SEARCH

SEARCH BY CITATION

Keywords:

  • renal neoplasm;
  • tumour staging;
  • pathology;
  • renal cell;
  • carcinoma;
  • kidney

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULT
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. REFERENCES

Study Type – Prognosis (case series)

Level of Evidence 4

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

We had known from former studies of RCC that the risks of high grade tumours increased with tumour size and probability of localized tumour decreased with tumour size increasing. Our study had provided large and detailed data about pathologic features of RCC. We also examined the exactly changing of probabilities of different subtypes with diameter increasing and evaluated the effects of hemorrhage, necrosis and cystic degeneration on pathologic subtypes.

OBJECTIVE

  • • 
    To investigate the correlation between tumour size and histopathological characteristics of renal cell carcinoma (RCC).

PATIENTS AND METHODS

  • • 
    A total of 1867 patients who underwent surgical operation between January 2002 and March 2010 due to RCC were included. According to 1997 WHO recommendation about Fuhrman nuclear grading of RCC which criteria we used, tumours were stratified by the largest pathologic diameter into 5 groups, the discrepancy of tumour grade between different groups and whether tumour size could predict histological subtype were analyzed.

RESULTS

  • • 
    The largest diameter (mean ±sd) of G1, G2, and G3 tumours were 3.27 ± 1.46 cm,4.87 ± 2.23 cm, and 7.39 ± 3.11 cm, respectively. The percentage of extracapsular extension tumours in 2 cm or less, 2.1 and 4.0 cm, 4.1 to 7 cm, 7.1 to 10 cm, and more than 10 cm group were 0.5%, 4.3%, 19.8%, 57.9%, and 91.9%, respectively. The distribution of G1 tumours shows a decreasing trend with the diameter becoming larger, while the G3 tumours shows an opposite trend (P < 0.05). Logistic regression analysis predicted that the odds of papillary, chromophobe, and other types vs clear cell decreased with increase in tumour size. If the tumour was complicated with hemorrhage or necrosis, the chance of being chromophobe was higher, while the probability of being papillary and chromophobe decreased when a tumour with cystic degeneration.

CONCLUSION

  • • 
    There was a significant correlation between tumour size and tumour grade and stage; Larger tumours were prone to have higher grade and stage, and the probability of being clear cell carcinoma grew higher as the tumour size increased.

Abbreviations
RCC

renal cell carcinoma

WHO

World Health Organization

CT

computed tomography

MRI

magnetic resonance imaging

NSS

nephron sparing surgery

RN

radical nephrectomy

AJCC

American Joint Committee on Cancer

OR

odds ratio

CI

confidence interval

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULT
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. REFERENCES

With the widespread use of advanced imaging techniques, such as ultrasonography, computed tomography (CT) and magnetic resonance imaging (MRI), the detection rate of renal solid tumours has grown substantially, and an increasing number of smaller renal masses have been detected incidentally [1,2]. At the same time, the treatment options for renal cell carcinoma (RCC) have become more abundant, such as active surveillance [3], cryotherapy [4], and radio frequency ablation [5]; although nephron sparing surgery (NSS) and radical nephrectomy (RN) are still the standard treatment for localized RCC. Treatment should be individualized and based on a combination of patient and tumour characteristics. For certain patients, the non-surgical choice might be more appropriate. Currently, tumour stage, nuclear grade, and pathologic subtype are thought to be the most important prognostic factors; so it has become increasingly important to obtain the histopathological features of the tumour preoperatively. Renal tumour biopsy has proven to be a safe and reliable technic in recent researches, and with low risk of tumour seeding and hemorrhage, but it can not provide detailed pathological information, such as subtype and grade, so is not routinely used in clinical setting [6,7]. Fortunately, tumour size based on imaging coincided highly with pathological tumour size [8], and perhaps more information can be derived from this finding. The objectives of this study were to examine the relationship between tumour size with histological subtype and nuclear grade using a large number of cases from a single institution.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULT
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. REFERENCES

A totoal of 1867 patients receiving NSS or RN for renal cell carcinoma (RCC) in the Peking University First Hospital between January 2002 and March 2010 were analyzed retrospectively. Ultrasonography and plain x-ray of the chest were done routinely, and computerized tomography (CT) or magnetic resonance imaging (MRI) were also performed preoperatively. Bone or brain scan, or any other additional investigations, were only performed as clinically indicated. The clinical and pathological features were collected and include age, gender, tumour size, grade, stage, and histological subtype. With the presence of multiple tumours in the same kidney, the largest dimension was calculated. If there were 2 nuclear grades in the same tumour, we assigned it to the higher grade. All pathological specimens were reviewed by a panel of experienced urological pathologists, and the histological subtype was evaluated according to the 2004 WHO classification. According to 1997 WHO recommendation about nuclear grading of RCC which criteria we used, Fuhrman 1 and Fuhrman 2 were merged to well differentiated tumour (G1), and Fuhrman 3 was considered to be intermediate differentiated tumour (G2), and Fuhrman 4 was considered to be poorly differentiated or undifferentiated tumour (G3). Tumour stage was determined according to the 2010 TNM classification of American Joint Committee on Cancer (AJCC). In this research, tumour size and T classification were obtained from the final pathology reports. The general data of patients are shown in Table 1, and the pathologic characteristics are shown in Table 2.

Table 1. Patients data
No. Patients1867
Median age (range)(years)55 (14 to 88)
Gender (%) 
 Male1303 (69.7)
 Female564 (30.3)
Side (%) 
 Left881 (47.1)
 Right952 (51.0)
 Bilateral34 (1.9)
Multifocal lesion (%) 
 Yes81 (4.3)
 No1786 (95.7)
Table 2. Pathologic characteristics
Nuclear grade (%) 
 G1 (Fuhrman 1 or 2)515 (27.6)
 G2 (Fuhrman 3)1002 (53.7)
 G3 (Fuhrman 4)350 (18.7)
Tumour complicated with 
 Hemorrhage887 (47.4)
 Necrosis697 (37.2)
 Cystic degeneration520 (27.8)
TNM staging 
 T1N0M01306 (70.0)
 T1aN0M0752 (40.3)
 T1bN0M0554 (29.7)
 T2N0M0187 (10.0)
 T2aN0M0152 (8.1)
 T2bN0M035 (1.9)
 T3N0M0254 (13.6)
 T3aN0M0226 (12.1)
 T3bN0M024 (1.3)
 T3cN0M04 (0.2)
 T4N0M07 (0.4)
 TxN1M039 (2.0)
 TxNxM174 (4.0)
 Subtype 
 Clear cell1672 (89.6)
 Papillary72 (3.9)
 Chromophobe56 (3.0)
 Others67 (3.5)

According to the largest diameter, all the tumours were stratified into 5 groups to facilitate data analysis, including 2 cm or less, 2.1 to 4.0 cm, 4.1 to 7 cm, 7.1 to 10 cm, and more than 10 cm in diameter. The discrepancy of tumour grade and stage between the different groups was analyzed by analysis of variance. Logistic regression analysis using tumour diameter as a continuously scaled variable was used to determine whether size predicted histological subtype. Odds ratios (ORs) were calculated to estimate the likelihood of having different subtypes based on diameter. Analyses were performed using SPSS V15.0 software with P value less than 0.05 considered statistically significant.

RESULT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULT
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. REFERENCES

The median patient age at surgery was 55 years (range 14 to 88). Mean ±sd tumour size was 4.9 ± 2.3 cm. There were 1303 male (69.7%) and 564 female (30.3%). Of 1867 pathologic blocks studied, metastatic tumour and multifocal lesion tumour were 50(2.7%) and 81(4.3%), respectively. The most common subtype was clear cell (89.6%), followed by papillary (3.9%), chromophobe (3.0%), and other types (3.5%). The overall nuclear grade distribution was 27.6%, 53.6% and 18.8% for G1, G2 and G3 tumours, respectively. The distribution of the number of G1 tumours shows a decreasing trend with the diameter becoming larger in the different group; while the distribution of G3 tumours shows an opposite trend, and the changes between different groups were mostly significant (P < 0.05) (Table 3). The mean ±sd diameters of G1, G2, and G3 tumours were 3.27 ± 1.46 cm, 4.87 ± 2.23 cm, and 7.39 ± 3.11 cm, respectively; and the discrepancies between every two groups were significant (P < 0.05). The distribution documented that tumours with a larger diameter were prone to have higher nuclear grade. The percentages of extracapsular extension tumours (stage T3 or higher) in 2 cm or less, 2.1 and 4.0 cm, 4.1 to 7 cm, 7.1 to 10 cm, and more than 10 cm group were 0.5%, 4.3%, 19.8%, 57.9%, and 91.9%, respectively (Table 4). This meant that the probability of it not being a localized tumour increased as the diameter became larger.

Table 3. Nuclear grade distribution according to tumour size
Tumour sizeG 1 (Fuhrman 1 or 2)G 2 (Fuhrman 3)G 3 (Fuhrman 4)Total
2 cm or less115 (59.9%)74 (38.5%)3 (1.6%)192 (100%)
2.1 to 4.0 cm276 (38.1%)398 (55.0%)50 (6.9%)724 (100%)
4.1 to 7 cm109 (17.4%)377 (60.3%)139 (22.3%)625 (100%)
7.1 to 10 cm15 (6.3%)127 (52.9%)98 (40.8%)240 (100%)
>10 cm0 (0%)26 (30.2%)60 (69.8%)86 (100%)
Total515 (27.6%)1002 (53.7%)350 (18.7%)1867 (100%)
Table 4. Tumour status distribution according to tumour size
Tumour sizeLocalized (T1&T2)Locally advanced (T3&T4)Metastatic (N1 or M1)Total
2 cm or less191 (99.5%)1 (0.5%)0 (0%)192 (100%)
2.1 to 4.0 cm693 (95.7%)17 (2.3%)14 (2.0%)724 (100%)
4.1 to 7 cm501 (80.2%)86 (13.8%)38 (6.0%)625 (100%)
7.1 to 10 cm101 (42.1%)93 (38.8%)46 (19.1%)240 (100%)
> 10 cm7 (8.1%)64 (74.4%)15 (17.5%)86 (100%)
Total1493 (80.0%)261 (14.0%)113 (6.0%)1867 (100%)

Logistic regression analysis predicted that the odds of papillary, chromophobe, and other types vs clear cell decreased as the tumour size increased. Using clear cell carcinoma as a reference, the OR of papillary, chromophobe, and other types was 0.806, 0.935, and 0.514, respectively (P < 0.05, 95%CI: 0.786 to 0.823, 0.841 to 1.015, 0.412 to 0.617), when the tumour size increased by 1 cm.

We also detected the effects of hemorrhage, necrosis and cystic degeneration on pathologic subtypes. If the tumour was complicated with hemorrhage or necrosis, the chance of being chromophobe was higher, and the OR was 1.433 and 2.828 (P < 0.05, 95%CI: 1.225 to 1.833; 1.353 to 5.911), respectively. If cystic degeneration was absent, the probability of being papillary and chromophobe RCC increased, and the OR was 32.428 and 4.417 (P < 0.05, 95%CI: 4.363 to 241.009; 1.760 to 11.089), respectively.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULT
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. REFERENCES

The increased use of modern imaging technique, such as ultrasonagraphy, CT and MRI, has led to an increase of renal solid tumour incidence, and the weight of asymptomatic small tumours has become heavier and heavier [1,2]. Up to 82%-88% of renal tumours were RCC [9,10]. Presently, RN or NSS was considered to be the standard treatment for clinically localized RCC; but it was not always the optimal choice in some special circumstances, especially in patients with worse medical conditions and performance status. Active surveillance [3], cryotherapy [4], and radio frequency ablation [5] had proven to be effective treatment options in selected patients. Therefore, doctors must balance the risk and the benefit, and individualize the treatment option. The best clinical decision should be based on a combination of patient condition, treatment availability, and tumour characteristics. Currently, tumour stage, nuclear grade, and pathologic subtype were thought to be the most important prognostic factors [11,12]. Therefore, it was helpful to know the histopathological features of the tumour preoperatively. Renal tumour biopsy had proven to be safe and reliable technic in recent researches, and with low risk of tumour seeding and hemorrhage, but it could not provide detailed pathological information, such as subtype and grade [6,7,13]. In these studies, they mainly evaluated sufficiency and accuracy of core needle biopsy in small renal tumour (≤4 cm), so it was unsuitable to use these conclusions in big tumours, and they didn't provide completely information of pathologic subtype and grade in most cases [6,7]. Although the reported histopathologic accuracy of biopsy was extremely high compared with final surgical pathological finding, the proportion of patients who underwent surgical procedure after biopsy was very low in the whole cohorts [6,7]. So it was not the time for routinely use of tumour biopsy currently. Fortunately, tumour size based on imaging highly coincided with pathological tumour size [8]. So, we could obtain the tumour size from CT or MRI image.

Our study indicates that the distribution of the number of G1 tumours shows a decreasing trend with the tumour size becoming larger in different groups; while the distribution of G3 tumours shows an opposite trend (Table 3); which was consistent with the results of other researchers [10,14–16]. The mean ±sd diameters of G1, G2, and G3 tumours were 3.27 ± 1.46 cm, 4.87 ± 2.23 cm, and 7.39 ± 3.11 cm, respectively. Our study documented that tumours with larger diameters were prone to have higher nuclear grade, and became more aggressive potentially, which coincided with the results of former studies [8–10,14–16,17]. In the study of Thompson et al., 16% of tumours less than 3 cm and 59% of tumours 7 cm or greater were high grade [10]. In our study,the percents of high grade tumour in group less than 4 cm and 7 cm or greater were 15.1% and 45.1%, respectively. The change of high grade tumour distribution in our study was not as sharp as Thompson's finding. We also noted that there were only 3 cases were Grade 3 tumour in 192 tumours equal to 2 cm or less. A total of 6.9% of renal tumours 2.1 to 4.0 cm in diameter and 22.3% of those of 4.1 to 7 cm in diameter were of Fuhrman Grade 3. As a result, 4 cm was a key point in the dramatic change of tumour aggressiveness reflected by the difference in nuclear grade. The study also documented the feasibility of choosing 4 cm as the cutoff for an active surveillance end point, and as a cutoff for the choice of a proven procedure used in today's clinical practice. Remzi et al. reported that 3 cm was the cutoff for the changing point in tumour aggressiveness [17]. Such differences may be due to its relatively small sample and different method of grouping. Recently, several large studies about chromophobe RCC had challenged the use of Fuhrman grading system in this subtype [18–20], but there was not a criteria being widely accepted by most urologic pathologists. And in Przybycin et al's research of chromophobe, they also validated that there was a significant relationship between tumour size and occurence of disease specific events, so the bigger tumour meant worse prognosis, and should have higher tumour grade, which coincided with our finding in this aspect [20]. Although the application of Fuhrman to chromophobe RCC was controversial, but it was still the most widely used grading system which we used in current study.

We also found that the percentages of extracapsular extension tumours (stage of T3 or higher) rose gradually with the increase in tumour size. Of 192 patients, only 1 patient had a locally advanced tumour in the 2 cm or less tumour size group, so the probability of no localized tumour was merely 0.5%. In the 2.1 to 4.0 cm tumour size group, the percentage of localized tumours was as high as 95.7%. Additionally, the percentage of G3 tumours was only 5.8% in the group with tumour size less than or equal to 4 cm. These results provided a pathological basis for the use of relatively conservative strategies in the treatment of small renal masses in patients considered poor surgical candidates.

The results of our research showed that as the tumour size increased, there was a significantly greater probability of it being clear cell carcinoma. In other words, the chance of no clear cell carcinoma became smaller and smaller. These results confirmed several studies published [9,21,22]. But in the research of Frank et al. and Rothman et al., they reported a reverse tendency of chromophobe distribution to our study [15,16]. In Frank's study, no chromophobe were less than 1 cm compared to 7.0% of tumours 7 cm or greater [16]. It maybe due to the small sample of our chromophobe tumours, which could not reflect intrinsic characteristics of the tumour, and it also probably duo to the ethnological difference. Further more, different from the studies cited above, we also examined the exactly changing of probabilities of different subtypes with each 1 cm increasing in our research, and it might be more helpful in clinical decision making than arbitrary fixed cutoff of tumour size. An important feature of our study, which made it differ from others was that we had also evaluated the effects of hemorrhage, necrosis and cystic degeneration on pathologic subtypes. If the tumour was complicated with hemorrhage and/or necrosis, the chance of being chromophobe was higher. If cystic degeneration was observed, there was a decrease in the probability of being papillary and chromophobe RCC. These features would increase the accuracy of our judgement about tomor pathologic characteristics.

There were several limitations in this study that merit discussion. Our data represent a retrospective review of findings at a single center. Consequently, the results of our research are subject to the inherent biases of a retrospective study. More importantly, our data represent a group of patients who were treated surgically and had pathologic findings. Therefore, patients who were not treated surgically, as in the case of widespread metastases or unresectable tumours, were not included in our study.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULT
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. REFERENCES

There was a significant correlation between tumour size and tumour grade, stage, and invasiveness; and larger malignant tumours were prone to have higher grade, higher stage, and invasiveness. The probability of detection of particular histological subtypes varies with an increase in tumour size, and can also be affected by the existence of hemorrhage, necrosis and cystic degeneration. With an increase in tumour size, the probability of it being clear cell carcinoma grew higher, especially the tumour with cystic degeneration. However, if the tumour evidenced hemorrhage and necrosis, the probability of it being clear cell carcinoma decreased. These findings are important when counseling patients on treatment options for renal cell carcinoma.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULT
  6. DISCUSSION
  7. CONCLUSION
  8. CONFLICT OF INTEREST
  9. REFERENCES