Bone marrow involvement in systemic lupus erythematosus.

BACKGROUND: Besides peripheral cytopenias, bone marrow abnormalities, such as fibrosis, pure red cell aplasia and aplastic anemia have been reported in patients with systemic lupus erythematosus (SLE), suggesting that bone marrow may be a 25 target organ in SLE. AIM: Our objective was to describe this bone marrow involvement. METHODS: This registry is a nationwide retrospective study. Centers provided data concerning medical history, SLE manifestations, type of hematologic disorder, treatments and outcome. Bone marrow aspirations and/or biopsies were transferred for centralized review. RESULTS: Thirty patients from 19 centers were included. Central hematologic manifestations comprised bone marrow fibrosis (n = 17; 57%), pure red cell aplasia (n = 8; 27%), myelodysplastic syndrome (n = 3; 10%), aplastic anemia and agranulocytosis (n = 1; 3% each). Bone marrow involvement was diagnosed concomitantly with SLE in 12 patients. Bone marrow biopsies showed fibrosis in 19 cases, including one case of pure red cell aplasia and one case of agranulocytosis and variable global marrow cellularity. Treatments included corticosteroids (90%), hydroxychloroquine (87%), rituximab (33%), intravenous immunoglobulins (30%), mycophenolate mofetil (20%) and ciclosporine (20%). After a median follow-up of 27 months (range: 1-142), 24 patients manifested complete improvement. No patient died. CONCLUSIONS: This registry comprises the largest series of SLE patients with bone marrow involvement. It demonstrates the strong link between SLE and bone marrow fibrosis. Patients with atypical or refractory cytopenia associated with SLE should undergo bone marrow examination to enable appropriate, and often effective, treatment. Long-term prognosis is good.

Cyclin-dependent kinase 1 expression is inhibited by p16(INK4a) at the post-transcriptional level through the microRNA pathway.

The p16(INK4a) protein regulates cell cycle progression mainly by inhibiting the activity of G1-phase cyclin-dependent kinases (CDKs) 4 and 6, the subsequent retinoblastoma protein (pRb) phosphorylation and E2F transcription factor release. The p16(INK4a) protein can also repress the activity of other transcription factors, such as c-myc, nuclear factor-kappaB and c-Jun/AP1. Here, we report that, in two p16(-/-), pRb(WT) and p53(WT) cell lines (MCF7 and U87), p16(INK4a) overexpression induces a dramatic decrease in CDK1 protein expression. In response to p16(INK4a), the decreased rate of CDK1 protein synthesis, its unchanged protein half-life, unreduced CDK1 mRNA steady-state levels and mRNA half-life allow us to hypothesize that p16(INK4a) could regulate CDK1 expression at the post-transcriptional level. This CDK1 downregulation is mediated by the 3'-untranslated region (3'UTR) of CDK1 mRNA as shown by translational inhibition in luciferase assays and is associated with a modified expression balance of microRNAs (miRNAs) that potentially regulate CDK1, analyzed by TaqMan Human microRNA Array. The p16(INK4a)-induced expression of two miRNAs (miR-410 and miR-650 chosen as an example) in MCF7 cells is confirmed by individual reverse transcription-qPCR. Furthermore, we show the interaction of miR-410 or miR-650 with CDK1-3'UTR by luciferase assays. Endogenous CDK1 expression decreases upon both miRNA overexpression and increases with their simultaneous inhibition. The induction of miR-410, but not miR-650 could be related to the pRb/E2F pathway. These results demonstrate the post-transcriptional inhibition of CDK1 by p16(INK4a). We suggest that p16(INK4a) may regulate gene expression by modifying the functional equilibrium of transcription factors and consequently the expression balance of miRNAs.

Prognostic value of miR-16 expression in childhood acute lymphoblastic leukemia relationships to normal and malignant lymphocyte proliferation.

miR-16, a miRNA involved in cell proliferation and apoptosis regulation, may interfere with either oncogenic or tumor-suppressor pathways and is implicated in leukemogenesis. We then explored its expression in 93 childhood acute lymphoblastic leukemia (ALL) cases. A high miR-16 expression was associated with hyperleukocytosis and poor cytogenetic groups. In the whole group and in B-cell ALLs, disease-free survival (DFS) was significantly shorter for miR-16 above quartile 75. In T-cell ALLs, for both DFS and overall survival, a significant trend was found with a survival shortening from the lowest to the highest miR-16 levels. miR-16 expression neither significantly correlated with normal and malignant lymphocyte proliferation nor varied according to lymphocyte differentiation. The prognostic value of miR-16 in childhood ALL highlighted the complexity of miR-16 functions.

Disseminated nontuberculous infections with Mycobacterium genavense during sarcoidosis.

Sarcoidosis is a chronic disease characterised by the development and accumulation of granulomas in multiple organs. We report two observations of disseminated Mycobacterium genavense infection in patients with proven sarcoidosis. High fever and abdominal pain appeared at 8 and 18 months following the initiation of immunosuppressive therapy. Abdominal computed tomography scans of the patients showed diffuse mesenteric lymphadenitis and splenomegaly. The diagnosis was obtained on bone marrow specimens for both patients with numerous acid-fast bacteria at direct examination and positive specific mycobacterial identification by nucleic acid amplification test. Despite prompt antimycobacterial therapy, occurrence of complications (peritonitis post-splenectomy surgery and lung carcinoma) resulted in a fatal outcome for both patients. These cases highlight that opportunistic infections like M. genavense or other nontuberculous mycobacterial infections should be considered for long-standing immunocompromised patients with sarcoidosis.

[Marginal zone B-cell lymphoma affecting the skin: histological and phenotypic study of 49 cases]. / Lymphome cutané de la zone marginale : étude histologique et immunophénotypique de 49 cas.

BACKGROUND: No histological or clinical criteria allow distinction between primary cutaneous marginal zone B-cell lymphoma (MZL) and secondary cutaneous forms of systemic marginal zone B-cell lymphoma. Consequently, staging alone can indicate the origin of lymphoma. Lymphoma is considered as primary cutaneous only if no other extracutaneous sites are found. We studied the histological appearance of 49 cutaneous lymphomas in order to find distinctive criteria indicative of an extracutaneous origin. MATERIALS AND METHODS: This was a retrospective descriptive study of histological appearance for 49 patients with cutaneous marginal lymphoma: 29 cases of the primary form and 20 cases with extracutaneous involvement. RESULTS: Comparison of histological criteria did not reveal any differences between the primary cutaneous form and others forms. No prognostic criteria for relapse were found. DISCUSSION: A cutaneous tropism of primary MZL suggested the hypothesis of specific cutaneous receptors on B-cells that could have led to different histological appearances depending on the origin of the lymphoma. However, our study did not confirm this hypothesis.

Inhibition of normal lymphocyte proliferation by Indirubin-3'-monoxime: a multifactorial process.

Indirubin-3'-monoxime (IO) is a derivative of Indirubin, compound of a Chinese medicinal recipe used to treat various diseases including leukemia. In this study, we investigated to what extent IO inhibits the growth of normal human lymphocytes. We defined various experimental conditions of peripheral blood lymphocyte treatment: IO introduced (i) on unstimulated lymphocytes, (ii) or on stimulated lymphocytes at the time of phytohemagglutinin stimulation (L1 protocol), (iii) 48 h after the beginning of stimulation (L2 protocol), and (iv) after nocodazole synchronization of stimulated lymphocytes. IO induces a concentration dependent cytotoxic effect yielding a characteristic sub-G1 peak in normal stimulated lymphocytes. Cell death was partly due to necrosis and apoptosis. Normal unstimulated lymphocytes remained insensitive to the cytotoxic effect of 10 microM IO treatment. A cell cycle inhibition was observed after IO treatment, stronger for the L1 than for the L2 protocol, without induction of polyploidy after Nocodazole synchronization. These cellular consequences were associated with a decrease in CDK activity, and with CDK and cyclin gene expression modifications. The inhibition of lymphocyte proliferation by IO indicates that indirubin derivatives may be potent immunosuppressive agents.

[Thrombocytosis and essential thrombocythemia in childhood]. / Thrombocytoses et thrombocytémies essentielles de l'enfant.

Thrombocytosis is frequently observed in pediatric patients. Among them the secondary thrombocytosis are the most frequent and result from several causes. The rarely primary thrombocytosis can be either constitutive (and often familial) or acquired (essential thrombocythemia). The purpose of this article is to give diagnostic orientation and to suggest which biological tests should be performed.

Bcl-2, cell cycle regulatory proteins and corticosensitivity in childhood acute lymphoblastic leukaemia.

The results of treatment in childhood acute lymphoblastic leukemia (ALL) remain incompletely satisfactory because of relapses observed even with high dose chemotherapy. The aim of this study was to evaluate the role of bcl-2 or cell cycle regulatory protein expression in peripheral blood cells before and during the first 48 hours of corticotherapy, and corticosensitivity criteria for predicting relapse and prognosis. Fifty two children presenting with ALL were studied at diagnosis and during the first 48 hours of treatment for the level of cell proliferation by measurement of DNA content, and for expression of several cell proliferation regulatory proteins by Western blot. Two criteria for corticosensitivity were used: 1--the number of blast cells present after seven days of treatment with a threshold at 1 G/L (usual criterion), 2--the D8/D1 blast cell ratio, which is independent of the initial leucocytosis. Relapse in the total patient population or in B-cell ALL could only be predicted by the level of leucocytosis before treatment or by p27kip1 expression during the first 48 hours of treatment. Disease free survival was significantly longer when the D8/D1 blast cell ratio was under the 0.75 quartile in the entire patient population (p = 0.03). Among the proteins analyzed, bcl-2 expression before treatment and p27kip1 expression analyzed after 48 hours of corticotherapy were the sole variables associated with significant differences in disease free survival duration in the entire patient population (p < 0.01 and p = 0.04 respectively) or in the B-cell ALL subgroup (p < 0.01). Comparable results were obtained for the overall survival data. The significance of these results is discussed but such a study on blood blast cells needs to be validated in a larger series.

[Cyclin dependent kinase inhibitors and replicative senescence]. / Inhibiteurs des kinases dépendantes des cyclines et sénescence réplicative.

Replicative senescence is defined for human diploid fibroblasts in culture as a cell growth arrest appearing beyond 50 +/- 10 population doublings and associated with telomeres' shortening. This phenomenon shows an increased expression of growth cell inhibitors: p21Waf1 described as an universal CDK inhibitor and p16INK4a as a specific inhibitor for both G1 phase kinases CDK4 and CDK6. The cell proliferation inhibitor p14ARF, product of INK4a/ARF locus is involved in replicative senescence too. Overexpression or homozygotic deletion of these inhibitors demonstrated their role in senescence induction. These proteins are involved in two different metabolic pathways, the first including p53, represented by E2F, ARF, MDM2, p53, p21Waf1, and the second concerning pRb and p16INK4a. These two pathways present numerous interactions and the polymerase (PARP) in relation with p53 and activated by telomere shortening might represent via p21Waf1 a link between this shortening and cell cycle control. An another metabolic pathway involving PTEN and p27KIP1 is discussed in senescent-like phenotype induction, but its activity in replicative senescent is uncertain.

Role of BCL-2 and cell cycle regulatory proteins for corticosensitivity assessment in childhood acute lymphoblastic leukaemia.

Results of treatment in childhood acute lymphoblastic leukaemia (ALL) remain unsatisfactory because relapses occur even after high-dose chemotherapy. Corticosensitivity is used in numerous therapeutic trials as a prognostic factor for treatment choice. The aim of this study was to evaluate the role of cell cycle regulatory protein expression before and during the first 48 h of corticotherapy for predicting corticosensitivity. Fifty-two children presenting with ALL were studied at diagnosis and during the first 48 h of treatment for cell proliferation and apoptosis level by measurement of DNA content, and for expression of several cell proliferation regulatory proteins by means of Western blot. Glucocorticoids induced a significant decrease in the percentage of cells in S-phase and in CDK1, CDK4 and CDK6 expression and an increase in the percentage of cells in subG1 peak. Two criteria for corticosensitivity were used: (i) the number of blast cells after 7 d of treatment with a threshold at 1 x 109/l (usual criterion), (ii) the J8/J1 blast cell ratio, which is independent from initial leucocytosis. Bcl-2 expression at diagnosis was the best predictive variable for the usual corticosensitivity criterion in B- and T-cell ALL. For the second criterion, in B-cell ALL, p21waf1 expression at diagnosis was the sole (albeit poorly) predictive variable, whereas bcl-2 remained of high interest in T-cell ALL. Interestingly, these proteins, bcl-2 and p21waf1, are associated with prolonged cell lifespan and their increased expression is often linked to poor response to cytotoxic drugs. Such preliminary results call for subsequent studies on large independent sets of T-cell and B-cell lineage ALL in order to confirm the J8/J1 blast cell ratio value as well as the role of bcl-2 and p21waf1 expression in predicting corticosensitivity.

CD40 ligation alters the cell cycle of differentiating keratinocytes.

CD40 is expressed in normal human keratinocytes, especially in the basal cell layer. We have recently reported that CD40 ligation strongly inhibits keratinocyte proliferation and induces their differentiation. In this study, the CD40 pathway that prevents keratinocyte growth was investigated. We first reported that interferon-gamma treatment potentiated the CD40-mediated inhibition of keratinocyte proliferation. CD40-CD40 ligand interactions, in the presence or absence of interferon-gamma, neither enhanced spontaneous keratinocyte apoptosis, nor did it enhance apoptosis induced by various agents. More importantly, we showed that CD40 signaling altered the keratinocyte cell cycle, as demonstrated by a decreasing number of cells in the G1 and S phases and an accumulation in G2/M phase of the cell cycle. Furthermore, western blot analysis of cell cycle regulatory proteins, showed a decrease in cyclin A and E expression in CD40-activated keratinocytes. Collectively, these results indicate that CD40 ligation inhibits keratinocyte renewal by a mechanism independent of cell apoptosis and that modulation of the keratinocyte cell cycle is an additional outcome of CD40 signaling.

CDK1 and cyclin A expression is linked to cell proliferation and associated with prognosis in non-Hodgkin's lymphomas.

Cellular proliferation is regulated by several kinasic complexes associating cyclins and their catalytic subunits cyclin-dependent kinases (CDKs). In order to gain insight into the mechanisms underlying proliferation in non-Hodgkin's lymphoma (NHL), we examined the expression of certain cell cycle regulatory proteins normally expressed in lymphoid cells, cyclins A, B, D3 and E and cdk1, 2, 4, and 6. In 70 patients presenting a previously untreated lymphoma, cyclins and CDKs were studied by Western blotting and quantified by densitometry. Flow cytometry study of DNA content was carried out for all patients in order to study cell proliferation and level of ploidy. The results were analysed according to the histological types, the immunological phenotypes of the lymphomas and the outcome of the patients. Cdkl and cyclin A were correlated with the percentage of cells in S and S+G2/M phases, and significantly different according to the grade of malignancy, with the lowest expression in low-, and the highest in high-grade NHL according to the Working Formulation. In B-NHLs, cdk1, cyclin A, as well as cdk2, cyclin D3 and E expression was higher in the aneuploid than in the euploid group. Our results point to some particularities of cell cycle regulation in two lymphoma sub-types: 1) a low expression of cyclin D3 and cdk6 in mantle cell lymphomas and 2) a discrepancy between the high proliferative activity and the level of protein expression in Burkitt's lymphomas. CDK1 and cyclin A showed a significant prognostic value for achievement of complete remission (Cdk 1) and for both disease free (cyclin A) and overall survival (cyclin A and cdk1): low protein level was associated with the best prognosis in B-NHLs. Our results show that differential cell cycle regulating protein expression may be associated with different biological and clinical behaviour of NHLs and confirm the usefulness of the study of cell cycle regulation as a tool for understanding lymphoid malignancies.

Granulocyte-macrophage colony-stimulating factor (GM-CSF) to increase efficacy of intensive sequential chemotherapy with etoposide, mitoxantrone and cytarabine (EMA) in previously treated acute myeloid leukemia: a multicenter randomized placebo-controlled trial (EMA91 Trial).

The EMA86 study showed efficacy of intensive sequential chemotherapy with mitoxantrone, 12 mg/m2 day on days 1-3, etoposide, 200 mg/m2/day as a continuous infusion on days 8-10 and cytarabine (araC), 500 mg/m2/day as continuous infusion on days 1-3 and 8-10 (EMA regimen) in previously treated patients with AML. The goal of the EMA91 study was to determine whether administration of GM-CSF between the two sequences of EMA chemotherapy and during the second sequence could increase therapeutic efficacy by potentially increasing leukemic cell recruitment into the S phase of cell cycle before the second sequence. One hundred and ninety-two patients aged less than 65 years with previously treated AML received GM-CSF, 5 microg/kg/day or placebo from day 4 to day 8 of EMA chemotherapy. One hundred and twenty were refractory and 72 were in first relapse after a complete remission (CR) of more than 6 months duration. CR rates after one course of chemotherapy were 65% in the GM-CSF group (refractory: 51%; first relapse: 89%), not significantly different from the 59% CR rate (refractory: 46%; first relapse: 81%) in the placebo group. Median time to recovery of neutrophils was 38 and 37 days and median time to last platelet transfusion 32 and 32 days respectively in the GM-CSF and placebo groups. WHO grade > or = 3 non-hematologic toxicities were mainly sepsis (45% and 51%, respectively) and mucositis (34% and 31%) and did not differ between the two groups. Toxic death rate was 5% and 8%, respectively, in the GM-CSF and placebo groups. Patients achieving CR were scheduled to receive six courses of maintenance with reduced-dose EMA. Time to progression tended to be longer in the GM-CSF group (median 154 vs 115 days, progression-free rate at 18 months 33% vs 19%, P = 0.08), particularly in refractory patients (P = 0.06). However, at the current follow-up, this did not translate into a significantly longer disease-free survival and survival. Cell cycle studies showed increased recruitment of cells in the S phase between day 4 and day 8 in the GM-CSF group compared to placebo (P = 0.006). However, this did not significantly relate to prognosis in this cohort of patients. GM-CSF might marginally increase efficacy of sequential chemotherapy without increasing its toxicity in the absence of any detected relationship between this effect and observed leukemic cell recruitment into the cell cycle.

Involvement of p27Kip1 in the G1- and S/G2-phase lengthening mediated by glucocorticoids in normal human lymphocytes.

Glucocorticoids inhibit cell proliferation by inducing cell cycle lengthening. In this report, we have analyzed, in normal peripheral blood lymphocytes, the involvement of p27Kip1 in this slowing of proliferation. Following dexamethasone (DXM) treatment, p27Kip1 expression and regulation varied differently with the level of lymphocyte stimulation. In quiescent cells, DXM inhibited p27Kip1 protein expression by decreasing its rate of synthesis, whereas its half-life and mRNA steady state remained constant. In contrast, in stimulated lymphocytes, DXM increased p27Kip1 expression by enhancing its mRNA steady state. This increase is not only a consequence of the DXM-induced interleukin 2 inhibition: we also found an increase in p27Kip1 mRNA stability that was not observed in quiescent lymphocytes. Cyclin/cyclin-dependent kinase (CDK) complexes immunoprecipitated with p27Kip1 are differentially modified by DXM addition: (a) G1 kinasic complexes (cyclin D/CDK4 or CDK6) associated with p27Kip1 are strongly decreased by DXM, (b) S-phase complexes (CDK2/cyclin E and A) remained stable or increased, and (c) the association of p27Kip1 with the phosphorylated forms of CDK1 is increased by DXM. In addition, CDK2 kinase activity was decreased in DXM-treated cells: we suggest that p27Kip1 might participate in inhibiting its catalytic activity. These results indicated that, in normal lymphoid cells, p27Kip1 may be involved in DXM antiproliferative effects. The increase of p27Kip1 expression and a decrease in G1 mitogenic factors, together with the redistribution of p27Kip1 to S/G2-M regulatory complexes, may explain the lengthening of G1 and S/G2 after DXM treatment in lymphocytes.

ckshs expression is linked to cell proliferation in normal and malignant human lymphoid cells.

Cyclin kinase sub-units (CKS) are known to interact with cyclin-dependent kinases (CDKs), but their functions are not completely understood and their expression in human tissues is not documented. For analyzing relationships of CKS with cell proliferation and/or with differentiation, we investigated the expression of ckshs1 and ckshs2 in normal and malignant human lymphoid cells. ckshs1 and ckshs2 expression appeared to be related to cell proliferation: (i) mRNAs increased with stimulation of normal peripheral-blood lymphocytes, and from the G1 to the SG2M phase in elutriated cells; (ii) P9 proteins were also induced by lymphocyte stimulation and were localized in nucleus where phosphorylated forms of CDK1 were also found; (iii) in vitro, the phosphorylated forms of CDK1 and CDK2 were preferentially linked to CKS. Among 45 patients presenting acute or chronic lymphoid malignancy, ckshs1 and ckshs2 mRNAs varied in a similar way and were significantly correlated to cell proliferation (p < 0.0001). When analysis was restricted solely to acute lymphoblastic leukemia (ALL) this correlation was still found and ckshs1 and ckshs2 were significantly more expressed in T-cell ALL than in B-cell-lineage ALL. These results confirm relationships between ckshs expression and cell proliferation, and pose the question of a link with cell differentiation.

Enhanced expression of p16ink4a is associated with a poor prognosis in childhood acute lymphoblastic leukemia.

The tumor suppressor gene p16ink4a is homozygously deleted in numerous T as well as in some B lineage acute lymphoblastic leukemia (ALL). We therefore analyzed the clinical and biological implications of this feature by studying p16ink4a expression in 58 cases of childhood ALL. mRNA and protein were significantly correlated and both appeared more highly expressed in B than in T lineage ALLs: 13 out of the 15 T cell ALLs did not show any p16ink4a expression. The main result of this study is the strong prognostic value of p16ink4a expression. When stratifying the patients in three groups according to p16ink4a expression, we observed in univariate analysis: (1) the shortest disease-free survival for patients presenting a high p16ink4a level; (2) contrasting with the good prognosis in the group of patients expressing p16ink4a at low level; (3) while cases without any expression of the inhibitor were associated with a medium course of the disease (P=0.0165). This prognostic value was confirmed by the multivariate analysis showing therapeutic regimen and p16ink4a protein expression as the only variables retained in the model. A specific metabolic profile related to cellular survival and proliferation was observed in each of the three p16ink4a expression groups. Among the cell cycle-related proteins we analyzed, only p21waf1 bcl-2 and CDK4 were significantly and positively correlated to p16ink4a. Furthermore, CDK6 was also strongly expressed in the group of cases with high p16ink4a level. An enhancement of p16ink4a, p21waf1 and bcl-2 was previously described in prolonged cellular survival, while aging cells showed a decrease in CDK4 expression. The concomitant high expression of the oncogenic protein CDK4 (and of CDK6), we observed, may amplify the leukemic advantage of prolonged lifespan blast cells by favoring cell progression through G1 phase. These data suggest that at least two mechanisms may be associated in the oncogenesis of very aggressive ALLs, ie deregulation of cell multiplication and prolonged blast lifespan.

Glucocorticoids induce G1 as well as S-phase lengthening in normal human stimulated lymphocytes: differential effects on cell cycle regulatory proteins.

In order to analyze dexamethasone effects on peripheral blood lymphocyte proliferation, we defined various experimental conditions: dexamethasone introduced (i) at the time of phytohemagglutinin stimulation, (ii) 48 h after the beginning of phytohemagglutinin stimulation, and (iii) on unstimulated lymphocytes. In stimulated lymphocytes, we observed an early G1 accumulation (P < 0.005), a delayed increase in the duration of S-phase (P < 0.03), and a consequent increase in cell-cycle duration. The expression of several cyclins, cyclin-dependent kinases (CDKs), and CDK inhibitors (CKIs) was modified. Cyclin D3, CDK4, and CDK6 involved in G1-phase control were significantly decreased under dexamethasone treatment whatever the level of stimulation of lymphocytes (stimulated or unstimulated PBL). Cyclin E and CDK2, acting in G1/ S-phase transition and S-phase regulation, decreased in stimulated lymphocytes before any modification of S-phase (P < 0.002). The expression of CKIs, mainly of p27Kip1, appeared to vary with the degree of cell stimulation: a decrease was observed on treated unstimulated lymphocytes, while p27Kip1 increased in dexamethasone-treated cells during stimulation. Our results indicate sequential modifications of the cell-cycle regulation by dexamethasone starting with an action on G1 followed by S-phase control modifications. The protein analysis pinpoints the major complexes concerned: CDK4 and CDK6/cyclin D are mainly involved in G1-phase modifications, while CDK2 and its partner, cyclin E, might be specifically involved in the lengthening of S-phase. The variations observed for p27Kip1 might amplify the functional effects of dexamethasone on kinasic complexes.

Phase I study of interleukin-6 in children with solid tumours in relapse.

The aim of this study was to evaluate the feasibility, toxicity and efficacy of escalating doses of subcutaneous recombinant interleukin-6 (IL-6) in children with solid tumours in relapse. Recombinant IL-6 was administered subcutaneously once daily for 14 consecutive days, with a 14 day follow-up period. The starting dose for IL-6 was 1 microgram/kg/day and was escalated in subsequent patients groups until 10 micrograms/kg. Doses were escalated every 3 patients, provided that grade III or IV organ toxicity did not occur at the preceding dose level. Twelve patients were treated, three at each dose level. No grade 3-4 major organ toxicity was observed. Flu-like symptoms and fatigue were the most common side effects. All these symptoms resolved after the end of IL-6 administration. Significant increases in acute-phase proteins (CRP [C reactive protein], fibrinogen) and ESR (Erthrocyte sedimentation rate) were observed in all patients. Stimulatory effects on thrombocytopoiesis were observed at every dose level, and were maximal at 5 micrograms/kg and 10 microgram/kg. There was no tumour response observed during IL-6 administration. Pharmacokinetic profiles performed in 3 patients are consistent with previous reports in adults. IL-6 is a promising new cytokine for paediatric oncology, in particular to increase thrombocyte counts. We recommend that further studies in children proceed at a dose of 5-10 micrograms/kg/day in a once or, better, twice daily administration.