Cytokines are the messengers of the immune system.3 Cytokines are substances, either proteins or glycoproteins, secreted by immune cells. They have autocrine and paracrine functions, so that they function locally or at a distance to enhance or suppress immunity. In cancer therapy, we generally use cytokines to enhance immunity. Video    Download Free RealPlayer

Cytokines regulate the innate immune system: natural killer (NK) cells, macrophages, and neutrophils. They also regulate the adaptive immune system, the T and B cell immune responses. In the immune system, cytokines function in cascades. Thus clinical trials of individual cytokines are rarely useful, since cytokines tend not to work individually. Some of the individual cytokines that have been tested and found ineffective for cancer treatment include interleukin 1 beta (IL-1 beta), although it may be useful because it helps to mediate the severe toxicity of interleukin 2 (IL-2). Tumor necrosis factor (TNF) certainly sounded promising, but in fact caused severe hypotension when used systemically. Interleukin 4 (IL-4) showed minimal anti-cancer activity and was toxic. Interleukin 6 (IL-6) had some activity against cancer cells, but turned out to be a growth factor for myeloma cells. Granulocyte-macrophage colony-stimulating factor (GM-CSF), used primarily in stem cell transplant to reconstitute the myeloid series, has been studied for melanoma with controversial results.

Which cytokines are important for cancer? IL-2 and interferon-alfa 2b are two cytokines approved by the FDA for treatment of cancer. IL-2 has demonstrated activity against renal cell, melanoma, lymphoma, and leukemia. Interferon has activity in the same histologies but also in Kaposi's sarcoma, chronic myelogenous leukemia, and hairy cell leukemia. Overall, cytokines are substances that appear to have application in the treatment of hematologic malignancies or immunogenic tumors. Here we will examine the major cytokines currently in use or under evaluation for cancer therapy: interferon alfa, IL-2, GM-CSF, and interleukin-12 (IL-12).

Interferon alfa

Interferon alfa is actually a family of molecules rather than a single molecule. They are encoded by closely related genes on chromosome 9, encoding proteins that are variably glycosylated. These are comprised of about 150 amino acids, and they bind to certain receptors on the surface of immune cells. They are known to have profound and diverse effects on gene expression.

Interferon alfa has many roles. It upregulates genes like MHC class I, tumor antigens, and adhesion molecules. It is also an anti-angiogenic agent. It is extremely active in the immune system, promoting B and T cell activity. Interferon alfa can stimulate macrophages, even dendritic cells, and upregulates Fc receptors.

Interferon was isolated in 1970 by investigators looking for antiviral substances. The substance we now know as interferon was isolated from white cells, and called interferon because it interfered with viral infection. It turns out type 1 and 2 interferons do have some antitumor activity, but not the hoped-for level.

Interferon's activity in cancer has been well documented, however. In kidney cancer we have seen small but consistent response rates in a number of studies. In a randomized controlled study of interferon alfa versus megestrol acetate, investigators found a minimal advantage to interferon in terms of survival (8.5 months vs. 6 months). However disease-free survival at 2 years was poor. With a response rate in the 5% to 10% range, it is somewhat surprising that interferon is now considered standard practice in the treatment of kidney cancer, especially in patients who are relatively sick and cannot tolerate IL-2. Optimal dose, route, and schedule are unknown. Most investigators would agree that for any therapeutic effect in kidney cancer, interferon doses of 5 million to 10 million units 3 times a week subcutaneously are reasonable. Still, in looking through most of the established literature on the uses of interferon alone in phase 2 trials, the median response rate is only about 11% with a range of 8% to 29%. Clearly, interferon alfa is not very active in kidney cancer and is in fact not FDA-approved for this use. It is considered second-line therapy for renal cell carcinoma, and then as a palliative maneuver. Eligible patients should be offered interleukin-2, which is currently the only FDA-approved drug for renal cell carcinoma. Video

Interferon has a long history in metastatic melanoma. A series of small phase 2 trials with high-dose interferon (up to 100 million units IV daily) reported response rates of 20% to 22%.43,44 While not markedly different than any single chemotherapy in metastatic melanoma, this response rate was significant and spurred continued interest. Results in combination with cimetidine were also disappointing. Five trials of interferon alfa with cimetidine showed no advantage.

The definitive test of whether adding any agent to interferon enhances its antitumor effect was an ECOG trial with dacarbazine alone, dacarbazine with tamoxifen, dacarbazine with interferon, or interferon with tamoxifen. In the approximately 260 patients, median time to progression was a brief 3.7 months. Overall survival in the interferon plus dacarbazine arm was 9.1 months, which was identical to the dacarbazine only arm.

Interferon was then evaluated in the adjuvant setting.45 In this setting, interferon's activity in melanoma is high. Patients with stage 3 and 2B resected high-risk melanoma were randomly allocated to observation or high-dose interferon for one year (N = 287). Interferon was administered intravenously every day for 4 weeks, at a dose of 20 million units/m2, followed by a maintenance regimen of 10 million units/m2 3 times a week subcutaneously. The group treated with interferon showed definite disease-free and overall survival benefits, which ultimately led to FDA approval of this regimen. Survival with interferon was 3.8 years versus 2.78 with observation – essentially a year more with interferon. Relapse-free survival also showed a 1-year advantage with interferon. The real advantage was 27% augmentation of survival at 7 years of follow-up, and at 9 years of follow-up the P value was between 0.03 and 0.04. It may not be a huge increment, but it is significant.

The follow-up confirmatory trial included 642 patients who received either a year of high-dose interferon, 2 years of lower dose interferon given subcutaneously at 3 million units a week, or observation.46 No differences in survival were shown. The E1690 trial did not confirm the results of the earlier trial. The hazard ratio for survival for high-dose interferon was only 1.0, meaning it was not different from observation. Amazingly, the median survival in the observation arm was 6 years. The median survival in the observation arm before was, as we said, a little less than 3 years. How could the survival change?

A third trial was initiated, comparing high-dose interferon versus a vaccine.47,48 One year of interferon was compared to 18 months of vaccination with ganglioside vaccine. The vaccine arm was not terribly toxic, while interferon had a fair degree of toxicity. This trial accrued 774 patients, but was terminated early. A clear disease-free and survival advantage for the interferon arm was evident early in the study. Relapse-free survival was now again prolonged with a hazard ratio of 1.6. Overall survival was prolonged with interferon with a hazard ratio of 1:3; stage 2B patients benefited and the P value for overall survival with a little over 2 years of follow-up was 0.04. At 3 years of follow-up the P value was even better at 0.03. The third trial essentially replicated the results of the first.

The appearance of improvement in relapse-free survival with the same treatment regimen over time is probably a function of better supportive care, better staging -- CT scans, PET scans, and MRIs lead to more accurate diagnosis. The treatment has not changed, but the patient population has.

These studies laid the foundation for interferon as standard adjuvant therapy for stage III resected melanoma. The toxicity remains an issue. When we treat patients with this regimen, we prescribe an antidepressant to decrease some of the toxicity, the depression, and decreased energy level. Vigorous hydration is essential as well with high-dose interferon, as patients tend to become dehydrated.

Interferon has been approved as treatment in other histologies as well. Interferon is definitely active in hairy cell leukemia.49 Nine complete and 17 partial responses were documented by bone marrow core biopsies. Peripheral blood hematologic indices improved or normalized in all patients. Previously untreated patients showed higher complete remission (P = 0.02) than in those who had undergone splenectomy. Therapy was well tolerated, with most patients experiencing tumor remission reporting improved quality of life.

Another study found similar results in a small population of patients, with an overall response rate of 93%.50 Peripheral blood counts returned to normal levels. This study also assessed natural killer cell activity and immunologic surface markers, and noted normalization of both measures after therapy.

A significant survival benefit of more than 89 months in a phase 2 trial in patients with chronic myelogenous leukemia suggests interferon has activity here as well. 51,52 This survival advantage was independent of cytogenetic improvement with interferon, which was also noted. We may see future approval for interferon alfa in hematologic cancers as well as solid tumors.


Interleukin-2 (IL-2) is a T cell growth factor that binds to a specific tripartite receptor on T cells. In dose escalation studies, patients treated with high doses of IL-2 showed clinical responses, although severe toxicity was seen. Response rates were as high as 24% at the highest IL-2 dose in patients with renal cell carcinoma. However, the toxicities of treatment were limiting. These toxicities stem from what is known as a capillary leak syndrome. Giving IL-2 in high doses is comparable to inducing a controlled state of septic shock. Low blood pressure, low systemic vascular resistance, high cardiac output, grade 3/4 hematologic toxicity, hepatic toxicity, renal toxicity, and pulmonary edema have all been documented. Toxicity is nearly always reversible. Video

The typical IL-2 regimen – 600,000 to 720,000 IU/kg, an average of 50 million units of IL-2 per dose given three times a day as a bolus over 15 minutes. The maximum most patients can tolerate is 14 doses. A rest period of 5 to 9 days between cycles is recommended, and patients must be treated in a step-down situation or ICU. This regimen was evaluated in 255 patients with renal cell carcinoma.53,54 Objective responses were initially reported in 15% of patients, with 17 (7%) complete responses. These figures were confirmed in follow-up. These numbers were not particularly impressive, but the median duration of response (54 months) and median duration for complete responses (80 months) demonstrated the potential benefit of IL-2 in these patients. Median survival for all patients was 16.3 months, with 10% to 20% estimated to be alive 5 to 10 years after treatment.55 The toxicity associated with IL-2 therapy is a major consideration. These patients require intensive management. Dopamine is administered prophylactically and therapeutically for low-urine output, neosynephrine is given for hypotension, acetaminophen for fever and chills, cimetidine for gastric acidity, and oxacillin for neutrophil dysfunction that leads to sepsis. Concomitant antibiotic therapy is essential, as the rate of sepsis without antibiotic therapy is as high as 27%.

Low-dose IL-2 has shown activity in renal cell cancer as well.56,57 Objective response rates of 18% to 23% were reported, without the toxicity of high dose IL-2.

A larger study of 425 patients evaluated the activity of low-dose IL-2 in combination with interferon-alfa, as well as each agent alone.58 Interferon-alfa or IL-2 alone had low response rates, but the response rate for the combination was significantly higher (P<0.01), with significantly improved 1-year event-free survival (P=0.01). However, no difference in overall survival was shown.

IL-2 is indicated for use in renal cell carcinoma and melanoma on the basis of duration of response, rather than achievement of response. A substantial number of patients with renal cancer who had an objective response to treatment are still alive, and stable or progression-free, 1 year post-treatment. With melanoma, we see patients54 who have responded alive and progression-free 5 to 15 years later. In metastatic melanoma this duration of survival is uncommon when chemotherapy is used.

Clearly, patient selection is critical for IL-2 therapy. How can we predict who will respond to high-dose IL-2? As expected, melanoma patients with high performance status do well, as do patients who develop vitiglio or autoimmune thyroiditis. These findings support the idea that IL-2 generates a T cell response that reacts with melanocytes.

Another important determinant of response was the amount of IL-2 given during the first cycle. The immunologic effect is manifested by the height of rebound leukocytosis. Lymphocyte levels plunge during initial treatment with IL-2, and rebound when treatment stops. This rebound is an indication of the response to IL-2.

IL-2 has shown activity in non-Hodgkin's lymphoma and leukemia and lymphoma post-stem cell transplant. Low-dose IL-2 was also evaluated in combination with histamine, but no differences in response were observed compared with IL-2 alone.

An interesting modification of the IL-2 molecule, known as BAY 50-4798, has shown potential in animal models. This agent is the same as IL-2 with two modified amino acids. It appears to have the therapeutic effect of IL-2 without the toxicity. Phase I dose escalation studies are underway. What we have seen so far is potential for clinical response with very little toxicity compared to IL-2.

Granulocyte-monocyte colony stimulating factor

Granulocyte-monocyte colony stimulating factor (GM-CSF) is a well-known cytokine, approved for use in stem cell and bone marrow transplant to reconstitute the myeloid series. We have data suggesting it also might have application as monotherapy in melanoma.59 In this trial, 48 patients with stage III and IV melanoma were treated with long-term, chronic, intermittent GM-CSF after surgical resection. The theory is that GM-CSF would reconstitute antigen-presenting cells, and thus the ability to mount an immune response. Overall and disease-free survival were significantly prolonged by GM-CSF therapy in patients who were clinically disease-free. Median survival was 37.5 months versus 12.2 months in matched controls (P < 0.001). Treatment was well tolerated.

However, the strength of these observations is questionable. The matched controls (disease, stage, age, sex) came from a database of patients from 1960 to 1988, before PET and MRI scans, sentinel node biopsy, sophisticated staging and treatment. Whether the therapeutic effect was due to general improvements in treatment over the past 41 years or to GM-CSF therapy will have to be confirmed in studies with true control arms.


Interleukin-12 (IL-12) is a very exciting cytokine. It is a heterodimeric protein that promotes NK and T cell activity and is a growth factor for B cells. It has demonstrated antitumor activity in mouse models. Alone, IL-12 shows minimal potential for therapeutic effect.60

However, IL-12 may have value as a vaccine adjuvant. When IL-12 was paired with peptide vaccines in patients with resected stage 3 and 4 melanoma, IL-12 appeared to boost the response to the vaccine.61 GM-CSF is also being evaluated as an adjuvant for vaccine therapy.

Cytokines such as IL-2 have modest antitumor activity in metastatic renal cell carcinoma and melanoma. The effect is remarkable for its duration rather than strength, and its application may have greatest significance for long-term survival. IL-2 at high doses is toxic. Low doses may have promising activity post-transplant as an immune restorative agent, and that effect is under evaluation in large randomized cooperative group studies. IL-12 and GM-CSF are unlikely to be stand-alone agents for any histology, but show promise as adjuvant therapy and in combination with other cytokines.

Continue to Vaccine Approaches Against Cancer