Preclinical studies of Interleukin-4 in humans have shown a variety of 
effects on B-cells, T-cells, large granular lymphocytes, and monocytes. IL-4 
induces proliferation of B-cells activated by cross-linking surface IgM (13,14). 
It also induces low affinity receptors for IgE (CD23) (15) and class II MHC 
antigens on normal and malignant B-cells (16) Many of the effects of IL-4 on B- 
cells , as well as other cells, are inhibited by Interferon gamma. IL-4 itself 
inhibits IL-2 induced effects includng the IL-2 induced proliferation of 
monoclonal B-cells. IL-4 enhances the proliferation of T-cells activated by 
antigen or mitogen. It enhances the induction of cytolytic T-cells from mixed 
leukocyte responses, and enhances the growth of clones or lines of tumor 
infiltrating lymphocytes isolated from human tumors (4) . In contrast to the 
murine studies, IL-4 decreases the IL-2 induced generation of cells with LAK 
activity from the peripheral blood, spleen, bone marrow, and tumor but is capable 
of generating cells with LAK activity from cells exposed to IL-2 for as little 
as 4-24 hours before exposure to IL-4 or from the peripheral blood of patients 
recently receiving IL-2 (3). On monocytes, IL-4 serves as a potent Inducer of 
CDll, CD23, and down regulator of CD14 and IgG-Fc receptors, including CD32 and 
CD64. It has been suggested that it might act as a potential anti-inflammatory 
agent in that it suppresses human monocyte production of tumor necrosis factor 
alpha. Interleukin- 1 , and prostaglandin E-2 (17). 
The human IL-4 gene product is not active on murine cells. Only limited 
studies of murine IL-4 have been conducted in vivo (Mule, unpublished 
observations). Recently, however, gene transfection studies utilizing IL-4 
introduced into two murine tumors by Tepper, Leder, and colleagues, have shown 
that IL-4 transfection is associated with tumor regression (18). Similarly, 
soluble IL-4 receptors have been prepared and, when administered to animals 
bearing allografts, have been shown to prolong allograft survival (Widmer, 
unpublished information). Based on these observations, as well as findings that 
IL-4 enhances specific growth of human tumor infiltrating lymphocytes, we 
initiated a phase I/II clinical protocol employing recombinant IL-4 in the fall 
of 1988. 
Recombinant yeast derived IL-4 has been administered alone or in 
conjunction with IL-2 to 73 cancer patients in 84 treatment courses. The maximal 
tolerated dose of IL-4 administered thrice daily appears to be 20 ug/kg and in 
conjunction with IL-2 at 7.2 x lO^IU/kg to be also 20 ug/kg. IL-4 administration 
is also associated with the development of vascular leak syndrome with end organ 
dysfunction (kidney) presumably secondary to edema, similar to that observed with 
IL-2. The weight gain observed with IL-2 or IL-4 administration alone appears 
to be increased when both of these cytokines are given together. (Median weight 
f ain 8.6 - 9.1% with high dose of IL-4 alone, 7.5% with high dose IL-2 alone and 
5.2% with IL-4 and IL-2). Unusual side effects occuring prominently during IL-4 
treatment include mild to severe nasal congestion and endoscopically proven 
gastritis and antral gastric ulceration (12 episodes) . We observed no responses 
(partial or complete) in 48 patients receiving IL-4 alone during 56 treatment 
courses including 23 patients with melanoma or renal cell cancer previously 
untreated with IL-2. Five patients in this group with stable disease were 
retreated safely, and received 4 courses (one patient) , 3 courses (1 patient) and 
two courses (3 patients) of therapy. Responses have been noted in patients 
previously failing IL-4 on subsequent IL-2 based regimens. A total of six 
responses were noted in 28 patients receiving combinations of IL-2 and IL-4 
including 1 complete response in a patient with renal cell cancer, 3 partial and 
1 mixed response in patients with melanoma and 1 partial response in a woman with 
breast cancer. 
We have also identified and characterized a more potent type of killer 
cell obtained directly from tumors, called the tumor infiltrating lymphocyte 
(TIL) (19). These cells are cytolytic T lymphocytes. In both mice and humans, 
TIL can develop the ability to specifically lyse the syngeneic or autologous 
tximor and not normal cells or allogeneic tumors (20, 21). We demonstrated that 
TIL were from 50-100 times more potent on a per cell basis than were LAK cells 
Recombinant DNA Research, Volume 14 
[837] 
