Changes in the stage and surgical management of renal tumours during 1995–2005: an analysis of the Dutch national histopathology registry

Authors


M. Pilar Laguna, Academic Medical Centre, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands. e-mail: m.p.lagunapes@amc.uva.nl

Abstract

OBJECTIVE

To evaluate changes in the pathological characteristics, stage of primary renal tumours and their surgical management in the Netherlands during the period 1995–2005.

METHODS

Extracts from the records of all patients who had surgery for primary renal tumours in the Netherlands during the period 1995–2005 were reviewed. Data were collected from PALGA, the nationwide network and archive of histocytopathology. The 2002 Tumour-Node-Metastasis and the three-tier Fuhrman grade were used for staging and grading.

RESULTS

In all there were 12 471 operations for primary renal masses during the study period. The incidence of surgically removed renal cancers increased from 6.2 in 1995 to 7.5 cases per 100 000 inhabitants (P = 0.005) in 2005. The mean (sd, median) age of the patients was 63.3 (11.9, 65.0) years, with a male-to-female ratio of 3:2. The mean (sd) tumour size of malignant tumours decreased from 7.3 (3.6) to 6.9 (3.7) cm (P = 0.301). The percentage of benign removed tumours remained relatively stable (P = 0.056), with a mean of 5.4% of all resected tumours. There was an increase of grade 1 tumours; the incidence of T1 tumours increased from 36.6% to 44.2%, and advanced tumours decreased from 46.4% to 33.7%, respectively. The percentage of nephron-sparing surgery increased from 3.5% in 1995 to 10.1% (P = 0.003) in 2005, mainly in the T1a tumours.

CONCLUSIONS

During the last decade there was an increase in the incidence of surgically treated renal tumours in the Netherlands. Tumours with favourable histopathological characteristics, low stage and grade, accounted for most of this increase. The percentage of surgically removed benign tumours remained stable. The use of nephron-sparing surgery increased during the last decade, especially in T1a tumours.

Abbreviations
NCR

Netherlands Cancer Registry

NSS

nephron-sparing surgery

SEER

Surveillance, Epidemiology and End Results.

INTRODUCTION

The GLOBOCAN series of the International Agency for Research on Cancer shows a worldwide increasing incidence of renal cancer, with the highest incidence rates in North America [1]. In the USA, data from the Surveillance, Epidemiology and End Results (SEER) of the National Cancer Institute shows that tumours of ≤4.0 cm account for most of the increasing incidence [2]. Parallel to this rise in incidence, the number of surgically removed benign tumours is also increasing [3]; clinical surveys showed that up to 14% of all resected tumours are benign [4,5].

However, in the USA the increase of small tumours did not lead to a significant difference in stage presentation in the last two decades [6]. The lack of a decrease in advanced stages implies that other factors than incidental discovery might contribute to the increasing incidence of renal cancer, suggesting a true increase in both localized and advanced renal tumours.

Data from the SEER has served well to characterize the practice patterns in the surgical management of small renal masses in the USA [7]. Although still relatively uncommon, there was a trend to the more frequent use of nephron- sparing surgery (NSS) during the last two decades for tumours of <7 cm, more evident for tumours of ≤4 cm. Concurrently adjusted rates of renal surgery increased during the last two decades, most notably for those small tumours [2].

Data from The Netherlands Cancer Registry (NCR) show that the incidence of kidney cancer has remained stable in the Netherlands since 1989 [8], in contrast with data from other countries [1,9,10]. The NCR, established in 1989, compiles data collected by the regional comprehensive cancer centres, and it is affiliated to all hospitals, radiological centres and pathology laboratories. All 64 pathological departments in the country archive their pathological reports centrally at PALGA (Pathologisch Anatomical Landelijk Geautomatiseerd Archief; Pathological Anatomy National Automated Archive; http://www.palga.nl), a nationwide histocytopathology network and archive [11]. Laboratories of origin generate the excerpts and transfer them to the central data bank.

While data from The NCR provide a clear picture of the incidence and mortality of kidney cancer in the Netherlands, it does not explore other pathological characteristics, e.g. stage and type of surgical management of kidney tumours. The aim of the current study was to determine possible changes in the pathological stage and trends in the surgical management of primary kidney tumours treated surgically during the last decade in the Netherlands.

METHODS

Abstracts of pathology reports (excerpts) from patients who had surgery for a primary renal tumour during 1995–2005 were reviewed. The excerpts contained the first 300 characters of the conclusion of the pathology reports until the year 2000. From that year on, this was extended to the first 1000 characters. Data were abstracted from PALGA using the procedural terminology derived from the Systematized Nomenclature of Medicine from the College of American Pathologists. The following terms were retrieved: kidney, oncocytoma, angiomyolipoma, metanephric adenoma, chromophobe adenoma, eosinophilic adenoma, adenopapilloma, adenoma and all malignancies with the exception for carcinoma in situ. Tumour biopsies, primary renal pelvic tumours, ureteric tumours and metastases were excluded.

The following variables were evaluated for the consecutive years: age, gender, tumour side, tumour size, tumour type (malignant/benign), stage, grade and operation type (radical/partial nephrectomy). For patients with multiple tumours in the same kidney, only the largest tumour was included for analysis. The Fuhrman grade was converted into a three-tier grading system for simplification of the analysis, merging grade 1 with grade 2 [12]. Pathological stage was determined from the available data in the excerpts and according to the 2002 TNM classification. Surgical management was regarded as a radical nephrectomy, apart from excerpts where a partial nephrectomy or enucleation was explicitly mentioned.

A database was created checked for consistency. The variables age and male-to-female ratio had no missing data. The nature of the missing data was explored by comparing mean and median age and sex ratio within cases with and without missing data for the other variables (tumour diameter, stage and grade). In the absence of significant differences between cases with and without missing data, cases with missing data can be consider as ‘completely at random’. Differences in age between cases with and without missing data were calculated using the unpaired (two-sample) t-test with P < 0.05 being considered to indicate statistical significance. To exclude the completeness of the pathological report (number of characters) as a source of missing data, the variables age and male-to-female ratio between cases with and without missing data between 1995–2000 and 2001–2005 were further compared.

Because age and tumour size were normally distributed, the unpaired (two-sample) t-test or one-way anova was used to evaluate differences of the mean age or tumour diameter between the years. The number of cases of the variables ‘benign tumour’, and ‘surgical management’ were divided by the total number of cases per year for trend analysis, using the chi-square test (linear-by-linear association). For all analyses, P < 0.05 was considered to indicate statistical significance.

The incidence of surgically treated tumours was calculated by determining the number of cases each year, dividing by the corresponding population count provided by the Statistics Netherlands (Central Bureau of Statistics) and multiplying by 100 000 [13]. The incidence and mortality of kidney cancer in the Netherlands was provided by The NCR [8].

RESULTS

In all, 12 477 primary renal tumours were operated on in 1995–2005; a bilateral synchronic nephrectomy was used in six patients with no specification of the individual tumours, and thus these six bilateral nephrectomies were regarded as six individual nephrectomies, and consequently 12 471 cases were included for analysis.

Figure 1 shows the incidence, mortality and the incidence of the surgically treated malignant renal tumours in the Netherlands during the past decade. The incidence of surgically treated malignant renal tumours increased from 6.2 to 7.5 cases per 100 000 people (linear-by-linear association, P = 0.005). The incidence of surgically treated benign renal tumours varied, at 0.3–0.6 cases per 100 000 people. The incidence of all (malignant and benign) surgically treated renal tumours increased from 6.5 to 8.0 cases per 100 000 (linear-by-linear association, P= 0.005). There were multiple tumours in 680 of 12 471 specimens, for an incidence of multiplicity of 5.5%.

Figure 1.

The incidence, mortality and the incidence of surgically treated renal tumours in the Netherlands over the past decade.

Table 1 shows the missing data from 1995 to 2005, for tumour side, size, grade and stage. The mean (sd) age for cases with (2886/12 471) and without (9585/12 471) missing data for tumour diameter was 63.3 (12.5) and 63.3 (11.8) years, respectively (t-test, P = 0.901). Of the cases with (1819/11 800) and without (9981/11 800) missing data for stage, the mean age was 63.1 (12.5) and 63.4 (11.7) years, respectively (t-test, P = 0.314). The cases with (5578/11 800) and without (6222/11 800) missing data for grade had a mean age of 63.5 (11.9) and 63.2 (11.7) years, respectively (t-test, P = 0.161). All groups had a median age of 65.0 years and a male-to-female ratio of 3:2. The comparisons of the demographic data between the population with and without missing data between 1995–200 and 2001–2005 also showed no statistically significant differences in mean age or sex ratio. These values confirmed that missing data were distributed completely at random.

Table 1.  The number and percentage of records with missing data 1995–2005
YearTotal cases, nTumour side*Tumour size*GradeStage§
  • *

    Percentages calculated by the number of cases with missing data divided by the total number of cases;

  • †three-grade or the Fuhrman grading system;

  • ‡percentages calculated by the number of cases with missing data divided by the number of malignant and inconclusive cases;

  • §

    §Stage was assigned by the investigators according 2002 TNM.

1995 1 006 97 (9.6) 412 (41.0) 605 (63.5) 265 (27.8)
1996 1 034 112 (10.8) 363 (35.1) 570 (57.8) 234 (23.7)
1997 1 076 94 (8.7) 347 (32.2) 635 (62.1) 208 (20.3)
1998 1 077 68 (6.3) 279 (25.9) 565 (55.0) 180 (17.5)
1999 1 013 76 (7.5) 239 (23.6) 458 (47.6) 169 (17.5)
2000 1 085 81 (7.5) 252 (23.2)  511 (49.8) 134 (13.1)
2001  1 113 52 (4.7) 202 (18.1) 506 (47.4) 138 (12.9)
2002  1 216 67 (5.5) 220 (18.1) 474 (41.2) 150 (13.0)
2003 1 249 68 (5.4) 208 (16.7) 470 (40.1)  121 (10.3)
2004 1 300 80 (6.2) 198 (15.2) 406 (33.7)  116 (9.6)
2005 1 302 78 (6.0) 166 (12.7) 378 (30.8) 104 (8.5)
1995–2005 12 471873 (7.0)2886 (23.1)5578 (47.3)1819 (15.4)

The mean (sd, median) age for all cases (12 471) was 63.3 (11.9, 65.0) years, with no statistically significant differences in age (one-way anova, P = 0.068) between the years. The male-to-female ratio was 3:2 for all years. Tumours were equally frequent in the right and left kidney between the years (Pearson chi-square test, P = 0.617).

Table 2 shows the number and percentage of benign and malignant tumours during the study period. The distinction between benign and malignant tumour (i.e. inconclusive) was not possible in 0.2% of cases. The percentage of benign lesions operated remained stable; 5.4% of all resected tumours were benign (linear-by-linear association, P = 0.056). The mean (sd) diameter was 5.2 (3.3) and 7.0 (3.7) cm for benign and malignant lesions, respectively (t-test, P < 0.001). The mean size of malignant tumours decreased statistically insignificantly from 7.3 (3.6) in 1995 to 6.9 (3.7) cm in 2005 (one-way anova, P = 0.301).

Table 2. 
The number and percentage of benign and malignant tumours from 1995 to 2005
YearN casesBenign tumourMalignant and inconclusive tumours
1995 1 006 53 (5.3)   953 (94. 7)
1996 1 034 47 (4.5)   987 (95.5)
1997 1 076 53 (4.9) 1 023 (95.1)
1998 1 077 50 (4.6) 1 027 (95.4)
1999 1 013 50 (4.9)   963 (95.1)
2000 1 085 59 (5.4) 1 026 (94.6)
2001 1  113 45 (4.0) 1 068 (96.0)
2002 1  216 66 (5.4)  1 150 (94.6)
2003 1 249 76 (6.1)  1 173 (93.9)
2004 1 300 96 (7.4) 1 204 (92.6)
2005 1 302 76 (5.8) 1 226 (94.2)
1995–200512 471671 (5.4) 11 800 (94.4)

Table 3 shows the number and percentage of the graded malignant tumours; there was an increase in grade 1 tumours, from 39.4% to 56.6%, between 1995 and 2005. Subsequently, there was a decrease from 39.9% to 29.3% for grade 2 tumours, and from 20.7% to 14.1% of grade 3 tumours, respectively.

Table 3.  The number and percentage of grade distribution between 1995 and 2005
YearNumber of cases*Grade 1Grade 2Grade 3
  • *

    Number of cases with tumour grade stated in the excerpts; grading according to the three-tiers Fuhrman grade.

1995 348 137 (39.4) 139 (39.9) 72 (20.7)
1996 417  181 (43.4) 150 (36.0) 86 (20.6)
1997 386 163 (42.2) 137 (35.5) 86 (22.3)
1998 458 203 (44.3) 185 (40.4) 70 (15.3)
1999 504 247 (49.0) 173 (34.3) 84 (16.7)
2000 514 264 (51.4) 173 (33.7) 77 (15.0)
2001 558 300 (53.8) 180 (32.3) 78 (14.0)
2002 672 359 (53.4) 228 (33.9) 85 (12.6)
2003 703 412 (58.6) 193 (27.5) 98 (13.9)
2004 798 429 (53.8) 250 (31.3) 119 (14.9)
2005 841 476 (56.6) 246 (29.3) 119 (14.1)
1995–20056199 3171 (51.2)2054 (33.1)974 (15.7)

Tumour stage is shown in Table 4; there was an increase in T1 tumours from 36.6% in 1995 to 44.2% in 2005. The percentage of advanced (T3/T4) tumours treated surgically decreased from 46.4% to 33.7% during the study period, while the percentage of T2 remained stable.

Table 4.  The number and percentage of tumour stages between 1995 and 2005
YearT1aT1bT2T3aT3bT3*T4Total
  • *

    T3-staged tumour not otherwise specified;

  • †total number of cases where the stage could be determined.

1995  111 (16.1)  141 (20.5)  117 (17.0) 127 (18.5)  161 (23.4)12 (1.6) 19 (2.8) 688
1996  113 (15.0) 169 (22.4) 136 (18.1) 125 (16.6)  191 (25.4) 4 (0.4) 15 (2.0) 753
1997 126 (15.5) 209 (25.6) 137 (16.8)  131 (16.1) 190 (23.3)12 (1.5) 10 (1.2) 815
1998  161 (19.0) 214 (25.3) 135 (15.9) 128 (15.1) 182 (21.5) 11 (1.3) 16 (1.9) 847
1999 169 (21.3) 186 (23.4) 136 (17.1) 124 (15.6) 163 (20.5) 6 (0.6) 10 (1.3) 794
2000 172 (19.3) 217 (24.3) 150 (16.8) 144 (16.1)  191 (21.4) 4 (0.3) 14 (1.6) 892
2001 201 (21.6) 228 (24.5) 165 (17.7) 122 (13.1) 188 (20.2)12 (1.2) 14 (1.5) 930
2002 213 (21.3) 239 (23.9) 188 (18.8) 139 (13.9) 196 (19.6) 6 (0.6) 19 (1.9)1000
2003 239 (22.7) 273 (26.0) 187 (17.8) 145 (13.8) 177 (16.8)19 (0.6) 12 (1.1)1052
2004 232 (21.3) 273 (25.1) 195 (17.9) 148 (13.6) 210 (19.3) 9 (0.6) 21 (1.9)1088
2005 274 (24.4) 261 (23.3) 209 (18.6) 147 (13.1) 199 (17.7) 8 (0.6) 24 (2.1) 1122
1995–2005 2011 (20.1)2410 (24.1)1755 (17.6)1480 (14.8)2048 (20.5)93 (0.8)174 (1.7)9981

Table 5 shows the type and percentage of surgery in the last decade. The absolute number and percentage of NSS increased steadily; the percentage of T1a tumours operated by NSS increased from 9.0% in 1995 to 27.0% in 2005 (linear-by–linear association, P = 0.013). The corresponding values for T1b tumours were 1.4% and 5.4%, respectively (P = 0.006).

Table 5. 
The number and percentage of NSS and radical nephrectomies, 1995–2005
YearRNNSSTotal
1995   971 (96.5) 35 (3.5) 1 006
1996   993 (96.0) 41 (4.0) 1 034
1997  1 041 (96.7) 35 (3.3) 1 076
1998 1 032 (95.8) 45 (4.2) 1 077
1999   948 (93.6) 65 (6.4) 1 013
2000 1 027 (94.7) 58 (5.3) 1 085
2001 1 035 (93.0) 78 (7.0)  1 113
2002  1 106 (91.0) 110 (9.0)  1 216
2003  1 143 (91.5)106 (8.5) 1 249
2004  1 193 (91.8)107 (8.2) 1 300
2005 1 170 (89.9)132 (10.1) 1 302
1995–2005 11 659 (93.5)812 (6.5) 12 471

Information on nodal status was available in 1683/11 800 (14.3%) of the malignant lesions, and from that, 453/1683 (26.9%) of the lymph nodes were positive for malignancy. Positive surgical margins were reported in 9.2% (1145/12 471) of cases. From the 812 tumours removed by NSS, 160 (19.7%) had positive tumour margins, of which 16 (10%) were benign cases and 144 (90%) malignant tumours.

DISCUSSION

In the present study we assessed the changes in pathological stage and surgical policies for kidney tumours in a European country, and this could be regarded as the European counterpart on the SEER data from the USA. Demographic data for patients treated surgically for renal tumours in the Netherlands remained stable during the last decade. The mean and median age, and the male-female ratio were similar to those reported for the USA population throughout the 15 years [14].

As observed in data from SEER, there was also a steady decrease in the size of the operated tumours in the Netherlands [14]. While the mean diameter decrease (from 6.7 cm in 1988 to 5.9 cm in 2002) alone did not modify staging in USA, the statistically insignificant 4 mm of mean diameter decrease of the operated tumours might have had clinical implications in the present series. According to the 2002 TNM, in the absence of fat or vascular invasion, the 4 mm variation in mean tumour size implies a shift to a more favourable stage. However, the stage shift confirmed by an increasing incidence of T1 tumours at the end of the period was mainly at the expenses of a decrease in advanced tumours (T3/T4), with the incidence of T2 remaining stable. These results could confirm not only the contribution of abdominal imaging to the incidental diagnosis of renal masses [15], but also the potential benefit of the incidental discovery. Conversely, an increase in both localized and advanced tumours was reported in the USA [6]. Discordance with the present data can be explained by the different nature of the registries. While SEER is a cancer registry, including a few States but considered as representative of the USA population [16], our source, PALGA, only includes surgical specimens. Although data on the true incidence of kidney tumours cannot be inferred from PALGA, it represents a good view of the staging changes at surgery and on surgical patterns during the 11-year period.

During the study period there was an increase in grade 1 tumours and a decrease in grade 2 and 3 tumours. This is consistent with the increase in stage T1 tumours, as it is known that decreasing tumour size is accompanied by increasing odds of a low-grade tumour [5,17,18]. However, the overall results on grade changes should be viewed with caution, as there was a high percentage of missing data for this variable. As grade is not a component of the TNM staging system, the lack of information did not hinder the staging process in the present series. With the increasing acknowledgement of the prognostic significance of grade, increased reporting of this variable would be expected during the study period; indeed, grade reporting improved from 36.5% in 1995 to 69.2% in 2005.

The absolute number of NSS increased during the study period, with values of 3.5% for 1995 and 10.1% for 2005, suggesting under-use of the technique and confirming data from other studies [7]. A more detailed analysis showed that most of this increase was in T1a tumours. Among the different reasons for this marginal increase in the number of partial nephrectomies are the potential fear of synchronous ipsilateral tumour, but most likely the widespread use of laparoscopic radical nephrectomy and the inherent difficulties of a partial nephrectomy. As other ablative techniques were introduced late in the Netherlands, with only 27 kidney tumours treated by cryoablation between 2003 and 2005, and even fewer treated by radiofrequency ablation, the inclusion of these treatments in our data do not significantly modify the results for NSS. As other techniques develop, and NSS techniques are implemented, changes in the surgical management, with a trend to a more conservative approach, can be expected in the future.

Data from The NCR show a stable incidence of and mortality from kidney tumours between 1989 and 2003 [19]. There was a slight increase in the incidence of surgically treated kidney cancers, from 6.2 per 100 000 people in 1995 to 7.5 per 100 000 in 2005. Whether this more aggressive policy, at the expense of increasing surgical removal of small tumours, will result in increased survival remains unanswered, as the response of an intervention for cancer can take a long time before definitive conclusions can be reached.

As we were aware of the limitations of the registry used for this study, we tried to analyse them thoroughly. The data from the PALGA registry was prospectively collected but retrospectively reviewed for this study. However, it is a well organised and valid registry which provides data for the NCR.

Although the presence of missing data was unavoidable, all the excerpts included the mandatory data relating to demography (age and sex). These demographic characteristics were compared between those with and those without missing data for each of the variables with missing data (tumour size, grade and stage). As there were no statistical differences in demographic data for the two groups for each of the different variables evaluated, missing data were considered completely at random, which leads to unbiased parameter estimates, but slightly underpowers the present study.

The presence of missing data might reflect the truncation of the reports, due to the limitations of the registry that restricted the report to the first 300 words of the conclusion until 2000 and expanded it to 1000 words from then on. However, an improvement in the data reported by pathologists over the years is most likely the main contributor to the better definition of tumour characteristics. An increased awareness of the prognostic risk factors and the adoption of different classifications and grading and staging systems could have resulted in considerably less missing data, with an inflexion point around 2001–2002. Furthermore, as the information contained in the excerpts was the same as in the conclusion of the definitive pathological report, the most important pathological characteristics contributing to the stage of the tumour were probably included.

The demographic data between populations with and without missing data between 1995–2000 and 2001–2005 were further compared, resulting in a statistically insignificant difference, and reinforcing the complete randomness of the missing data. Consequently, it can be safely concluded that the results of the present study reflect variations in the stage of the operated kidney tumours and the surgical policies of the last decade in the Netherlands.

In conclusion, there was an increase in the incidence of surgically treated renal tumours during the last decade in the Netherlands. Tumours with favourable histopathological characteristics, e.g. low stage and grade, accounted for most of the increase. There was a shift in stage at the expense of a decrease in advanced stages and this was unrelated to the small decrease in mean tumour diameter. The percentage of surgically removed benign tumours remained stable. Although there was some under-use of NSS in the last decade, there is an increasing trend for NSS at the expense of T1a tumours.

ACKNOWLEDGEMENTS

To the PALGA foundation for providing the data for this study and to the Department of Clinical Epidemiology and Biostatistics of the AMC for their support in the analysis.

Conflict of interest

None declared.

Ancillary