Abstract
Background: High-dose interleukin-2 (HD IL-2) is known to produce durable responses in metastatic melanoma. The purpose of this study was to evaluate the response of metastatic melanoma to treatment with HD IL-2. Patients and Methods: A retrospective analysis was performed on all adult patients with stage IV melanoma treated with HD IL-2 from January 2000 to October 2008 at the University of Minnesota. HD IL-2 was given intravenously every 8 hours at 600,000 IU/kg for a maximum of 14 doses per course. Results: Fifteen patients with metastatic melanoma had been treated with HD IL-2. There were 4 patients exhibiting some response, with 1 complete response (CR), 1 partial response (PR), 1 mixed response (MR) and 2 stable disease (SD). Average time to disease progression (TTDP) was 5.67 months. Two patients had complete resolution of brain lesions after HD IL-2 therapy. One of these patients experienced CR and is disease free 34 months after stopping therapy. The other patient experienced MR and is currently alive with disease, but without recurrence of brain lesions. Twelve out of the 15 patients received 2 courses of therapy. Common grade (G) 3 and 4 adverse events included: hyperbilirubinemia (G 3=26.67%), hypotension (G 3=6.67%, G 4=6.67%), peripheral edema (G 3=26.67%), and pulmonary edema (G 3=13.33%). Conclusion: We propose further evaluation of HD IL-2 in patients with brain metastases because this patient population is typically considered ineligible for HD IL-2 therapy.
Metastatic melanoma carries a grim prognosis, with a 5-year survival rate of less than 10% (1). Metastatic tumors can invade virtually any organ and are unresponsive to most standard chemotherapeutic agents. The drugs dacarbazine, temozolomide and paclitaxel, with or without combination cisplatin or carboplatin, have shown activity in metastatic melanoma. The response rates from these drugs are minimal, with most patients dying of the disease. No consensus exists on the preferred regimen for patients with metastatic melanoma. As a result, the majority of patients are enrolled in clinical trials as a first-line therapy.
Another approach to treating metastatic melanoma is with immunotherapeutic agents, such as interleukin 2 (IL-2). IL-2 is a cytokine produced by human T lymphocytes and is involved in regulating immune reactions. It is hypothesized that IL-2 stimulates the activation and expansion of immune T-cells with T-cell receptors capable of recognizing tumor antigens (2). High-dose (HD) IL-2 was found to have activity in metastatic melanoma patients when used alone or with lymphokine-activated killer cells (3). A phase II study in 1990 showed that HD IL-2 yielded a significant overall response rate (ORR) in patients with metastatic melanoma (4). A retrospective analysis of 250 patients with metastatic melanoma revealed an ORR of 16% (95% confidence interval, 12-21%) (5). In these studies, complete responses (CR) are commonly reported. The durability of complete responses has been reported as 6.6% after a median follow-up of 7 years (6).
Although these response rates are encouraging, HD IL-2 treatment is associated with significant acute toxicities, mainly revolving around vascular leak syndrome. Vascular leak syndrome causes a sepsis-like syndrome with widespread capillary leakage causing decreased intravascular volume resulting in hypotension, pulmonary edema, systemic edema and renal failure (7). Despite this, HD IL-2 therapy is one of the only treatments that results in durable CR, and therefore it continues to be utilized in metastatic melanoma.
Management of patients with melanoma metastatic to the brain has been an area of concern. Brain metastases are a major cause of morbidity and mortality in metastaticmelanoma. In one major study of 702 metastatic melanoma patients, the median overall survival of patients with brain metastases was 3.8 months (8). Treatment of these lesions is limited as there are few effective therapies. HD IL-2 has typically been avoided due to concerns over vascular leak syndrome and the possibility of increased intracranial pressure. HD IL-2 can also cause thrombocytopenia which raises the risk for hemorrhage from brain lesions. Additionally, neurologic side effects from HD IL-2 include confusion and behavioral changes, making it difficult to discern symptoms from worsening brain metastases and side effects of therapy. For these reasons, patients with brain lesions have been excluded from clinical trials in the past. Now data are emerging that suggest that HD IL-2 may be useful for patients with melanoma brain metastases (9).
The purpose of this retrospective review was to evaluate the outcomes of patients with metastatic melanoma treated with HD IL-2 at a single institution. The outcomes were analyzed for associations with metastatic site, age, sex, lactate dehydrogenase (LDH) levels and previous therapies.
Pantients and Methods
Study design and patient selection. A retrospective analysis was performed on all metastatic melanoma patients treated with HD IL-2 between January 2000 to October 2008 at the University of Minnesota Medical Center, Fairview in Minneapolis, Minnesota. The purpose of the analysis was to compare clinical response to HD IL-2 to previously published data. All patients in the study had a histologically documented melanoma with metastatic disease, stage IV, based on tumor, node, metastasis (TNM) staging criteria. All patients had an Eastern Cooperative Oncology Group (ECOG) performance score between 0-2 due to the toxicity of the treatment. Minimum age of all participants was at least 18 years. The retrospective chart review was approved by the University of Minnesota's Institutional Review Board (IRB).
Treatment schedule. HD IL-2 was given intravenously every eight hours as a bolus over 15 minutes at a dose of 600,000 IU/kg. The therapy was given in a non-intensive care hospital setting. A maximum of 14 doses for each cycle was given. Patients were closely monitored for side effects which centered around capillary leak syndrome including hypotension, systemic edema and pulmonary edema. After receiving each cycle, patients were given 10-21 days' rest before starting the next cycle. If patients had a response to therapy, they continued with further cycles. Therapy was halted if patients could no longer tolerate the side-effects or if their disease progressed. Patients with a CR also had their therapy stopped.
Evaluating response. After every 2 treatment cycles, computed tomography (CT) imaging of the chest, abdomen and pelvis was performed to assess response. In patients with brain metastases, magnetic resonance imaging (MRI) of the brain was performed to evaluate response. CR was defined as disease that had disappeared and stayed in remission for at least one month. Partial response (PR) was defined as disease that had regressed by greater than 30% in tumor size. Stable disease (SD) was classified as disease that had regressed in size by less than 30% or grown by no more than 20%. Disease progression was indicated when the tumor had increased by more than 20% in size. Mixed response (MR) was defined as having at least one site of tumor progression and one site of tumor response.
Patient characteristics (N=15).
Data collection. Data were collected from patient medical records in the following areas: sex, age, primary site, prior treatments, metastatic sites, LDH at diagnosis, response to therapy, number of doses per cycle, number of treatment cycles, time to disease progression (TTDP) and toxicity.
Toxicity evaluation. Toxicity of therapy was evaluated for each patient. The NCI Common Terminology Criteria for Adverse Events v3.0 was used when grading toxicity (10). Grades (G) 3 and 4 toxicity are reported in this study.
Statistical methods. For the purposes of these analyses, patients were grouped based on clinical benefit of therapy. Overall response rate (ORR) was calculated by adding patients with a PR, MR and CR.
Fisher's exact test was used to determine if gender, age, prior treatment or elevated LDH were related to clinical benefit from therapy. TTDP was calculated from study entry date to the date of progression or was censored at the date of last contact for patients who remained disease free. TTDP was summarized using Kaplan-Meier methods (11). All analyses were performed using SAS version 9.1 (SAS Institute Inc., Cary, NC, USA). P-values less than 0.05 were considered statistically significant.
Results
Patient characteristics. A total of 15 patients met the retrospective review criteria and were included in this retrospective study (Table I). The most common sites of metastasis were the lungs (N=9, 60%), lymph nodes (N=7, 46.67%), skin/subcutaneous (N=5, 33.3%) and the liver(N=5, 33.3%). Other sites of metastasis included the brain, the heart, muscle, adrenal glands, parotid glands, pancreas and the chest wall. Gender was virtually equally distributed with 7 females and 8 males. A total of 9 of the 15 patients (60%) had received prior systemic treatment. Average age at diagnosis was 48.8 years. Average LDH at initiation of therapy was 627.7 U/L.
Treatment response. PD was seen in 66.67% (N=10) of patients. CR was seen in 6.67% (N=1), PR in 6.67% (N=1), MR in 6.67% (N=1) and SD in 13.33% (N=2) of patients. ORR was seen in 20% (N=3) of patients. The median TTDP for patients with SD and MR was 6.5 months and 3 months, respectively. The CR patient is alive and disease free after 34 months.
Twelve out of the 15 identified patients (80%) completed two cycles of HD IL-2 therapy. Due to toxicity or response, no patients received a third or additional cycle of therapy. A total of 20% (N=3) of patients were only able to tolerate one cycle. The average number IL-2 injections per cycle for patients with CR, PR, SD and MR were 9, 9, 5, and 10, respectively. The average number of injections given during the first cycle was 10.47. The average number of injections given to patients receiving two cycles was 8.83.
Correlative studies. There was no significant association between benefit of therapy based on sex (p=0.608) or prior treatment (p=1.0). The average LDH at diagnosis for responders and non-responders was 591.6 (SD=194.9) and 645.7 (SD=201.7), respectively. LDH at diagnosis was not found to be a statistically significant predictor of response to therapy (p=0.629). The average age at diagnosis for responders and non-responders was 53.4 years (SD=5.1) and 46.5 years (SD=9.3), respectively. The association between age and response to therapy was not found to be statistically significant (p=0.151).
Treatment-related toxicity. Toxicity of therapy was evaluated in all patients. Significant G 3 and 4 adverse events included hyperbilirubinemia (G 3=26.67%), hypotension (G 3=6.67%, G 4=6.67%), peripheral edema (G 3=26.67%), fever (G 3=13.33%) and pulmonary edema (G 3=13.33%). These adverse events are further outlined in Table II.
Response of metastatic brain tumors to HD IL-2. Further analysis of study participants revealed that two patients with brain metastases were treated with HD IL-2. The characteristics of these two patients are outlined in Table III. Neither of these patients had neurologic symptoms from their brain metastases prior to therapy. One patient achieved a CR with therapy and is disease free 34 months following therapy. This patient had received prior treatment with whole brain radiation due to the development of multiple-brain metastases. The whole-brain radiation decreased the size of most of the lesions to the point that there were only two measureable lesions at the time of initiation of HD IL-2. A brain MRI performed prior to initiation revealed a 10 mm diameter lesion in the left frontal lobe and a 17 mm diameter lesion in the left parietal lobe. Innumerable smaller brain metastases were also noted on MRI. Follow-up brain MRIs after HD IL-2 therapy showed complete resolution of all brain metastases. The response has been durable for over 34 months.
Adverse events in patients treated with HD IL-2.
The second patient with brain metastases received an MR from HD IL-2, with complete resolution of a single brain metastasis. Prior to being treated with HD IL-2, the patient had been treated with temozolomide. A brain MRI prior to initiation of HD IL-2 revealed a 2 mm diameter lesion in the medial left parietal lobe. Following treatment with the HD IL-2, the lesion resolved. There has been no recurrence of the brain metastasis after 21 months; however, the patient has had disease recurrence at other sites.
Two patients with brain metastases.
Discussion
The purpose of this retrospective study was to compare the response of patients with metastatic melanoma treated with HD IL-2 at the University of Minnesota to the patient responses reported in other studies. In our study, the ORR (20%) and durable CR (6.67%) are comparable to previously published data (4, 5). There were no partial responses in our population; however, patients with MR or SD had rates of TTDP that are similar to those reported in other studies. The reported G 3 and G 4 toxicities in our study are similar to those reported in other studies, but the frequency of specific toxicities is variable.
Perhaps one of the most interesting findings of this study was that two patients with brain metastases had favorable responses to HD IL-2. One patient actually had a durable CR, with complete resolution of two measurable brain metastases. This patient remains disease free after 34 months and has not received further therapy. Prior to receiving HD IL-2 this patient had been treated with the large multivalent immunogen (LMI) vaccine and whole-brain radiation. The patient had progression of disease on the LMI vaccine. After whole-brain radiation, these lesions were noted to stabilize and the amount of surrounding edema decreased substantially. Prior to administration of HD IL-2, these lesions had been stable but persisted at a diameter of 10 mm and 17 mm. It was only after the patient received HD IL-2 that the lesions resolved completely.
The second patient had been treated with temozolomide (TMZ). Although this drug has been noted to show activity in brain metastases, this patient's brain lesion had not responded to TMZ. The lesion was noted to resolve completely following HD IL-2 and has not recurred on follow-up brain MRI studies. Despite the fact that TMZ shows activity in melanoma brain metastases, it is felt that the CNS response was attributed to HD IL-2 in this case.
Typically HD IL-2 is not used in patients with brain metastases due to concerns over neurologic complications resulting from capillary leak syndrome. Additionally, there is concern with thrombocytopenia related hemorrhaging from metastatic brain lesions. Furthermore, some concern has been raised as to whether the brain is an immune-privileged site (12). This would suggest that HD IL-2 would not be an effective therapy for brain lesions. These concerns in conjunction with the other toxicities of HD IL-2 therapy have limited its use in patients with melanoma brain metastases.
Detailed studies have evaluated the array of toxicities with HD IL-2. Information from these studies has resulted in measures to improve safety and reduce the adverse effects associated with HD IL-2 (13). A retrospective study in 2002 evaluated the use of HD IL-2 in patients with brain metastases from melanoma and renal cell carcinoma, in which patients with previously treated and untreated brain metastases prior to HD IL-2 therapy were evaluated for safety and efficacy (9). These patients had a limited number of small lesions with little or no surrounding edema. The study found that patients with and without brain metastases had similar G 3 and 4 toxicities, and neurologic adverse events were similar in both groups. Overall, there was no statistical difference between response rates in the patients with brain metastases and those without. Specifically, 2 out of 36 (5.6%) patients with previously untreated brain metastases had both intracranial and extracranial response to HD IL-2 therapy. These data suggest that the brain is not an immune-privileged site and thus HD IL-2 may be an option for carefully selected patients with brain metastases.
Brain metastases are common in patients with metastatic melanoma and methods to effectively treat brain metastases are few. Several acceptable methods exist for treating melanoma brain metastases including surgical resection (14), whole brain radiation therapy (15), stereotactic radiosurgery (gamma-knife) (16), chemotherapy or a combination of these modalities (17). Surgical or stereotactic radiosurgery techniques are effective in patients with few lesions and minimal systemic disease. These two therapies have been shown to provide definitive local control of brain metastases in this select group of patients. Patients with multiple or surgically inoperable lesions are typically treated with whole brain radiation with or without chemotherapy. Chemotherapeutic agents shown to have activityin melanoma brain metastases include TMZ (18), and TMZ (19). Despite their activity, these agents have minimal response rates and are not definitively proven to significantly prolong survival. Immunotherapies, such as HD IL-2, have typically been avoided for use in patients with brain metastases and few trials analyzing these therapies exist (17).
Our retrospective review is limited due to the small size of the study, but nevertheless it does add important information to the prior study. Findings from this prior study as well as our study raise the question as to whether HD IL-2 should be utilized in patients with brain metastases. Although the number of patients included in both of these studies was small, the fact that there was evidence of an intracranial response to HD IL-2 is intriguing. These two patients had been treated with prior therapies that are known to have activity in melanoma brain metastases, but based on the sequence of therapies and responses, it is felt that HD IL-2 was behind their response. Brain metastases are quite difficult to treat in patients with melanoma and lead to significant morbidity and mortality. The need for more therapies to adequately treat this patient population is obvious. Based on data from these studies, we believe that further prospective and retrospective studies looking at HD IL-2 in carefully selected melanoma patients with brain metastases should be developed.
Acknowledgements
We thank Michael Franklin for editorial support.
Footnotes
-
Sources of Support: Experimental Therapeutics Fund, Minnesota Medical Foundation.
- Received July 23, 2009.
- Accepted August 18, 2009.
- Copyright© 2009 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
