Granulocyte colony stimulating factor (G-CSF) producing renal cell carcinoma

A 88-year-old male patient with G-CSF producing renal cell carcinoma is reported. The patient was admitted to our hospital complaining of macrohematureia. The laboratory examination showed marked leukocytosis of 18,200/microliter (neutrophil 92%) in the peripheral blood and high levels of G-CSF (120 pg/ml) in the serum. An abdominal CT scan revealed a right renal tumor. The neutrophil count rose to 38,700/microliter with increasing of tumor size. Histopathological diagnosis was renal cell carcinoma (Grade 3) and immunohistochemical staining using Histo anti-G-CSF antibody demonstrated cancer cells produced G-CSF. This is the second case of G-CSF producing renal cell carcinoma diagnosed by immunohistochemical staining in the literature.     

Clinical benefits of diagnosing incipient AA amyloidosis in pediatric rheumatic diseases as estimated from a retrospective study

We describe here a newly established cell line from an eccrine carcinoma which produced an abundant amount of granulocyte colony-stimulating factor (G-CSF). An eccrine carcinoma of the scalp of a 69 year-old-Japanese female had metastasized to the pleura. Clinically, she had marked neutrophilia (up to 60,000/mm3), and a high level of G-CSF (38.7 x 10(3) pg/ml) was detected in the pleural effusion, as determined by enzyme-linked immunosorbent assay (ELISA). We established a cell line in vitro and maintained the cells in culture for 30 months in 90 subcultures. We investigated whether these tumor cells were able to produce G-CSF in culture and found that they were. We also found that the amount of G-CSF produced paralleled the rise in cell number (26.5 x 10(3) pg/ml at confluency). When culture media were administered to rabbits (25 ml/rabbit), the amount of circulating neutrophils increased until the number was equal to or greater than that resulting from injection of recombinant human G-CSF (rhG-CSF)(75 micrograms). This effect persisted for 7 days. When tumors were induced in SCID and nude mice by injecting cultured cells (1 x 10(7) cells/mouse), the number of circulating neutrophils also correlated well with tumor size in these mice (200,000/mm3, 3 cm tumor). After tumor removal, the neutrophil number returned to normal within 30 days. G-CSFmRNA in cultured, cells was detected by RT-PCR. Based on these results, it was confirmed that the marked neutrophilia observed in the patient was caused by the tumor-generated G-CSF. This is the first G-CSF-producing cell line developed from a cancer of the skin.     

Levels of serum granulocyte colony-stimulating factor in patients with cerebrovascular diseases

The role of leukocytes in the pathogenesis of cerebrovascular disease, in particular, cerebral ischemic disease has recently become a focus of research. Several studies have reported that a positive correlation between increased functional activities of neutrophils and the risk of cerebral ischemic disease. Granulocyte colony-stimulating factor (G-CSF) is known to be not only a granulocyte proliferating factor but also a potent activator of mature neutrophils. In this study, we measured the serum G-CSF levels in 143 patients with cerebrovascular diseases and in 100 patients with other diseases, using our established enzyme-linked immunosorbent assay (ELISA) for G-CSF The minimal detection level was 20 pg/ml G-CSF. In patients with cerebral infarction, G-CSF could be detected in 18.3% and in patients with cerebral hemorrhage, it could be detected in 9.8% of analyzed samples. On the other hand, 6% of the patients with other diseases had measurable levels of G-CSF. The differences among these three groups were statistically significant according to the chi 2 test (p < 0.01). Our findings that there was a significantly high frequency of elevated levels of G-CSF among patients with cerebrovascular diseases, may indicate that the action of G-CSF as a potent activator of neutrophils plays some role in the occurrence of cerebrovascular disease, in particular, cerebral infarction.     

The history of occupational health service in Korea

Dialysate and serum levels of granulocyte-colony stimulating factor (G-CSF), granulocyte macrophage colony stimulating factor (GM-CSF) and leukemia inhibitory factor (LIF) were analyzed in patients with continuous ambulatory peritoneal dialysis (CAPD). Samples from the peritoneal effluent and from serum were obtained during the first months of dialysis and during peritonitis from the first three dialysate bags drained on the day of admittance and form nightbags on days three and ten. Serum samples were drawn on days one and ten. On the first day of infection G-CSF was detected in twelve out of fifteen samples in the dialysate and reached its peak median level, 443 pg/ml, in the first drained bag and thereafter decreased significantly. Also in serum a peak, 190 pg/ml, was observed on the first day. LIF was found in six of ten analyzed dialysate samples, with a peak median level of 77 pg/ml on day one, while only four of ten patients had detectable GM-CSF. Peripheral blood mononuclear cells from non-infected CAPD patients were stimulated with lipopolysaccharide and G-CSF levels in the supernatants increased significantly (P < 0.05) after 6 h stimulation. We conclude that G-CSF is produced locally in the dialysate during the acute stage of peritonitis and to a lesser extent also systemically. These findings are in line with G-CSF production after LPS stimulation of peripheral blood mononuclear cells.     

Production of granulocyte-macrophage colony-stimulating factor in a patient with metastatic chest wall large cell carcinoma

Recent reports of cancers that produce colony-stimulating factors (CSF) and which are associated with leukocytosis indicate that most are granulocyte CSF-producing tumors. A 71-year-old man with metastatic chest wall tumors from large cell lung cancer with marked leukocytosis and eosinophilia was reported. His maximal leukocyte count was 48300/microliter with 37.5% eosinophils. Granulocyte-macrophage CSF (GM-CSF) activity detected by enzyme-linked immunosorbent assay (ELISA) in serum was 112 pg/ml (normal range < 2.0 pg/ml), but G-CSF was normal. Immunohistochemical detection of GM-CSF protein on a chest wall tumor sample was positive. Irradiation of the chest wall tumor was performed and the leukocyte count decreased temporally. However, he died of respiratory failure due to progressive tumor growth 56 days after admission. Based on these results it appears that autocrine production of GM-CSF is a possible cause of this leukemoid reaction.     

Endogenous interleukin-8 (IL-8) surge in granulocyte colony-stimulating factor-induced peripheral blood stem cell mobilization

The relationship between stem cell mobilization with granulocyte colony-stimulating factor (G-CSF) and the endogenous production of interleukin-8 (IL-8), macrophage inflammatory protein-1alpha (MIP-1alpha), tumor necrosis factor-alpha (TNF-alpha), and interferon-gamma (IFN-gamma) was studied in normal donors for allogeneic peripheral blood stem cell (PBSC) transplantation. G-CSF was administered to 20 normal donors at a dose of 10 microg/kg/d for 5 days with aphereses on days 5 and 6 of G-CSF treatment. Cytokine serum levels were measured using an enzyme-linked immunosorbent assay (ELISA) before and during G-CSF treatment. Before treatment, the average level of IL-8 was 7.1 pg/mL, increasing to 207.0 pg/mL on day 5 and 189.1 pg/mL on day 6. Serum IL-8 levels correlated CD34(+) cell numbers (P =.0151 and P =.0005 on days 5 and 6, respectively) and colony-forming unit-granulocyte-macrophage (CFU-GM) numbers (P =.0019 and P =.0010 on days 5 and 6, respectively). Furthermore, preapheresis serum IL-8 levels correlated with the yield of CD34(+) cells (P =.0027). In contrast, before treatment, the average levels of MIP-1alpha, TNF-alpha, and IFN-gamma were 70.1, 4.03, and 3.84 pg/mL, respectively, and no significant changes in the levels of these cytokines were observed during G-CSF treatment. These studies suggest that IL-8 production may be critical to G-CSF-induced stem cell mobilization, although the underlying mechanism could not be clarified.     

Crossover study of the haematological effects and pharmacokinetics of glycosylated and non-glycosylated G-CSF in healthy volunteers

A cross-over study of glycosylated and non-glycosylated G-CSF was performed in 20 healthy male volunteers to compare the effects of the different forms of G-CSF, the extent of inter-individual progenitor cell mobilization and to determine whether any differences observed were related to the serum concentrations of G-CSF attained. The peak WBC achieved during 6 d of G-CSF administration at a dose of 5 microg/kg/d was significantly higher with the glycosylated than the non-glycosylated product (P = 0.02) as was the peak level of granulocyte-monocyte colony forming cells (GM-CFC) (P=0.03). The average GM-CFC count on days 5, 6 and 7 was 28% higher with the glycosylated product (P=0.003). Serum concentrations of G-CSF achieved were significantly higher with the non-glycosylated G-CSF, however, suggesting that the difference in bio-efficacy was not due to a difference in G-CSF stability. Marked inter-individual variation in progenitor mobilization was observed, but this was not related to serum G-CSF levels. The G-CSF concentrations on day 6 were approximately one third of those on day 1, with both forms of G-CSF.     

Treatment of early stages of ovarian carcinoma. Evaluation of experience at the Institute of Radiation Oncology over a 10-year period (1983-1993)

The granulocyte colony-stimulating factor (G-CSF) regulates neutrophil differentiation and function. Serum levels of G-CSF increase during acute infectious processes. The levels of G-CSF were measured in 59 surgical intensive care unit (ICU) patients. In general, G-CSF was only elevated during the first 2 days after admission to the ICU. Higher G-CSF levels were more frequently observed in patients without infectious complications and in patients who survived. Later on, G-CSF levels were below 100 pg/ml in almost all patients studied. The highest G-CSF level (20,000 pg/ml) was observed in one patient with septic shock 36 h after leukopenia. The patient recovered from septic shock and multiple organ failure and was discharged. It is proposed that surgical ICU patients with low or undetectable G-CSF serum levels may benefit from exogenous G-CSF substitution protocols.     

GH dependence and GH withdrawal syndrome in GH treatment of short normal children: evidence from growth and cardiac output

BACKGROUND: Cytokines and cytokine receptors are involved in the systemic and local inflammatory response in patients with urinary tract infections. METHODS: We examined urine and serum concentrations of soluble IL-6 receptor (sIL-6R), IL-10 and granulocyte colony-stimulating factor (G-CSF) in 29 women with acute pyelonephritis caused by Escherichia coli 2 weeks after the infection, during the subsequent episode of cystitis or asymptomatic bacteriuria and also later when the same patients were free from bacteriuria. Concentrations of sIL-6R, IL-10 and G-CSF were related to the expression of five virulence markers of E. coli and to glomerular filtration rate (GFR) after pyelonephritis. RESULTS: On admission because of acute pyelonephritis the serum concentration of sIL-6R was similar to that of 12 healthy controls. Two weeks after the infection when all patients had received antibiotic treatment, the serum concentration of sIL-6R was significantly higher compared to that on admission (p < 0.001) and also higher compared to healthy controls (p = 0.001). Patients with increased concentrations of sIL-6R in serum 2 weeks after infection had significantly lower GFR at follow-up (p < 0.05). Patients with acute pyelonephritis had higher concentrations of G-CSF and IL-10 in serum compared to healthy subjects (p < 0.001 and p = 0.06, respectively). G-CSF in serum was higher in patients infected by E. coli producing cytotoxic necrotizing factor (p < 0.05). Patients infected by strains producing hemolysin had lower concentrations of sIL-6R (p < 0.001). Patients with detectable levels of the anti-inflammatory cytokine IL-10 in serum had significantly higher concentrations of IL-6 and the soluble tumor necrosis factor receptors I and II in serum as compared to patients in whom IL-10 was not detectable (p < 0.001, p = 0.001 and p < 0.05, respectively. CONCLUSION: These investigations, together with our previous findings summarized in this paper, contribute to an increased understanding of the local and systemic inflammatory response arising in response to acute pyelonephritis.     

Giant cell tumour of the retropharynx

INTRODUCTION: Few pancreatic carcinomas (5-22%) are resectable at the time of diagnosis because this lesion is seldom diagnosed in an early stage. Unresectability is mainly due to the presence of metastases to the liver, peritoneum and lymph nodes and to tumor spread especially to the portal mesenteric trunk where it can invade, compress, reduce, or occlude the vessels. We investigated the diagnostic yield of multiplanar and 3D spiral CT in the assessment of pancreatic carcinoma resectability. MATERIAL AND METHODS: Twenty-seven patients with histologically confirmed pancreatic head cancer were submitted to spiral CT and color Doppler US in the Surgical Clinic I of the Bologna University. The examination results were correlated with the intraoperative findings of careful inspection and palpation and of US studies of the pancreatic mass and adjacent structures. The tumors were classified in relation to some CT parameters: tumor size (T), infiltration of the stomach (S) and/or duodenum, lymph nodes (N) or distant (M) metastases, involvement of vascular structures (V), particularly of portal or superior mesenteric vein, or superior mesenteric artery. Five grades of vascular involvement were considered. The results of these techniques were correlated with intraoperative findings from careful inspection and palpation and with US studies of the pancreatic mass and adjacent structures. RESULTS: Spiral CT revealed vascular involvement in 19 of 27 cases (70.4%): involvement of portal and superior mesenteric vein was found in 14 (73.6%), superior mesenteric vein was involved in 2 (10.6%), the portal vein in one (2%) and, finally the portal, superior mesenteric vein and superior mesenteric artery in 2 cases (10.6%). The spiral CT results were confirmed intraoperatively in 26 of 27 cases (96.3%); spiral CT did not reveal hepatic metastasis only in one case. Spiral CT with multiplanar reconstructions had very high specificity and sensitivity (100%) in the assessment of vascular involvement, while color Doppler US had the same specificity but lower sensitivity (84.2%). Spiral CT was less sensitive (80%) in the detection of liver metastases. CONCLUSIONS: We believe that spiral CT is currently the best technique for pancreatic carcinoma staging, providing useful information for correct surgical planning.     

Changes in serum granulocyte colony-stimulating factor (G-CSF) in burned patients

Serum G-CSF level was assayed in 20 burned patients (TBSA > or = 30%) 2 weeks postburn. In 11 of them plasma endotoxin was also measured. The results showed serum G-CSF level was increased in 87.5% (7/8) of burned patients with exceeding 50% TBSA and in 58.3% (7/12) of patients with 30%-50% TBSA, and in the former it was earlier increased (1-5 day postburn) than the latter (5-11 day postburn). Serum G-CSF was increased in 90% of the burned patients with wound sepsis. Increased of serum G-CSF in burned patients, especially in those exceeding 50% TBSA, indicates the occurrence of sepsis.     

A case of multiple myeloma producing granulocyte colony-stimulating factor

A case of multiple myeloma (IgA-lambda) with marked granulocytosis, which measured up to 9.9 x 10(4)/mm3, is described. Matured neutrophils were predominant and blasts were not found in the peripheral blood. The serum granulocyte colony-stimulating factor (G-CSF) was notably elevated. The disease ran a chronic course and granulocytosis and elevated serum G-CSF continued. The patient developed atelectasis and bronchopneumonia, and died of respiratory failure. At autopsy, bone marrow showed marked myeloid hyperplasia in varying states of differentiation. The enlarged spleen also disclosed numerous myeloid cells of varying differentiation. Small aggregations of atypical plasma cells were present in the marrow and spleen. Immunohistochemically, atypical plasma cells were positive for anti-G-CSF antibody, which indicated G-CSF secretion from the myeloma cells. To our knowledge, this is the first reported case of G-CSF-producing multiple myeloma.     

Expression of granulocyte colony-stimulating factor receptor correlates with prognosis in oral and mesopharyngeal carcinoma

Granulocyte colony-stimulating factor receptors (G-CSFRs) have been observed on the surface of not only hematopoietic cells but also several cancer cells. The stimulation of G-CSF has been demonstrated to induce proliferation and activation of G-CSFR-positive cells. In this study, we investigated the expression of G-CSFR on the surface of tumor cells and G-CSF production in oral and mesopharyngeal squamous cell carcinoma (SCC) by an immunohistochemical approach. Of 58 oral and mesopharyngeal SCCs, 31 cases (53.4%) and 36 cases (62.1%) were positive for G-CSFR and G-CSF, respectively. There was no association between G-CSFR expression and G-CSF staining. In the group positive for G-CSFR expression, relapse was significantly more likely after primary treatment (P = 0.0069), whereas there was no association between G-CSFR expression and age, sex, tumor size, lymph node metastasis, and clinical stage. Also, the G-CSFR-positive groups had a significantly lower disease-free and overall survival rate than the G-CSFR-negative groups (P = 0.0172 and 0.0188, respectively). However, none of the clinical markers correlated significantly with G-CSF staining, nor did the status of G-CSF production influence the overall survival. The results imply that assessment of G-CSFR may prove valuable in selecting patients with oral and mesopharyngeal SCC for aggressive therapy.     

Production of granulocyte colony stimulating factor, granulocyte macrophage colony stimulating factor, and tumor necrosis factor alpha during remission and infections in patients with acute leukemia

We measured circulating serum levels of granulocyte colony stimulating factor (G-CSF), granulocyte macrophage colony stimulating comparable to the levels of these factors in 12 children with acute febrile infections without malignancy or hematological disorders and 15 age matched healthy controls. There were significantly elevated levels of G-CSF, GM-CSF and TNF alpha, in 12 children with infections without leukemia, as compared with controls. Also in 18 leukemic children with infections serum G-CSF and TNF alpha levels were significantly higher than those in the leukemic children without infection and healthy controls, whereas no significant difference was noted in the GM-CSF levels in these groups. Although elevation in TNF alpha levels in response to infections were similar in the children with and without leukemia, in the G-CSF levels lower elevation was noted in the leukemic children with infections as compared to the children with infections without leukemia. Despite leukopenia enhanced the production of G-CSF, even in leukopenic children with leukemia and infections, serum G-CSF levels were still lower than those for the children with infections without leukemia. We concluded that, the production of G-CSF and GM-CSF as a response to infection was deficient in the patients with acute leukemia in remission, probably due to the maintenance and reinforcement chemotherapy. Therefore, the use of recombinant G-CSF may be recommended in the infections of these patients.     

Three cases of resected primary mediastinal yolk sac tumor following six courses of bleomycin, etoposide and cisplatin (BEP) combination chemotherapy

We experienced three cases of primary mediastinal yolk sac tumor which were resected after 6 courses of BEP chemotherapy with G-CSF support. All cases had high levels of AFP. CT scan revealed an anterior mediastinal tumor infiltrating the surrounding tissue in all cases, and multiple pulmonary nodules in one case. The serum AFP level decreased markedly, but did not return to normal after the chemotherapy was completed. The markedly decreased mediastinal tumor masses were removed with en-bloc resection of the lung and pericardium. Viable tumor cells were not present in the resected tumors. These three cases remain free of disease at present, 3 years, 9 months and 5 months after operation.     

Cytokine-specific activation of distinct mitogen-activated protein kinase subtype cascades in human neutrophils stimulated by granulocyte colony-stimulating factor, granulocyte-macrophage colony-stimulating factor, and tumor necrosis factor-alpha

To clarify the differences of the signaling pathways used by granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), and tumor necrosis factor- (TNF), we investigated activation of mitogen-activated protein kinase (MAPK) subtype cascades in human neutrophils stimulated by these cytokines. G-CSF exclusively tyrosine-phosphorylated extracellular signal-regulated kinase (ERK). GM-CSF tyrosine-phosphorylated ERK strongly and p38 MAPK weakly, whereas TNF tyrosine-phosphorylated p38 MAPK strongly and ERK weakly. Consistent with these findings, MEK, an upstream kinase of ERK, was phosphorylated by G-CSF, GM-CSF, and TNF, whereas MKK3/MKK6, an upstream kinase of p38 MAPK, was phosphorylated by GM-CSF and TNF, but not by G-CSF. The potency of these cytokines to phosphorylate ERK and MEK was GM-CSF > G-CSF > TNF, whereas that to phosphorylate p38 MAPK and MKK3/MKK6 was TNF > GM-CSF. C-Jun amino-terminal kinase (JNK) was not tyrosine-phosphorylated by any cytokine despite the existence of JNK proteins in human neutrophils, whereas it was tyrosine-phosphorylated by TNF in undifferentiated and all-trans retinoic acid-differentiated HL-60 cells. Increased phosphorylation of ERK or p38 MAPK was detected within 1 to 5 minutes after stimulation with each cytokine and was dependent on the concentrations of cytokines used. MEK inhibitor (PD98059) reduced tyrosine phosphorylation of ERK, but not p38 MAPK, induced by G-CSF, GM-CSF, or TNF. GM-CSF- or TNF-induced superoxide (O2-) release was inhibited by p38 MAPK inhibitor (SB203580) in a dose-dependent manner, suggesting the possible involvement of p38 MAPK in GM-CSF- or TNF-induced O2- release. The results indicate that G-CSF, GM-CSF, and TNF activate the overlapping but distinct MAPK subtype cascades in human neutrophils and suggest that the differential activation of ERK and p38 MAPK cascades may explain the differences of the effects of these cytokines on human neutrophil functions.     

Interleukin-1 beta and tumor necrosis factor-alpha stimulate granulocyte colony-stimulating factor production by placental villous core mesenchymal cells

OBJECTIVE: To test the hypothesis that interleukin-1 beta (IL-1 beta) and tumor necrosis factor-alpha (TNF-alpha) regulate granulocyte colony-stimulating factor (G-CSF) production by human placental villous core mesenchymal cells. METHODS: Villous core mesenchymal cells were isolated from placentas at 14-20 weeks' gestation and cultured in vitro. Cells were treated with IL-1 beta or TNF-alpha in dose-response and time-course studies. We measured G-CSF mRNA expression by Northern blot analysis and G-CSF protein production by enzyme-linked immunosorbent assay of the conditioned media. RESULTS: Unstimulated mesenchymal cells expressed negligible G-CSF. Steady-state G-CSF mRNA expression was maximal 3-6 hours after IL-1 beta treatment and 6-18 hours after TNF-alpha treatment. Each cytokine induced G-CSF protein production in dose-and time-dependent manners, with IL-1 beta more potent than TNF-alpha. The G-CSF mRNA expression and G-CSF protein production induced by the combination of both cytokines exceeded that induced by either cytokine alone. CONCLUSIONS: Interleukin-1 beta and TNF-alpha stimulate G-CSF production by placental villous core mesenchymal cells in vitro. These results identify a potential mechanism by which villous core mesenchymal cells mediate, in part, the placental response to these two cytokines.     

Human CD34(+) bone marrow cells regulate stromal production of interleukin-6 and granulocyte colony-stimulating factor and increase the colony-stimulating activity of stroma

Cytokines produced by stromal cells induce the proliferation and differentiation of hematopoietic cells in the marrow microenvironment. We hypothesized that cross-talk between hematopoietic cells at different stages of differentiation and stromal cells influences stromal cytokine production and is responsible for maintaining steady-state hematopoiesis and responding to stress situations. We show that coculture of primitive CD34(+) cells in contact with or separated by a transwell membrane from irradiated human bone marrow stromal layers induces a fourfold to fivefold increase in interleukin-6 (IL-6) and granulocyte colony-stimulating factor (G-CSF) levels in the stromal supernatant (SN) during the first week. Levels of both cytokines decreased to baseline after coculture of CD34(+) cells for 3 to 5 weeks. Coculture of more mature CD15(+)/CD14(-) myeloid precursors induced only a transient 1.5- to 2-fold increase in IL-6 and G-CSF at 48 hours. Neither CD34(+) nor CD15(+)/CD14(-) cells produced IL-6, G-CSF, IL-1beta, or tumor necrosis factor alpha. When CD34(+) cells were cultured in methylcellulose medium supplemented with cytokines at concentrations found in stromal SN or supplemented with stromal SN, a fourfold to fivefold increase in colony formation was seen over cultures supplemented with erythropoietin (EPO) only. When cultures were supplemented with the increased concentrations of IL-6 and G-CSF detected in cocultures of stroma and CD34(+) cells or when CD34(+) cells were cocultured in methylcellulose medium in a transwell above a stromal layer, a further increase in the number and size of colonies was seen. The colony-forming unit-granulocyte-macrophage-stimulating activity of stromal SN was neutralized by antibodies against G-CSF or IL-6. These studies indicate that primitive CD34(+) progenitors provide a soluble positive feedback signal to induce cytokine production by stromal cells and that the observed increase in cytokine levels is biologically relevant.     

Primary infection of dogs with Echinococcus granulosus: systemic and local (Peyer's patches) immune responses

Local and systemic lymphocyte proliferation and antibody production were tested in five dogs 35 days after primary experimental infection with Echinococcus granulosus. A significant cell proliferation was demonstrated by [3H] thymidine incorporation in mesenteric, popliteal and/or Peyer's patches (PPs) cells in response to E. granulosus protoscolex or adult worm antigen in three of five infected dogs, but not in five control animals. In contrast, blood mononuclear cells responded very weakly in only two of the infected dogs to parasite antigens. Elevated levels (compared with preinfection status) of protoscolex- and adult worm antigen-specific serum IgG were detected (ELISA) in four of the five dogs 35 days after infection. Furthermore, slightly elevated levels of parasite-specific IgE and IgA were observed in the sera of three and four in four infected dogs, respectively. Specific serum IgM was not significantly higher 35 days after infection than before infection. Local antibody production was studied in vitro using PPs, mesenteric and popliteal cells isolated from three infected and three uninfected dogs by ELISA using adult worm antigen. In two of three cultures of unstimulated PPs cells of infected dogs, parasite-specific IgG was detectable. Parasite-specific IgA and IgM were detected in one of the unstimulated PPs cell culture derived from an infected dog. Following in vitro stimulation with parasite antigen, PPs cells from two infected dogs showed increased parasite-specific IgG and PPs cells of all three infected dogs produced parasite-specific IgA. PPs cells from uninfected dogs did not produce significant quantities of parasite-specific antibodies and cells from mesenteric and popliteal lymph nodes of infected or uninfected dogs neither produced antibodies whilst in in vitro cultures.