Consensus panel members included the following: Michael B. Atkins, MD (Beth Israel Deaconess/Harvard Medical School, Boston, MA); Ronald M. Bukowski, MD (The Cleveland Clinic Foundation, Cleveland, OH); Robert A. Figlin, MD (David Geffen School of Medicine, Los Angeles, CA); Bonni Lee Gearhart, MD (Overlook Hospital Cancer Center, Summit, NJ); Thomas E. Hutson, DO, PharmD (Texas Oncology, PA/Baylor Sammons Cancer Center, Dallas, TX); Khuda Dad Khan, MD, PhD (American Health Network Oncology/Hematology, Indianapolis, IN); Robert J. Motzer, MD (Memorial Sloan-Kettering Cancer Center, New York, NY); Robert G. Uzzo, MD (Fox Chase Cancer Center, Philadelphia, PA); and Christopher G Wood, MD (The University of Texas M.D. Anderson Cancer Center, Houston, TX).
New developments in the treatment of patients with metastatic renal cell cancer (MRCC) have suggested a need to reevaluate the role of systemic therapies. The authors convened a panel of medical and urologic oncologists to rate the appropriateness of the main options.
The authors used the RAND/University of California-Los Angeles Appropriateness Method to evaluate systemic therapy options and cytoreductive nephrectomy. After a comprehensive literature review, an expert panel rated the appropriateness of systemic options (108 permutations) and cytoreductive nephrectomy (24 permutations) for patients with MRCC.
The appropriateness evaluation indicated that 27.3% of permutations were rated “appropriate,” 46.9% were rated “inappropriate,” and 25.8% were rated “uncertain.” There was a high rate of agreement (95%). Sunitinib and sorafenib were rated appropriate for patients with low-to-moderate risk regardless of prior treatment. Temsirolimus was rated appropriate for first-line therapy for higher risk patients. Interferon-α and low-dose interleukin-2 were rated inappropriate or uncertain. In patients who received prior immunotherapy, cytokines were rated inappropriate. In all permutations for evaluating systemic therapy, enrollment into an investigational trial was considered appropriate, treatment with bevacizumab was uncertain, and thalidomide was inappropriate regardless of risk status or prior therapy. For good surgical risk patients with planned immunotherapy, nephrectomy was rated appropriate in patients who had limited metastatic burden regardless of tumor-related symptoms and in symptomatic patients regardless of metastatic burden. Only the most favorable combination of surgical risk, metastatic burden, and symptoms generated an “appropriate” rating for patients with planned targeted therapy.
Renal cell carcinoma (RCC) is the most common malignancy of the kidney and accounts for approximately 3% of adult tumors. Annual RCC incidence estimates have increased steadily, and approximately 25% of patients present with advanced disease at the time of diagnosis.1 The management of metastatic RCC (MRCC) remains a significant challenge both for the medical oncologist and for the urologist. Treatment with standard cytotoxic chemotherapeutic agents and radiotherapy have produced disappointing outcomes, with low overall response rates (<7%) and limited survival improvement. Several issues have focused attention on treatments that target immune manipulation, including reports of spontaneous regression of metastatic lesions, the presence of cytotoxic T lymphocytes in renal tumors, reports of prolonged stable disease in the absence of systemic therapy, and the recent description of tumor-associated and human leukemic antigen-restricted antigens on renal cancer cells.2, 3 A variety of cytokines have been studied in patients with MRCC. Two of these agents, interleukin-2 (IL-2) and interferon α (IFN-α), have antitumor effects that are reproducible and that can be durable in some patients.4, 5 Numerous clinical trials evaluating these agents have been conducted over the past 2 decades; however, controversy remains regarding which cytokine is optimal.
We are approaching a turning point in the treatment of MRCC. New molecular therapies aimed at vascular endothelial growth factor or its receptors on endothelial cells may offer new options with potentially fewer adverse effects than cytokine-based therapy. Two of these agents, sunitinib malate (SU11248)6 and sorafenib tosylate (BAY 43–9006)7, have been approved recently by the U.S. Food and Drug Administration,8, 9 but the majority of available evidence for these agents has yet to be published in a peer-reviewed setting. It is beginning to appear that these targeted therapies may offer a different type of benefit that is characterized by brief to possibly extended periods of treatment-maintained disease stabilization for the majority of patients rather than durable remissions off treatment for a small minority of patients. The identification of biomarkers promises to alter the therapeutic landscape, offering the potential to identify the patients who are most likely to benefit from specific therapeutic options. These developments also may affect the way we view the other major treatment modality in MRCC, cytoreductive nephrectomy.
Given these changes, it is useful to reevaluate key treatment decisions, taking into account new evidence as well as gaps in the evidence. Reliable data slowly are becoming available, but some definitive trial results with newer agents may be 1 or 2 years in the future. Meanwhile, decisions need to be made on a daily basis. For example, physicians and patients must consider potential trade-offs between the chances for long-term remission for a small minority of patients with cytokine therapy versus a delay in disease progression for the majority of patients with targeted therapies. We convened a multidisciplinary panel of medical and urologic oncologists to conduct a formal review of the available data on the main options for systemic treatment of patients with MRCC and to rate the appropriateness of each therapy by using the validated RAND/University of California-Los Angeles (UCLA) Appropriateness Method.
MATERIALS AND METHODS
The consensus panel method developed at RAND/UCLA combines an evidence-based review with practical experience from clinicians and leaders in the field. It has been demonstrated that appropriateness guidelines developed using this method are reproducible,10 are consistent clinically,11 and are correlated with clinical outcomes.12 Oncology applications have included breast cancer,13 melanoma,14 colorectal cancer,9, 15 and hematologic malignancies.16–19
We performed a comprehensive literature review to identify studies that assessed the use of systemic therapy in the treatment of MRCC. We searched the MEDLINE data base from 1994 to February 2005 for English-language articles by using search terms like metastatic renal cell carcinoma; carcinoma, renal cell; and clinical trial as keywords. A manual search of bibliographies of relevant articles and reviews was performed to identify additional publications. Abstracts from the website of the American Society of Clinical Oncology (ASCO) were searched, and we solicited references and opinions from experts in the field and conducted supplementary literature reviews as needed.
The treatments assessed included IL-2 in high-dose and low-dose regimens, IFN-α, combination immunotherapy (IL-2 and IFN-α), bevacizumab, thalidomide, IFN-γ, and temsirolimus (CCI-779). The main outcomes of interest were efficacy measures (specifically, treatment response and disease survival) and safety measures (specifically, Grade 3 and 4 toxicities). Data abstraction was performed by 2 physician reviewers who were trained in health services research (R.J.H. and M.B.). Evidence tables summarizing the benefits and risks of each therapy were included in a literature review and were provided to each member of the consensus panel.
Based on the literature review and consultation with experts in the field, a set of clinical scenarios was developed to represent the situations most likely to arise in clinical practice. For systemic therapy, clinical factors included the following: the risk classification systems developed by the Memorial Sloan-Kettering Cancer Center (MSKCC) for patients with MRCC20–22; the type of prior immunotherapy; and the expression of carbonic anhydrase IX (CAIX) for individuals who received treatment with IL-2. For cytoreductive nephrectomy, clinical factors included surgical risk, symptoms related to the primary tumor, and metastatic burden. For this scenario, all patients were considered to have synchronous metastatic disease, which rendered the disease-free interval irrelevant as a prognostic factor. For all scenarios, consideration was limited to patients with clear cell histology and no active central nervous system metastases. Scenarios, clinical factors, and treatment options are described in Table 1. Key terms and definitions are provided in Table 2.
Table 1. Clinical Indications and Treatment Options Rated for Patients with Metastatic Renal Cell Cancer*
MSKCC indicates Memorial Sloan-Kettering Cancer Center; CAIX, carbonic anhydrase IX; IL, interleukin; IFN, interferon; CNS, central nervous system.
Note that all scenarios were limited to patients who had clear cell histology and no active CNS metastases.
Treatment options: bevacizumab, IFN-α alone, high-dose IL-2 alone, low-dose IL-2 alone, sorafenib tosylate (BAY 43-9006), sunitinib malate (SU11248), temsirolimus (CCI-779), thalidomide, and investigational trial. Note that temsirolimus was not considered a treatment option for patients who received prior immunotherapy.
Scenario 3: Patients with primary tumor in place (no prior immunotherapy)‡
Good risk (good performance status, no major comorbid conditions)
Poor risk (poor performance status, major comorbid conditions, and/or life expectancy <6 months)
Symptoms related to primary tumor
Lung metastases only
Limited burden (limited lung metastases or limited bone involvement)
Extensive burden (lung and bone metastases, or any liver or CNS involvement)
Table 2. Definitions of Key Terms
MSKCC indicates Memorial Sloan-Kettering Cancer Center; MRCC, metastatic renal cell carcinoma; LDH, lactate dehydrogenase; CAIX, carbonic anhydrase IX; IL-2, interleukin 2; CNS, central nervous system.
I. “Appropriateness” of the therapeutic intervention
An intervention is considered “appropriate” if: the expected clinical benefits to the patient (e.g., complete response/partial response, increased survival) outweigh the clinical risks (e.g., treatment toxicity) by a sufficiently wide margin to justify the intervention
II. MSKCC risk classification for survival in untreated patients with MRCC (Motzer et al., 199920 and Motzer, 200321)
Five pretreatment patient related- prognostic factors associated with a shorter survival:
1. Karnofsky performance status <80%
2. High LDH level (>1.5 × upper limit of normal)
3. Low hemoglobin level (below lower limit of normal)
4. High corrected calcium level (>10 mg/dL)
5. Absence of nephrectomy
Based on these defined risk factors, patients are classified into 3 risk groups:
1. Favorable risk: Patients with 0 risk factors
2. Intermediate risk: Patients with 1 or 2 risk factors
3. Poor risk: Patients with >3 risk factors
III. MSKCC risk classification for survival in previously treated patients with MRCC (Motzer et al., 200422)
1. Karnofsky performance status <80%
2. Low hemoglobin level (<13 g/dL for males; <11.5 g/dL for females)
3. High corrected calcium level (>10 mg/dL)
1. Favorable risk: Patients with 0 risk factors
2. Intermediate risk: Patients with 1 risk factor
3. Poor risk: Patients with 2 or 3 risk factors
IV. CAIX as molecular marker (in patients receiving IL-2)
1. High CAIX expression: Staining >85% in tumor tissue
2. Low CAIX expression: Staining ≤85% in tumor tissue
V. Surgical risk
1. Good surgical risk: Patient with good performance status and no major comorbid conditions
2. Poor surgical risk: Patient with poor performance status, major comorbid conditions, and/or life expectancy <6 months
VI. Symptoms related to primary tumor
1. Symptoms such as flank pain, hematuria, or gastrointestinal symptoms
VII. Metastatic burden
1. Lung metastases only
2. Limited metastatic burden: Limited lung metastases or limited bone involvement
3. Extensive metastatic burden: Lung and bone metastases or any liver or CNS involvement
A pool of potential panelists was identified by using rosters of specialty societies (ASCO, Society of Urologic Oncologists), the board of Medical Advisors of the Kidney Cancer Association, authors of key clinical trials, and recognized community-based practitioners. The panel was composed of 9 specialists (2 urologists and 7 oncologists) from diverse practice settings (5 university-based settings and 3 community-based settings) and geographic sites (9 states) within the U.S.
Panelists received the literature review and an initial set of scenarios by mail. The first round of ratings was completed prior to the panel meeting. At a subsequent 1-day meeting, panelists reviewed the first-round ratings, revised the structure of the clinical scenarios, modified the definitions of key terms, discussed reasons for the degree of agreement or disagreement in the ratings from the first round, and confidentially rerated all indications. A third round of ratings for selected agents was added after new data became available in 2006. Each indication was rated on a 9-point appropriateness scale (a rating of 1 indicated that a treatment was extremely inappropriate, 5 indicated that a treatment was uncertain, and 9 indicated that a treatment was extremely appropriate). A therapy was deemed appropriate when the expected benefits exceeded the expected risks by a sufficiently wide margin to justify prescribing the therapy given the financial constraints of the current health care environment.
The final rating was determined by calculating the median score of the 9 panelists. Indications were considered “inappropriate” for median ratings between 1 and 3 (without disagreement), “uncertain” for median ratings between 4 and 6 or if panelists disagreed, and “appropriate” for median ratings between 7 and 9 (without disagreement). We defined disagreement as occurring when at least 2 panelists rated the indication appropriate and at least 2 panelists rated the indication inappropriate.
We accepted 22 controlled clinical trials that included a total of 4289 patients with MRCC. Twenty studies were randomized trials, 1 was a prospective nonrandomized study, and 1 was a retrospective analysis. Treatments assessed in the accepted trials included IL-2 (6 studies),23–28 IFN-α (5 studies),29–33 combination immunotherapy (8 studies),34–41 bevacizumab (1 study),42 temsirolimus (1 study),43 and IFN-γ (1 study).44 We did not identify any clinical trials of thalidomide that met our inclusion criteria. However, we located a literature review that evaluated thalidomide as a treatment option for patients with MRCC.45 Among the studies reviewed, the most consistent inclusion criteria were disease stage (MRCC or advanced RCC, Stage IV), age (≥18 years), and good-to-acceptable performance status (Eastern Cooperative Oncology Group scale, 0–2; Karnofsky score, >70%). Most studies excluded patients with brain metastases and set limits on organ dysfunction. Most patients had undergone previous cytoreductive nephrectomy (median, 85%; range 30–100%). However, 5 studies included explicit assessments of nephrectomy as a therapeutic option.23–31 New results for sunitinib and sorafenib were presented at the 2004 and 2005 ASCO meetings in addition to recent information regarding cytokine treatment for patients with an intermediate prognosis.46–51 At the suggestion of the panelists, this material was summarized at the panel meeting. Finally, a third round of ratings was carried out to incorporate new results for sunitinib and temsirolimus that were presented at the 2006 ASCO meeting.52, 53
Efficacy findings are summarized in Table 3, and common toxicities associated with systemic therapies for MRCC are summarized in Table 4. Based on the literature review and advice of the panelists, IFN-γ was removed from further consideration because of its lack of efficacy.
Table 3. Summary of the Efficacy of Systemic Therapy for Metastatic Renal Cell Carcinoma*
No. of study arms
Total no. of patients
Median objective response rate (range), %
Median CR rate (range), %
Median PFS rate (range), months
Median overall survival (range), months
CR indicates complete response; PFS, progression-free survival; NR, not reported; IFN, interferon; IL, interleukin.
This table does not include 13 treatment arms from 5 studies that assessed nephrectomy plus immunotherapy. No randomized, controlled trials were available for thalidomide.
Results were reported from a single study. Results for sorafenib were reported only in abstract form.
Fatigue, lethargy and constipation (>30%); neuropathy (with prolonged therapy; 25%); deep vein thrombosis (20%); and pulmonary embolism (15%)
The rating structure resulted in 3 high-level clinical scenarios, resulting in 108 permutations that addressed systemic therapy and 24 permutations that addressed cytoreductive nephrectomy. For all scenarios, consideration was limited to patients with clear cell histology and no active central nervous system metastases. Although to our knowledge the expression of CAIX has not been validated in randomized prospective studies, it was considered important enough by the panel to warrant its inclusion in the scenarios. The final ratings were collapsed for simpler presentation when the categorization of appropriateness did not differ by clinical factors. Statistical testing is not provided because of the descriptive nature of the analysis and the small sample size.
Of the 132 possible permutations of therapy for patients with MRCC, 27.3% of indications were rated appropriate; 25.8% were rated uncertain, and 46.9% were rated inappropriate. Using our definition of disagreement, the panel disagreed on only 5.3% of the indications in the final ratings, a reduction from 22.1% in the first round of ratings. Specific results for each scenario are described below. In Figures 1 through 3, green represents “appropriate” ratings, yellow represents an “uncertain” rating or disagreement among the panelists, and red represents an “inappropriate” rating.
The first scenario involved the use of systemic therapy in patients who did not receive prior immunotherapy (Fig. 1). The panel considered high-dose IL-2 appropriate for patients with favorable risk status (no MSKCC risk factors) when the expression of CAIX is high or unknown and for patients with intermediate risk status (1 to 2 MSKCC risk factors) when the expression of CAIX is high. All other use of immunotherapy was considered uncertain or inappropriate. Among targeted therapies, both sunitinib and sorafenib were rated appropriate for favorable-risk and intermediate-risk patients, and temsirolimus was rated appropriate for poor-risk patients. All other use was considered uncertain.
The second scenario involved the use of systemic therapy in patients who received prior immunotherapy (Fig. 2). Overall, immunotherapies were considered inappropriate for patients who already had failed other immunotherapy. However, there was uncertainty regarding the use of high-dose IL-2 for favorable-risk patients (0 MSKCC risk factors) and intermediate-risk patients (1 MSKCC risk factor) who had high CAIX expression and who previously had failed IFN-α treatment. Sunitinib and sorafenib were considered appropriate for any patients with MRCC who had failed previous immunotherapy.
Several systemic therapy options were consistent across Scenarios 1 and 2. Enrollment of MRCC patients into an investigational trial always was considered appropriate, treatment with bevacizumab always was rated uncertain, and treatment with thalidomide was rated inappropriate regardless of risk status or prior therapy.
The third scenario addressed the use of cytoreductive nephrectomy in patients with primary tumor in situ who had received no prior immunotherapy (Fig. 3). Nephrectomy was rated appropriate for patients with good surgical risk, symptoms related to the primary tumor, and limited metastatic burden. For patients with planned immunotherapy and good surgical risk status, the appropriate ratings also included 1) patients with limited metastatic burden and no primary tumor-related symptoms and 2) patients with extensive burden who had symptoms present. Nephrectomy was considered inappropriate for patients with poor surgical risk, no primary tumor-related symptoms, and extensive metastatic burden. For patients with planned targeted therapy, the inappropriate rating also included individuals with poor surgical risk, no symptoms, and limited metastatic burden. All other permutations were rated uncertain.
The current results provide important insights for clinicians and investigators dealing with MRCC with clear cell histology. For clinicians, several trends emerge. First, our panel recognized that the therapeutic options available for MRCC still are imperfect. Therefore, enrollment on an investigational trial continues to be an appropriate option for all eligible patients. Second, empirically tested risk schema, such as those developed by the MSKCC, may be used to support both evidence-based and clinical decision-making. Third, the panel agreed that thalidomide, despite widespread use in clinical practice, has no role in the management of MRCC.
Among the newer targeted therapies, the panel supported an “appropriate” rating for sunitinib and sorafenib in lower risk patients, whether or not they had received previous treatment. The most recent results for sunitinib were considered particularly encouraging, with a reported objective response rate of >30%.52 The uncertainty for these 2 agents as first-line therapy for poor-risk patients was based on the lack of data in this population. However, recent (2006) results for temsirolimus led to an “appropriate” rating for this difficult group of poor-risk patients.53 Temsirolimus was not considered a treatment option for second-line therapy. The situation with bevacizumab was more complex. Although some panelists considered that bevacizumab was a reasonable option in the second-line setting for immunotherapy-resistant patients, others considered this agent less appropriate, leading to a combined panel rating of “uncertain.” The panel did not believe that sufficient data were available at the time to make a recommendation with regard to the use of bevacizumab in previously untreated patients. Whether treatment with targeted therapies may limit downstream treatment options with immunotherapy was not addressed.
These results help clarify the role of immunotherapy in relation to the newer available therapies. The most notable change involves the rating of IFN-α as inappropriate for the treatment of MRCC, a marked change from prior practice. These ratings resulted from head-to-head studies presented in 2006 that reported superior results with sunitinib in patients with better risk status and with temsirolimus in poor-risk patients.52, 53 Low-dose IL-2 also was rated inappropriate for most patients, although there was some uncertainty regarding better risk patients with unknown or high expression of CAIX. Some panelists stated that high-dose IL-2 still offers the best chance for a durable response, albeit to a minority of patients. At experienced referral centers, treatment with high-dose IL-2 can be administered safely, and adverse events can be predicted and managed.54 This is based on careful patient selection as well as appropriate monitoring and management. Despite these advances, immunotherapy generally is inappropriate for patients who have failed previous immunotherapy and those with poor risk status. Combination immunotherapy was not considered explicitly by the panel.
The panel held that the available data on the use of CAIX expression as a biomarker for response to IL-2 were very promising, but more prospective validation is needed. Particularly, more information is needed to integrate CAIX status with other prognostic factors, such as histologic subtype and MSKCC score. Information on the appropriate treatment options for patients with low expression of CAIX and the role of CAIX in predicting benefit for other targeted therapies also is needed. Panelists noted that this test is not yet available routinely for clinical use; therefore, at the moment, these recommendations are speculative.
The panel held that cytoreductive nephrectomy was most appropriate for patients with lower surgical risk, lower metastatic burden, and symptoms related to the primary tumor, especially in the context of planned immunotherapy. For good surgical risk patients with planned immunotherapy, nephrectomy was rated appropriate for patients with limited metastatic burden regardless of whether there were symptoms related to the primary tumor and for symptomatic patients regardless of their metastatic burden. Only the most favorable combination of good surgical risk, limited metastatic burden, and symptoms related to the primary generated an “appropriate” rating in patients with planned targeted therapy. These results suggest a need for additional studies to quantify the risks and benefits of cytoreductive nephrectomy better for patients with poor surgical risk and symptoms related to their primary tumor. In addition, clarification is needed regarding the role of nephrectomy in patients for whom targeted therapy is planned.
The RAND/UCLA Appropriateness Method is a well studied approach that blends evidence from the literature with an expert consensus process. However, no panel can deduce the results of a definitive randomized controlled trial before it has been conducted.
The scenarios rated by our panelists represent a simplification of the clinical decision-making process. Several variables—for example, tumor grade and the presence of specific comorbid conditions—were not addressed explicitly by this panel. These variables and others can and should influence the choice of therapy in particular circumstances. Nothing in these guidelines should be construed as contradicting the accepted use of any therapy, as indicated on the product labeling.
The current results should be viewed in the context of a rapidly changing field. New agents currently are being tested, and new results may become available at any time, potentially altering the appropriateness of any given therapeutic option. Our results represent a single point in a changing process and should be treated as such.
In conclusion, the emergence of newer therapies and biologic markers to enhance the use of existing therapies has created areas of uncertainty regarding the appropriate management of patients with MRCC. For clinicians, these results help begin the process of defining an appropriate role for cytokines, newer targeted therapies, and surgery in the treatment of this challenging disease. For investigators, these results highlight the gaps between currently available data and the evidence needed for clinical decision making. Future trials should be designed to address these needs.