Hormonal therapy has been investigated as therapy for RCC based upon the finding of estradiol and progesterone receptor expression on RCC cells [24]. Conflicting results regarding the utility of progestational therapy were reported in early, small, retrospective series [24–26]. A randomized trial of 1 year of medroxyprogesterone as adjuvant therapy was subsequently conducted in Italy, enrolling 136 patients with Robson stage I–III disease [27]. No difference in relapse rate was detected between the treated and observation groups. The 5-year disease-free survival rate was 67.1 % in the medroxyprogesterone group and 67.3 % in the observation group. Side effects included loss of libido in men and weight gain. No significant relationship between sex steroid receptor expression and relapse was detected. Further study of hormonal therapy in the adjuvant setting has not been pursued.

RCC has traditionally been characterized as insensitive to traditional cytotoxic chemotherapy agents. The fluoropyrimidines have been one minor exception to this generalization, with low levels of activity reported in the literature [28]. UFT is a combination of tegafur (a 5-fluorouracil prodrug) and uracil developed in Japan that has predominantly been used in colorectal carcinoma and is approved in many countries outside of the USA. A Japanese single-arm study of adjuvant UFT in combination with vinblastine and doxorubicin reported 96 % 5-year survival among the 31 enrolled patients [29]. A subsequent Japanese trial randomized 71 patients with Robson I–II disease to observation or to 2 years of daily UFT after nephrectomy [30]. No difference in 5-year recurrence rate or overall survival was detected. Side effects were relatively mild and predominantly gastrointestinal in nature. The study included a relatively low risk, early-stage population, as reflected by an 80.5 % 5-year nonrecurrence rate in the UFT arm.

For many years, immunomodulatory agents including interferon-α (IFN-α) and interleukin-2 (IL-2) were the basis of treatment for metastatic kidney cancer. Modest survival benefit with IFN was suggested in two randomized trials [31, 32], while the efficacy of IL-2 was evidenced by low but reproducible response rates [33]. High-dose IL-2 remains an option for select patients with metastatic disease based on its association with complete and durable responses in a minority (5–7 %) of patients. Given the vantage of cytokines as the only active therapies for RCC in the 1980s–1990s, a number of randomized trials investigated the adjuvant utility of IFN and IL-2 during this period (Table 14.3).

Several trials have evaluated the efficacy of single-agent IFN given postoperatively. An Italian study randomized 264 patients with Robson stage II–III RCC to IFN-α-2b three times per week for 6 months or to observation [34]. There were no differences in 5-year overall or event-free survival, the primary end points of the study. Subset analysis suggested an improvement in relapse rate among the small number of patients with extensive nodal disease (pN2–pN3) but also suggested a harmful effect of IFN among patients with N0 disease. An Eastern Cooperative Oncology Group/Intergroup trial randomized 283 patients with locally advanced or node-positive disease to 12 cycles of lymphoblastoid IFN-α-NL administered daily for 5 days every 3 weeks or to observation [35]. No statistically significant difference in overall survival was observed, but there was a trend toward better survival in the observation arm (median 7.4 vs. 5.1 years, p = 0.09).

Combination cytokine regimens incorporating IFN and subcutaneous IL-2 were reported to have greater response rates than single-agent therapy in the metastatic setting [36]. While later randomized studies would fail to show a benefit of combination therapy over single-agent cytokines [37–39], early investigations of combination therapy were undertaken in the adjuvant setting. The Italian Oncology Group for Clinical Research reported preliminary results of a randomized trial of subcutaneous IL-2 and IFN-alpha vs. observation in patients with tumors >2.5 cm and more advanced local disease [40]. This low-dose immunotherapy regimen was given intermittently with twelve 4-week cycles administered over 5 years. This regimen was hoped to be less toxic and with the potential for a prolonged immune stimulatory effect. Approximately one-third of patients were low risk by the UISS system. At a median follow-up of 52 months, there was no difference in RFS (HR 0.81; 0.51–1.27 p = 0.36) or overall survival (HR 1.07; 0.64–1.79 p = 0.79).

As discussed above, 5-fluorouracil (5-FU) is one of the few chemotherapeutic agents with a reproducible albeit low response rate in RCC [28]. Some of the highest response rates of the cytokine era were reported with regimens combining IFN and IL-2 with 5-FU [41, 42]. The German Renal Carcinoma Chemoimmunotherapy Group conducted a randomized adjuvant trial using this approach in patients with tumor extending into the renal vein or invasive beyond Gerota’s fascia, node-positive patients, and patients after complete surgical resection of solitary metastatic disease [43]. Two hundred three patients were randomized to 8 weeks of treatment with subcutaneous IL-2, IFN-α-2a, and 5-FU or to observation. The primary end point was relapse-free survival. No significant difference was seen between the treatment and observation arms. Overall survival was significantly decreased in the treatment arm compared with the observation arm (5-year survival 58 vs. 76 %; p = 0.0278). While no mention of side effects was reported in this publication, the possibility that treatment-related toxicity contributed to the worse survival must be considered. A second randomized trial using a very similar regimen was conducted by the EORTC and NCRI (UK) [44]. Three hundred nine patients with locally advanced or node-positive disease or exhibiting positive microscopic margins or microscopic vascular invasion were randomized to either observation or subcutaneous IL-2, IFN-α-2a, and 5-FU. There was no significant difference in the primary outcome measure, disease-free survival at 3 years (50 vs. 61 %), nor in overall survival at 5 years (63 vs. 70 %).

High-dose, human recombinant IL-2 was the first agent approved for metastatic renal cancer in the USA based on non-randomized, pooled data from 255 patients yielding a response rate of 15 % (95 % CI, 11–20 %) including 7 % complete responders [33]. While the complete response rate in the metastatic setting would suggest potential utility as adjuvant therapy, the significant side effect profile of high-dose IL-2 therapy precludes the ability to conduct a blinded study, poses difficulty in subject recruitment, and greatly limits its widespread use as an adjuvant. An attempt was made by the Cytokine Working Group in studying one course of high-dose IL-2 in the adjuvant setting. This was a randomized trial with observation as the control arm [45]. The trial included patients with locally advanced tumors and was expanded to include patients with M1 disease resected to no evidence of disease. The study was closed for futility after interim analysis suggested minimal likelihood that the study would meet its primary end point of a 30 % absolute improvement in disease-free survival. While side effects were as expected, 88 % of patients in the IL-2 arm experience grade 3–4 toxicity including hypotension requiring vasopressor support in 52 %.

Given the remarkable ability of high-dose IL-2 to occasionally induce complete and durable responses, its use as an adjuvant therapy remains a provoking concept. However, further investigation of the drug in the adjuvant setting would necessitate the existence of a reliable method of predicting responders in order to limit the exposure of those unlikely to benefit. Unfortunately, the ability to identify such patients to a high degree of certainty in the metastatic arena remains an enigma [46].

Adoptive immunotherapy involves the harvest of a patient’s T lymphocytes and ex vivo activation, followed by reinfusion in attempt to engender an immune response against the tumor. The use of this technique as adjuvant therapy was studied in a small, randomized study in patients with node-positive disease after nephrectomy. Forty-five patients were randomized to adjuvant therapy with ex vivo activated T cells plus cimetidine (to reduce in vivo suppressor T-cell function) or to cimetidine alone. The median time to recurrence was 16.4 months for the adoptive immunotherapy-treated patients and 6.5 months for controls (p = 0.0360) [47]. A subsequent 100-patient phase II trial of adjuvant activated T-cell therapy in high-risk patients (including metastatic patients resected disease-free) showed favorable survival compared to institutional historical controls [48]. Despite these promising preliminary results, adoptive immunotherapy has not been pursued further in definitive studies.

Autologous vaccination strategies are based on the premise that RCC cells express antigens capable of eliciting a T-cell response. The sensitivity of metastatic renal cancer to immunostimulatory interventions such as cytokines is evidence of the immunogenic nature of RCC. Vaccine approaches in RCC have included whole-cell vaccines, lysates of cancer cells, and heat-shock proteins [49]. The post-nephrectomy setting – when tumor burden is at its lowest and the immune system has potentially been relieved of suppression – may be the most opportune time to instigate an immune response through vaccination. The appeal of tumor-derived vaccine strategies has led to a number of such trials in the adjuvant setting (Table 14.4).

An early report of adjuvant tumor vaccination strategy investigated autologous irradiated tumor cells admixed with BCG administered by intradermal and endolymphatic injections [50]. This trial included 43 post-nephrectomy patients of all stages who were randomized to either hormonotherapy with a progestogen (Primostat) or to hormonotherapy in combination with the vaccine. While there was a trend toward improved disease-free interval in the vaccinated patients, no statistically significant difference was seen in this small study.

Another such “active specific immunotherapy” approach was reported by Galligioni et al. [51]. Patients with pT1–3b pN0 or pN+ disease at nephrectomy were randomized to immunization (n = 60) or to observation (n = 60). The vaccine was prepared by irradiation of autologous tumor cells and was mixed with Bacillus Calmette–Guèrin (BCG) for the first two of three vaccinations. After a median follow-up of 61 months, there was no difference in 5-year disease-free survival or overall survival between the two groups. Delayed-type cutaneous hypersensitivity response to autologous tumor cells was 1 month after immunization was detected in 70 % of patients, but was not observed in control patients.

Favorable results have been reported in trials using an autologous tumor-derived lysate vaccine (Reniale) developed in Germany [52]. This process involves obtaining tumor cells at the time of nephrectomy followed by incubation with IFN-γ and devitalization by rapid repeated freezing. A large series of T2–3N0 patients received adjuvant therapy with the vaccine in initial studies, with higher 5-year progression-free survival and overall survival rates as compared with historical controls [53, 54]. A subsequent randomized trial was performed to confirm the activity of the vaccine in post-nephrectomy patients as compared to observation [52]. Those randomized to the vaccine received an intradermal injection every 4 weeks for a total of six injections. The primary end point of the trial was progression-free survival. Among 379 patients evaluable for the intention-to-treat analysis, the risk of tumor progression was significantly less in the vaccine group (HR = 1.59, p = 0.0204). The majority of patients were N0 (96 %), and only 30 % had T3 disease. Subgroup analysis revealed the greatest potential benefit among patients with T3 tumors. Several methodological flaws limit interpretation of this study: randomization was performed prior to nephrectomy, and as a result, 32 % of enrolled subjects were lost prior to starting treatment leading to an ultimate imbalance in the study arms. A follow-up intent-to-treat analysis of 477 patients did not indicate an overall survival advantage (p = 0.1185), although a secondary per-protocol analysis of 352 patients did suggest an overall survival benefit (p = 0.0356) [55].

Subsequent to the randomized trial, data from a compassionate use program with Reniale were analyzed to estimate potential survival benefit [56]. Six hundred and ninety-two patients with T2–3 N0–2 M0 (1992 classification) disease who had been treated with the vaccine between 1993 and 1996 were matched with 661 controls who had undergone nephrectomy between 1992 and 2006 at a single center in Germany. The matching criteria included a number of prognostic variables including pT stage, but tumor size was not used due to missing data. Seventy-nine percent of patients had pT2, and 21 % had pT3 disease. Ten-year survival was 69 % in the vaccine group compared with 62 % in the control group (p = 0.066). On subgroup analysis, improved survival was seen among patients with pT3 tumors (p = 0.022) but not among those with pT2 disease (p = 0.365). On multivariate analysis of the whole study group, treatment with the vaccine was associated with improved survival (HR = 1.28, p = 0.030). Interpretation of these data is limited by the retrospective nature of the analysis, selection of controls from a single institution, and the absence of one important prognostic factor (tumor size) from the matching criteria.

Vitespen (Oncophage) is a heat-shock protein (glycoprotein 96)–peptide complex derived from autologous tumor. Heat-shock proteins are involved in protein folding and are upregulated in response to stress. They bind cellular peptides and are highly immunogenic. In a study of 818 patients with cT1b–T4 N0 M0 or N1–2 M0 clear cell RCC, patients were randomized to vitespen or observation [57]. Vitespen was administered by weekly intradermal injections for 4 weeks, followed by every 2-week injections until depletion of vaccine supply or disease progression. Among 728 patients included in the intent-to-treat analysis, no difference was seen in recurrence rate between the vitespen (38 %) and observation (40 %) groups after a median follow-up of 1.9 years (HR = 0.923, 95 % CI 0.729–1.169; p = 0.506). Subgroup analysis suggested a trend toward improved relapse-free survival in patients with stage I–II disease, with recurrence noted in 15 % of vitespen-treated patients and 27 % of observation patients (HR = 0.576, 95 % CI 0.324–1.023; p = 0.056). No overall survival difference was seen after an additional 17 months of follow-up with approximately 88 % patients alive in both groups. The trial had a number of limitations, including the inability to prepare a vaccine for 8 % of patients and a large number of subjects who were not eligible upon blinded review of the intent-to-treat population. Exclusion of these subjects in a full analysis data set resulted in a greater difference in outcomes between the vitespen and control groups but still did not meet statistical significance. Longer-term follow-up of 294 of the patients enrolled in a follow-up registry continued to demonstrate a trend toward improved outcome in lower-stage disease but without statistical significance [58].