Repetitive weekly cycles of 4-day continuous infusion of recombinant interleukin-2: a phase I study

Factors associated with response to high-dose interleukin-2 in patients with metastatic melanoma

PURPOSE

The present study attempted to identify characteristics that correlated with clinical response to interleukin (IL)-2 therapy in patients with metastatic melanoma.

PATIENTS AND METHODS

We retrospectively evaluated laboratory and clinical characteristics of 374 consecutive patients with metastatic melanoma treated with high-dose intravenous bolus IL-2 (720,000 IU/kg) from July 1, 1988, to December 31, 1999, at the Surgery Branch of the National Cancer Institute.

RESULTS

The overall objective response rate was 15.5%. Pretreatment parameters such as patient demographics, laboratory values, and prior therapy did not correlate with response; however, 53.6% of patients with only subcutaneous and/or cutaneous metastases responded, compared with 12.4% of patients with disease at other sites (P2 =.000001). During therapy, patients who were responders tended to have received more doses during course 1 (16.2 +/- 0.3 doses v 14.5 +/- 0.2 doses; P2 =.0095); however, when limited to patients who were able to complete both cycles of course 1, there was no statistically significant difference (P2 =.27). Responders had a higher maximum lymphocyte count immediately after therapy compared with nonresponders (P2 =.0026). The development of abnormal thyroid function tests and vitiligo after therapy was associated with response (thyroid-stimulating hormone, P2 =.01; free T4, P2 =.0049; vitiligo, P2 < 10(-6)), although thyroid dysfunction may have been related more to the length of IL-2 therapy than to response.

CONCLUSION

The presence of metastases only to subcutaneous and/or cutaneous sites, lymphocytosis immediately after treatment, and long-term immunologic side effects, especially vitiligo, were associated with antitumor response to IL-2 therapy.

Decreased tolerance to interleukin-2 with repeated courses of therapy in patients with metastatic melanoma or renal cell cancer

High-dose interleukin-2 (IL-2) therapy has a response rate of approximately 20% in patients with metastatic melanoma and renal cell cancer. Animal models have shown that the anti-tumor effects of IL-2 are dose and schedule dependent, and one report has shown that patients with melanoma who responded to IL-2 therapy received more doses of IL-2 than did those who did not respond. The current study evaluated patients' tolerance to IL-2 over multiple courses of therapy and the factors that affected the number of doses delivered. Patients with metastatic melanoma or renal cell cancer who received at least two consecutive courses of high-dose intravenous IL-2 alone from October 1, 1985 through December 31, 1996 were evaluated. Patients served as their own controls in paired analyses. The number of doses tolerated from one course to the next and the reasons for stopping therapy were analyzed. One hundred fifty-nine patients received two or more courses of therapy during the study. The median number of doses of high-dose IL-2 decreased from course 1 (15 doses) to course 2 (12 doses) (p2 = 0.0001). Pretreatment factors were not found to significantly influence the decrease in the number of doses delivered. Only 2 of 33 separate toxic effects resulting in discontinuation of IL-2 dosing were found to be significantly different between courses. After adjusting for multiple tests of statistical significance, serum aspartate aminotransferase elevations were more likely to stop course 1 (p2 = 0.0033) and creatinine elevations were more likely to stop course 2 (p2 = 0.00007). The influence of renal toxicity was further assessed by comparing the median creatinine value at the time IL-2 dosing was discontinued. This difference was found to be significant when cycle 1 of course 1 (1.5 mg/dL) was compared with cycle 1 of course 2 (1.8 mg/dL; p2 = 0.0001). When pretreatment factors were analyzed, male sex (p2 = 0.006), a diagnosis of renal cell cancer (p2 = 0.008), previous nephrectomy (p2 = 0.001), and older age (p2 = 0.0055) were significantly associated with the development of renal toxicity that resulted in discontinuation of IL-2 therapy. Furthermore, the same four patient subsets had higher baseline creatinine values in individual univariate analyses. This study identified subsets of patients who tolerated less IL-2 with repeated courses. The decreasing tolerance to IL-2 was associated primarily with elevations in creatinine. Finding ways to ameliorate the renal toxicity seen during IL-2 therapy in this patient population may allow an increase in the number of IL-2 doses administered and potentially an increase in clinical response.

Cytokine-induced autoimmune disorders

Cytokines are now commonly used in the treatment of many conditions, especially cancer, haematological malignancies and chronic viral hepatitis. With some of these cytokines, clinical induction and/or exacerbation of autoimmune manifestations have been observed. This has been the case with interferon-alpha and interferon-gamma, interleukin-2 and some colony-stimulating factors. All known biological and clinical autoimmune features have been observed but thyroid abnormalities have been particularly frequent. Some of these manifestations appear to be related to the effect of these exogenous cytokines on the T helper cell (TH) 1/TH2-cytokine balance and the regulatory properties of these cells. Patients with a history of underlying autoimmune disease or baseline serological abnormalities should be monitored for autoimmunity when treated with certain TH1 or TH2-inducing cytokines.

Activation of the immune system of cancer patients by continuous i.v. recombinant IL-2 (rIL-2) therapy is dependent on dose and schedule of rIL-2

The effect of dose and schedule of continuous i.v. rIL-2 infusions on leucocyte subset counts, activation status of CD56+CD3- natural killer (NK) and CD3+ T lymphocytes, and cytolytic activities of peripheral blood mononuclear cells (PBMC) was studied. A single 4-day course of rIL-2 in escalating doses (0.9-11.5 x 10(6) U/m2 per day) was given to 18 patients with various types of metastatic cancer. The serum IL-2 concentration during rIL-2 therapy ranged between 23 and 64 U/ml and was proportional to the administered rIL-2 dose, as was the rebound lymphocytosis following therapy. Before therapy, the CD56+CD3- NK cells expressed low levels of the p75 chain of the IL-2 receptor (IL-2R) and virtually no IL-2R(p55). Most CD3+ T cells were IL-2R(p55-,p75-). Between 2 and 4 days following therapy, i.e. at the time of lymphocytosis, the percentage of CD56+,CD3- NK cells among the lymphocytes had increased proportional to the administered rIL-2 dose. The levels of IL-2R(p75) expression by the CD56+,CD3- NK cells had increased. The percentages of CD3+ T cells expressing IL-2R(p55), HLA-DR and CD45RO had increased proportional to the administered rIL-2 dose. The level of lymphokine- activated killer (LAK) activity against Daudi cells was also positively correlated with rIL-2 dose. Subsequently, seven patients received 4-weekly cycles of rIL-2 (2.9-4.4 x 10(6) U/m2 per day) during 4 consecutive weeks. This schedule led to marked increments in lymphocyte and eosinophil counts, and to increased cytolytic activities compared with pretreatment. We conclude that CD56+,CD3- NK and CD3+ T cells are activated differentially by continuous i.v. rIL-2 proportional to dose and duration of treatment.