Lithium is an ineffective therapy for human cyclic hematopoiesis

Cyclic hematopoiesis is a rare disease in man in which severe neutropenia recurs at 21-day intervals with associated illness. Because lithium carbonate therapy has been shown to eliminate cyclic hematopoiesis in grey collie dogs, we examined the effects of lithium treatment on five patients with this disease. With lithium levels maintained between 0.5 and 1.0 meq/liter, these patients showed no change in the fluctuations of their neutrophil counts. We conclude that lithium carbonate is not a simple cure for human cyclic hematopoiesis.     

Cyclic hematopoiesis: effects of lithium on colony-forming cells and colony-stimulating activity in grey collie dogs

The cycling of blood cell counts in grey collie dogs with cyclic hematopoiesis can be eliminated by treatment with oral lithium carbonate. To explore the mechanism by which lithium alters this stem cell disorder, studies of bone marrow granulocyte-macrophage progenitor cells (CFU-C), neutrophil colony-forming cells (neutrophilic CFU-C), and colony-stimulating activity (CSA) were performed. In untreated dogs, the proportions of CFU-C were found to fluctuate cyclically, but the cyclic fluctuations in neutrophil colony-forming cells were even more marked, with numbers decreasing to undetectable levels during each period of neutrophilia. Dogs on lithium, however, did not cycle the numbers of total or neutrophilic CFU-C. Tritiated thymidine suicide rates were not altered by treatment with lithium. Serum CSA levels and bone marrow cell elaboration of CSA were not increased by lithium. These studies suggest that lithium corrects cyclic neutropenia by a direct effect on the differentiation and proliferation of CFU-C; normalization of the proportion of CFU-C that enter neutrophilopoiesis appears to be an important effect of the lithium therapy.     

Lithium augments GM-CSA generation in canine cyclic hematopoiesis

Cyclic hematopoiesis in gray collie dogs can be cured by lithium treatment. We examined the mechanism of lithium's effect by developing an assay for the canine equivalent of GM-CSF (called GM-CSA). Phytohemagglutinin (PHA)-stimulated canine blood mononuclear cells produce GM-CSA in a dose-dependent manner; this GM-CSA stimulates more neutrophil-containing colonies than does endotoxin-treated dog serum. Production of GM-CSA by PHA-stimulated normal dog cells was not altered by lithium. However, cells from gray collies during their neutrophilic period increased their GM-CSA when lithium (2 mEq/L) was added to low doses of PHA, whereas neutropenic gray collie cells did not. These data suggest that lithium could modulate cyclic hematopoiesis by increasing intramedullary GM-CSA at the time when marrow neutrophilic progenitor cells are at their nadir.     

Effect of adenine nucleotides on granulopoiesis and lithium-induced granulocytosis in long-term bone marrow cultures

Studies were undertaken to evaluate the role of adenine nucleotides in regulating hematopoiesis using a long-term liquid culture system. In contrast to early investigations using clonogenic stem cell assays, where inhibitory effects were observed, adenosine and adenosine-5'-monophosphate (AMP) were found to stimulate myelopoiesis whereas the dibutyryl derivative of cyclic adenosine-3',5'-monophosphate (dcAMP) had either a modest inhibitory effect or no effect on long-term hematopoiesis. Dose effects for AMP enhancement of hematopoiesis were relatively narrow. When cultures were exposed to a broad range of concentrations (10 mM-10 nM), stimulation was only seen at a molar concentration of 1 X 10(-4) M. Stem cell assays revealed stimulation of multipotent stem cells (CFU-S), as well as committed progenitor cells (CFU-C). Lithium chloride has been shown to cause granulocytosis both in vivo and in vitro. Reductions in intracellular cAMP levels resulting from adenylate cyclase inhibition is a proposed mechanism for this stimulatory effect. However, lithium-induced granulocytosis in long-term cultures could not be blocked by the addition of dcAMP. Measurement of nucleotide levels on spent medium revealed rapid utilization and/or degradation of these reagents. This suggests that failure to abrogate the lithium effect with dcAMP may have been related to the inability to maintain constant intracellular concentrations. The varied observations regarding adenine nucleotide effects on hematopoiesis, as well as the reproducible stimulation by lithium, may be explained by our current appreciation of the complex adenylate cyclase system, which contains both inhibitory and stimulatory subunits for nucleotides and monovalent cations.     

Effect of lithium on stem cell and stromal cell proliferation in vitro

Lithium is recognized as a potent stimulator of hematopoiesis both in vivo and in vitro. Previous work has suggested that this stimulation is mediated as an indirect, humoral effect by the action of lithium upon the stromal cell population. In the present study, the effects of lithium on the stromal population were investigated using a long-term liquid marrow culture model. These findings indicate that exposure of in vitro cultures to lithium results in an increase in the total cellularity and in the number of various hematopoietic progenitor cells residing within the stromal layer. A distinct morphologically recognizable cell has not been identified as the target cell responsible for the indirect stimulation of hematopoiesis by lithium. However, two candidate radioresistant stromal cells believed to be active in the production of humoral mediators of hematopoiesis did proliferate in response to lithium exposure.     

Effects of lithium on vasopressin-dependent cyclic AMP in rat renal medulla.

A vasopressin resistant urinary concentrating defect has been described in patients receiving lithium salt for affective disorders. For the pathogenic mechanism of the concentrating defect it has been postulated that lithium inhibits the vasopressin-dependent cyclic AMP system. However, the results of indirect studies on the lithium effect are equivocal. Therefore, the effect of lithium specifically on the vasopressin-dependent cyclic AMP system was investigated in rat renal medulla. The increase of cyclic AMP concentration by vasopressin was inhibited by lithium. But lithium had no effect on the PTH-dependent cyclic AMP concentration in renal cortical slices. Regardless of magnesium concentrations from 0-10 mM in the incubation media, 10 mM lithium had no moeasurable effect on the vasopressin-dependent adenylate cyclase of rat renal medulla. However, 10 mM lithium augmented the cyclic AMP-phosphodiesterase activity in renal medulla in the high Km system. These results suggest that lithium inhibits the vasopressin-dependent cyclic AMP concentration in renal medulla via the augmentation of its catabolism, rather than via the inhibition of cyclic AMP generation.     

Role of neutrophil elastase in bone marrow failure syndromes: molecular genetic revival of the chalone hypothesis

Two forms of inherited deficiency of neutrophil numbers are cyclic hematopoiesis and severe congenital neutropenia. In cyclic hematopoiesis, neutrophil counts oscillate opposite monocytes in a 3-week cycle. Severe congenital neutropenia consists of static neutropenia and a predisposition to myelodysplasia and acute myelogenous leukemia. All cases of cyclic neutropenia and most cases of severe congenital neutropenia result from heterozygous germline mutations in the gene encoding neutrophil elastase, ela2. Recent work extends the list of neutropenia genes to include WASp, Gfi-1, adaptin, and tafazzin. Studies of mosaic patients suggest that ela2 mutations act in a cell-autonomous fashion. A hypothetical feedback circuit potentially interconnects these genes. Genetic dissection of signaling in model organisms along with experimental hematology implicate C/EPBepsilon, RUNX1/AML1, Notch family members, LEF1, and Cdc42 as additional nodes in this pathway. The authors propose that neutrophil elastase acts as an inhibitor of myelopoiesis, substantiating a chalone hypothesis proposed many years ago.     

Abnormal responsiveness of granulocyte-committed progenitor cells in cyclic neutropenia.

The mechanism(s) driving cyclic hematopoiesis in human cyclic neutropenia remains unknown. Clinical trials suggest that an abnormal responsiveness of bone marrow progenitor cells to hematopoietic growth factors might cause oscillatory blood counts. Studies were performed to determine whether an abnormal responsiveness to multiple growth factors exists in this disorder and whether the defect could be shown in highly enriched populations of marrow progenitor cells. Bone marrow mononuclear cells from patients with congenital cyclic neutropenia required higher concentrations of added granulocyte-colony-stimulating factor (G-CSF) to achieve half-maximal colony growth than cells from normal subjects (478 +/- 90 pmol/L v 53 +/- 12 pmol/L, P less than .01). Patients also differed in requirement for granulocyte-macrophage-CSF (P less than .05), but not for interleukin-3 (P greater than .30). CD34+ bone marrow cells from three patients also showed this difference in G-CSF responsiveness (P less than .05). These data suggest that the defect in congenital cyclic hematopoiesis lies in growth factor receptor binding or the postreceptor signal transduction system that drives granulocytopoiesis.     

Effect of lithium on catecholamine-induced antidiuresis and cyclic AMP in dog kidneys.

Patients treated with lithium salt have an inability to concentrate urine, possible due to the inhibition of the antidiuretic effect of vasopressin. Since beta adrenergic stimulation also induces antidiuresis, a possible effect of lithium on the catecholamine-induced antidiuresis was investigated in dog kidneys. The urinary concentrating ability induced by the iv injection of isoproterenol 0.1 mug/kg was markedly inhibited in the lithium-treated animals (plasma lithium 1.13 plus or minus 0.10 mM). The increase of cyclic AMP concentration by 1 muM isoproterenol was also significantly less in the renal medullary slices obtained from the lithium-treated animals than in those obtained from the control animals. These findings suggest that the inability to concentrate urine in the patients treated with lithium salt is probably due to the inhibition of the antidiuretic effect of catecholamine as well as that of vasopressin; and the inhibitory mechanism of lithium on the catecholamine-induced antidiuresis is possibly through the inhibition of the catecholamine-dependent cycle AMP system in renal medulla.     

The effect of lithium carbonate on chemotherapy-induced neutropenia and thrombocytopenia

Lithium carbonate ameliorates neutropenia associated with cancer chemotherapy. The effect of lithium on platelet suppression has not, however, been well established. In the present study, five patients with ovarian carcinoma received daily lithium during alternate cycles of treatment with hexamethylmelamine, cyclophosphamide, adriamycin, and cis-platinum. Analysis of myelosuppression was performed on 24 paired consecutive cycles given at identical doses, one with and one without lithium. During lithium cycles, nadir leukocyte, neutrophil, and platelet counts were significantly higher (P less than 0.01, less than 0.01, less than 0.05 respectively) and the interval between treatments was shorter (P less than 0.01). One patient who has received 11 cycles of chemotherapy continues to receive 100% doses owing to the beneficial effect of lithium on chemotherapy-induced thrombocytopenia. Lithium was poorly tolerated by some patients because of either tremor or nausea and vomiting, in spite of nontoxic serum lithium levels. The amelioration of drug-induced platelet suppression as well as neutrophil suppression noted in this study suggests that lithium's effect on hematopoiesis is not limited to stimulation of neutrophil production. The ability of lithium to decrease chemotherapy-induced myelosuppression suggests that lithium administration may facilitate escalation of chemotherapy doses in selected patients.     

Cyclic Hematopoiesis: animal models

The four existing animal models of cyclic hematopoiesis are briefly described. The unusual erythropoietin (Ep) responses of the W/Wv mouse, the Sl/Sld mouse, and cyclic hematopoietic dog are reviewed. The facts reviewed indicate that the bone marrow itself is capable of influencing regulatory events of hematopoiesis.     

A comparison of treatment of canine cyclic hematopoiesis with recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF), G-CSF interleukin-3, and canine G-CSF.

Cyclic hematopoiesis in gray collie dogs is a stem cell disease in which abnormal regulation of cell production in the bone marrow causes cyclic fluctuations of blood cell counts. In vitro studies demonstrated that recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-3 (IL-3), and granulocyte colony stimulating factor (G-CSF) all stimulated increases in colony formation by canine bone marrow progenitor cells. Based on these results, gray collie dogs were then treated with recombinant human (rh) GM-CSF, IL-3, or G-CSF subcutaneously to test the hypothesis that pharmacologic doses of one of these hematopoietic growth factors could alter cyclic production of cells. When recombinant canine G-CSF became available, it was tested over a range of doses. In vivo rhIL-3 had no effect on the recurrent neutropenia but was associated with eosinophilia, rhGM-CSF caused neutrophilia and eosinophilia but cycling of hematopoiesis persisted. However, rhG-CSF caused neutrophilia, prevented the recurrent neutropenia and, in the two animals not developing antibodies to rhG-CSF, obliterated periodic fluctuation of monocyte, eosinophil, reticulocyte, and platelet counts. Recombinant canine G-CSF increased the nadir neutrophil counts and amplitude of fluctuations at low doses (1 micrograms/kg/d) and eliminated all cycling of cell counts at high doses (5 and 10 micrograms/kg/d). These data suggest significant differences in the actions of these growth factors and imply a critical role for G-CSF in the homeostatic regulation of hematopoiesis.     

Abnormal GM-CSA generation by lymphocytes in patients with cyclic hematopoiesis

Human cyclic hematopoiesis (CH) is a disease characterized by regular 21-day cyclic fluctuations of blood cell counts due to fluctuations in bone marrow cell production. The regular periodicity of the fluctuations suggests a defect in a regulatory feedback control loop. We examined the production of monocyte-derived recruiting activity (MRA) by monocytes and the response to MRA of lymphocytes from three patients with CH. MRA production was normal or increased in patients' monocytes, but granulocyte-macrophage colony-stimulating activity (GM-CSA) production in response to MRA was decreased in lymphocytes from patients with CH (p = 0.005). These data suggest that the regulatory defect in CH may involve defective lymphocyte generation of GM-CSA, resulting in deficient production of mature neutrophils.     

Mechanism of canine cyclic hematopoiesis: the role of prostaglandin E in feedback regulation

Prostaglandin E inhibits granulocyte-macrophage colony formation in vitro in man and mouse, suggesting that it plays a role in feedback regulation of granulocyte production in vivo. Therefore, we examined the role of PGE in normal canine hematopoiesis and its potential role in the pathogenesis of cyclic hematopoiesis in grey collie dogs. The prostaglandin synthesis inhibitors indomethacin and ibuprofen (10(-5) M) increased CFU-C growth to 194 and 160% of control, respectively, while PGE2 addition caused a dose-dependent inhibition of bone marrow CFU-C growth in both normal and grey collie dogs. These concentrations of indomethacin and ibuprofen decreased bone marrow cell elaboration of PGE measured by radioimmunoassay to less than 5% of control values. The levels of PGE in leukocyte conditioned medium prepared from grey collies correlated with the number of monocytes in the conditioning cell suspension (r = 0.78, n = 10, p less than 0.05) so that PGE production per monocyte was no different in normal and grey collie dogs. The effect of PGE2 upon CFU-C was to inhibit formation of macrophage, but not neutrophil colony subtypes. These findings make aberrant PGE-mediated inhibition of precursor cells an unlikely mechanism to cause cyclic hematopoiesis, and show that PGE produced by monocytes acts as a feedback inhibitor for precursor cells destined to produce monocytes but not for those destined to form neutrophils.     

Lithium does not inhibit the parathyroid hormone-mediated rise in urinary cyclic AMP and phosphate in humans.

To test the hypothesis that lithium is a general inhibitor of hormone-activated adenylate cyclase, we infuse parathyroid hormone (PTH) into human subjects prior to and during lithium carbonate administration. PTH infusion caused a significant increase in urinary cyclic AMP and urinary phosphate excretion. There was no significant difference in these responses in the lithium compared to the control period. In four patients with primary hyperparathyroidism, lithium had no significant effect on serum calcium or phosphate or on tubular reabsorption of phosphate. The data do not substantiate the hypothesis that lithium (at therapeutic concentrations) is a general inhibitor of hormonally-activated adenylate cyclase, nor do they support its potential clinical utility in primary hyperparthyroidism.     

Adult-onset cyclic bicytopenia: a case report and review of treatment of cyclic hematopoiesis

A unique case of adult-onset synchronous cyclic neutropenia and thrombocytopenia occurring at six-week intervals is presented. Periods of cytopenia were associated with fever, myalgias, gastrointestinal symptoms, and mild mucocutaneous bleeding. Alternate-day steroid treatment failed to correct the periodic fluctuations in peripheral blood counts but ameliorated symptoms during cytopenia. The treatment of cyclic hematopoiesis is reviewed.     

Progenitor cells in canine cyclic hematopoiesis.

Granulocytic (colony-forming units in culture, or CFU-c) and erythrocytic (erythropoietin-responsive cells, or ERC) progenitor cells in canine cyclic hematopoiesis (CH) have been shown to fluctuate over the cycle and in the same phases as one another. The ERC cycle is preceded by 3 or 4 days by a rise in serum erythroid-stimulating activity and is followed by a reticulocytosis. During the cycle CFU-c show a differential sensitivity to two sera, one normal and one containing elevated amounts of colony-stimulating activity. The proliferation rate of CFU-c also fluctuates from well above normal to considerably less than normal over the cycle. These results are discussed in the light of present knowledge of the pathogenesis of canine CH. We suggest that these results support the contention that CH is a disorder of hemopoietic stem cells and that the cycling of humoral factors and peripheral blood cells may follow as a consequence.     

Cyclic neutropenia: amplification of granulocyte oscillations by lithium and long-term suppression of cycling by plasmapheresis

A patient with well documented cyclic neutropenia (CN) was given chronic lithium therapy as well as a short course of plasmapheresis with therapeutic intention. While on lithium therapy, an increase in the amplitude of the granulocyte oscillations was observed, but recurrent agranulocytotic periods persisted. A 2-wk course of plasmapheresis (total exchange 17 liters) resulted in a gradual decrease of granulocyte oscillations, with the granulocyte count remaining above 500/cu mm at all times, and the patient became asymptomatic. The improvements in the clinical state of the patient have now persisted for more than 9 mo. Quantitative analysis of the oscillation period and the phase lags between peak counts of the hematologic cell lines revealed a 20-day cycle for granulocytes, monocytes, reticulocytes, and platelets, but not for lymphocytes. The observed phase lags, together with the literature values for the intramarrow maturation times of the hematologic cell lines, suggest that the feedback regulation abnormality, believed to be at the basis of CN, primarily affects the stem cell that is still capable of differentiating into granulocytes and monocytes, but that the oscillations of reticulocytes and platelets are the result of a stem cell competition for uncommitted precursor cells by the primarily oscillating granulocyte and monocyte production.     

Modifications in the phagocytosis of human neutrophils induced by vinblastine and cytochalasin B: the effects of lithium

The effects of lithium on the microtubular and microfilament systems of neutrophil granulocytes is described. The phagocytic capacity of human neutrophils was reduced or inhibited by using vinblastine or cytochalasin B. Lithium at concentrations ranging from 0.5 to 2.0 mEq/l proved to be capable of antagonising the action of vinblastine, whereas the action of cytochalasin B was found to be unaffected by lithium. Based on the antagonising effect of lithium on vinblastine-induced inhibition of phagocytosis, it is suggested that lithium acts--at therapeutically achievable levels--on the microtubular system of neutrophils, possibly through modulation of cyclic AMP levels.