Differences in presentation characteristics of renal cell carcinoma in the last 25 years: a single center experience

Abstract

Purpose

The aim of this study was to examine trends in clinicopathological characteristics of renal cell carcinoma (RCC) cases at presentation in a single institution over a 25-years period.

Patients and methods

The medical files of 505 patients with histologically confirmed primary RCC from 1981 to 2006 were retrospectively reviewed. Host and tumor characteristics at presentation were compared following stratification by hospitalization period (1981–1990, 1991–2000, and 2001–2006).

Results

Age at presentation did not change significantly over time. The incidentally diagnosed cases increased significantly by time (10.2, 40.5, 62.7%), in proportion to small (<4 cm) tumors (8.6, 17.3, 30.6%), while tumor diameter decreased significantly (8.5 ± 3.8, 7.4 ± 3.5, 5.8 ± 2.9). The rate of organ-confined tumors increased significantly (42.1, 63.6, 68.9%), followed by a less pronounced decrease of metastatic cases (12.3, 8.9, 6.8%).

Conclusions

The evolution of tumor characteristics at presentation in a single institution is apparent within the last 25 years. Major changes were noticed within organ-confined and small tumors and call for familiarization of urologists with nephron-sparing techniques and novel ablation technologies.

Introduction

Renal cell carcinoma (RCC) is the third most common genitourinary tumor, representing about 3% of all cancers, with the highest incidence occuring in the more developed countries (Jemal et al. 2006). Estimates of the worldwide incidence, mortality and prevalence in the GLOBOCAN series of the International Agency for Research on Cancer in the year 2002 resulted in approximately 208,000 new cases and 102,000 deaths (Parkin et al. 2005). The steadily increasing incidence rates in all age groups within the last three decades (Mathew et al. 2002) that resulted in an average increase of 126% since 1950 in the United States (Pantuck et al. 2001), have been initially attributed to a more prevalent use of ultrasonography and CT scan for the evaluation of abdominal complaints (Chow et al. 1999) that led to early, incidental, detection. However, a real increase in RCC incidence at least in central Europe has been documented (Wunderlich et al. 1998).

Typically, RCC predominates in the sixth and seventh decades of life; a median age at diagnosis of 66 years has been reported to remain constant throughout the 1990s among patients registered in the national surveillance, epidemiology, and end results (SEER) database (Ries et al. 2000), while a 5-years increment between 1982 and 1997 along with a twofold increase in the percentage of patients older than 65 years has been reported elsewhere (Luciani et al. 2000). The male-to-female predominance of 3–2 does not show any changes between 1988 and 2002 in the SEER database (Nguyen et al. 2006), although more rapid increases in incidence in females than males have also been observed (Chow et al. 1999).

Incidentally diagnosed RCCs are more likely to be smaller and of lower grade and stage in comparison to the symptomatic ones (Pantuck et al. 2001; Chow et al. 1999). Indeed, the greatest increase in RCC incidence since the 1970s has occurred for localized tumors (Chow et al. 1999). Mean size of resected tumors has decreased significantly (Nguyen et al. 2006; Lee et al. 2000). A most notable down-ward shift has been observed in disease stage at diagnosis; 38% of patients had organ-confined disease in the classic report of Robson (1969) while the corresponding percentage shifted to 78% in the late 1990s (Lee et al. 2000).

Histologically, RCC is currently recognized as a family of cancers derived from the renal tubular epithelium with different morphological features resulting from distinct genetic abnormalities rather than a single entity expressing different histological appearances. The incidence of the various subtypes has been found to correlate with tumor size (Frank et al. 2003) and age at presentation (Taccoen et al. 2007), but is comparable for men and women in the incidentally detected and symptomatic patient groups without changes over time (Gudbjartsson et al. 2005).

We examined the trends in host and tumor characteristics at presentation (sex, age, symptomatic or incidental diagnosis, stage, tumor size and histological type) during the last 25 years by analyzing three time cohorts of RCC patients referred to a single institution, including 1981–1990, 1991–2000 and 2001–2006.

Patients and methods

This retrospective study included all RCC patients treated surgically from 1981 to 2006 in our department, a total of 505 consecutive patients. All patients underwent either partial (37 patients, 7.3%) or radical (468 patients, 92.7%) nephrectomy and had a histologically confirmed diagnosis. There were 339 men and 166 women (male/female ratio 2.04), with a mean age of 62.4 ± 11.8 years (range: 27–89 years). Using clinical information from hospital records, diagnosis was assigned as either incidental (asymptomatic patient; tumor detection from imaging studies performed during regular check-up or for unrelated abdominal complaints) or symptomatic (hematuria, colic or dull flank pain). Tumor size (largest diameter) was estimated from resected specimens.

All histologic findings were reviewed retrospectively according to the 1997 UICC/AJCC classification of renal neoplasms (Storkel et al. 1997), while the Fuhrman four-grade scale (Fuhrman et al. 1982) was used to assess the histopathologic nuclear grade. In all patients, tumor spread was re-estimated using the most recent (2002) revision of the TNM staging system (Sobin and Wittekind 2003).

To investigate trends, the 25-year study period was divided into three intervals, 1981–1990 (143 patients), 1991–2000 (211 patients), and 2001–2006 (151 patients). ANOVA and χ² test for trend analysis in proportions were applied to assess changes of age at presentation and tumor size, or those of rates over time, accordingly (Armitage and Berry 1987).

Results

Patient and tumor characteristics are shown in Table 1, while data stratified by time cohorts are summarized in Table 2. Figure 1 depicts the main trends that were found to be significant. Mean age at diagnosis did not vary by time period (61.8 ± 10.9, 61.9 ± 11.6, and 63.4 ± 12.7, respectively, P = 0.46). Similarly, the proportion of young (≤40 years old) patients was statistically similar among groups (3.4, 3.5, and 6.6%, respectively, P = 0.21). Variation of male to female ratio was also statistically insignificant (1.95, 2.51, 1.62, respectively, P = 0.45). The proportion of incidentally diagnosed cases increased significantly by time (10.2, 40.5, 62.7%, respectively, P < 0.001) while tumor size decreased significantly (8.5 ± 3.8, 7.4 ± 3.5, and 5.8 ± 2.9, respectively, P < 0.001). The proportion of small tumors (maximum diameter <4 cm) increased significantly over time (8.6, 17.3, and 30.6%, P < 0.001) in parallel with a decrease in large (>7 cm) tumors (57.8, 42.3, and 23.3%). A down-ward migration of stage at diagnosis was clearly evident; the incidence of organ-confined tumors (T_1–2N₀M₀) increased impressively from 42.1% to 63.6 and 68.9% (P < 0.0001), and those of locally advanced (T_3–4N₀M₀) and metastatic (T_anyN₊M₀ or T_anyN₀M₊) disease decreased, in a less pronounced mode, from 45.6 to 27.7 to 24.3% and from 12.3 to 8.7 to 6.8%, respectively. Regarding histological types, a slight decrease in the proportion of papillary RCCs and a slight increase of the chromophobe neoplasms was noted, while the rate of clear cell and sarcomatoid RCC remained statistically stable over time. Low grade (Fuhrman’s grade I and II) was assigned to approximately two thirds of tumors during the first two time cohorts while in the last one it was considered in half of the corresponding cases.

Table 1.

Patient and tumor characteristics

Age (range, mean ± SD)	27–89, 62.4 ± 11.8
Young (≤40 years old) patients (%)	4.4%
Male/female ratio	2.05
Incidental diagnosis (%)	51.7%
Tumor size (range, mean ± SD)	0.9–25, 7.2 ± 3.6
Small (<4 cm) tumors (%)	19.2%
Tumor spread (%)
Organ-confined	58.7
Locally advanced	32
Metastatic	9.3
Histology (%)
Clear cell	82.3
Papillary	5.8
Chromophobe	9.7
Sarcomatoid	2.2
Fuhrman’s grade (%)
I/II	58.8
III/IV	41.2

Table 2.

Evolution of host and tumor characteristics in 505 patients over a 25-years period

	Time period
	1981–1990	1991–2000	2001–2006
Number of patients	142	211	152
Age (mean ± SD)	61.8 ± 10.9	61.9 ± 11.6	63.4 ± 12.7
Young (≤40 years old) patients (%)	3.4	3.5	6.6
Male/female ratio	1.95	2.51	1.62
Incidental diagnosis (%)	10.2	40.5	62.7
Tumor size (cm; mean ± SD)	8.5 ± 3.8	7.4 ± 3.5	5.8 ± 2.9
Small (<4 cm) tumors (%)	8.6	17.3	30.6
Tumor spread (%)
Organ-confined	42.1	63.6	68.9
Locally advanced	45.6	27.7	24.3
Metastatic	12.3	8.7	6.8
Histology (%)
Clear cell	87.3	83.9	75.6
Papillary	9.8	5.7	2.0
Chromophobe	0.7	8.5	19.8
Sarcomatoid	2.2	1.9	2.6
Fuhrman’s grade (%)
I/II	62.5	65.7	49.6
III/IV	37.5	34.3	50.4

Fig. 1.

Evolution of the rates of incidental tumors, tumors less than 4 cm, and organ-confined tumors over a 25-year period

Discussion

The global steadily increasing incidence rates in the last three decades seem to affect all age groups (Mathew et al. 2002); hence, a change in the mean age at diagnosis should not have been expected unless an effective screening test had been widely applied. Similarly to the report from the SEER program (Ries et al. 2000), no variation of either mean age at diagnosis or the proportion of young patients over the period studied was found. However, in a study of 1,092 patients, Luciani et al noted an increase of mean age at presentation from 57 years in 1982 to 1983 to 62.6 years in 1996 to 1997, along with a twofold increase in the percentage of patients older than 65 years, from 24.7 to 48.7% that was attributed to a possible higher number of radiologic examinations performed in the elderly population and could result in higher detection rates of asymptomatic tumors (Luciani et al. 2000). A similar tendency to an older age at diagnosis was in a report from a single center in France; median age shifted from 63 years in 1984–1992 to 66 years in 1993–1997 and 65 years in 1998–2003 (Patard et al. 2004).

The predominance of males over females did not change by time in our series, in accordance to other reports (Nguyen et al. 2006; Patard et al. 2004), although incidence in females has been recently reported to increase more rapidly than that in males (Chow et al. 1999). Nowadays, it is widely accepted that the classic triad of symptoms (flank pain, gross hematuria and palpable abdominal mass) should not be longer mentioned as the usual mode of presentation of RCC; as observed in previous reports (Luciani et al. 2000; Patard et al. 2004), the incidental detection rates in our series increased steadily from 1981 to 2006.

As previously published (Nguyen et al. 2006; Lee et al. 2000; Patard et al. 2004), mean tumor size decreased steadily in our series over time as well. Consistent with this, a higher proportion of small tumors (<4 cm) was observed in the latter cohorts of our study as well as in others. This evolution has resulted in increasing application of nephron-sparing surgery or even more conservative treatments (Rini et al. 2006). In agreement with the Memorial Sloan-Kettering Cancer Center data (Lee et al. 2000), we noticed a considerable increase in organ confined tumors and a less pronounced decrease in locally advanced and metastatic tumors. The latter was also observed in France where the percentage of T₃ and metastatic tumors remained stable from 1998 to 2002 (Patard et al. 2004) and in the Netherlands where an increased incidence of low stage RCC and no significant difference in T₃–T₄ tumors was observed in 2004 compared to 1995 (Kummerlin et al. 2006). In a nationwide study in Iceland, changes in stage at presentation were mostly confined to stage I (increase) and stage IV (decrease) tumors and the last 10 years of the 30-year study (Gudbjartsson et al. 2005). However, an increasing incidence of late stage RCC has also been observed (Mathew et al. 2002), while RCC-specific mortality does not appear to have decreased significantly and may be even rising (Murai and Oya 2004).

The most common subtype of RCC is clear cell carcinoma; the odds of having clear cell compared to papillary RCC increase with tumor size (Frank et al. 2003) but the reverse is observed with patient age (Taccoen et al. 2007). However, any possible changes in histological subtype incidence rates as the ones, albeit slight, observed in our study should not be considered as such since they concern a surgical series from a single center. The distribution of histologic subtypes did not change during a 30-year period in the Icelandic nationwide study (Gudbjartsson et al. 2005). As far as tumor grade is concerned, no changes occurred between periods 1 and 2 after which, paradoxically, a decrease in the rate of low grade tumors was observed. Patard et al also expected a shift toward lower grades as it could have been expected from the increase in small incidental tumors (Patard et al. 2004). Conversely, prominent trends towards increase of grade 2 and decrease of grade 3 lesions, especially during the last 10 years of the study, were observed in Iceland (Gudbjartsson et al. 2005).

The evolution of tumor characteristics along with better understanding of disease biology has resulted in major changes in the treatment of localized RCC. Since 1969, the standard of care has been open radical nephrectomy (Robson et al. 1969). However, in nowadays, the standard recommendation for T_1a tumors is open nephron-sparing surgery (Ljungberg et al. 2007), while other minimally invasive renal sparing techniques are considered as the standard treatment in specialized centers (Permpongkosol et al. 2006). Certain patients in our series could have qualified for partial nephrectomy; the fact that instead, they underwent radical nephrectomy can be attributed to either patients’ and/or surgeons’ attitudes and preferences or to the lack of experience in this type of surgery.

The evolution of tumor characteristics at presentation in a single institution is apparent within the last 25 years. Major changes were noticed within organ-confined and small tumors and call for familiarization of urologists with nephron-sparing techniques and novel ablation technologies.