Unclassified renal cell carcinoma: a report of 56 cases

Authors


Antonio Lopez-Beltran, Unit of Anatomical Pathology, Department of Surgery and Pathology, University of Cordoba, Faculty of Medicine, Avda. Menendez Pidal s/n, E-14004 Cordoba. Spain. e-mail: em1lobea@uco.es

Abstract

Study Type – Therapy (case series)

Level of Evidence 4

What's known on the subject? and What does the study add?

Unclassified RCC represents 0.7–5.7% of renal tumours. Limited reported data from two series suggests that unclassified RCC is an aggressive form of RCC, mainly because most cases are at an advanced stage at presentation, but overall and cancer-specific survival were not significantly different between unclassified and clear-cell RCC in an additional series of 38 patients.

Our study of 56 cases of unclassified RCC describes the pathological features that can be applied to predict prognosis on a daily basis. In particular nuclear grade, TNM classification, tumour coagulative necrosis, tumour size, microvascular invasion and 2004 WHO histotype are independent predictors of disease-free and cancer-specific survival.

OBJECTIVE

  • • To evaluate the clinicopathological features and outcomes of 56 patients with unclassified renal cell carcinoma (RCC) meeting 2004 World Health Organization diagnostic criteria.

PATIENTS AND METHODS

  • • Urological pathology files of the participating institutions were reviewed and cases of unclassified RCC that met the inclusion criteria were retrieved.
  • • Nuclear grade, pT status, tumour size, regional lymph node involvement, distant metastases, coagulative tumour necrosis, mucin and sarcomatoid differentiation were evaluated in radical nephrectomy or nephron-sparing specimens.
  • • Significant factors in univariate analysis were then assessed by a multivariate analysis of independent prognostic factors using Cox proportional hazard regression analysis.

RESULTS

  • • Fifty-six cases met the histological criteria for unclassified RCC. Thirty-four (61%) cases were categorized as unrecognizable cell type (mean overall survival 47 months; median 36 months), 20 (36%) as composites of recognized types (mean overall survival 36 months; median 26 months), and two (4%) (mean survival 16 months; median 16 months) as pure sarcomatoid morphology without recognizable epithelial elements.
  • • Cox multivariate analysis showed nuclear grade (P= 0.020), stage (P < 0.001), tumour coagulative necrosis (P= 0.018), tumour size (P < 0.001), microvascular invasion (P < 0.001) and tumour histotype (P= 0.028) to be independent predictors of disease-free survival, with tumour size being the most significant (hazard ratio [HR] 9.068, 95% confidence interval [CI] 3.231–25.453).
  • • Nuclear grade (P= 0.026), stage (P < 0.001), tumour coagulative necrosis (P < 0.001), tumour size (P= 0.044), microvascular invasion (P < 0.001), tumour recurrence after surgery (P < 0.001) and tumour histotype (P= 0.056) were independent predictors of cancer-specific survival, with tumour recurrence after surgery being the most significant (HR 14.713, 95% CI 5.329–40.622).

CONCLUSION

  • • The prognosis of patients with unclassified RCC seems to be related to clinicopathological features known to be relevant in common forms of RCC.
Abbreviations
UCT

unrecognizable cell type

CRT

composite of recognized types

HR

hazard ratio.

INTRODUCTION

Unclassified RCC, a subtype of kidney cancer introduced in 1997, is a diagnostic category for those tumours that do not fit into any of the four known common categories of RCC: clear-cell, papillary, chromophobe, or collecting duct carcinoma [1–8]. According to the 2004 WHO classification of urological tumours [3], the features that might prompt assignment of a tumour to this category include: composites of recognized subtypes, mucin production, rare mixture of epithelial and stromal elements, pure sarcomatoid morphology without recognizable epithelial elements, and unrecognizable cell types (Table 1). The frequency of this type of RCC is quite variable, accounting for 0.7–5.7% of all RCCs in some larger contemporary series [1–17]. Studies of unclassified RCC suggest that this category is associated with unfavourable histological features and aggressive behaviour [3–8,12,13,15–17].

Table 1.  Features that support assignment to unclassified RCC category in the current WHO classification of kidney tumours
• Composites of recognized types
• Pure sarcomatoid morphology without recognizable epithelial elements
• Mucin production
• Rare mixtures of epithelial and stromal elements
• Unrecognizable cell types

Tumours assigned to this category tend to be heterogeneous and most often are of high grade [3,4]. Although the effect of histological subtype on the prognosis of patients with RCC has been investigated among the more common histological subtypes, few series have evaluated the significance of the unclassified RCC category on RCC-specific outcomes [3–17].

The aim of the present study was to describe the clinicopathological features and specific outcomes of 56 patients with unclassified RCC meeting 2004 WHO diagnostic criteria.

PATIENTS AND METHODS

PATIENT SELECTION

For this retrospective, observational study, the sample series was obtained through an international and multi-institutional search including centres from the USA and Europe. Urological pathology files of the participating institutions were reviewed and cases of unclassified RCC meeting the 2004 WHO diagnostic criteria were included (Table 1). At least two dedicated uropathologists, blinded, reviewed each case for inclusion in the study. The first case was diagnosed in February 1993 and the last in December 2006. Follow-up ended in June 2010. In the larger series of 14 cases at one of the participating institutions (the University of Cordoba), unclassified RCC represented 4.8% out of 290 consecutive cases. The study received Institutional Review Board approval from the University of Cordoba.

CLINICOPATHOLOGICAL FEATURES

Clinical and demographic information obtained through the hospital records included clinical manifestations and follow-up. The present study included patientswho underwent open radical nephrectomy or nephron-sparing surgery (Table 2). Most were patients with metastases, who received immunotherapy and were treated with recombinant interleukin-2 based regimens; none of the cases received targeted therapies. The clinical features included age at nephrectomy, gender and symptoms at presentation. Patients that were defined as symptomatic presented with: palpable flank or abdominal mass, discomfort, gross haematuria, acute onset varicocele, or constitutional symptoms including rash, sweats, weight loss, fatigue, early satiety and anorexia. Patients that were defined asymptomatic had no symptoms and the diagnosis was incidental or had symptoms not related to RCC.

Table 2.  Patient and tumour characteristics in 56 cases of unclassified RCC
Variable 
  1. NED, no evidence of disease; DFD: died from disease. *TNM classification 2010.

Mean (sd; range) age, years63 (13.29; 29–91)
Age, n (%) 
 ≤40 years3 (5.5)
 41–50 years7 (12.5)
 51–60 years14 (25)
 >60 years32 (57)
Gender, n (%) 
 Male31 (55)
 Female25 (45)
Clinical presentation, n (%) 
 Incidental22 (39)
 Symptomatic34 (61)
Surgery, n (%) 
 Radical42 (75)
 Nephron-sparing14 (25)
Nuclear grade, n (%) 
 Grade 11 (2)
 Grade 224 (43)
 Grade 318 (32)
 Grade 413 (23)
*pT stage, n (%) 
 T1a23 (41)
 T1b10 (18)
 T2a6 (11)
 T2b1 (2)
 T3a7 (12)
 T3b7 (12)
 T3c2 (4)
*pN status, n (%) 
 Nx15 (27)
 N029 (52)
 N112 (21)
Necrosis, n (%) 
 Yes18 (32)
 No38 (68)
Microvascular invasion, n (%)
 Yes18 (32)
 No38 (68)
Tumour size, n (%) 
 ≤4 cm23 (41)
 5–6 cm13 (23)
 ≥7 cm20 (36)
Tumour recurrence after surgery, n (%)
 Yes20 (36)
 No36 (64)
Histotype, n (%) 
 UCT34 (61)
 CRT20 (36)
 Pure sarcomatoid2 (4)
Survival status, n (%) 
 NED31 (55)
 DFD25 (45)

Gross features were obtained in each case from pathology reports and photographs of the specimens. Tumour size was measured according to the longest diameter and tumour stage was based on the 2010 revision of the TNM system. Pathological evaluation consisted of reassessing all original histological material by at least two dedicated uropathologists blinded to the clinical outcomes of the patients. Nuclear grade was determined according to the Fuhrman system. The presence of coagulative tumour necrosis, mucin and sarcomatoid differentiation was evaluated in all cases. Immunohistochemistry was performed to confirm renal epithelial origin and exclude true sarcoma of the kidney and secondary tumours in selected cases. Radiological and operative surgical information and pathology reports were used to assess the extent of disease (Table 2). The endpoints of the study were disease-free and cancer-specific survival determined from the date of nephrectomy to the date of death or the last follow-up. The newly described RCC subtypes, namely tubulocystic RCC and clear-cell papillary carcinoma were excluded from the study [3,4,8].

STATISTICAL METHODS

To determine variables significant for survival (disease-free and cancer-specific) a univariate analysis was carried out using the Kaplan–Meier method. Comparisons of groups were performed using the log-rank test to assess the significance of the Kaplan–Meier curves. Features that were significant in the univariate analyses were then used in a multivariate analysis of independent prognostic factors using Cox proportional hazard regression analysis, and the relative risk with 95% CI was calculated. The statistical analysis was performed using SPSS (version 15.0) for Windows software (SPSS Inc, Chicago, IL). A P value of ≤0.05 was considered to indicate statistical significance.

RESULTS

CLINICOPATHOLOGICAL FEATURES

Table 2 shows the clinical and pathological variables of the 56 cases that met the histological criteria of unclassified RCC according to the current WHO classification of kidney tumours. The mean (range) age at diagnosis was 63 (29–91) years. Unclassified RCC was found incidentally during evaluation for unrelated conditions in 39% of the cases. Twenty-five percent of patients received nephron-sparing surgery. All cases were unilateral at diagnosis. One patient had multifocal tumours.

Five (9%) patients developed lung metastases. On follow-up, mean (sd) overall survival of 56 patients was 42 (38.33) months (median 29 months; range 2–179 months), 31 patients (55%) were alive with no evidence of disease (mean [sd] 60 [41.99] months; median 43 months; range 15–179 months), and 25 died from the disease (mean [sd] 19 [13.82] months; median, 17 months; range, 2–59 months).

Thirty-four (61%) cases were classified as unrecognizable cell type (UCT) (mean survival 47 months; median 36 months), 20 (36%) as composites of recognized types (CRT) (mean survival 36 months; median 26 months), and two (4%) as pure sarcomatoid morphology without recognizable epithelial elements (mean survival 16 months; median 16 months [Fig. 1]). Mucin was present in four cases (range 5–20% of tumour) and because of other concurrent histological features, these cases were included as UCT (two cases) or CRT (two cases). None of the 56 unclassified RCC cases was categorized as rare mixture of epithelial and stromal elements. Tumour coagulative necrosis was present in 18 (32%) cases. Microvascular invasion was seen in 18% of cases.

Figure 1.

Histological features in unclassified RCC: A, mixed histology with clear and cromophobe cell types; B, pure sarcomatoid carcinoma histology; and C, unrecognizable cell type histology with prominent lipoid change.

PATIENT SURVIVAL ANALYSIS

Univariate survival analysis showed that nuclear grade (P < 0.001), tumour classification (TNM 2010 revision; P < 0.001), microvascular invasion (P < 0.001), tumour coagulative necrosis (P < 0.001), tumour size (P < 0.001), and 2004 WHO tumour histotype (UCT vs. CRT vs. pure sarcomatoid; P= 0.025), were related to disease-free survival (Fig. 2). No significant differences in patients' age (P= 0.339), gender (P= 0.854) and tumour histotype (UCT vs. CRT; P= 0.141) were found in disease-free survival analysis (Table 3). Univariate survival analysis also showed that nuclear grade (P= 0.001), tumour stage (TNM 2010 revision; P < 0.001), microvascular invasion (P < 0.001), tumour coagulative necrosis (P < 0.001), tumour size (P < 0.001), tumour recurrence after surgery (P < 0.001), and 2004 WHO tumour histotype (UCT vs. CRT vs. pure sarcomatoid; P= 0.042) were related to cancer-specific survival (Fig. 3). No significant differences in patients' age (P= 0.457), gender (P= 0.667) and tumour histotype (UCT vs. CRT; P= 0.275) were found in cancer-specific survival analysis (Table 3).

Figure 2.

Kaplan–Meier plots showing differences in disease-free survival of 56 patients with unclassified RCC according to the current WHO criteria. (A) nuclear grade; (B) pT classification; (C) necrosis; (D) microvascular invasion; (E) tumour size; (F) histological subtype. (disease-free survival in months)

Table 3.  Disease-free and cancer-specific univariate survival analysis in 56 cases of unclassified RCC according to the Kaplan–Meier method and the log-rank test
Variable, N= 56 n (%)Disease-free survival, nlog-rank P Cancer-specific survival (%)log-rank P
  1. *Excluding two pure sarcomatoid carcinoma cases (N= 54). **TNM classification 2010.

Age, years  3.360.339 2.600.457
 ≤403 (5.5)3/3  3/3  
 41–507 (12.5)3/7  3/7  
 51–6014 (25)10/14  8/14  
 >6032 (57)20/32  17/32  
Gender  0.030.854 0.190.667
 Male31 (55)20/31  17/31  
 Female25 (45)16/25  15/25  
Nuclear grade  19.81<0.001 16.160.001
 Grade 11 (2)1/1  1/1  
 Grade 224 (43)17/24  16/24  
 Grade 318 (32)15/18  13/18  
 Grade 413 (23)3/13  2/13  
**pTstage  63.96<0.001 59.22<0.001
 T1a23 (41)22/23  22/23  
 T1b10 (18)8/10  6/10  
 T2a6 (11)3/6  1/6  
 T2b1 (2)0  0  
 T3a7 (12)3/7  2/7  
 T3b7 (12)0  0  
 T3c2 (4)0  0  
Microvascular invasion  42.04<0.001 35.89<0.001
 Yes18 (32)2/18  1/18  
 No38 (68)34/38  30/38  
Necrosis  23.23<0.001 32.92<0.001
 Yes18 (32)5/18  2/18  
 No38 (68)31/38  30/38  
Tumour size  25.09<0.001 42.35<0.001
 ≤4 cm23 (41)22/23  22/23  
 5–6 cm13 (23)10/13  8/13  
 ≥7 cm20 (36)4/20  2/20  
Tumour recurrence after surgery   44.71<0.001
 Yes20 (36)  1/20  
 No36 (64)  30/36  
Histotype  7.340.025 6.310.042
 UCT34 (60)24/34  21/34  
 CRT20 (36)12/20  11/20  
 Pure Sarcomatoid2 (4)0  0  
*Histotype  2.170.141 1.190.276
 UCT34 (63)24/34  21/34  
 CRT20 (37)12/20  9/20  
Figure 3.

Kaplan–Meier plots showing differences in cancer-specific survival, in months, of 56 patients with unclassified RCC according to the current WHO criteria. A, Nuclear grade; B, pT classification; C, necrosis; D, microvascular invasion; E, tumour size; F, tumour recurrence; G, histological subtype.

Cox multivariate analysis (Table 4) showed that nuclear grade (P= 0.020), stage (TNM 2010; P < 0.001), tumour coagulative necrosis (P= 0.018), tumour size (P < 0.001), microvascular invasion (P < 0.001) and tumour histotype (P= 0.028) were independent predictors of disease-free survival, with tumour size being the most significant (hazard ratio [HR] 9.068, 95% CI 3.231-25.453). Also, nuclear grade (P= 0.026), stage (TNM 2010; P < 0.001), tumour coagulative necrosis (P < 0.001), tumour size (P < 0.044), microvascular invasion (P < 0.001), tumour recurrence after surgery (P < 0.001) and tumour histotype (P= 0.056) were independent predictors of cancer-specific survival, with tumour recurrence after surgery being the most significant (HR 14.713, 95% CI 5.329–40.622)

Table 4.  Cox multivariate analysis for independent predictors of disease-free and cancer-specific survival in 56 patients with unclassified RCC
Variable P HR95% CI
LowerUpper
  1. *TNM classification 2010; **2004 WHO.

Disease-free survival     
 Nuclear grade0.0201.9711.1113.494
 *pT stage<0.0012.1121.5522.874
 Necrosis0.0183.4251.2309.533
 Tumour size<0.0019.0683.23125.453
 Microvascular invasion<0.0014.4570.88222.514
 **Histotype0.0282.1961.0874.437
Cancer-specific survival     
 Nuclear grade0.0261.7821.0732.960
 *pT stage<0.0011.9081.5172.401
 Necrosis<0.0017.8243.29618.576
 Tumour size0.0442.2490.7596.666
 Microvascular invasion<0.0019.6983.92923.939
 Tumour recurrence after surgery<0.00114.7135.32940.622
 **Histotype0.0561.947.9843.852

DISCUSSION

According to the current WHO classification of kidney cancer [3],the following features define unclassified RCC: composites of recognized types; pure sarcomatoid morphology without recognizable epithelial elements; mucin production; rare mixtures of epithelial and stromal elements; and unrecognizable cell types. Unclassified RCC represents 0.7–5.7% of renal tumours, and the majority are of high grade and stage at presentation [3–17]. One study [6] found that the proportion of unclassified RCC did not change with increasing decade of life. Although features that might place a carcinoma in this category are defined by the current WHO classification of kidney cancer, reported data suggest differences on how these criteria are applied, resulting in a more heterogeneous group than would be expected.

Limited reported data suggest that unclassified RCC is an aggressive form of RCC, mainly because most cases are at an advanced stage at presentation [1–17]. Results based on three relatively large series support this suggestion. Zisman et al. [6] reported 31 patients with unclassified RCC and compared the prognostic outcome with 317 matched patients with clear-cell RCC. The incidence of unclassified RCC in that series was 2.9%, slightly lower than in the present series, where the incidence was 4.8% out of 290 consecutive cases (based on 14 cases provided by one of the participating institutions). These differences are probably related to more restrictive inclusion criteria since Zisman et al. [6] included only pure anaplastic, pure sarcomatoid or mixtures of both types of RCC. On follow-up, 94% of the patients with unclassified tumours had metastatic disease vs 83% with matched patients with clear-cell RCC. Unclassified carcinomas were larger tumours, had increased risk of adrenal gland involvement (25% of cases), direct invasion to adjacent organs (42%), bone (52%), regional (52%) and non-regional lymph node (41%) metastases [6]. Unclassified histology was a significant indicator for poor prognosis on multivariate analysis [6]. The median survival in patients with unclassified RCC was 4.3 months [6]. Karakiewicz et al. [7] reviewed 85 cases of unclassified RCC to compare cancer-specific mortality in patients with unclassified RCC vs. clear-cell RCC after nephrectomy. Of the patients with unclassified RCC, 80% had Fuhrman grades III or IV vs. 37.8% for clear-cell RCC [7]. In addition, 36.5% of patients with unclassified RCC had pathologically confirmed nodal metastases vs. 8.6% with clear-cell RCC. Finally, 54.1% of patients with unclassified RCC had distant metastases at the time of nephrectomy vs. 16.8% with clear-cell RCC. Despite these differences in the overall analyses, after matching for tumour characteristics, the unclassified RCC specific mortality rate was 1.6 times higher in matched analyses and 1.7 times higher in multivariate analyses [7]. The studies by Zisman et al. [6]and Karakiewicz et al. [7] both provide evidence supporting the hypothesis that unclassified RCC is more aggressive than matched clear-cell RCC.

By constrast, Crispen et al. [14] more recently reported on 38 unclassified RCCs and showed that patients were more likely to have regional lymph node involvement, higher grade, tumour necrosis and sarcomatoid differentiation as compared to patients with clear-cell RCC, but overall and cancer-specific survival were not significantly different between patients with unclassified RCC and those with clear-cell RCC in either the unmatched or matched cohorts. In their study, Crispen et al. [14] included RCC cases with oncoytic/eosinophilic features, pure sarcomatoid/undifferentiated epithelial tumours, or tumours with pure rhabdoid differentiation. Again, the different results of the Crispen et al. [14] study can be explained by the selection criteria, which were different from those used in the series by Zisman et al. [6] and Karakiewicz et al. [7]. The latter two studies concluded that unclassified RCC is associated with distinct and highly aggressive biological behaviour, and poor clinical outcome; likewise, Crispen et al. [14] concluded that although unclassified RCC is more likely to present with advanced clinicopathological features compared with clear-cell RCC, no statistically significant differences in outcome were noted after adjusting for these features in a matched analysis.

The present study of 56 cases of unclassified RCC is different from previous large series mainly because our cases were selected by dedicated uropathologists following the diagnostic criteria reported by the 2004 WHO [3], and also because emphasis was placed on pathological features as prognostic predictors of disease-free and cancer-specific survival. The present study is original as it describes the pathological features that can be applied to predict prognosis on a daily basis. We showed that nuclear grade, TNM classification, tumour coagulative necrosis, tumour size, microvascular invasion [18,19] and 2004 WHO histotype are independent predictors of disease-free and cancer-specific survival in our cohort of 56 unclassified RCC cases, with tumour size and recurrence as main predictors; therefore, these histological markers could be included in nomograms to assess unclassified RCC in clinical practice [20–22]. With regard to 2004 WHO histotype, it is notable that pure sarcomatoid morphology portended a dismal prognosis in the present series.

Pathological studies on unclassified RCC are limited, with most reporting on small cases series [3,4]. Bruder et al. [10] reported 13 cases of unclassified RCC in a series of 41 cases of RCC in patients younger than 22 years. These tumours could not be assigned to types specified by the current WHO classification [3].Of these, 10 cases were grouped as unclassified not otherwise specified, including a unique RCC with prominently vacuolated cytoplasm and WT1 expression. Three additional unclassified RCCs occurred in combination with nephroblastoma. The study by Bruder et al. [10] adds new morphological features which, together with the WHO criteria, might place a carcinoma in the category of unclassified RCC. Pathologists must be aware of the spectrum of histological findings within unclassified RCC to ensure accurate diagnosis [3,4,8,10]. As recently suggested by Reuter [4], using the current WHO criteria, unclassified RCC represents a histologically and clinically heterogeneous category of tumours that does not fit neatly into any of the other well-defined categories.

In conclusion, unclassified RCC according to 2004 WHO diagnostic criteria represents a heterogeneous group of tumours whose prognosis seems to be related to histological features known to be relevant in more common forms of RCC.

ACKNOWLEDGEMENT

Supported in part by the grant SAF2007-64942, Ministry of Education. Madrid, Spain.

CONFLICT OF INTEREST

None declared.

Ancillary