Phase I trial of subcutaneous interleukin 3 in patients with refractory malignancy: hematological, immunological, and pharmacodynamic findings.

We conducted a Phase I trial of s.c. recombinant human interleukin 3 (rhIL-3) to evaluate the toxicity, maximal tolerated dose, pharmacokinetics, and in vivo biological effects of this cytokine. Thirty-one patients with refractory cancer were entered into the study between November 1991 and June 1993. Therapy consisted of s.c. rhIL-3 daily for 15 days administered to cohorts of three to nine patients at dose levels of 60-4000 microgram/m2/day. Cycles were repeated at intervals of 28 days. Seventy-five cycles of rhIL-3 were administered (median, two per patient) and the maximal tolerated dose was 2000 microgram/m2/day. Toxicity was moderate, with most patients developing chills, fever, and myalgia. Dose-limiting toxicity consisted of diarrhea (two patients) and headache (one patient). Hematological effects of rhIL-3 included significant dose-related increases of WBC (P < 0.001), neutrophils (P < 0.001), and eosinophils (P < 0.001). Platelet counts and absolute lymphocyte numbers also increased. Various CD3(+) lymphocyte subsets increased; however, lytic activity (natural killer and lymphokine-activated killer) of peripheral blood lymphocytes was not enhanced. Serum levels of the soluble IL-2 receptor increased in a dose-related fashion, and IL-2-induced lymphocyte proliferation also was increased variably. Pharmacokinetic studies were performed in 13 patients, and area under the curve and maximal concentration values increased with increasing rhIL-3 dose levels (P < 0.001) and correlated with maximal changes from baseline in WBC, neutrophils, and eosinophils. rhIL-3 antibodies were detected in 8% of patients by day 29 of cycle 1 but were not neutralizing. rhIL-3 is well tolerated when administered s.c. and has reproducible hematological and immunological effects. The pleiotropic effects of this cytokine on various in vivo biological parameters were demonstrated clearly. Further studies of its immunoregulatory effects are warranted.

Selective cytokine gene expression in renal cell carcinoma tumor cells and tumor-infiltrating lymphocytes.

The progression of tumors such as renal cell carcinoma (RCC), despite the presence of substantial lymphocytic infiltrates (TIL), suggests that the ability of the local immune response to control tumor growth is impaired. Cytokine gene expression was examined to further investigate the nature of this response. Initial studies were performed with frozen tumors using PCR-assisted mRNA amplification with cytokine-specific primers for interleukin 10 (IL-10), interleukin 2 (IL-2) and interferon-gamma (IFN-gamma). IL-2 mRNA was not detected, despite the presence of T cells as defined by the expression of CD3 gamma mRNA. In contrast, mRNA for IFN-gamma was expressed in 4/9 and for IL-10 in 5/9 tumors. To confirm this, 5 fresh tumor specimens were examined, and PCR demonstrated that IL-10 mRNA was detectable in 4/5 tumors from which RNA was isolated at the time of nephrectomy. In these experiments multiple cycles and dilutions were employed to semi-quantitate the expression of IL-10. To identify potential sources of this cytokine in the tumor bed, IL-10 mRNA expression in freshly isolated lymphocytes and tumor cells, TIL lines, cultured RCC and established RCC lines was examined. Our studies demonstrate that within the tumor TIL may be one source of IL-10. Lymphocyte-enriched populations from 4/5 tumors expressed IL-10 mRNA as did 4/6 freshly isolated tumor cell preparations. IL-10 gene expression was not detected, however, in tumor cells after one passage in vitro in short-term cultured RCC tumor cells (passages 2-5) or in established RCC tumor cell lines. Finally, 4/9 CD4+ and 2/5 CD8+ TIL lines expressed IL-10 mRNA either constitutively or after stimulation with anti-CD3 antibody. This finding was associated with IL-10 production in vitro. Our studies demonstrate that IL-10 mRNA is frequently present in RCC tumors and may originate from the tumor-infiltrating mononuclear cell population.

Defective granzyme B gene expression and lytic response in T lymphocytes infiltrating human renal cell carcinoma.

Granzyme B is a protein thought to play a pivotal role in the cytolytic functions of T cells. In view of this, the inducibility of this gene in freshly isolated T cells (T-TILs) infiltrating human renal cell carcinoma (RCC) in vitro was examined by using the reverse transcriptase-polymerase chain reaction (RT-PCR). A reduction in granzyme B messenger RNA (mRNA) expression in stimulated T-TILs from five of nine patients with RCC compared with autologous peripheral blood T cells was noted. The reduced expression was observed after multiple stimuli including anti-CD3 antibody, interleukin-2 (IL-2), and phytohemagglutinin (PHA). Because CD8+ T cells represent the predominant cytotoxic population, the ability of this cell population to express granzyme B mRNA after stimulation also was examined. When compared with CD8+ peripheral blood lymphocytes (T-PBLs) from patients with RCC and normal donors, the induction of granzyme B mRNA was reduced in CD8+ T-TILs. CD8+ T-TILs also had lower non-major histocompatibility complex (MHC)-restricted cytotoxic activity than did CD8+ T-PBLs against both Daudi cells and allogeneic RCC cell lines. These results show that in a subset of patients with RCC, depressed lytic activity of CD8+ TILs compared with CD8+ PBLs is present. Reduced granzyme B mRNA expression also was noted in selected patients.