Proteus syndrome with renal involvement

We report on a case of an 11 year old girl with Proteus syndrome involving both kidneys morphologically and functionally. The patient had hemihypertrophy, scoliosis, overgrowth of lower extremities, partial macrodactyly, subcutaneous lipoma, and pigmented nevi compatible with the Proteus syndrome. She also had overgrowth of the right kidney with mildly impaired function and a small left kidney with severe dysfunction. Morphological involvement of visceral organs has previously been described; however, deteriorating renal function has not previously been reported in this syndrome.

Transport mechanisms of enoxacin in rat brush-border membrane of renal cortex: interaction with organic cation transport system and ionic diffusion potential dependent uptake

The mechanism of the renal transport of enoxacin (ENX) has been investigated using brush-border membrane vesicles (BBMVs) isolated from the rat renal cortex. The initial rate and time-course of ENX uptake were quite dependent upon the medium pH (pH 5.5 > pH 7.5). The pH dependence was in accordance with the degree of cationic form. Carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP) affected the transient uphill transport of ENX across the renal brush-border membrane in the presence of an outward-directed H(+)-gradient. The initial uptake was saturable, and transport kinetic parameters were given for a Km and Vmax of 0.59mM and 1.37nmol/(mg protein)/30s, respectively. On the other hand, an outward H(+)-gradient (pHin = 5.5, out = 7.5) dependent uptake of ENX was partially decreased by the voltage-clamped BBMVs. Furthermore, a valinomycin-induced K(+)-diffusion potential (interior negative) was found to increase the uptake of ENX at pH 5.5, which is cationic form-rich. These results suggest that ENX uptake participates in not only the H+/organic cation antiport system for organic cation secretion but also the ionic diffusion potential (interior negative) dependent permeation through the membrane.

[Extramedullary relapse in childhood acute lymphoblastic leukemia: experience at the Kyushu Yamaguchi Children's Cancer Study Group]

Since October 1984, children with acute lymphoblastic leukemia (ALL) were treated with six protocols of the Kyushu-Yamaguchi Children's Cancer Study Group (KYCCSG). We reviewed cases in whom the first relapse sites of ALL were CNS, testis and ovary. Between October 1984 and July 1993, 302 children with ALL were registered. Ten children (3.3%) developed CNS-L. After treatment for CNS-L, four cases were alive, however one of them has had several episodes of CNS-L. Testicular relapse occurred in 8 boys and three of them are alive. One girl developed ovarian and uterine involvement 7 months after completion of treatment and she survived after intensive chemotherapy for an additional 3 years. Before September 1990, children with ALL received 18Gy cranial irradiation in standard risk group and 24Gy in high risk group. Since October 1990, children with ALL received no cranial irradiation in low risk group, 15Gy in intermediate risk group and 18Gy in high risk group. Incidence of CNS-L has not increased in children treated with protocols, in which cranial irradiation was reduced.

Fas antigen-mediated DNA fragmentation and apoptotic morphologic changes are regulated by elevated cytosolic Ca2+ level

More than 80% of cells of the human B cell line FMO, which expresses the Fas Ag, underwent apoptosis within 2 h after addition of an anti-Fas mAb. Rises in cytosolic Ca2+ concentration ([Ca2+]i) and their effects were investigated by imaging individual cells continuously and introducing Ca2+ chelators to the outside and/or inside of the cell. The typical Ca2+ response consisted of 1) an early [Ca2+]i rise (basal [Ca2+]i, 85 to 110 nM; peak, 140 to 200 nM; duration, 15 to 20 min), 2) sustained elevation of [Ca2+]i at 140 to 150 nM, 3) a second Ca2+ rise (200 to 500 nM, 15 to 30 min) between 1.5 and 2 h, and 4) a large [Ca2+]i rise (1.2 to 2.0 microM) between 2 and 4 h after addition of mAb. Responses 1, 3, and 4 were mainly caused by Ca2+ entry, and response 2 involved intracellular Ca2+ release from stores. Apoptosis could be induced by mAb even in Ca(2+)-deprived external medium. Chelating cytosolic Ca2+ revealed that the [Ca2+]i rise is a prerequisite for fragmentation of DNA and chromatin, and is also necessary for fragmentation of cells. The critical [Ca2+]i was 140 to 150 nM and a sustained [Ca2+]i rise was more effective. A [Ca2+]i rise itself (without mAb) was ineffective. About 20% of mAb-treated cells showed chromatin condensation at the periphery of the nucleus (possibly an earlier stage of nuclear change) and a bubble-like cell shape even when [Ca2+]i was held below 100 nM. Thus, Ca2+ is mobilized immediately after Fas stimulation and functions as a key factor causing advanced apoptotic changes. Response 4 was related to secondary necrosis.

Fas antigen-mediated DNA fragmentation and apoptotic morphologic changes are regulated by elevated cytosolic Ca2+ level

More than 80% of cells of the human B cell line FMO, which expresses the Fas Ag, underwent apoptosis within 2 h after addition of an anti-Fas mAb. Rises in cytosolic Ca2+ concentration ([Ca2+]i) and their effects were investigated by imaging individual cells continuously and introducing Ca2+ chelators to the outside and/or inside of the cell. The typical Ca2+ response consisted of 1) an early [Ca2+]i rise (basal [Ca2+]i, 85 to 110 nM; peak, 140 to 200 nM; duration, 15 to 20 min), 2) sustained elevation of [Ca2+]i at 140 to 150 nM, 3) a second Ca2+ rise (200 to 500 nM, 15 to 30 min) between 1.5 and 2 h, and 4) a large [Ca2+]i rise (1.2 to 2.0 microM) between 2 and 4 h after addition of mAb. Responses 1, 3, and 4 were mainly caused by Ca2+ entry, and response 2 involved intracellular Ca2+ release from stores. Apoptosis could be induced by mAb even in Ca(2+)-deprived external medium. Chelating cytosolic Ca2+ revealed that the [Ca2+]i rise is a prerequisite for fragmentation of DNA and chromatin, and is also necessary for fragmentation of cells. The critical [Ca2+]i was 140 to 150 nM and a sustained [Ca2+]i rise was more effective. A [Ca2+]i rise itself (without mAb) was ineffective. About 20% of mAb-treated cells showed chromatin condensation at the periphery of the nucleus (possibly an earlier stage of nuclear change) and a bubble-like cell shape even when [Ca2+]i was held below 100 nM. Thus, Ca2+ is mobilized immediately after Fas stimulation and functions as a key factor causing advanced apoptotic changes. Response 4 was related to secondary necrosis.

In vitro and in vivo activity of DQ-2556, a new cephalosporin

The antibacterial activity of DQ-2556, a new cephalosporin, was compared with that of cefepime, cefpirome, and other antibiotics. DQ-2556 was more active than cefepime and less active than cefpirome against gram-positive bacteria. DQ-2556 was the most active compound against most members of the Enterobacteriaceae, such as Escherichia coli, Klebsiella, Proteus, Salmonella spp., and Shigella spp., and Haemophilus influenzae and Neisseria gonorrhoeae. In vivo, treatment with DQ-2556 was as or more efficacious than that with reference compounds in both systemic and respiratory tract infections in mice. DQ-2556 achieved higher concentrations than cefepime and cefpirome in lungs and kidney of mice.

Evidence for inositol tetrakisphosphate-activated Ca2+ influx pathway refilling inositol trisphosphate-sensitive Ca2+ stores in hamster eggs

To identify the Ca2+ influx pathway responsible for maintaining Ca2+ oscillations in hamster eggs, changes in intracellular Ca2+ concentration ([Ca2+]i) were recorded using the Fura-2 fluorescent imaging technique during iontophoretic injection of inositol phosphates under voltage clamp. Both inositol 1,4,5-trisphosphate (InsP3) and 1,3,4,5-tetrakisphosphate (InsP4) caused repetitive Ca2+ transients when injected continuously into eggs, although the latter was much less effective. These Ca2+ transients were inhibited by the monoclonal antibody 18A10 to the InsP3 receptor/Ca2+ channel. In Ca(2+)-free medium, InsP4-induced Ca2+ transients were absent or much less frequent than in normal medium. A small but persistent increase in [Ca2+]i during InsP4 injection was revealed when Ca2+ uptake into InsP3-sensitive Ca2+ stores was suppressed by thapsigargin. This Ca2+ rise is due to Ca2+ entry, but not Ca2+ release, because it was: (i) increased by raising the extracellular Ca2+ concentration and abolished in Ca(2+)-free medium; (ii) larger at more negative membrane potentials which provide greater electrical driving force for Ca2+ entry; and (iii) not affected by 18A10. A moderate dose of InsP3 did not cause substantial Ca2+ entry, as tested in thapsigargin- and 18A10-treated eggs. InsP4 facilitated the restoration of Ca2+ stores after Ca2+ releases induced by pulsatile InsP3 injections. Thus, we obtained evidence for a Ca2+ influx pathway activated by InsP4 which provides Ca2+ to refill InsP3-sensitive Ca2+ stores in intact cells.

Production and expression of granulocyte- and macrophage-colony-stimulating factors in newborns: their roles in leukocytosis at birth

We analyzed the production and expression of three colony-stimulating factors (CSFs) in neonates to clarify the mechanism of leukocytosis at birth. Serial blood samples (n = 23) were collected from mothers, cord blood, and from newborn infants on days 1, 5, and 30 after birth. The serum levels of granulocyte-CSF (G-CSF), granulocyte/macrophage/CSF (GM-CSF) and macrophage-CSF (M-CSF) were measured by ELISA. The G-CSF levels on day 1 after birth were significantly higher than those thereafter, and they were also higher in the mothers than those on days 5 and 30 after birth. The GM-CSF levels did not change significantly during the neonatal period. The serum M-CSF levels were higher on postnatal day 1 than at other times, and gradually decreased thereafter. To confirm the production sites of G-CSF and M-CSF, the mRNA for these CSFs in peripheral mononuclear cells (MNCs) from healthy adults, mothers, and cord blood were analyzed by PCR. The expression of G-CSF and GM-CSF mRNA was undetectable in MNCs from adults, mothers, and cord blood, while these cells expressed low levels of M-CSF mRNA. After stimulation with lipopolysaccharide or phorbol myristate acetate, the MNCs expressed high levels of G-CSF and GM-CSF mRNA. The levels of G-CSF PCR products in cord MNCs were lower than those in adult and maternal MNCs. The expression of M-CSF mRNA was virtually unchanged by stimulation. To detect the localization of G-CSF and M-CSF in the placenta and umbilical cord, these tissues were immunocytochemically stained with anti-G-CSF and anti-M-CSF antibodies, G-CSF and M-CSF were expressed in trophoblasts and decidual stromal cells, whereas the umbilical cord did not express these CSFs. Moreover, large amounts of G-CSF and M-CSF were detected in the supernatant of cultured trophoblasts and decidual stromal cells. The expression of G-CSF and M-CSF in these cells was confirmed by PCR. These findings suggested that G-CSF and M-CSF produced in the placenta (trophoblasts and decidual stromal cells) are the major factors that induce leukocytosis in newborn infants at birth.

[A case of suprasellar arachnoid cyst followed up for a long time]

We followed a case of suprasellar arachnoid cyst for 12 years. The patient was a sixteen-year-old girl without particular problems in her general condition. She showed optic atrophy in both eyes and optic nerve hypoplasia with an inferotemporal quandranopsia in the left eye. A suprasellar arachnoid cyst communicating with the tubarachnoid space was found to extend into the sella turcica as an empty sella. A cyst wall was resected and a cyst-peritoneal shunt performed. After 12 years from the operation, sensitivity was slightly depressed in the visual field where it had already been disturbed. Although there are few reports in the literature on involvement of the optic nerves and chiasma by suprasellar arachnoid cysts, papilledema and optic atrophy are often found in children, and infero-temporal quandranopsia or homonymous hemianopsia have been reported. Visual field defects were most likely caused by compression of the optic nerve by cyst or prolonged papilledema. We also suspect that some kind of disturbance to the optic nerve occurred during extension of the arachnoid cyst as an empty sella, or during formation of arachnoid cyst in the fetus stage.

Cerebellar ataxia in patients with Leber's hereditary optic neuropathy

We report the cases of a mother and son with Leber's hereditary optic neuropathy (LHON), where a point mutation of mitochondria DNA from guanine to adenine on nucleotide position 11778 was verified. Both also had cerebellar ataxia and dysarthria and in both cases cerebellar atrophies were detected by computed tomography or magnetic resonance imaging. It was not possible to elucidate the relationship between LHON and the cerebellar atrophy, but it should be kept in mind that various neurological complications may occur in LHON.

Effects of 1-amino-5-bromouracil on the benzodiazepine-GABAA receptor complex

We investigated the effects of 1-amino-5-bromouracil on the benzodiazepine-gamma-aminobutyric acid (GABA)A receptor complex to elucidate its central action. 1-Amino-5-bromouracil neither displaced nor enhanced [3H]muscimol, [35S]t-butylbicyclophosphorothionate (TBPS), or [3H]dehydroepiandrosterone sulfate binding to the rat brain synaptosomal membranes. The anesthesia induced by 1-amino-5-bromouracil was potentiated by diazepam, pentobarbital, and muscimol, and was antagonized by picrotoxin but not by bicuculline. 1-Amino-5-bromouracil protected mice from picrotoxin-induced seizure and slightly ameliorated TBPS-induced seizure, but did not antagonize bicuculline-induced seizure. Diazepam antagonized both the bicuculline- and the picrotoxin-induced seizure, and pentobarbital antagonized the picrotoxin- and the TBPS-induced seizure. Our in vivo studies suggest that part of the central action of 1-amino-5-bromouracil is concerned with the benzodiazepine-GABAA receptor complex including the chloride channel.

Effects of a non-xanthine adenosine antagonist, CGS 15943, and a phosphodiesterase inhibitor, Ro 20-1724, in a light/dark test in mice

We investigated the effects of a non-xanthine adenosine receptor antagonist, CGS 15943, and a non-xanthine phosphodiesterase inhibitor, Ro 20-1724, in a light/dark test in mice. CGS 15943 at a dose of 50 mg/kg had no significant effects on any parameter, although at 10 mg/kg it significantly antagonized decreases in locomotion and rearing behavior induced by treatment with A1 and A2 selective agonists, N6-cyclopentyladenosine and CGS 21680 respectively. On the other hand, Ro 20-1724 decreased locomotion and rearing behavior in the light and dark zones, number of shuttle crosses between both zones, and the time spent in the light zone dose-dependently at doses ranging from 1-10 mg/kg. In conclusion, the phosphodiesterase inhibitor decreased all parameters in the light/dark test, while the adenosine antagonist showed no effect.

Effects of xanthine derivatives in a light/dark test in mice and the contribution of adenosine receptors

We investigated the effects of adenosine receptor antagonists, caffeine, theophylline, 8-phenyltheophylline, and 8-cyclopentyl-1,3-dipropylxanthine (DPCPX), in a light/dark test in mice. All antagonists decreased the time spent in the light zone in this test, which suggested that these compounds have anxiogenic effects. The anxiogenic effects of theophylline were reduced by pretreatment with CGS 21680, an A2-selective agonist, but not by N6-cyclopentyladenosine (CPA), an A1-selective agonist. However, the antagonism of the theophylline-induced anxiogenic effects by CGS21680 was only observed in the time spent in the light zone, and DPCPX-induced anxiogenic effects were neither reversed by CGS 21680 nor by CPA. Finally, it is notable that xanthine-derived adenosine antagonists tested here commonly showed anxiogenic effects in the light/dark test in mice. It is suggested that there is a minor contribution of adenosine receptors to these effects, although theophylline-induced anxiogenic effects were antagonized by an A2 receptor agonist.

Effect of 1-amino-5-bromouracil on brain monoamine metabolism in rats

The effect of 1-amino-5-bromouracil (ABU), a novel central-acting agent, on monoaminergic neurotransmitter levels of rat brain was investigated. Under the nonstressed condition, ABU (20 and 30 mg/kg intraperitoneally [IP]) did not affect monoamine metabolism, whereas diazepam (5 mg/kg IP) increased the 3-methoxy-4-hydroxyphenylethylene glycol (MHPG)/noradrenaline (NA) ratio. One-hour immobilization stress increased the MHPG/NA ratio in various brain regions of drug-naive rats, but did not increase the homovanilic acid (HVA) plus 3,4-dihydroxyphenylacetic acid (DOPAC)/dopamine (DA) ratio or the 5-hydroxyindole acetic acid (5-HIAA)/serotonin (5-HT) ratio. Pretreatment with ABU or diazepam suppressed the activation of noradrenergic neurons induced by immobilization stress. By contrast, electric foot shock stress increased the MHPG/NA and HVA+DOPAC/DA ratios. Pretreatment with ABU or diazepam suppressed the activation of noradrenergic and dopaminergic cortical neurons by electric foot shock stress. These results indicate that these two physiologic stresses affected monoaminergic neurons differently and that their effects were suppressed by ABU and diazepam.

1H-NMR analysis of nerve edema in the streptozotocin-induced diabetic rat

To define the existence of intracellular hydration caused by metabolic derangements in the excised sciatic nerves of diabetic rats quantitatively, relaxation times (T1, T2) and fraction of intracellular water content were measured with 1H-nuclear magnetic resonance (NMR) spectroscopy in normal rats (control group, n = 10), streptozotocin (STZ)-induced (50 mg/kg, i.v.) diabetic rats (DM group, n = 10), STZ-induced diabetic rats treated with an aldose reductase inhibitor (ARI, Epalrestat, 100 mg/kg) (ARI group, n = 8), and STZ-induced diabetic rats treated with insulin (insulin group, n = 4). For selective measurement of intracellular relaxation times, the inversion recovery (IR) method for conventional T1 and Carr-Purcell-Meiboom-Gill method for T2 were used with an aqueous chemical shift reagent, 10 mmol/L dysprosium triethylenetetramine-N,N,N',N",N"',N"'-hexaacetic acid, resulting in distinct separation of intracellular (Schwann cell, axon, endothelial cell, and pericyte) and extracellular waters under the isotonic condition of the rat sciatic nerves. Furthermore, a new method of driven-equilibrium single-pulse observation of T1 (DESPOT) was used for rapid measurement of T1 for the purpose of clinical application on magnetic resonance imaging (MRI). T1 values measured by the DESPOT and IR methods were significantly correlated (p < 0.001). Total and intracellular water contents, sorbitol contents, and relaxation times of the sciatic nerve taken from the DM group were significantly elevated (p < 0.01), while myoinositol (p < 0.01) and extracellular water (p < 0.05) contents were significantly decreased as compared with the control group. Both insulin and ARI treatments significantly improved relaxation times as compared with those in the DM group (p < 0.05-0.01). Relaxation times correlated positively with total water (T1, p < 0.05-0.01; T2, p < 0.01), intracellular water (T1, p < 0.001; T2, p < 0.001), and sorbitol (T1, p < 0.001; T2, p < 0.001) contents of the excised nerve. Sorbitol content correlated positively with total and intracellular water contents (p < 0.01) but negatively with extracellular water content (p < 0.05). These findings indicated that sorbitol itself and/or secondary sodium accumulation caused by an increase in sorbitol may be a major contributor to the increase in intracellular hydration and prolonged relaxation times associated with hyperglycemia, which are reversible with insulin or ARI treatment. It was also suggested that rapid T1 measurement would provide new insights into the pathogenesis of human diabetic neuropathy as a non-invasive evaluation method on MRI.

A variant form of hypobetalipoproteinaemia associated with ataxia, hearing loss and retinitis pigmentosa

A six-year-old Japanese boy had ataxia, mental retardation, peripheral neuropathy, proximal myopathy, hearing loss, retinitis pigmentosa and deficiencies in apolipoprotein AI, B, CII and CIII. His clinical features except for hearing loss resembled those of abetalipoproteinaemia or symptomatic hypobetalipoproteinaemia, but his apolipoprotein abnormalities were distinct from these disorders. He had apolipoprotein B-100 with a normal molecular weight. Although most of his neurological manifestations were compatible with those of vitamin E deficiency, their early onset and the presence of hearing loss was unusual for that condition. There has been slight deterioration of ataxia during two years follow-up despite high-dose vitamin E supplementation. Other abnormalities in lipid metabolism might be associated with the neurological damage in this case.

[Enhancement of rectal absorption of rifampicin by sodium para-aminosalicylate dihydrate in human subjects]

The suppositories of rifampicin (RFP) containing sodium para-aminosalicylate dihydrate (PAS-Na) were prepared in order to enhance the rectal absorption of RFP. By the addition of PAS-Na, the in vitro release of RFP from the suppositories was enhanced and the hardness of the suppositories decreased. The rectal absorption of RFP from the suppositories containing no PAS-Na (control suppositories) was significantly lower compared to oral administration of it (26%) in human subjects. When PAS-Na was added to the suppository (300 mg), both the area under the plasma concentration-time curve (AUC) and the maximum plasma concentration (Cmax) increased significantly compared to those of the control suppositories. The rectal absorption of PAS-Na itself from the suppositories seemed to be fast. PAS-Na might increase the absorption of RFP dissolved in the rectal fluid from the suppositories, but not affect the undissolved RFP.

Discriminative stimulus properties of diazepam and the novel anxiolytic agent 1-amino-5-bromouracil in rats

A stimulus cue of 1-amino-5-bromouracil (ABU, CAS 127984-93-4) was compared with that of diazepam (DZP) using a drug discrimination paradigm in rats. Groups of rats were trained to discriminate DZP (1 mg/kg i.p.) or ABU (20 mg/kg i.p.) from vehicle. Generalization of the cue of the trained drug to pentobarbital was shown in DZP- and ABU-trained rats at a dose of 5 mg/kg. The stimulus cue of ABU showed a tendency to generalize to DZP in ABU-trained rats but generalization of that of DZP to ABU in DZP-trained rats was only partial. Also partial generalization of that of DZP to imipramine and clonidine was found but not to 8-hydroxy-2-(di-n-propylamino)tetraline (8-OH-DPAT) in DZP-trained rats. Full generalization of the stimulus cue of ABU to imipramine and partial generalization of that of ABU to clonidine and 8-OH-DPAT was observed in ABU-trained rats. The results suggest that the discriminative stimulus properties of ABU differ from those of DZP.

N-methyl-D-aspartate receptor-mediated, prolonged afterdischarges of CA1 pyramidal cells following transient cerebral ischemia in the rat hippocampus in vivo

We previously reported the post-ischemic potentiation (PIP) of synaptic efficacy in hippocampal Schaffer collateral/CA1 responses of the rat beginning at 6-8 h following 12 min transient cerebral ischemia in vivo. The present study demonstrated that repetitive stimulation with a relatively low frequency (5 Hz, 6 s), which produced short-lasting afterdischarges (ADs; duration, 4.49 +/- 4.26 s; n = 7) in sham-controls, resulted in prolonged ADs (duration, 26.33 +/- 12.63 s; n = 6; P < 0.001) at the same period after ischemia. The PIP was not affected by 2-amino-5-phosphonovalerate (APV) administered via microdialysis at 7 h post-ischemia. The prolonged ADs in response to repetitive stimulation were, however, reversed to short-lasting ADs (duration, 7.13 +/- 1.44 s; n = 4; P < 0.02) by the same procedure, leaving the response to single stimulation unaffected. These findings suggest that, during the reperfusion period, Ca2+ influx into the CA1 pyramidal cells can be greatly increased through N-methyl-D-aspartate (NMDA) receptor-coupled ion channels if appropriately timed multiple synaptic inputs bombard these cells. Such Ca2+ influx may contribute to delayed death of CA1 pyramidal cells after transient cerebral ischemia if synaptic activity is maintained at relatively high levels during the reperfusion period.

[Clinical trial of protocol-AL851 for children with high-risk acute lymphoblastic leukemia. Kyushu Yamaguchi Children's Cancer Study Group (KYCCSG)]

Fifty five children diagnosed as having high-risk acute lymphoblastic leukemia (ALL) between 1985 and 1988 were treated with protocol AL851. The agents used in the protocol were as follows: induction therapy: vincristine (VCR), prednisolone, daunorubicin (DNR) and l-asparaginase, consolidation therapy: an intermediate-dose methotrexate (MTX), central nervous system (CNS) leukemia prophylaxis: intrathecal MTX and 24Gy cranial irradiation, reinduction therapy: VCR, adriamycin, dexamethasone and high dose cytarabine (AraC), maintenance therapy: 6-mercaptopurine, cyclophosphamide, MTX, DNR, VCR and AraC. Patients received chemotherapy for 3 years after achieving complete remission (CR). CR was obtained in 51 patients (92.7%). Twenty-four of them relapsed after achieving CR (bone marrow 16, CNS 3 and testis 5). At median follow-up of 79 (range 64-102) months, the estimated 8-year disease free survival rate was 49.1 +/- 6.7%. Four patients relapsed at bone marrow during the first 6 months of the treatment, indicating that more intensive combination chemotherapy should be included in earlier stage of the protocol. The high incidence of testicular relapse (14.3% in boys) suggests that high-dose MTX or AraC should be needed for improvement of the prognosis of high-risk ALL patients.