their feasibility for the adoptive immunotherapy of CEA(+) carcinomas.

Functional balance between T cell chimeric receptor density and tumor associated antigen density: CTL mediated cytolysis and lymphokine production.

Genetically engineered expression of tumor-specific single chain antibody chimeric receptors (ch-Rec) on human T lymphocytes endow these cells with the parental monoclonal antibody (mAb) dictated tumor specificity and may be useful for clinical immuno-genetherapy. Therefore it was of importance to assess how the densities of tumor-specific receptors and tumor associated antigens (TAA), respectively, affect primary human T lymphocyte functions in relation to target cell susceptibilities to lysis. We therefore studied the functional balance between ch-Rec densities on human T lymphocytes and TAA on tumor cells. The gene construct encoding a ch-Rec derived from (1) a renal carcinoma cell (RCC) specific mouse mAb (G250), and (2) the human signal transducing Fc(epsilon)RI gamma-chain was used. To obtain ch-RecHIGH-POS and ch-RecLOW-POS T lymphocytes, two distinct retroviral vectors were used to introduce the gene constructs into primary human T lymphocytes. Levels of ch-Rec-redirected T lymphocyte mediated tumor cell lysis, as well as lymphokine production were determined using RCC lines as target/stimulator cells, which express either no or increasing densities of the TAA. A functional and dynamic balance between ch-Rec densities on cytotoxic T lymphocytes (CTLs) on the one hand and TAA densities on RCCs on the other, was found. In short, ch-RecHIGH-POS CTLs are triggered by TAAHIGH-POS as well as TAALOW-POS RCCs to lyse tumor cells and produce (IFN-gamma and TNF-alpha) lymphokine. In contrast, ch-RecLOW-POS T lymphocytes are only triggered for cytolysis and lymphokine production by relatively TAAHIGH-POS RCCs. In conclusion, (1) the activation of T lymphocyte responses is co-determined by the expression levels of the ch-Rec on T lymphocytes and the TAA on tumor cells and (2) at relatively high T lymphocyte ch-Rec expression levels the CTLs lyse tumor cells with a wide range of TAA densities. Gene Therapy (2000) 7, 35-42.

Restoration of expression of signal-transduction molecules in lymphocytes from patients with metastatic renal cell cancer after combination immunotherapy.

A decrease in lymphocyte signal-transduction molecules, described in cancer patients and patients with chronic infectious diseases, has been proposed as a possible mechanism leading to an impaired immune response in cancer patients. Here we report the effects of combination immunotherapy on the levels of T cell receptor zeta chain and p56lck tyrosine kinase in a retrospective study of cryopreserved lymphocytes from 26 metastatic renal cell carcinoma patients treated with high-dose interleukin-2 (IL-2), interferon alpha (IFNalpha) and ex vivo IL-2-activated lymphocytes. Of the 26 patients, 12 were responders (5 complete and 7 partial) and 14 were non-responders (6 stable and 8 with progressive disease). Prior to treatment, 21 of 26 patients (81%) and 13 of 21 patients (62%) respectively expressed zeta chain and p56lck at less than 50% of the levels observed in healthy controls. During therapy, this low zeta chain and p56lck expression increased to at least 50% of normal in 13 of the 21 patients (62%) and in 6 of the 13 patients (46%) respectively; in the remaining patients expression levels remained at 50% of normal or more, or declined. Although, in this limited study, pretreatment levels of and p56lck did not show significant correlation with antitumor response, 4 of 5 patients that achieved a complete response (80%) corrected both zeta chain and p56lck levels to at least 50% of normal, while restoration of both signal-transduction molecules to such levels was only observed in 3 of 7 partial responders (43%), 1 of 5 patients with stable disease (20%) and 2 of 7 patients with progressive disease (29%). Thus, these results suggest that analysis of changes in signal-transduction molecules may a be useful tool for immunological monitoring of patients throughout immunotherapy, and could provide important information for designing new clinical trials that restore impaired signal transduction while activating T cell responses.

Quality control of flow cytometric immunophenotyping of haematological malignancies.

Immunophenotyping of haematological malignancies has developed as a clinically valuable but technically complicated diagnostic procedure. It involves a variety of methodological features, in-process strategic judgements and an extensive knowledge of clinical, morphological and other laboratory features of the disease processes under study. We discuss the various internal quality control steps necessary to guarantee reliable results with respect to instrument set-up and calibration; sample preparation; selection and validation of monoclonal antibody panels; and flow cytometric data acquisition, analysis and interpretation of results. The quality of the entire procedure is documented by the analysis of representative specimens in the setting of an external quality assurance programme.

Large-scale production of natural cytokines during activation and expansion of human T lymphocytes in hollow fiber bioreactor cultures.

We studied the large-scale production of a variety of natural cytokines during the activation and expansion of human T lymphocytes in a hollow fiber bioreactor culture system. Peripheral blood mononuclear cells (PBMC) were activated using phytohemagglutinin plus recombinant interleukin-2 (IL-2). Phytohemagglutinin was either present in the hollow fiber bioreactor during the entire 15-16-day culture period or only during the 20-h preactivation of the PBMC in culture bags. The expanding T lymphocytes were mainly CD3+,8+ and exerted maximal natural, activated, bispecific monoclonal antibody-redirected and lectin-dependent cytolytic activities between days 9 and 13 of culture. IL-1 and IL-4 were only produced in low amounts. IL-8 and lymphotoxin were primarily produced during the first week of culture. Harvest of the hollow fiber bioreactor culture supernatant at the time of peak cytokine concentration would have yielded per 10(8) PBMC input between 3.7 and 4.9 micrograms of IL-8 (at days 2 or 3), and between 0.02 and 0.5 microgram of lymphotoxin (at days 6 or 7). Tumor necrosis factor-alpha and IL-6 were produced during the entire culture period of 15 or 16 days: per 10(8) PBMC input, between 0.1 and 0.4 microgram of tumor necrosis factor-alpha (at days 2 or 3) and between 0.03 and 0.5 microgram of IL-6 (at days 15 or 16). Production of interferon-gamma and granulocyte-macrophage colony-stimulating factor started from initiation of cultures onwards to reach peak levels at the end of the 15- or 16-day culture period, yielding at that time between 2.1 and 17.7 micrograms/ml of interferon-gamma and between 0.4 and 4.2 micrograms of granulocyte-macrophage colony-stimulating factor per 10(8) PBMC input. The production of tumor necrosis factor-alpha, IL-6, interferon-gamma, and granulocyte-macrophage colony-stimulating factor was proportional to the extent of lymphocyte multiplication. These results demonstrate the usefulness of hollow fiber bioreactor cultures to produce natural cytokines during the activation and expansion of predominantly CD3+,8+ T lymphocytes.

Expression of T cell activation antigen CD134 (OX40) has no predictive value for the occurrence or response to therapy of acute graft-versus-host disease in partial T cell-depleted bone marrow transplantation.

CD134 (OX40) is a member of the tumor necrosis factor family which is expressed by activated T lymphocytes. CD134 expression on T cells was monitored during the first 35 days post-transplant in 14 patients, receiving either an HLA-identical sibling bone marrow transplant (BMT), a matched unrelated transplant (MUD-BMT) or an autologous peripheral blood progenitor cell transplant (PBPCT). The sibling and unrelated grafts were partially depleted of T cells. CD134 expression on CD4+ T cells peaked between 7 and 14 days after BMT, with a mean peak value of 45% of CD4+ cells (range 26-70%) over all three patient groups. The observed pattern of CD4+ CD134+ expression, an increase during the first 2 weeks post-BMT followed by a gradual decline towards values of 15-40%, was similar in all groups. No difference in the kinetics of CD134 expression by CD4+ T cells was observed between the patients that did or did not develop graft-versus-host disease (GVHD), nor did the clinical effect of any treatment given for GVHD correlate with alterations in CD134 expression by CD4+ T cells. Absolute CD4+,CD134+ T cell numbers showed a more rapid increment after autologous PBPCT than after sibling or MUD transplants. We conclude that expression of CD134+ by CD4+ T lymphocytes cannot serve as a surrogate marker for allo-reactivity. CD134+ expression may reflect lymphocyte regeneration, rather than alloreactivity.

Level of anti-mouse-antibody response induced by bi-specific monoclonal antibody OC/TR in ovarian-carcinoma patients is associated with longer survival.

More than 60% of cancer patients injected with intact murine monoclonal antibody (MAb) develop a humoral response against the antigen even after a single dose. Analysis of a series of 35 ovarian-cancer patients entered in phase-I and -II clinical studies of T-cells retargeted with the bi-specific F(ab')2 OC/TR revealed: (i) a detectable human anti-mouse antibody (HAMA) response in 31/35 (88%) patients, with high HAMA levels (> or = 150 ng/ml) in 18/31 (58%) cases by the end of the treatment; (ii) no correlation between HAMA levels and the form of delivery of the mAb (OC/TR bound to T cells or bound plus soluble), time schedule or cumulative dose; (iii) an association between high HAMA levels and favorable clinical parameters and response to immunotherapy; and (iv) a significantly longer median survival probability in patients with high HAMA levels than in patients with lower HAMA levels, even when the sub-group of non-responder patients was considered. Evaluation of the anti-idiotypic response in HAMA-positive sera indicated that 11/17 sera showed high-titer (>6000) binding of OC/TR, as evaluated by a specific radioimmunoassay, and 15/18 and 16/16 sera specifically inhibited the binding of the MOv18 and anti-CD3 parental MAbs to ovarian-carcinoma cells and T lymphocytes respectively. Of 7 patients evaluated for duration of the HAMA response, 5 showed stable or even increased HAMA levels. The long-lasting HAMA response maintained an anti-idiotypic component, directed mainly against the alphaCD3 idiotype of bi-MAb OC/TR in 2 out of 3 cases tested.

Phase I study of subcutaneously administered recombinant human interleukin 12 in patients with advanced renal cell cancer.

A phase I study was conducted to characterize the maximum tolerated dose (MTD), dose-limiting toxicity (DLT), and pharmacokinetics of a single dose followed by three times weekly s.c. injections of recombinant human interleukin 12 (rHuIL-12). The study encompassed 28 patients with advanced renal cell carcinoma. rHuIL-12 was administered on day 1, followed by an observation period of 7 days. Starting on day 8, repeated s.c. injections were administered 3 times a week for 2 weeks. The MTD of the initial injection was evaluated at dose levels of 0.1, 0.5, and 1.0 microg/kg. DLT was observed at 1.0 microg/kg and consisted of fever, perivasculitis of the skin, and leukopenia. The MTD of the subsequent repeated injections after 1 week of rest was studied at dose levels 0.5, 1.0, and 1.25 microg/kg. DLT at 1.25 microg/kg comprised deterioration of performance status, fever, vomiting, mental depression, and leukopenia. Other notable toxicities were oral mucositis and elevation of hepatic enzymes. Fever, leukopenia, and elevation of hepatic enzymes were more severe after the initial injection than after repeated injections at the same dose level. At dose level 0.5 microg/kg, the mean area under the plasma concentration-time curve decreased from 7.4 ng/h/ml after the first injection to 3.3 ng x h/ml (P = 0.034) after repeated administrations, and at dose level 1.0 microg/kg, it ranged from 31.8 ng/h/ml to 6.0 ng x h/ml (P = 0.041). One patient had a partial response and seven had stable disease. The MTD of a single s.c. injection of rHuIL-12 was 0.5 microg/kg, and the MTD of three subsequent administrations per week was 1.0 microg/kg. In comparison with a single administration, the three times weekly administrations at the same dose level was accompanied with a milder pattern of side effects and a reduction of the area under the plasma concentration-time curve.

Chimeric scFv/gamma receptor-mediated T-cell lysis of tumor cells is coregulated by adhesion and accessory molecules.

Adhesion and accessory molecules play a critical role in T-cell activation and effector function in general and in tumor cell recognition and lysis in particular. We investigated the contribution of CD2, CD3, CD11a/CD18, CD54 and CD58 molecules in T lymphocyte-tumor cell interactions mediated by chimeric immunoglobulin receptors. The chimeric receptor is composed of a single chain antibody binding site and a gamma-chain signal transducing molecule (scFv/gamma). T lymphocytes expressing such scFv/gamma receptors recognize the G250 Ag on renal cell carcinoma (RCC) in an major histocompatibility complex (MHC)-unrestricted manner and exert RCC selective cytolysis. A coregulatory role for CD2, CD3 and CD11a/CD18 molecules in scFv/gamma-mediated cytolysis was demonstrated using monoclonal antibody (MAb)-induced inhibition of scFv/gamma-mediated cytolysis. The inhibition of lysis was not due to inhibition of cytotoxic T lymphocyte (CTL)-target cell conjugation but rather to a post-conjugate signaling event. Binding of CD54 and CD58 MAbs to the RCC did not inhibit cytolysis of RCC cells that expressed high levels of both CD54 and the G250 antigen (Ag) (A75), whereas cytolysis of RCC expressing intermediate levels of CD54 and G250 Ag (SK-RC-17 cl.4) was partly inhibited by the CD54 MAb. Binding of low concentrations of G250 MAb to RCC (A75) rendered these cells sensitive to CD54 MAb inhibition, demonstrating a direct functional relation between G250 Ag expression level and adhesion molecules. Taken together, our findings indicate a coregulatory role for CD2, CD3 and CD11a/CD18 molecules in the scFv/gamma-mediated cytolysis of tumor cells and show that the requirement of CD11a/CD18-CD54 interactions is dependent on the level of free Ag. This make these gene-transduced T lymphocytes attractive tools for adoptive immunogene therapy of cancer.

A phase I/II study of multicyclic dose-intensive chemotherapy supported with G-CSF, or G-CSF and haematopoietic progenitor cells in whole blood, in two consecutive cohorts of patients.

We investigated the reconstitutive potential of haematopoietic progenitor cells collected in autologous whole blood during multicycle dose-intensified chemotherapy. Forty patients with metastatic solid tumours were treated with up to six cycles of cisplatin and escalating doses of ifosfamide every 14 days. Cisplatin was administered in 3% sodium chloride over 3 h, followed by ifosfamide over 24 h and mesna over 36 h. The first cohort of patients received granulocyte colony-stimulating factor (G-CSF) days 4-14. Once dose-limiting toxicity was reached in cohort 1, the study continued with a second cohort of patients, in whom, in addition to G-CSF on days 4-14, 500 ml of G-CSF and chemotherapy-'primed' whole blood was collected on day 15, i.e. on day 1 of treatment cycles two to six, before cisplatin administration. This volume of blood was kept unprocessed at 4 degrees C and reinfused 20-24 h after the completion of ifosfamide. In cohort 1, dose-limiting toxicity (DLT) was reached at ifosfamide 6.0 g m(-2) with two out of six of the patients developing neutropenic fever. Although in cohort 2 no neutropenic fever was encountered, neither the frequency nor the duration of grade 4 neutropenia and thrombocytopenia were reduced. Cumulative asthenia resulted in DLT at 7.0 g m(-2). The median number of CD34+ cells in 500 ml of whole blood after the first cycle (i.e. at start of cycle 2) was 1.15 x 10(6) kg(-1). This number was significantly greater after the second cycle (2.06 x 10(6) kg(-1), P = 0.01) and then gradually decreased after cycles three to six. After storing whole blood, the number of CD34+ cells had not decreased (median + 10%). We conclude that the method of combined bone marrow support by G-CSF and haematopoietic progenitor cells in autologous whole blood collected before each cycle of a 2-weekly regimen of cisplatin-ifosfamide does not result in clinically measurable reduced bone marrow toxicity compared with what can be expected by the use of G-CSF alone.

A retroviral vector system 'STITCH' in combination with an optimized single chain antibody chimeric receptor gene structure allows efficient gene transduction and expression in human T lymphocytes.

Genetic engineering of T lymphocytes for adoptive clinical immunotherapy calls for efficient gene transduction methods. Therefore, a transient retroviral gene transduction system 'STITCH' was developed comprising pSTITCH retroviral vector encoding the transgene, plasmids encoding Moloney murine leukemia virus gag/pol and gibbon ape leukemia virus envelope, and the human kidney cell line 293T as a packaging line. Cotransfection of retroviral vector and packaging plasmids in 293T cells results in the production of GALV env pseudotyped viral particles with a titer of 10(7) infectious units per milliliter. The 'STITCH' gene transduction system efficiently transduces genes into activated human T lymphocytes derived from healthy donors and cancer patients. The efficacy of gene transduction is donor-independent. A direct application of the 'STITCH' gene transduction system is the genetic engineering of activated human T lymphocytes to induce expression of antibody based chimeric receptors in their membrane. Introduction of these chimeric receptors into activated human T lymphocytes graft these cells with specificity for, for example, renal cell carcinoma. In order to study the effect of the chimeric receptor gene structure on the processes ultimately leading to functional membrane expression, we designed a number of different chimeric receptor gene structures and subsequently compared their membrane expression on 293T cells and activated human T lymphocytes. Distinct membrane expression densities were observed on 293T cells and human T lymphocytes for the different chimeric receptor gene constructs. Gene transduction of activated human T lymphocytes with four out of five chimeric receptor gene constructs resulted in functional expression of chimeric receptor as demonstrated by specific recognition and cytolysis of renal cell carcinoma.