Abnormalities of adherent layers grown from bone marrow of patients with myelodysplasia.

Myelodysplastic syndromes (MDS) are characterized by a clonal disorder of haemopoiesis with defective growth in vitro. The long-term culture system was used to examine aspects of stromal function in MDS patients. Primary long-term cultures of MDS bone marrow showed poor myelopoiesis with progenitors being detected for a median 3.5 weeks (n = 12) compared with 18 weeks in cultures of normal marrow (n = 10; P < 0.0001). The haemopoietic function of adherent layers was assessed in secondary co-cultures seeded with 5 x 10(6) cord blood mononuclear cells on irradiated normal (n = 27; aged 38-82 years) or MDS (n = 32; aged 41-86 years) adherent layers (> 60% confluent). The median myeloid progenitor number/cord blood co-culture was 135 in 5-week-old cultures with normal adherent layers and 22 in those with MDS layers (P < 0.0001). Myeloid colonies were detectable for a median 11 weeks with normal adherent layers and 6 weeks with MDS adherent layers (P < 0.0001); erythroid colonies were detectable for 7 weeks (normal) compared with 5 weeks (MDS) (P < 0.01). The differences in granulocyte-macrophage colony forming unit (CFU-GM) generation were not related to patient age. Cells from adherent layers of at least half of the primary normal (n = 48) and MDS (n = 26) long-term cultures expressed cytokines [interleukin (IL)-3, IL-1 beta, thrombopoietin (Tpo) and erythropoietin (Epo)] and receptors for retinoic acid (RAR alpha) [IL-2, IL-3, macrophage colony stimulating factor (M-CSF) (Fms) and Tpo (Mpl)]. Only IL-1 beta expression was reduced in week-5 MDS cultures compared with those from normal marrows (P < 0.05). There was also a highly significant decline in IL-1 beta expression in normal (but not MDS) adherent layers between week 5 and week 10. Thus, the adherent layers in cultures grown from MDS patients were haemopoietically defective and showed abnormal IL-1 beta expression.

Control of pH in human long-term bone marrow cultures with low-glucose medium containing zwitterion buffer lengthens the period of haemopoietic activity.

Primary long-term bone marrow cultures grown in 40 mM HEPES-buffered McCoy's 5A medium produced granulocyte-macrophage colony-forming units (CFU-GM) for a median of 9 weeks compared with 7 weeks with CO2/bicarbonate-buffered cultures. Reducing the medium glucose concentration (from 12.5 to 2.75 mM) extended the culture longevity to 17 weeks. The median period of erythroid colony detection increased from 6 to 8 weeks. Secondary cultures (5 x 106 cord blood mononuclear cells seeded on irradiated stroma) showed statistically similar myeloid and erythroid longevity to primary cultures. Improved control of medium pH significantly improved the capacity of long-term stromal layers to maintain stem cells in vitro.

A novel CSF-1 binding factor in a patient in complete remission following cytotoxic therapy for lymphoma.

A novel colony stimulating factor-1 (CSF-1) binding factor present in the serum from a patient in remission from lymphoma is described. Radioimmunoassay (RIA) repeatedly failed to detect circulating levels of CSF-1 in the peripheral blood system of this patient. Molecular analysis showed a normal CSF-1 gene structure by Southern blot analysis and a 46,XX karyotype by cytogenetic analysis. CSF-1 mRNA expression in peripheral blood leucocytes was confirmed using reverse transcriptase polymerase chain reaction analysis. Morphological analysis of bone marrow cells was normal and peripheral blood progenitor cell colony assays showed a pattern of growth within the normal range in response to CSF-1 alone and in combination with other cytokines. Analysis of the patient's plasma and conditioned media prepared from peripheral blood mononuclear and granulocytic cell fractions for their ability to bind 125Iodine-labelled CSF-1 revealed the presence of a plasma CSF-1 binding factor. This binding factor was not present in the patient's urine, because CSF-1 was detected by RIA and production of the binding factor by the patients peripheral blood white cells could not be demonstrated in vitro. To our knowledge, this is the first reported case of a soluble CSF-1 binding factor.

Non-dysplastic myelodysplasia?

Patients successfully treated for a malignancy with cytotoxic therapy have an increased risk of developing secondary myelodysplasia (MDS) and acute myeloid leukemia (AML). We report a patient in remission from Hodgkin's disease (HD) who remains hematologically normal 4 years after combination chemotherapy, but who has biological and genetic abnormalities characteristic of myelodysplasia. X-inactivation analysis using a 5' phosphoglycerate kinase (PGK) probe demonstrates polyclonal hematopoiesis, but cytogenetic analysis reveals a clonal population with a minority of metaphases having a 7q-deletion. NRAS mutations are not detectable 1 year after treatment, but are present in two separate clones (at codons 12 and 15) analyzed by single-stranded conformational polymorphism (SSCP), followed by cloning and sequencing 4 years after treatment. The presence of an activated NRAS with the same codon 12 mutation was independently confirmed by the nude mouse tumorigenicity assay. In vitro peripheral blood granulocyte-macrophage colony-forming units (CFU-GM) have changed from normal to undetectable levels while erythroid burst forming units (BFU-E) were significantly reduced on two occasions during the period of observation. These abnormalities are characteristic of MDS. Continued clinical follow-up will determine whether these evolving genetic and biological abnormalities pre-date the onset of clinical and morphological features of MDS.

Growth response to cytokines of circulating myeloid progenitors from myelodysplastic patients at diagnosis and more than 600 days after diagnosis.

Myeloid colony growth from the peripheral blood of myelodysplastic (MDS) patients was assessed for abnormal in vitro response to haemopoietic growth factors (granulocyte colony-stimulating factor (G-CSF), macrophage colony-stimulating factor (M-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-1 (IL-1), interleukin-3 (IL-3)). Abnormal colony growth, increased or reduced, was observed with each of the factors. No specific growth pattern was related to any of the French-American-British classification (FAB) types of disease. MDS patients who had survived greater than 600 days after diagnosis (n = 34) showed significantly fewer abnormalities than patients assayed at the time of diagnosis (n = 37), the major difference being less frequent stimulation of colony growth. These findings indicate that the time of sampling relative to diagnosis needs to be considered when interpreting the in vitro response to growth factors of myeloid colonies from MDS patients.

Colony-cluster ratio and cluster number in cultures of circulating myeloid progenitors as indicators of high-risk myelodysplasia.

Circulating myeloid progenitors were assayed in 172 normal subjects and 147 patients with myelodysplastic syndrome (MDS). Patients whose cultures had colony/cluster ratios (CCR) less than 0.3 had significantly shorter survival periods than comparable patients with CCR greater than 0.3. A second prognostic indicator, which complemented CCR, was identified in patients with less than 5% blasts. Median survival was significantly reduced in patients with greater than 15 clusters/ml blood despite colony and cluster numbers being predominantly within the normal range. Characteristic differences were found in three FAB groups large enough to allow statistical analysis. Survival amongst patients with refractory anaemia with excess of blasts (RAEB) was related to CCR and was independent of cluster number. Amongst sideroblastic patients (SA) survival related only to cluster number. Refractory anaemia (RA) patients included individuals in both high-risk groups with only three patients out of 64 showing both features. Amongst all the MDS patients, those with CCR greater than 0.3 and less than 15 clusters/ml blood formed a low-risk group (n = 60) with a relatively good prognosis of whom 85% survived the study period (median duration 938 d) including 94% of those in this group with less than 5% marrow blast cells.

Stimulation of myeloid colony growth from peripheral blood by medium with an initially low osmolality.

The clonal growth of myeloid colonies from peripheral blood was maximal when cultures were established with an initial osmolality of 220 mosmol/kg which increased during incubation as a result of partial drying. When osmolality was stabilized by secondary humidification, the optimum osmolality was 270 mosmol/kg, but growth was always two- to fivefold less than similar cultures established at low osmolality and incubated on an open shelf. Cultures established at 270 mosmol/kg or above were statistically similar whether or not drying was eliminated. Maximum colonies were apparent after 14 days incubation under both conditions; addition of conditioned medium did not alter the pattern of growth. The greater sensitivity of cultures established at 220 mosmol/kg is advantageous when assaying circulating progenitors in pathological conditions where a low number of granulocyte/macrophage colony-forming units is common.

Effect of 5637-conditioned medium on peripheral blood granulocyte-macrophage progenitors in normal subjects and patients with the myelodysplastic syndrome.

There was no overall increase in PB-CFU-GM from normal subjects, or MDS patients, when exogenous CSA (5637-CM) was added to the culture medium. However, there was a sub-group of MDS patients (seven of 35) whose PB-CFU-GM numbers were significantly stimulated by 5637-CM. In addition, there were 11 (out of 48) MDS patients with undetectable PB-CFU-GM in assays without exogenous CSA but only two when 5637-CM was added (p less than 0.01). This sub-group is of particular interest as it is known that those without detectable PB-CFU-GM tend to have significantly shorter survival times than others. The mechanism of the functional abnormality is yet to be determined.

Undetectable peripheral blood CFU-GM as a prognostic indicator in myelodysplastic syndrome.

The survival of MDS patients without detectable circulating CFU-GM (median = 188 days) was significantly lower than those in whose peripheral blood CFU-GM were detected (median greater than 1000 days) (p less than 0.01). About half of those with detectable PB-CFU-GM died within 2 yr whilst the remainder survived more than ca 3 yr. There was no significant difference in the distribution of patients having 0-5% and greater than 5% marrow blast cells within the three groups.

Peripheral blood granulocyte-macrophage progenitors in patients with the myelodysplastic syndromes.

Peripheral blood (PB) granulocyte-macrophage progenitor (CFU-GM) growth was measured in 47 normal subjects and, together with bone marrow CFU-GM and blast cell numbers, in 45 newly diagnosed patients with myelodysplastic syndromes (MDS). Both PB colony and cluster numbers were significantly reduced in MDS patients. In patients with greater than 5% marrow blasts there was a negative correlation between blast cell numbers and PB colony growth. Bone marrow and PB colony growth were also well correlated in this group. Poor growth of PB CFU-GM appears to be more closely related to prognosis than does bone marrow progenitor growth.

Effect of medium osmolality on the growth of murine continuous marrow cultures.

The effect of medium osmolality was examined in primary, continuous bone-marrow cultures established from TO strain mice. The non-adherent cell population increased exponentially between weeks 2 and 5 and thereafter declined steadily. The number of CFU-GM followed a similar pattern but showed greater variability. The optimum osmolality in 4 week old cultures was found to be about 345 mosmol/kg which was higher than the plasma osmolality (n = 20; mean = 323.3 mosmol/kg; range = 313-331). Maximum non-adherent cell numbers were found at about 345 mosmol/kg (better than half-maximum between 320 and 370 mosmol/kg). CFU-GM numbers in the culture supernatant were maximal at about 355 mosmol/kg (better than half-maximum between 320 and 400 mosmol/kg). An adherent layer developed over a wider range of osmolality than supported granulopoiesis (better than half-maximum between 258 and 402 mosmol/kg). It was necessary to increase the osmolality of Fischer's medium in order to obtain maximum growth.

Acute depression of serum folate in surgical patients during preoperative infusion of ethanol-free parenteral nutrition.

Total parenteral nutrition for 48 h with Aminoplex solutions and a calorie source led to statistically similar, severe depression of the serum folate concentration (by 63-77%) irrespective of whether the calorie source contained ethanol, sorbitol, lipid or glucose. Folate levels were also depressed following infusion of other ethanol-free regimens (Freamine and glucose, 59%; Vamin and glucose, 44%). Least depression (31%0 followed use of Aminosol and glucose, a regimen based on casein hydrolysate. It seems important, therefore, to give prophylactic folate to all patients receiving parenteral nutrition, regardless of whether the calorie source is ethanol.

Aminoacid infusion induced depression of serum folate after cholecystectomy.

Aminoacid solutions were given to cholecystectomy patients for a period of 48 h starting 24 h after the operation. Infusion of the aminoacid component of parenteral nutrition regimens, without calorie source, frequently led to an acute fall in the serum folate concentration. The degree of depression varied with different commercial aminoacid solutions. (Aminoplex 14 = c.60%, Vamin = 30%, Aminosol = c.0%). The folate depression by Vamin infusion was unaffected by supplementation with L-methione but was alleviated by the addition of glycine. As a consequence of surgical operation itself, folate values fell by 37% during the 24 h following cholecystectomy. This seems likely to increase the risk of haematological toxicity (i.e. acute leukopenia and thrombocytopenia) in patients given folate-free parenteral nutrition after a surgical operation.

Double-blind randomised controlled trial of folate treatment before conception to prevent recurrence of neural-tube defects.

A randomized controlled double-blind trial was undertaken in south Wales to prevent the recurrence of neural-tube defects in women who had had one child with a neural-tube defect. Sixty women were allocated before conception to take 4 mg of folic acid a day before and during early pregnancy and 44 complied with these instructions. Fifty-one women were allocated to placebo treatment. There were no recurrences among the compliant mothers but two among the non-compliers and four among the women in the placebo group. Thus there were no recurrences among those who received supplementation and six among those who did not; this difference is significant (p = 0.04). It is concluded that folic acid supplementation might be a cheap, safe, and effective method of primary prevention of neural-tube defects but that this must be confirmed in a large, multicentre trial.

Effect of Tween 80 on incubation period of Lactobacillus casei assay of serum folate.

Contamination of folic acid assay medium with oleate led to a lag phase when Lactobacillus casei cultures were grown on folic acid but not when whole serum provided the folate source. This led to high assay values when using 20-hour incubation periods but satisfactory results were obtained after 48 hours' incubation. The addition of Tween 80 to the medium eliminated the lag phase and the assay results were consistent after 13, 24, and 48 hours' incubation.

Acute folate deficiency associated with intravenous nutrition with aminoacid-sorbitol-ethanol: prophylaxis with intravenous folic acid.

Preoperative folate levels were initially normal in 30 patients with gastrointestinal tract disease but fell within 48 h by 60-95% in 20 patients who received intravenous nutrition for 6-12 d with aminoacid-sorbitol-ethanol (ASE). This depression persisted in patients not given folate supplements. Folate levels in 10 control patients not given ASE showed only minimal decline. Haematological changes were reduced to a minimum in 10 patients given 0.5 mg i.v. folic acid daily whilst eight unsupplemented patients showed evidence of megaloblastic haemopoiesis. Three of these eight patients developed thrombocytopenia and/or leukopenia which was fatal in one patient.

Continuous-flow automation of the Lactobacillus casei serum folate assay.

A method is described for the continuous-flow automation of the serum folate assay using Lactobacillus casei. The total incubation period is approximately four hours. The growth response of the organism to folate is estimated by measuring the rate of reduction of 2,3,5-triphenyl tetrazolium chloride (TTC). A simple continuous culture apparatus is used to grow the inoculum. Supplementation of the assay medium is necessary to obtain parallel results. A statistical assessment shows a favourable comparison with the whole-serum tube assay using a chloramphenicol resistant strain of L. casei. The method is less sensitive to inhibitory substances than the tube assay.