A common genetic feature signature of sporadic clear cell RCC is the loss of chromosome 3p, suggesting the presence of one or more RCC tumor suppressor genes at this site. The von Hippel-Lindau tumor suppressor gene (VHL), which resides on chromosome 3p25, is mutated or silenced in greater than 50% of sporadic clear cell RCCs.16,17 Germline VHL mutations give rise to von Hippel-Lindau syndrome, which is characterized by an increased risk of blood vessel tumors (hemangioblastomas), endocrine tumors and RCC. The VHL gene product, pVHL, is the substrate recognition module of an E3 ubiquitin ligase that targets the hypoxia-inducible factor (HIF) α transcription factors (HIF1α, HIF2α, and HIF3α) for destruction in the presence of oxygen. Hypoxic cells, or cells lacking pVHL (“pseudohypoxic”), accumulate high levels of HIF, which activates the transcription of a variety of genes, including vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF-3) B, and transforming growth factor alpha (TGF-2). Restoration of pVHL function in VHL−/− mutant renal carcinoma cells suppresses their ability to form tumors experimentally by reducing HIFα levels.^18,19 Inhibition of HIFα is necessary and sufficient for tumor suppression by pVHL in RCC nude mouse xenograft assays. This provides a rationale for treating VHL−/− RCC with inhibitors of HIFα or its downstream targets. While the HIF1α isoform was initially believed to be more important, increasing literature supports HIF2α as the more important HIFα member in mediating tumorigenesis.^18,19 Although most investigation has focused on the role of HIFα isoforms, the pVHL protein also has several other targets in addition to HIFα, postulated by some to contribute to tumorigenesis. Elucidating these targets will lead to further knowledge of how pVHL suppresses tumor growth. Analysis of mutations in exon 3 of the VHL gene may be useful in refining the diagnostic criteria for conventional RCC versus chromophobe RCC with clear cell features.²⁰

A recent work from Dr. Rathmell’s group have generated gene expression microarray data using software that implements iterative unsupervised consensus clustering algorithms to identify the optimal molecular subclasses. A Consensus Cluster analysis identified two distinct subtypes of clear cell RCC within the training set, designated clear cell type A (ccA) and B (ccB). In each subtype the analysis of data defined a small, highly predictive gene set. A validation data set of 177 tumors was analyzed. Tumors designated ccA had improved disease-specific survival compared with ccB (median survival of 8.6 vs. 2.0 years). These preliminary data suggest that a cluster subtype classification is possible in RCC as performed in other diseases such as breast cancer. Prospective clinical studies will be necessary to validate these important findings.²¹

The genetic studies in familial RCC have led to the identification of specific molecular signatures in non–clear cell histotypes as well, such as Met hyperactivation in papillary type I, fumarate hydratase mutation in papillary type II, and c-Kit overexpression in chromophobe RCC (see “Papillary Types I and II Renal Cell Carcinoma”). Preliminary results from ongoing genetic studies have shown the possibility of clustering different histotypes based, for example, on kinase expression (Fig. 82-1A).²²

DNA, histones, and nonhistone proteins are condensed into a highly complex nucleoprotein structure known as chromatin that acts as a prototype for all of the genetic information. Chromatin conformation can be in a heterochromatin form, a highly compact structure, associated primarily with transcriptionally inactive genes, or a euchromatin form, consisting of a more open or relaxed structure, associated with transcriptionally active genes.²³ Epigenetic mechanisms mediated by histone deacetylases (HDACs), histone acetyltransferases, and histone methyltransferases play critical roles in various biological and cellular processes, including cell proliferation, angiogenesis, hypoxia-related effects, and cell-cycle regulation. Histones on the N-terminal are altered by acetylation, methylation, phosphorylation, ubiquitination, sumoylation, deamination, or adenosine diphosphate ribosylation.^24,25 The different histone residues and their modifications result in either transcriptionally active or repressive marks. For example, histone H3 lysine 4 (designated as H3K4), H3K36, and H3K79 are associated with active marks, whereas H3K9, H3K27, and H4K20 are associated with repressive marks.²⁶

Several studies have recently reported SETD2, a histone methyltransferase, as a tumor suppressor gene in RCC.27,28 An initial study consisting of a large-scale next-generation sequencing of primary RCC genomes identified somatic truncating mutations in SETD2 (approximately 3%).²⁸ These mutations are associated with a decrease in H3K36 trimethylation levels in several clear cell renal carcinoma cell lines, VHL mutations and/or hypoxic phenotype (determined by a panel of hypoxia-related genes) in clinical samples, as well as a chromatin remodeling complex protein PBRM1.²⁹ It is also important to note that SETD2 and PBRM1 are located on the short arm of the human chromosome 3 (3p21.31), in close proximity to the VHL gene and the 3p region is frequently altered in clear cell RCCs³⁰ (see Fig. 82-1B). Although mutations in enzymes governing chromatin remodeling are often found in advanced stage disease, these mutations have not been correlated to survival.³¹ The H3K27 methyltransferase EZH2 (which is associated with aggressiveness in breast cancer and overexpressed in metastatic prostate cancer)³² is overexpressed in renal tumor patient samples compared with their normal adjacent.³³ Conversely, histone demethylases governing H3K27 methylation status UTX and JMJD3 are overexpressed in clear cell RCCs compared with the adjacent nontumor tissue, which corresponds to the decreased H3K27 methylation in clear cell RCC.³⁴ These preliminary evidences suggest that alterations in histone-modifying genes may be associated with RCC and could be exploited for therapeutic interventions.

More recently, the BAP1 protein, a nuclear deubiquitinase, has been reported to be inactivated in 15% of clear cell RCCs. BAP1 mediates suppression of cell proliferation, ubiquitinates the histone 2A lysine 119, and its loss sensitizes RCC cells in vitro to genotoxic stress. Initial evidences suggest that BAP1 loss is associated with high tumor grade and poor prognosis. Interestingly, mutations in BAP1 and PBRM1 appear to be mutually exclusive. These preliminary results provide a rationale for future integrated pathological and molecular genetic classification of RCC that will likely lead to subtype specific treatments.³⁵

Intratumor heterogeneity has always been considered a potential major clinical hurdle to develop personalized-medicine strategies that depend on results from single-tumor biopsy samples. A recent report has now shown very elegantly that intratumor heterogeneity is a reality.³⁶ Different techniques, including exome sequencing, chromosome aberration analysis, and ploidy profiling on multiple spatially separated samples obtained from primary renal carcinomas and associated metastatic sites, determined that 63% to 69% of all somatic mutations were not detectable across every tumor region. Mutational intratumor heterogeneity was seen for multiple tumor suppressor genes and chromatin remodeling genes, including SETD2, PTEN, and KDM5C. Gene-expression signatures associated with good and poor prognosis were detected in different regions of the same tumor. These new intriguing findings, combined with the complex genomic landscape, make us rethink the common assumption that a single genomic test might guide therapy. The timing and location of the tumor sample acquisition remain a challenge, but there is still an opportunity to make personalized medicine a goal to achieve in the future for the treatment of cancer, and RCC in particular.

Familial Renal Cell Carcinoma

In a small percentage (5%) of cases, RCC is a feature of one of several hereditary syndromes.15,45 Such syndromes are associated with distinct histologic subtypes of RCC, and in each case patients have increased risk of multifocal tumor development.^15,45 Management is dependent on preservation of renal function. Close surveillance and minimization of surgical procedures constitute the mainstay of treatment.

von Hippel-Lindau syndrome is a disorder of autosomal dominant inheritance that occurs in 1 in 40,000 births. The mean age at onset is in the fourth decade of life. The syndrome is inherited as a result of a germline mutation in a single allele of the VHL gene tumor suppressor gene located on chromosomal band 3p25–26.15,45,46 Sporadic loss of the remaining wild-type VHL allele provides the “second hit” necessary for tumorigenesis, most commonly via chromosome 3p deletion. Multifocal tumorigenesis is observed in multiple organ systems, with each tumor harboring an independent second VHL mutation. Renal manifestations include cysts and clear cell RCC tumors. Both tend to be multifocal and bilateral, and are found in the majority of patients with von Hippel-Lindau disease.^15,45,46 Hundreds of independent clear cell cancers may be present in a single kidney, including dozens of macroscopic tumors. VHL syndrome patients are at high risk for chronic renal insufficiency because of the lifelong risk of multifocal RCC tumor development and need for repeat renal surgeries. As a result, VHL patients should undergo active surveillance until the largest tumor reaches 3 cm, at which time attempts may be made to resect all tumors in that kidney. Resection by enucleation without clamping of the main renal artery is recommended to maximize nephron sparing. Surgical candidates, particularly those with numerous tumors, are counseled as to the high possibility of local recurrence from de novo tumor formation and future ipsilateral surgery. The discovery of the VHL gene in hereditary clear cell RCC enabled the subsequent identification of a VHL mutation in sporadic clear cell RCC tumors (see above).⁴⁷

In addition to renal cysts and cancer, common manifestations of the VHL syndrome include benign vascular tumors of the spinal cord, cerebellum and retina, presenting as neurologic and visual symptoms. The endocrine system may also be affected by adrenal pheochromocytomas and pancreatic neuroendocrine tumors, in addition to pancreatic cysts. Benign papillary cystadenomas of the epididymis or broad ligament also develop in a minority of patients, which, when bilateral are pathognomonic for von Hippel-Lindau disease. Firm epididymal nodules on physical should prompt scrotal ultrasound and with a family history of VHL provides easy method to confirm the diagnostic suspicion. Endolymphatic sac tumors of the inner ear also may be seen.

Birt-Hogg-Dubé syndrome is a disorder of autosomal dominant inheritance. Signs and symptoms usually manifest in the fifth decade of life and include renal tumors and cysts, benign skin tumors (fibrofolliculomas) and pulmonary cysts, which can lead to spontaneous pneumothorax. The renal neoplasms may be multifocal and bilateral tumors and most often have pure chromophobe histology or a “hybrid” mixture of chromophobe and oncocytoma; infrequently, pure oncocytoma tumors may be present. Patients can present with several different tumor types within the same kidney and the presence of benign (oncocytoma) and malignant tumors within the same kidney should immediately prompt the suspicion of Birt-Hogg-Dubé (BHD) syndrome. The BHD gene mutated in this syndrome encodes the protein Folliculin and is located on chromosome 17p11.2.15,45,46,48,49 The BHD gene appears to have the characteristics of a loss-of-function tumor suppressor gene.⁵⁰ Folliculin has unknown function but is found in complexes with adenosine monophosphate-activated protein kinase, the major sensor of cell energy and a negative regulator of the mammalian target of rapamycin (mTOR) pathway. Recently, multiple studies have implicated Folliculin in adherens junction formation and signaling.^51,52

Tuberous sclerosis is a syndrome of autosomal dominant inheritance, with two genes identified, TSC1, located on 9q34, and TSC2, located on 16p13.3. It affects 1 in 6000 people and is usually diagnosed at birth.15,45,46 This syndrome encompasses multiple organ systems, including dermatologic, cardiac, pulmonary, and renal. Skins lesions include facial angiofibromas, periungual fibroma, shagreen patches, and hypopigmented macules. Patients also develop cardiac rhabdomyomas, pulmonary lymphangioleiomyomatosis, retinal hamartomas, subependymal nodules, and giant cell astrocytomas. The renal manifestations include bilateral and multifocal angiomyolipomas (AMLs) and less commonly clear cell renal carcinoma. In contrast with spontaneous AML, AML in this setting can be associated with a low risk of occult RCC (1%). The TSC1 and TSC2 gene products inhibit activation of mTOR signaling, a major promoter of protein synthesis and cell growth.

Hereditary papillary renal cell carcinoma (HPRCC), inherited as an autosomal dominant trait, is caused by mutations in Met protooncogene on chromosomal band 7q31–34.15,46,53 It is characterized primarily by bilateral, multifocal papillary type I RCC. These tumors are not aggressive and rarely metastasize. Age at onset is around the fifth decade. The Met oncogene encodes a membrane tyrosine kinase that, in HPRCC, harbors an activating mutation in the kinase domain. Hyperactivation of the Met oncoprotein leads to upregulation of several intracellular signaling pathways involving Akt, Rac, and MAP kinase.

Hereditary leiomyomatosis and renal cell carcinoma (HLRCC) is a disease of autosomal dominant inheritance. The gene for this disorder localizes to chromosomal band 1q42.3–43 and has been identified as fumarate hydratase (FH).15,45,46 Age at onset is between the third and fourth decades of life. The syndrome consists of cutaneous leiomyomas, uterine leiomyomas (fibroids), and papillary type II RCC tumors with high metastatic potential, even when small in size. Unlike the VHL patients, in whom delay in surgical treatment is usually the rule until the largest tumor reaches 3 cm, there should be no delay in treatment of solid renal lesions in these patients. Cystic lesions in these patients should be followed with close surveillance and early surgical intervention for any radiographic development of a potentially solid component. Thirty percent of patients have solitary and unilateral renal tumors.^45,46

The FH gene functions as a tumor suppressor, with loss of the second allele detected in kidney tumors. The wild-type gene encodes an enzyme in the Krebs cycle catalyzing fumarate conversion into malate. Loss of the FH enzyme leads to accumulation of fumarate, which has been suggested to promote carcinogenesis through indirect stabilization of transcription factors HIFα and NRF2.39,40,54

Distinct from HPRCC and HLRCC, Malchoff and colleagues have described a three-generation family with five cases of papillary thyroid carcinoma and two cases with papillary renal neoplasia.⁵³ With the use of linkage analysis, these investigators demonstrated that the fPTC/PRN phenotype was linked to 1q21.⁵³ They characterized a distinct inherited tumor syndrome that may establish a link between papillary RCC and familial papillary thyroid carcinoma.⁵³

Hereditary renal cell cancer associated with melanoma has been recently described. The TFE3 gene mutated in sporadic RCC (see “Diagnosis of Renal Cell Carcinoma”) is one of four members of the MiT family of transcription factors; although TFE3 mutations have not been identified in hereditary RCC syndromes, the related MiT member, MiTF, was shown to have a specific amino acid substitution associated with hereditary RCC tumors associated with melanoma.⁵⁵ This substitution confers hyperactivation of MiTF transcriptional activity by preventing its sumoylation and degradation. Histologic features of these MiTF renal cancers are yet to be characterized.

Radical Nephrectomy

Radical nephrectomy is the traditional gold-standard treatment for localized RCC. Long-term cancer-specific survival for pT1a, pT1b, pT2, and pT3 lesions can be approximated at greater than 90%, 80%, 70%, and 60%, respectively. Historically, a radical nephrectomy has included resection of the ipsilateral adrenal gland. However, adrenal resection increases the risk of life-threatening adrenal insufficiency (addisonian crisis) should the contralateral adrenal gland require resection or become otherwise functionally compromised. Today, it is common practice to limit adrenalectomy to the setting of large or upper pole renal tumors, given a low risk of adrenal involvement with smaller RCC tumors. Even in this setting, ipsilateral adrenalectomy has been challenged by recent studies suggesting that negative CT imaging of the ipsilateral adrenal gland can effectively rule out RCC involvement.⁸⁵ Although still controversial, in the setting of negative cross-sectional imaging, adrenal gland preservation should be considered to avoid long-term risk of addisonian crisis.

The role of regional lymphadenectomy at the time of radical nephrectomy is also controversial. The most common regional landing sites for RCC are the paracaval, interaortocaval, and retrocaval nodes for the right kidney, and the paraaortic, interaortocaval, and preaortic nodes for the left kidney. In the presence of radiographic or palpable lymphadenopathy, a lymph node dissection is justified. In the absence of such findings, the role of lymphadenectomy is prognostic only, with no level 1 evidence to support a therapeutic benefit. Some series suggest a therapeutic role for lymphadenectomy specifically in patients without lymphadenopathy if their primary tumors are otherwise high risk, but these series are limited by retrospective nature and sample size. The European Organization for Research and Treatment of Cancer (EORTC) 30881 trial has provided the only prospective randomized trial investigating radical nephrectomy (N = 772) with or without lymphadenectomy, but showed no benefit with regional lymphadenectomy at a median follow up of 12.6 years.⁸⁶ This trial has been criticized for its inclusion of patients with predominantly low tumor stage and thus a low rate of nodal positivity (4%), suggesting insufficient power to detect a therapeutic benefit. In the absence of level 1 evidence supporting such a role, most urologists continue to avoid lymphadenectomy at nephrectomy, given the significant risk for hemorrhage during dissection along the aorta and vena cava.

Nephron-Sparing Surgery

Radical nephrectomy has been shown to cause significant decreases in renal function, raising the risk of chronic renal insufficiency and dialysis, and chronic renal insufficiency, in turn, raises the risk of cardiovascular events, hospitalization, and overall mortality.⁸⁷ Acknowledging the detrimental impact of chronic renal insufficiency on overall health, urologists have increasingly strived over the past 2 decades to preserve renal function in RCC patients. With increasing sensitivity and use of imaging studies leading to detection of smaller renal tumors, partial nephrectomy has become an effective alternative to radical nephrectomy for patients with localized RCC, resulting in low morbidity and good oncologic outcomes.

Traditional indications for partial nephrectomy have included conditions in which radical resection would leave the patient anephric (bilateral RCC, a horseshoe kidney, or a solitary kidney), requiring immediate dialysis; and unilateral RCC with a contralateral kidney at risk for compromised function, as in diabetes, hypertensive nephrosclerosis, renal artery stenosis, renal calculi, and chronic pyelonephritis. Initial outcomes under such indications supported operative feasibility and oncologic safety of a nephron-sparing approach. Patient selection quickly expanded to include localized primary tumors with diameter of 4 cm or less and exophytic location, regardless of renal function.⁸⁸ Although the risk of local recurrence is necessarily increased with a nephron-sparing approach, risk of cancer-specific mortality appears similar to a radical approach. Perioperative complications are, however, more common with partial nephrectomy, due largely to increased bleeding/transfusion and urinary leaks related to collecting system injury.

Indications for partial nephrectomy have expanded to include select larger tumors.⁸⁹ Several studies have now retrospectively compared partial and radical nephrectomy for larger tumors up to 7 cm, demonstrating similar oncologic outcomes.^90,91 Leibovich and associates compared outcomes in 91 patients managed with nephron-sparing surgery and 841 patients who underwent radical nephrectomy for 4- to 7-cm tumors.⁹⁰ These investigators found no statistical difference in cancer-specific survival and distant metastases between the two groups. Becker similarly compared 196 patients undergoing partial (N = 45) or radical nephrectomy (N = 151) for clear cell RCC tumors of 4 to 7 cm. They observed no differences in oncologic outcomes between the two groups, but noted better postoperative renal function at 3 months for patients undergoing nephron sparing.⁹¹ With regard to feasibility of partial nephrectomy, tumor location (exophytic vs. endophytic) may be a better determinant than tumor size. With either larger size or more endophytic location, ischemic time can be expected to increase, along with complications related to bleeding and collecting system injury (urinary leak).^94–94 The indications for nephron-sparing surgery are partly dependent on the experience and skill of the surgeon.

The primary oncologic concern with nephron-sparing surgery is the risk or recurrence in the same kidney, avoided with radical nephrectomy. Ipsilateral intrarenal recurrence occurs in 1% to 6% of patients after partial nephrectomy, and may result from either primary tumor multifocality (including de novo tumor formation) or positive surgical margins.⁹⁵ However, it remains controversial whether positive surgical margins during partial nephrectomy increase the risk of RCC recurrence or have no prognostic significance. Many series support good local recurrence-free survival for patients with positive margins and intermediate-term follow up. Furthermore, immediate or delayed radical nephrectomy most often fails to reveal any residual disease in these patients. In a recent description of a large single-institutional series, Kwon et al. found no relation between margin status after partial nephrectomy and primary tumor risk (metastatic potential).⁹⁶ However, local recurrence occurred in 4% of patients with a positive compared to only 0.5% of patients with negative surgical margins, and all cases of local recurrence in both groups were from high-risk primary tumors. The authors concluded that local recurrence may be more likely when positive surgical margins occur in the specific setting of high-risk primary tumor histology, whereas positive margins with low-risk primary tumors may have no significance. Current practice generally assumes no prognostic significance of positive surgical margins.

Given the benefit of nephron preservation and potential insignificance of positive margins, the necessity of a 1-cm resection margin proposed by Vermooten in 1950 has come under question.⁹⁷ Margin width during partial nephrectomy has thus been increasingly reduced over time.^98,99 Several recent partial nephrectomy series show no association between disease recurrence and margin size.^100–100 Some urologists now advocate a zero-margin tumor “enucleation,” taking advantage of a natural plane between the renal parenchyma and tumor pseudocapsule, a histologically confirmed structure ranging from 0.04 mm to 0.79 mm in width.¹⁰¹ In a series of 232 patients undergoing enucleation of tumors smaller than 4 cm, Carini et al. reported 97% and 95% 5- and 10-year cancer-specific survival rates, respectively, with no cases of positive margins or recurrences in the resection bed.¹⁰² According to these investigators, even though pseudocapsular microinfiltration is frequent (33%) in enucleated tumors, positive margins are still avoided because of a surrounding chronic inflammatory layer that is, on average, 1 mm in width.¹⁰¹ In addition to optimizing renal preservation, tumor enucleation enables the surgeon to work in a relatively avascular plane and may reduce blood loss during resection and risk of collecting system injury. The reduction in bleeding may allow resection without clamping of the main renal artery, avoiding renal ischemia.¹⁰³

A large volume of existing retrospective literature now suggests survival benefits of partial nephrectomy over radical nephrectomy, but suffers from an uncertain selection bias. In a meta-analysis of more than 40,000 patients undergoing radical or partial nephrectomy, the latter was associated with a significant benefit in both survival and renal function; however, the quality of pooled studies was judged to be low.¹⁰⁴ Recently, a large multiinstitutional randomized trial from EORTC comparing partial and radical nephrectomy for tumors 5 cm or smaller has challenged current thinking on the role of nephron sparing.¹⁰⁵ A total of 541 patients were enrolled in this study and median follow-up was more than 9 years. Based on an intention-to-treat analysis, the radical nephrectomy group cohort showed better overall survival compared with patients assigned to partial nephrectomy. Because there were very few cancer-specific deaths in either cohort, the benefit in overall survival for radical nephrectomy was attributed to a reduction in non–cancer-related deaths. This unexpected outcome has come under intense scrutiny and the study has been heavily criticized because of underaccrual and its intention-to-treat analysis with frequent study arm crossover. Perhaps unduly, these findings have been largely dismissed by the urology community.

With the standardization of staging and earlier diagnosis of the disease, a more tumor-specific surgical management has proved to be advantageous to maximizing the preservation of functional renal tissue. With increased evidence of the need for long-term dialysis in patients undergoing total nephrectomy, maintenance of renal function in both kidneys represents the strongest argument for nephron-sparing surgery in patients with RCC who have the best chance for cure.

Surgical Approach

An open incision has provided the traditional approach to resection of localized kidney cancers. More recently, laparoscopy has gained widespread popularity as a result of its potential for lower morbidity and faster convalescence. Whether an open or laparoscopic surgical approach is performed depends in large part on the extent of localized disease, including tumor size and location, and the training and experience of the surgeon. For lesions that appear to be locally extensive with potential involvement of neighboring organs, open surgery is still preferred, particularly if a colectomy, splenectomy, or distal pancreatectomy is anticipated. In addition, local adenopathy or tumor thrombus extension into the renal vein and vena cava provides a relative indication for open surgery. Even in such situations, however, laparoscopic and robotic resection may be possible, but should be reserved for specialists with experience in such techniques, as the morbidity can be quite high.

For cancers confined to the kidney requiring radical nephrectomy, a laparoscopic approach is generally preferred based on its tendency for lower blood loss, faster inpatient and outpatient recovery, and similar oncologic outcomes relative to an open approach, although these differences are inferred largely from retrospective comparisons.106–110 In a prospective randomized comparison between these two approaches among patients with tumors up to 8 cm, Burgess et al. found that laparoscopic radical nephrectomy achieved significantly better postoperative pain scores and convalescence, but no reduction in hospitalization.¹¹¹

While laparoscopy is suitable for most radical nephrectomies, large renal tumors often are best approached in open fashion. In some instances, however, even large tumors can be approached laparoscopically if the local anatomy is favorable, although intraoperative mobilization of the kidney may become more difficult. In a retrospective comparison of open and laparoscopic radical nephrectomy for T2 tumors with a mean size of 10 cm, Hemal et al. observed lower blood loss, less narcotic use, and quicker recovery with laparoscopy, at the expense of a longer mean operative time.¹¹² In another retrospective study involving tumors larger than 7 cm, complications were more common with laparoscopic than open radical nephrectomy.¹⁰³ Conversion from laparoscopy to open can be expected at higher frequency with large tumors. Gong et al. reported open conversion in 12% of laparoscopic nephrectomies for T2 tumors compared with only 1% for T1 tumors, although postoperative recovery was no different.¹¹⁴

In contrast to radical nephrectomy, the gold standard approach to partial nephrectomy remains an open incision. However, with increasing usage of minimally invasive therapies and enhanced dexterity afforded by robotic assistance, the indications for laparoscopy are expanding to include partial nephrectomy in select patients. Today, the choice between open and robotic or laparoscopic partial nephrectomy depends on anatomic location of the tumor, body habitus, and ability to tolerate pneumoperitoneum. The optimal tumor for laparoscopic partial nephrectomy is small (<4 cm) and peripheral/exophytic, although experienced laparoscopists may take on larger and more centrally located lesions. In a study from a single high-volume laparoscopic surgeon, similar operative outcomes were observed for laparoscopic resection of tumors larger than 4 cm (N = 58) compared to smaller tumors (N = 367), with the exception of a 6- to 8-minute increase in ischemic time.¹¹⁵ Central tumors resected laparoscopically were associated with increases in both ischemic and overall operative times, but not bleeding. An experienced surgeon can thus perform technically challenging laparoscopy with good surgical and oncologic outcomes.¹¹⁶

The advent of the daVinci robot has simplified the most difficult technical portion of the operation, namely intracorporeal suturing and reconstruction of the resection defect. This has translated to improvements in several key operative outcomes for minimally invasive partial nephrectomy, in addition to a quicker learning curve. In one experienced minimally invasive surgeon’s series of 492 laparoscopic (N = 231) or robotic (N = 261) partial nephrectomies, the latter approach correlated with a significant reduction in operative time, ischemia time, positive margins, and perioperative complications, despite greater patient comorbidity and tumor complexity.¹¹⁷ Other retrospective series indicate reduced blood loss with the robotic approach as well.^118,119

The single remaining concern for laparoscopic and robotic nephron-sparing surgery is the amount of warm ischemic time and resulting loss of renal function. Precooling of the kidney prior to renal arterial clamping during open partial nephrectomy is believed to reduce ischemic injury but has thus far not been feasible with a laparoscopic approach. Warm ischemic times during laparoscopy in excess of 20 minutes have been suggested to result in irreversible injury. For this reason, complex resections, including large, endophytic or multifocal tumors, may require alternative methods, including either open resection using cold ischemia, or resection of the tumor without main renal artery clamping. The latter approach increases blood loss and the likelihood of transfusion, but optimizes preservation of renal function. Recently, renal hypoperfusion by pharmacologically induced hypotension has been proposed to avoid main renal artery clamping, but the safety and efficacy requires further study.¹²⁰ Newer methods include superselective segmental renal artery clamping, with or without administration of near infrared dyes to visualize clamping impact. Such strategies may allow resection without putting the entire kidney through an ischemic insult.^121,122 However, immediate and long-term efficacy of these technical innovations are still under active investigation.

When open surgery is required, the choice of anatomic approach also is dictated by the surgeon’s training and experience, and by the location and extent of disease. Resection of tumors with thrombus extending into the inferior vena cava above the level of the hepatic veins usually is performed through a midline incision to allow extension to a sternotomy, if required. Tumors with thrombus extending into the inferior vena cava, but below the hepatic veins, often can be accessed by a thoracoabdominal approach, if the primary lesion is on the right side. For other, less extensive tumors, laparoscopy is the preferred approach when nephrectomy is deemed necessary.

Although most laparoscopic surgeons prefer a transperitoneal (transabdominal) approach for this operation, many surgeons opt for a retroperitoneal approach. The difference in approach seems to matter relatively little to the patient, but rather is a technical issue for the surgeon. The transabdominal approach allows complete visualization of the viscera with multiple landmarks and visual cues for orientation, but does require more anterior and anteromedial dissection. Surgery through the retroperitoneal approach can be faster in the hands of an experienced laparoscopic surgeon, but many visual cues and landmarks to ensure dissection in the proper planes are lacking with this procedure. Regardless of approach, however, current surgical practice in the United States is to perform nephrectomy robotically or laparoscopically (with or without hand assistance), except when an excessively large tumor, locally advanced disease, local adenopathy, or tumor thrombus is present. Even in these situations, it is often beneficial to start a case laparoscopically and then complete it in a controlled open fashion. Such a strategy can result in a smaller incision, often in the midline, with less pain and morbidity than if the entire case is performed in open fashion.

The common benefits of cosmesis and pain control with minimally invasive surgery is partly compromised during laparoscopic radical nephrectomy by the substantial skin/fascial incision needed to extract the kidney and surrounding Gerota’s fat en bloc. Some laparoscopists have accordingly advocated morcellation of the specimen as an alternative to removing the kidney as a whole. An intact extraction has the benefit of maintaining kidney and tumor integrity for more precise histopathological evaluation, decreases the chance of tumor seeding of the port sites or renal fossa, and requires shorter surgery time.¹²³ Pautler and colleagues compared the histopathological diagnosis of their operative needle biopsy before morcellation with the morcellated specimen.¹²⁴ They found adequate histopathologic material for diagnosis but were unable to stage most cases.¹²⁴ Although it has been argued that lack of pathological tumor staging information does not affect management of the patient, the tumor stage is in fact the major determinant of current risk-based postoperative surveillance regimens.¹²⁵ Hence, although there was initially great enthusiasm among the laparoscopic surgeons to remove the kidneys and tumors through morcellation, this practice has been abandoned and nearly all urologic surgeons now prefer intact extraction of the kidney and mass for accurate histopathological diagnosis and staging.

Thermal Ablation for Renal Cell Carcinoma

Urologists pioneered minimally invasive surgery and have been constantly searching for less-invasive techniques. With the advent of energy-based ablative alternatives to open and laparoscopic surgeries in selected patients, it is now possible to achieve cancer-specific survival with decreased surgical morbidity. Two types of ablation are currently performed, cryoablation and radiofrequency ablation. Although either can be performed laparoscopically under direct vision, ablation is most often performed percutaneously under ultrasound or CT guidance. While the usefulness of ablative therapies is still being evaluated by many groups, the American Urological Association consensus guidelines considers the efficacy of ablative techniques to be slightly less than surgical resection and should be reserved primarily for patients who decline surgery or who are not good surgical candidates but who desire treatment.¹²⁶

Renal cryoablation has been developed for the treatment of small renal tumors as an alternative to nephron-sparing surgery, with the goal of reducing operative morbidity. Use of standard cryoprobes of 3 to 8 mm can cause rupture of the renal capsule and parenchyma, resulting in significant bleeding. The use of multiple 2.4-mm cryoprobes under intraoperative real-time ultrasound guidance may decrease the risk of bleeding and make this technique a more feasible approach.¹²⁷ Increasing evidence suggests that cryoablation is an acceptable surgical alternative to traditional open and laparoscopic nephrectomy.^128–132 In addition to shorter hospital stay and less morbidity, cryoablation has been shown to be an effective nephron-sparing cancer therapy.¹³² Patients selected for this option tend to have peripheral lesions 5 cm or less in diameter.^128–132 Minimally invasive therapy without the sequelae of open or laparoscopic surgery may be reasonable in these situations, particularly in morbid or elderly patients who are poor surgical candidates. Cryoablation surgery has been described in the literature with open, laparoscopic, and, more recently, imaging-guided percutaneous approaches.^133,134

Multiple theories have been proposed for the mechanism of action of cryoablative surgery. The most accepted theory postulates direct cellular injury leading to coagulative necrosis.130,131 Injury to the cancer cells occurs as a result of intracellular ice crystal formation during the freezing phase of the treatment.^130,131 The freezing process causes protein damage, which injures the cell membrane and essential enzymatic processes.¹³⁰ Ice crystals that form within the cell disrupt intracellular organelles and membranes. Additionally, indirect ischemic injury caused by occlusion of the microvasculature during the thaw phase creates stasis within blood vessels, leading to end-tissue infarction. These ensuing insults to renal tissue lead to liquefactive necrosis.¹³⁰

In most cases, the cryoprobes used are 2.4 mm in diameter. The usual procedure is percutaneous CT guidance cryotherapy, in which each lesion is treated with a double freeze–thaw cycle, with active freeze at 100% efficiency for a minimum of 10 minutes and a passive thaw of 8 minutes.¹²⁸ These cycle times consistently deliver the lethal freeze of −4° C within 1 cm of the ice ball edge.¹²⁸ The critical threshold for necrosis has been reported as between −19.4° C and −40° C.¹²⁸ The number of probes used is specific to the size and shape of the lesion, with freeze margins of 0.5 to 1 cm of the lesion.

Advantages to CT-guided cryotherapy include general availability and its excellent capability to differentiate among the visceral organs. It also provides visualization of the entire ice ball while differentiating between frozen and unfrozen tissue. In addition, it provides real-time guidance in the CT fluoroscopy mode. One major disadvantage with CT guidance is the constant radiation exposure to the patient and the surgeon, which can be avoided with adequate and appropriate protection.

Numerous studies of cryoablation support adequate oncologic efficacy with short- and intermediate-term patient follow-up. However, there are at present sparse long-term data available. Gill and colleagues recently provided the first 10-year oncologic outcomes for cryotherapy using a laparoscopic approach. The mean tumor size was 2.3 cm and cancer specific survival at 10 years was 83%, comparable but slightly lower than would be expected for surgical resection of renal cancers of this size.¹³⁵ Because the renal tumor is not excised and histopathological margins are unknown, whether the entire tumor has been extirpated remains uncertain.¹²⁸ This challenge forces frequent CT and MRI follow-up studies on patients who receive this therapy.

Radiofrequency ablation (RFA) provides an alternative method of renal tumor ablation. This approach employs a 150- to 200-W generator to create a high-frequency alternating current (460 to 500 Hz), causing frictional heating and protein denaturation. Electrode needles with or without tines are placed in the tumor laparoscopically or percutaneously under ultrasound or CT guidance. Temperatures of 60° C to 105° C are generated in the tumor, causing immediate cell death. Heating may be cycled, with each cycle approximately 5 to 12 minutes each. Probes may be inserted into the surrounding normal parenchyma for thermal monitoring. A margin of 0.5 to 1.0 cm is typically achieved.

Potential complications of RFA are similar to cryoablation, and include urinary leak/fistula, ureteral stricture or obstruction as a result of tissue sloughing, and injury to adjacent organs. Bleeding complications are infrequently encountered with RFA and make this procedure an attractive option for poor surgical candidates on blood thinning medication. Allaf and colleagues reported increased pain with RFA compared to cryoablation, attributing this to a potential analgesic effect of the latter.¹³⁶

As with cryotherapy, numerous studies now support effective short-to-intermediate–term oncologic efficacy of RFA; however, longer-term outcomes data are, at present, sparse.137–140 Psutka and colleagues documented oncologic outcomes for 185 RFA patients with a median tumor size of 3 cm and a median follow-up of 6.3 years. Only 2.2% and 1.6% of patients developed metastasis and died, respectively, although 8% and 24% of T1a and T1b patients, respectively, were not disease-free.¹⁴¹ The low rate of metastasis and death is promising, and validation of these longer-term outcomes is now awaited.

Active Surveillance

Active surveillance has emerged has an alternative approach to extirpative or ablative management of small renal masses. Approximately 30% of renal tumors are benign when 4 cm or less in diameter. Among small renal masses with malignant histology, the vast majority (75% to 85%) are low grade and clinically indolent. Hence, the American Urological Association now recommends consideration of active surveillance for small renal masses in select patients willing to accept a low risk of metastasis. Such management may be ideal for select patients with poor renal function or high operative risk, such as the elderly, or comorbid disease, and those with high tumor complexity.

The precise risk of metastasis with active surveillance is unclear, but appears to be around 1% to 2% based on current literature.142,143 However, follow up in these studies is mostly short- or intermediate-term only, thus this long-term rate may be an underestimate, particularly given the 5% to 10% rate of metastasis expected after surgical resection of pT1a tumors. In addition, reporting bias may further underestimate the rate of metastasis on active surveillance. During patient selection for active surveillance, the metastatic risk must be weighed in light of both the patient’s willingness to undergo treatment and the patient’s operative risk, including general health/performance status, renal function, and tumor complexity. Percutaneous tumor biopsy at present does not play a routine role during patient selection, probably because of concerns of common undergrading and a historically high nondiagnostic rate, although accuracy of 80% to 90% is described in most contemporary biopsy series.

Presently, there are no standardized guidelines for surveillance regimens and thresholds for implementation of delayed treatment. A number of studies have attempted to identify clinical predictors of metastasis for patients on active surveillance, but are limited by low metastatic case number. In many cases of metastasis on active surveillance, the primary tumor has progressed to greater than 4 cm. Accordingly, tumor size may provide a useful trigger for intervention. Increasingly, tumor growth rate is also being used as a trigger. Although growth rate of small renal masses does not differ between malignant and benign tumors (0.2 to 0.3 cm/year), high-grade and particularly metastatic primary tumors appear to grow faster, with the latter ranging from 1.3 to 2.9 cm/year.142,144–147 A growth rate of more than 0.5 cm/year has accordingly been implemented by some urologists as a threshold for intervention. This approach has been challenged by the fact that oncocytomas commonly demonstrate rapid growth rates.¹⁴⁸

von Hippel-Lindau Disease and Other Familial Renal Cell Carcinomas

Patients with von Hippel-Lindau RCC or other familial forms of RCC are at high risk for local recurrence after nephron-sparing surgery and require close lifelong surveillance. More than 80% of patients with von Hippel-Lindau disease treated with nephron-sparing surgical resection will have a recurrence in the ipsilateral kidney within 10 years, and lesions will develop in the contralateral kidney, if they have not done so already.¹⁶⁴ This is because multiple microscopic lesions are present throughout the kidneys, despite the grossly normal appearance of the parenchyma.¹⁶⁵ A diagnosis of von Hippel-Lindau disease should be considered in any patient with early onset or multifocal RCC or RCC in conjunction with the following: a history of visual or neurological symptoms; a family history of blindness, central nervous system (CNS) tumors, or RCC; or coexisting pancreatic cysts, epididymal lesions, or inner ear tumors.^166,167 Standard recommendations for surveillance in patients with von Hippel-Lindau disease include (a) CT of the abdomen and pelvis every 6 months in patients with solid renal lesions and every 12 months in patients without solid renal lesions, with alternating MRI and CT in younger patients to avoid secondary malignancies associated with prolonged use of the former, (b) annual physical and ophthalmologic evaluations, (c) estimation of urinary catecholamines every 1 to 2 years, (d) MRI of the CNS every 2 years, and (e) periodic auditory examinations. Molecular genetic and clinical screening should be offered to appropriate family members based on an autosomal-dominant pattern of inheritance.^164,167,168

Adjuvant Therapies

A role for adjuvant therapy after nephrectomy remains to be established, and to date, observation remains the standard of care outside a clinical trial. Clinical studies with immunotherapies have failed to demonstrate a clinical benefit. One randomized, multicenter, prospective study has reported on the efficacy of adjuvant interferon alfa-2b given after nephrectomy to patients with Robson stages II and III RCC.¹⁷³ Patients randomized to receive interferon alfa-2b were given a dose of 6 million IU intramuscularly 3 times per week for 6 months, starting within 1 month of surgery. Patients in the observation-only group who relapsed were given interferon alfa-2b, 10 million IU intramuscularly, 3 times per week, or the best available treatment at the time of relapse. Treatment groups were well matched at baseline for patient and tumor characteristics. No significant differences were observed in overall or event-free survival for patients randomized to management by observation (N = 124) and patients receiving interferon alfa-2b (N = 123). This study did not support the use of adjuvant interferon alfa-2b after nephrectomy. An analyzed and reported phase III study of interferon-α as adjuvant treatment for resectable RCC by the ECOG/Intergroup showed that adjuvant treatment with interferon did not contribute to better overall survival or relapse-free survival.¹⁷⁴

As new drugs have become available for the treatment of metastatic disease, a rationale has arisen for testing these agents in micrometastatic disease. A large, randomized, double-blind phase III study conducted by the ECOG (E2805) has been completed. Currently accruing patients who have undergone nephrectomy for kidney cancer. Patients with lesions staged as above T1c or Fuhrman grade of 3 or higher (or both) were randomly assigned to a placebo group or to either a sorafenib or sunitinib treatment group. Treatment was to continue for 1 year with close monitoring. The primary clinical end point was time to tumor progression among the three groups. The goal of this important study is to identify a possible role for VEGF receptor (VEGFR) TKI in the adjuvant setting; confirmation of a positive effect will establish a new standard of care after nephrectomy for patients with risk for recurrence. Additional adjuvant studies involving other antiangiogenic drugs are ongoing.

Resection of Metastases in Renal Cell Carcinoma

Resection of solitary metastases from RCC is associated with improved survival, although the selection criteria have been poorly defined. This issue was addressed in a retrospective, single-institution analysis of patients with recurrent RCC (N = 278) at the Memorial Sloan-Kettering Cancer Center.¹⁸³ Recurrent disease was defined as solitary if it recurred within the resected renal bed or if one organ system or site was involved. Multiple unilateral lesions in the lung were considered to represent a solitary site of metastasis. Bilateral lung involvement or recurrence in two or more sites was considered to constitute multiple sites. Surgery for metastatic disease was considered curative if metastases were curatively resected and noncurative if gross tumor was left behind. Of the 278 patients who underwent initial curative nephrectomy, recurrence was solitary in 155 patients and multiple in 123 patients, with a median time to first recurrence of 25 months. The overall 5-year survival rate of patients who underwent curative resection for the first recurrence was 44% (N = 141), 14% for those who received noncurative resection (N = 70), and 11% for those who were treated nonsurgically (N = 67). Favorable predictors of survival by multivariate analysis included a single site of first recurrence, curative resection of the first metastasis, a disease-free interval of more than 12 months, and a metachronous presentation with recurrence. Curative resection of isolated metastases to glandular tissue (thyroid, salivary gland, pancreas, adrenal, ovary) was associated with the best 5-year overall survival rate (63%), followed by resection of isolated lung metastases (54%). In this study, resection of solitary brain metastases, however, was associated with poor outcome, with an 18% 5-year overall rate. The 5-year overall survival rates of 46% and 44%, respectively, for patients who underwent second (N = 62) or third curative resection (N = 22) of subsequent metastases after initial curative metastasectomy were not significantly different from those for patients who received only initial curative metastasectomy (44%; N = 141). Thus patients who undergo complete resection of metastatic disease in a solitary site after a disease-free interval of longer than 12 months may experience long-term survival. As with previous reports, complete resection is the important factor, rather than the number of sites resected, even in patients who have undergone prior metastectomy.^184,185

Brain metastases from RCC raise specific therapeutic problems because they are relatively unresponsive to whole-brain radiation therapy and tend to bleed. Stereotactically guided high-precision irradiation (radiosurgery) has shown promising results in selected patients with brain metastases from RCC.¹⁸⁶ Radiosurgery appears attractive owing to its low risk of toxicity and minimal invasiveness. Multiple lesions can be treated at the same time, and retreatment can be performed for local or distant recurrences.

Lymphadenectomy in Locally Advanced Disease

The role of lymphadenectomy in RCC remains controversial. Nodal disease has been found to be a predictor of poor prognosis even with M0 disease, although, there is very little evidence to support the value of lymph node dissection.189–189 The decision for lymph node dissection should be based on multiple factors. The need for accurate staging, decreased local recurrence rates, and improved survival should be weighed against risk of morbidity and mortality.^187,190 One argument against lymph node dissection is that the kidney is drained hematologically and lymphatically through multiple pathways.¹⁸⁷ Owing to the unpredictable pathways of spread accurate staging may not be possible. From 58% to 95% of patients with lymph node disease also have synchronous metastasis.¹⁹¹ consequently, the likelihood of finding patients with localized disease is low at 2% to 9%. Canfield and associates did a retrospective study supporting the need for aggressive nodal resection in the presence of positive lymph nodes without evidence of metastasis.¹⁹¹ Blom and coworkers of the EORTC Genitourinary Group conducted a large randomized trial comparing radical nephrectomy with lymph node dissection and radical nephrectomy alone.¹⁸⁹ At 5-year follow-up evaluation, these workers found no significant difference in morbidity and mortality between the two groups after surgery. They also found no significant difference in survival. A small subset of patients may benefit from lymph node dissection. Pantuck and associates have shown improved survival in patients found to have disease-positive lymph nodes who undergo cytoreductive nephrectomy and postoperative immunotherapy.¹⁹⁰ The literature is variable on the benefits of lymph node dissection in patients with renal neoplasms. Overall, lymphadenectomy provides little staging information and confers no clear benefit in decreasing risk of recurrence or in survival.

Inpatient high-dose bolus IL-2 received FDA approval for treatment for patients with stage IV RCC in 1992 based on data presented on 255 patients who were entered into seven phase II clinical trials.192,193 In these studies, patients received 600,000 to 720,000 IU/kg of recombinant human IL-2 by 15-minute infusion every 8 hours during two 5-day courses (maximum: 14 doses per course) separated by 5 to 9 days of rest. Stable or responding patients received 2 to 5 courses of therapy at 8- to 12-week intervals and then were observed while not receiving any additional therapy. ORs were seen in 37 (15%) of the 255 patients, including 17 complete responses (CRs; 7%) and 20 partial responses (PRs; 8%). The median duration of response was 54 months for all of the responders and 20 months for partial responders; the median has not yet been reached for complete responders. The median survival was 16 months for all 255 patients. Most patients who achieved a CR that lasted longer than 30 months and those with PRs after resection resulting in “no evidence of disease” after a response to high-dose IL-2 were unlikely to experience disease progression and may actually be cured.

Although the inpatient high-dose bolus IL-2 regimen produces favorable outcomes in a handful of patients, it also is associated with significant toxic effects and cost and is not universally available. Low-dose IL-2 regimens (with or without interferon-α) have produced similar response rates and survival in nonrandomized phase II trials, but responses appeared to be less durable than those seen with high-dose IL-2.194–197

In an effort to determine the value of outpatient subcutaneous IL-2 and interferon-α relative to high-dose intravenous IL-2, the Cytokine Working Group did a phase III trial in which patients were randomized to receive either outpatient IL-2 and interferon-α every 6 weeks or standard high-dose inpatient IL-2 every 12 weeks.198,199 Of the 193 patients who were enrolled, 192 were evaluable for toxicity and tumor response. The response rate for high-dose IL-2 was 23%, versus 10% for IL-2 and interferon-α. Eight patients achieved a CR while taking high-dose IL-2, versus only three patients taking low-dose IL-2 and interferon-α. The median response durations were 24 months for high-dose IL-2 and 15 months for IL-2 and interferon-α. Median overall survival times were 17.5 and 13 months, favoring high-dose IL-2. Ten patients (nine major responders) who received high-dose IL-2 were progression free at 3 years, versus three patients (two major responders) who received IL-2 and interferon-α. Of note, responses to high-dose IL-2 were seen with equal frequency across the stratification criteria, whereas low-dose IL-2 and interferon-α seemed to produce fewer responses in patients with liver or bone metastases and in those who had not undergone prior nephrectomy to remove the primary tumor. For patients with bone or liver metastases, or with unresected primary tumor, survival was superior with high-dose IL-2 compared with IL-2 and interferon-α, whereas no significant survival differences between the two treatments were noted for patients who had undergone prior nephrectomy or who were without bone or liver metastases.

Taken together, these studies suggest that high-dose intravenous bolus IL-2 is superior in terms of response rate and possibly response quality to regimens that involve either low-dose IL-2 and interferon-α, intermediate- or low-dose IL-2 alone, or interferon-α alone.¹⁹⁹ The superiority of high-dose IL-2 is particularly apparent in patients with tumor metastases in immune-sequestered sites, such as liver or bone, or whose primary tumor has not been resected, or who fall into the intermediate-risk or poor-risk group defined by the French Immunotherapy Group. Consequently, although low-dose cytokine therapy has a limited role in mRCC, it must be concluded that high-dose intravenous IL-2 should remain the preferred therapy for appropriately selected patients with access to such therapy. In view of the toxicity and limited efficacy of high-dose intravenous IL-2 therapy, however, additional efforts should be directed at better defining the patient population for whom this therapy is appropriate.²⁰⁰

Use of interferon-α as a single agent has resulted in only a modest survival benefit in the randomized setting, although combination of this agent with low-dose IL-2 (and perhaps with fluorouracil) may lead to an improved response rate, but evidence of survival benefit in a prospective randomized trial is lacking.203–203 Although interferon-α has been the outpatient treatment of choice in most of Europe and the United States until recently, benefit again appears to be restricted to patients with good prognostic risk.²⁰⁴

Predictors of Response to Cytokine Therapies

Responses to immunotherapy most frequently are seen in patients with RCC of clear cell histology.205–208 This observation was detailed in a retrospective analysis of pathology specimens obtained from 231 patients (163 primary and 68 metastatic tumor specimens) who had received IL-2 therapy in Cytokine Working Group clinical trials.²⁰⁸ For patients with primary tumor specimens available for review, the response rate to IL-2 was 21% (30 of 146) for patients with clear cell histology primary tumors, compared with 6% for patients with non-clear cell histology (1 responder in 17 patients). Among the patients with clear cell carcinoma, response to IL-2 was also associated with the presence of alveolar features and the absence of papillary or granular features. The response rate in patients whose primary tumors had “good” predictive features (e.g., greater than 50% alveolar and no granular or papillary features) was 39% (14 of 36). Patients with tumors that contained “poor” predictive features (e.g., greater than 50% granular or any papillary features) had a response rate of 3% (1 of 33). When this model was then applied to the 68 patients with specimens from metastatic sites, those patients who were treated without resection of their primary tumors, five tumor responses were seen in the 20 patients with “good” predictive features, whereas no tumor responses were seen in the 16 patients in the “poor” predictive group, thus supporting the validity of the model developed from the primary kidney tumor specimens. As a result of these data, it may be appropriate for patients whose primary tumor is of non–clear cell histologic type, or of clear cell histologic type but with “poor” predictive features, to forgo IL-2-based treatment altogether. Even in the most favorable predictive group, however, more than 50% of patients failed to respond to IL-2 therapy, so additional investigations to identify tumor-associated predictors of responsiveness to IL-2 are still necessary.²⁰⁵

Some investigators have begun to examine tumor tissue to identify immunohistochemical markers that may predict the outcomes for patients with RCC. Carbonic anhydrase IX (CAIX) has been identified as one potential marker. Bui and colleagues used a monoclonal antibody designed to detect CAIX expression to perform an immunohistochemical analysis of paraffin-embedded RCC specimens. These investigators showed that greater than 90% of RCCs express CAIX, and that its expression decreases with advancing stage.²⁰⁹ In their analysis, high CAIX expression in primary tumors was seen in 79% of patients and was associated with improved survival and possibly improved response to IL-2-based therapy. Building on this work, Atkins and coworkers performed a nested case-control study within the larger cohort of patients for whom the histopathological findings were analyzed.²¹⁰ CAIX expression levels were correlated with response to IL-2, pathological risk categorization, and survival. Median survival times were 3 years and 1 year for high and low CAIX expressors, respectively. Although tumor response was seen in six patients with low CAIX staining, survival beyond 5 years was seen only in the patients with high-CAIX–expressing tumors. High-CAIX staining was associated with better histopathological features but remained an independent predictor of response. Additional studies to explain these preliminary observations and correlate results with previously described clinical features are necessary.²⁰⁵

The Cytokine Working Group has completed the high-dose IL-2 “Select” Trial. The primary objective of this study was to determine, in a prospective fashion, if the predictive model proposed by Atkins and associates can identify a group of patients with advanced RCC who are significantly more likely to respond to high-dose IL-2-based therapy (“good” risk) than a historical, unselected patient population. New factors (including baseline immune function, immunohistochemical markers, and gene expression patterns) that might be associated with response to high-dose IL-2 therapy have been explored in an attempt to more narrowly limit the application of IL-2 to those patients most likely to benefit.²⁰⁵

In this multicenter prospective study, patients with histologically confirmed RCC that was metastatic or unresectable, ECOG PS 0 to 1 and adequate organ function received high-dose IL-2 (600,000 U/kg/dose intravenously every 8 hours on days 1 through 5 and 15 to 19 [maximum 28 doses] every 12 weeks). The primary end point of the study was to determine the major response rate and associated “good” predictive features. One hundred twenty eligible patients were accrued between 2007 and 2009. The majority of patients (approximately 70%) had an ECOG performance status of 0 and were Memorial Sloan-Kettering Cancer Center intermediate risk. At the time of the analysis the investigator assessed, relative risk was 30% in clear cell RCC with six complete remissions. Response to IL-2 was not associated with any pretreatment clinical factor and was not seen in patients with non–clear cell histology. Preliminary analysis of CAIX staining did not show correlation with outcome. Additional analyses are still ongoing.²¹¹