Papillary carcinoma (PRCC) is thought to arise from either the proximal or distal convoluted tubules of the nephron and accounts for 10–15 % of localized RCC in most large series [13, 71, 80]. In approximately 10 % of PRCC cases, tumors are multifocal or bilateral [101]. Two morphologic subtypes of PRCC, type 1 with small cells and little cytoplasm and type 2 with large cells and eosinophilic cytoplasm, have been identified and shown to have different genetic profiles (Fig. 21.1a, b) [2, 25, 52, 71, 90]. Further work has suggested two separate molecular classes of PRCC: the first class, exhibiting excellent survival, has dysregulation of G1-S checkpoint genes and higher c-MET expression and combines morphologic type 1, low-grade type 2, and mixed type 1/low-grade type 2 tumors, and the second class, exhibiting poor survival, has dysregulation of G2-M checkpoint genes and is morphologically composed of high-grade type 2 tumors [129]. Although hereditary PRCC is associated with activating MET mutations [96, 100], histologically this tumor appears to be most similar to sporadic type 1 PRCC, and only about 14 % of patients with sporadic PRCC harbor a MET mutation [60, 97]. PRCC can also be seen in the hereditary leiomyomatosis and RCC syndrome due to fumarate hydratase (FH) tumor suppressor inactivation, and this appears to be histologically most similar to type 1 PRCC; FH mutations have not been described in sporadic cases [48, 55, 119]. Finally, vascular endothelial growth factor (VEGF) and VEGF receptors (VEGFR) have been shown to also be expressed in PRCC, but the clinical correlation remains unclear [28, 58]. Tubulocystic renal cell carcinoma and mucinous, tubular, and spindle cell carcinoma are two newer renal cancer subtypes that have been proposed to be additional variants of PRCC [59, 102].

Chromophobe RCC (CHRCC) is thought to originate from the intercalated cells in the renal collecting ducts and accounts for approximately 4–10 % of RCC (Fig. 21.2) [4, 9, 19]. In familial chromophobe RCC associated with Birt-Hogg-Dubé (BHD) syndrome, inactivation of a tumor suppressor, folliculin, has been identified and may activate the mTOR pathway, but folliculin alterations have rarely been found in sporadic CHRCC [43, 75, 123]. VEGF and upregulation of c-KIT have been noted in tumor specimens, although activating mutations of c-KIT have not been found [28, 95, 110, 127].

Collecting duct RCC (CDRCC) likely arises from the collecting (Bellini) ducts of the kidney in the renal medulla and is an aggressive RCC subtype with approximately one third of patients presenting with metastatic disease (Fig. 21.3) [69]. A relationship to urothelial cell carcinoma has been proposed [125].

Renal medullary carcinoma (RMC), a rare, aggressive, and usually fatal RCC variant, is a close relative of CDRCC (Fig. 21.4) [22, 128]. Almost all RMC occurs in children and young adults with sickle cell trait or disease.

MiT family translocation RCC involves different translocations of the TFE3, TFEB, TFEC, and MiTF transcription factors typically involving chromosome Xp11.2 but also including (6:11) translocations [6, 10, 102]. This was previously thought to be an extremely rare entity seen exclusively in children and young adults, but large series showed that 15 % of RCC patients under the age of 45 had this subtype of RCC. Although certain specific translocations can have indolent behavior, the majority of cases seen in adults are very aggressive [51].

Sarcomatoid dedifferentiation, first described by Fallow et al. in 1968, is not a separate histologic subtype but rather a variant that is observed with any RCC histology, can involve 1–100 % of a given tumor, and is seen in 5–10 % of RCC based on a number of large surgical series; although, it has been described to occur up to 30 % in CDRCC [12, 23, 35]. It exhibits a spindle cell pattern of growth, is always a high-grade tumor, and has been associated with the expression of VEGF, c-Kit, PDGFR-alpha, and S6 kinase as well as p53 mutations and is associated with poor prognosis (Fig. 21.5) [24, 78, 83, 117].

PRCC is in general resistant to conventional chemotherapy [17]. A retrospective study of 38 patients with metastatic PRCC, 30 of whom were treated with systemic therapy of which six received conventional chemotherapy, showed no objective responses [93]. In a study of 153 patients treated with gemcitabine and 5-fluorouracil, two had definite PRCC and neither showed an objective response [107]. A phase I study of gemcitabine, capecitabine, IFN-α, and thalidomide in 12 patients included two with PRCC, one of whom had a partial response [3]. An analysis of 18 patients with PRCC treated with various agents including conventional chemotherapy showed no significant responses [73]. A phase II study of 45 patients with mRCC receiving S-1 (an oral formulation combining tegafur, 5-chloro-2,4-dihydropxypyridine, and potassium oxonate) included eight patients with PRCC; ORR was 12.5 % and PFS 5.9 months [76]. Another phase II study of carboplatin and paclitaxel for 16 PRCC patients showed no responses [11]. Finally, a phase II study of single-agent capecitabine in previously untreated metastatic NCCRCC patients enrolled 51 individuals (39 PRCC, 7 CHRCC, 5 CDRCC) most of whom had an intermediate MSKCC risk score and all of whom had a prior nephrectomy and surprisingly high response rates of 2 complete responses (CR), 11 partial responses (PR), and 24 with stable disease (SD) with a median PFS, and OS of 10.1 months and 18.3 months, respectively, was reported [120]. One of the CRs occurred in a PRCC patient.

CHRCC has rarely been evaluated in regard to conventional chemotherapy use, but in 12 individuals in a series of 64 patients with metastatic NCCRCC, none had a response to conventional chemotherapy, although specific agents used were not specified [73]. One CR out of seven CHRCC patients was observed in a series of 51 NCRCC patients treated with single-agent capecitabine [120]. No responses were seen in two patients with CHRCC receiving pemetrexed plus gemcitabine [91].

CDRCC and RMC treatment with conventional chemotherapy has been evaluated in a number of retrospective series and case reports due to histologic similarities between CDRCC and urothelial carcinoma. In a series of 64 patients with NCCRCC, 26 had collecting duct/medullary histology, and one had a 5-month PR to gemcitabine plus cisplatin therapy [73]. A series of 12 patients with CDRCC treated most commonly with methotrexate, vinblastine, doxorubicin, and cisplatin (MVAC) reported only one response lasting 5 months [27]. The largest retrospective series of CDRCC included 81 patients from Japan and included 26 patients with distant metastatic disease. Almost everyone was initially treated surgically, and 17 individuals were treated postoperatively with chemotherapy with only a single response to combination of gemcitabine and carboplatin [118]. In a separate study of nine patients with CDRCC treated with gemcitabine and cisplatin, two CRs were noted [84]. This was followed by a prospective phase II trial of gemcitabine and cisplatin or carboplatin for 23 treatment-naive metastatic CDRCC patients from six French centers with an objective response (CR + PR by Response Evaluation Criteria in Solid Tumors) rate of 26 % (one CR and five PR) and a median PFS and OS of 7.1 months and 10.5 months, respectively [79]. The previously mentioned small phase II trial of pemetrexed and gemcitabine had no activity in three CDRCC patients and in NCCRCC in general [91]. Finally, a series of five patients treated with gemcitabine, platinum, and bevacizumab with bevacizumab maintenance led to three cases of PR, one case of SD, and one CR after a metastasectomy; mPFS was 15.1 months and mOS 27.8 months [81].

Sarcomatoid dedifferentiation leads to an aggressive growth pattern and uniformly poor prognosis. Early studies of MAID (mesna, adriamycin, ifosfamide, dacarbazine), gemcitabine/docetaxel/carboplatin, and doxorubicin-based CYVADIC treatments showed limited success [7, 20, 45, 98]. A phase II trial of doxorubicin and ifosfamide in 25 patients with metastatic sarcomatoid RCC showed no objective response and a median OS of 3.9 months, although case reports of CRs to the same chemotherapy can be found [33, 87]. Eighteen patients with sarcomatoid features treated with a combination of doxorubicin and gemcitabine had two CRs, five PRs, and one SD (ORR = 39 %), and long-term follow-up of four patients found the two complete responders alive 6 and 8 years later and the other two surviving over 3 and 6 years [30, 77]. An ECOG phase II trial of doxorubicin and gemcitabine in 39 patients with sarcomatoid features showed a 16 % response rate (five PRs and one CR) and 10 (26 %) with SD with a median OS of 8.8 months and a median PFS of 3.5 months [41]. It also appeared that those tumors with a higher sarcomatoid component responded better to the chemotherapy. Of note, these trials made no attempt to distinguish the histologic origin or specific renal cancer subtype. A trial of sunitinib with or without gemcitabine for mRCC patients with sarcomatoid features is ongoing (clinicaltrials.gov NCT01164228).

In summary, nucleoside analog therapy appears to have similar low level activity in PRCC and CHRCC as it does in CCRCC, with the caveat that the clinical significance is unknown. CDRCC occasionally responds to agents typically utilized for urothelial cancer, but complete responses are extremely rare, and response rates and response durations appear to be less than in typical urothelial cancer. Reports of complete responses with gemcitabine and doxorubicin in RCC with sarcomatoid differentiation remain intriguing but have not been uniformly replicated, and it has not been possible to determine whether such responses are limited to any specific RCC subtype.

In the randomized discontinuation trial of sorafenib, 15 patients with PRCC were included and two attained a PR [88]. A retrospective analysis examining 41 metastatic PRCC patients treated with sorafenib or sunitinib between 2002 and 2006 at four European and one American center showed a PFS of 7.6 months, with a response rate of 4.8 % (two patients, both treated with sunitinib and achieving a PR) and a PFS of 11.9 vs. 5.1 months in those treated with sunitinib vs. sorafenib (P < 0.001) [15]. Two studies, the EU-ARCCS (Advanced Renal Cell Carcinoma Sorafenib) and the North American ARCCS, examined sorafenib use in a community-wide, expanded-access manner and analyzed its data on NCCRCC [8, 101]. The North American ARCCS trial evaluated 1,891 patients for RECIST response of whom 107 had PRCC. The CR, PR, SD, and CR+PR+SD rates for all patients and PRCC were CR = <1 %, 0 %; PR = 4 %, 3 %; SD = 80 %, 81; CR+PR+SD = 84 %, 84 %, respectively. The EU-ARCCS looked at 1,150 patients, with a 4 % PR rate and a PFS of 6.6 months; a subset analysis of those with papillary features (n = 112) found the PR and PFS to be similar to the clear-cell subset [8]. An expanded-access trial of sunitinib of over 4,300 patients, 68 % of whom had prior cytokine therapy and 3,464 of whom were evaluable, showed an overall objective response (CR +PR) of 17 % (1 % CR) and SD of >3 months in 59 % [39]. In the study, there were 437 (13 %) NCCRCC (not further subclassified) patients able to be evaluated, and an ORR of 11 % (2 CRs) and SD of 57 % was achieved, both comparable to the entire cohort. A phase II sunitinib trial enrolled 25 PRCC patients of whom 74 % had either poor or intermediate risk by MSKCC criteria and no PRs or CRs were seen, and median PFS and OS were only 1.6 and 12.6 months, respectively [110]. Another phase II trial of sunitinib in patients with NCCRCC included 22 evaluable patients, eight of whom had PRCC and achieved a PFS of 5.6 months with OS not reported [70]. A phase II trial of sunitinib in five patients with type 1 PRCC and 23 patients with type 2 PRCC showed a PR in one type 2 patient with 13 SD and 3/5 patients with SD in the type 2 [86]. A multicenter phase II trial from South Korea reported on 31 NCCRCC patients, of whom 22 had PRCC, treated with sunitinib; for the entire cohort, the median PFS was 6.4 months, and for the PRCC patients, the ORR was 36 % [55]. Another multicenter trial from the Turkish Oncology Group reported on 63 NCCRCC patients treated with sunitinib with 88 % having PRCC [126]. Overall study RR and disease control rate was 11.1 % and 63.5 %; median PFS and OS were 7.6 months and 22 months. Finally, two prospective trials of sunitinib versus everolimus in NCCRCC have been completed, and one has been reported in preliminary fashion [109]. This trial was aborted early due to improved overall survival in the sunitinib arm. Notably, 27 of 68 enrolled patients had PRCC, and their estimated median OS in the sunitinib and everolimus arms was 16.6 and 14.9 months, respectively; notably, no objective responses with first-line therapy in PRCC patients were seen. Differences in outcomes seen between the expanded-access trials and the phase II data may be related to lack of central pathologic review in the large trials as well as a selection bias and possibly ethnic differences in response to the phase II trials.