Fax: (214) 648-8786
Reassessment of the 1997 TNM classification system for renal cell carcinoma
A 5-cm T1/T2 cutoff is a better predictor of clinical outcome
Article first published online: 16 OCT 2003
Copyright © 2003 American Cancer Society
Volume 98, Issue 11, pages 2329–2334, 1 December 2003
How to Cite
Elmore, J. M., Kadesky, K. T., Koeneman, K. S. and Sagalowsky, A. I. (2003), Reassessment of the 1997 TNM classification system for renal cell carcinoma. Cancer, 98: 2329–2334. doi: 10.1002/cncr.11806
- Issue published online: 17 NOV 2003
- Article first published online: 16 OCT 2003
- Manuscript Revised: 22 AUG 2003
- Manuscript Accepted: 22 AUG 2003
- Manuscript Received: 11 MAR 2003
- renal cell carcinoma (RCC);
- size cutoff
The 1997 TNM staging classification for renal cell carcinoma (RCC) defined Stage I tumors as organ-confined tumors measuring up to 7 cm in size. The authors evaluated the validity of this cutoff size by assessing the survival of patients with Stage I RCC according to a series of alternative size cutoff values. In addition, the authors determined how these size cutoffs affected the risk of having nonorgan-confined tumors, regional lymph node involvement, and metastatic disease.
A database containing the records of 1324 patients with RCC who underwent open radical nephrectomy between 1960 and 1991 was evaluated. Patients with Stage I disease were stratified by size cutoffs ranging from 2.5 to 7.0 cm in 0.5-cm increments. Five-year disease-specific survival (DSS) rates were estimated using the Kaplan–Meier method. The log-rank test was used to compare survival curves. The survival of patients with tumors smaller than a specified size cutoff was compared with the survival of patients with tumors larger than that cutoff and the most discriminating cutoff was identified. The same size cutoffs were used to compare the incidence of local nonorgan-confined, lymph node-positive, and metastatic disease for all patients with tumors 7.0 cm or smaller.
Of 544 evaluable patients, 351 patients had tumors 7.0 cm or smaller and 233 of these patients had 1997 Stage I (T1N0M0) disease. When patients with 1997 Stage I tumors were separated using the various size cutoffs, survivals were most different using a 5.0-cm cutoff. The 5-year DSS rates for patients with Stage I tumors 5 cm or smaller versus those with tumors measuring 5.1–7 cm were 94.6% versus 79.2% (P = 0.003). Furthermore, the survival of patients with Stage I RCC lesions measuring 5.1–7.0 cm was the same as for patients with 1997 Stage II (T2N0M0) RCC. The difference in probability of having local nonorgan-confined disease was also greatest with a 5.0 cm cutoff value. Nonorgan- confined disease was reported to be present in 16.2% of the patients with tumors smaller than 5.0 cm compared with 36.8% of the patients with tumors measuring 5.1–7.0 cm in size. The difference in the probabilities of having lymph node-positive or metastatic disease did not change significantly using any of the cutoffs, although the probability of both of these increased with increasing tumor size.
Survival and disease recurrence analysis in a large group of patients with RCC who underwent radical nephrectomy showed that the 1997 TNM cutoff of 7.0 cm used to separate Stage I from Stage II disease was too high. A size-related survival difference was found among patients with organ-confined 1997 Stage I disease and a 5.0-cm cutoff best stratified this difference. This finding was in general agreement with the changes made in the 6th edition of the American Joint Committee on Cancer cancer staging manual. Patients with tumors measuring between 5.1 cm and 7.0 cm were found to have the same survival as patients with Stage II disease. Thus, subclassification of T1 into T1a and T1b, as in the 6th edition of the AJCC cancer staging manual, may not be optimal. The 5-cm cutoff also best stratified the risk of developing nonorgan-confined disease. This finding may have an impact on nephron-sparing surgery in selected patients. The findings of the current study, as well as those of others, supported an upper size cutoff of 4–5 cm for patients with Stage I RCC. Cancer 2003. © 2003 American Cancer Society.
The TNM staging system is the most widely used staging system for classifying renal cell carcinoma (RCC). The method categorizes the status of the primary tumor, the regional lymph nodes, and metastases. The classification scheme is modified as needed to reflect changes in clinical presentation, diagnosis, and treatment outcomes. In the 5th edition of the TNM classification of malignant tumors,1 the size cutoff between T1 and T2 organ-confined RCC was changed from 2.5 cm to 7.0 cm. This was based on a review of clinical series, which suggested a similar survival rate for patients with organ-confined RCC up to 7.0 cm in size.1–3 During the past 5 years, several investigators have suggested that the 1997 T1/T2 cutoff is too high and does not best stratify for clinical outcome.4, 5 We performed a retrospective review of a large group of patients with RCC who underwent radical nephrectomy to address the staging criterion for the size cutoff of T1 tumors. The findings are discussed in light of the the 6th edition of the American Joint Committee on Cancer (AJCC) cancer staging manual,6 which was published while the current study was being conducted.
MATERIALS AND METHODS
The medical records of 1324 patients with RCC who underwent open radical nephrectomy between 1960 and 1991 were reviewed. Data were collected by review of surgery, pathology, and radiology reports and clinic notes. Patients with incomplete preoperative clinical stage, pathologic stage, or pathologic tumor size were excluded from analysis. Patients who received any form of adjuvant therapy or who did not have at least one postoperative follow-up note available for review also were excluded from analysis. Tumor size and T classification were obtained from the final pathology report and the N and M components were determined by a combination of clinical and pathologic data. Patients who died of other causes or who were lost to follow-up were censored from analysis.
Disease-specific survival (DSS) based on tumor size and according to 1997 TNM staging criteria was determined using the cohort of evaluable patients. The outcomes of patients with 1997 Stage I disease were analyzed according to tumor size using 0.5-cm increments from 2.5 cm to 7.0 cm. The 5-year DSS rate was calculated using the Kaplan–Meier method and the log-rank test was used to compare survival curves. The survival of patients with tumors smaller than a specified size cutoff was compared with the survival of patients with tumors larger than that cutoff for each 0.5-cm increment. The most discriminating size cutoff between 2.5 cm and 7.0 cm was identified. The 5-year DSS rate of patients with 1997 Stage I RCC larger than the most discriminating cutoff was compared with the survival rate of patients with 1997 Stage II (T2N0M0) disease. Overall stage-specific survival using the most discriminating cutoff to separate Stage I and Stage II RCC was determined and the effect of this change was assessed.
The incidence of local nonorgan-confined, lymph node-positive, and metastatic disease was analyzed using the same size cutoffs for all patients with tumors 7.0 cm or smaller.
Five hundred forty-four patients met all criteria for evaluation. The mean follow-up period was 74 months (range, 2–239 months) for the 281 patients who were alive at last follow-up, including 16 patients with disease recurrence. One hundred forty-five patients died of RCC with a mean time to death of 33.3 months (range, 1–176 months). The remaining 118 patients died of other causes at a mean of 48.3 months (range, 1–216 months). Three hundred thirty-seven patients were male (62%) and 207 were female (38%). The mean age at diagnosis was 58 years (range, 10–87 years). The mean tumor size was 6.7 cm (range, 0.5–32 cm). In this cohort of patients, 233 patients, 89 patients, 125 patients, and 97 patients, respecitively, had 1997 Stage I, Stage II, Stage III, and Stage IV disease. Three hundred fifty-one patients had tumors measuring 7.0 cm or smaller.
The 1997 TNM stage-specific 5-year survival rates were 89.6%, 82.7%, 57.7%, and 18.3%, respectively, for patients with Stage I, Stage II, Stage III, and Stage IV disease (Fig. 1). The 5-year DSS rates of patients with Stage I RCC by 0.5-cm increments between 2.5 cm and 7.0 cm are shown in Table 1. A 5.0-cm cutoff provided the greatest difference in survival (log-rank score = 8.41; P = 0.003). Patients with tumors measuring 5.0 cm or smaller had a 5-year DSS rate of 94.6% whereas patients with tumors measuring 5.1–7.0 cm had a 5-year DSS rate of 79.2%. Figure 2 shows the recalculated survival curves using a 5.0-cm cutoff.
|Size cutoff (cm)||Smaller than cutoff (%)||Larger than cutoff (%)||Log-rank score||P value|
As shown in Figure 3, the survival rate of the 71 patients with tumors 5.1–7.0 cm in size was no different than the survival rate of patients with 1997 T2N0M0 (Stage II) tumors (79.2% vs. 82.7%; P = 0.89). The 5-year DSS rates of patients with Stage I and Stage II RCC using a 5.0-cm cutoff were 94.6% and 81.1%, respectively (P = 0.002). In contrast, the 5-year DSS rates for Stage I and Stage II patients using 1997 size criteria were 89.6 % and 82.7%, respectively. Figure 3 shows that a 5.0-cm T1/T2 cutoff better separates the survival of Stage I from Stage II without altering Stage II survival from that observed with 1997 Stage II criteria.
The probability of having local nonorgan-confined, lymph node-positive, or metastatic disease increased with tumor size. A 5.0-cm cutoff best stratified for organ-confined versus nonorgan-confined tumors. Local nonorgan-confined disease was present in 16.2% of patients with tumors smaller than 5.0 cm and in 36.8% of patients with tumors 5.1–7.0 cm in size (Table 2). The probability of positive lymph nodes or metastatic disease also appeared to increase with tumor size but was not found to be significant for any of the size cutoffs tested (data not shown).
|Size cutoff (cm)||Smaller than cutoff (%)||Larger than cutoff (%)||Difference|
The pathologic stage of RCC remains the most important determinant of disease recurrence and death. In 1997, the TNM classification for RCC was changed to increase the upper size limit for organ-confined T1 tumors from 2.5 cm to 7.0 cm based on several series which suggested similar survival among these patients.1–3 Despite a consensus statement from the AJCC and International Union Against Cancer (UICC), controversy has surrounded the 1997 classification changes.2, 7 Recently, it has been suggested that the increase in the size cutoff between T1 and T2 tumors to 7.0 cm overcorrected the 1987 TNM classification and diminished its ability to stratify survival.4 In the current study, a large cohort of patients who underwent radical nephrectomy over a long period of time were analyzed to reassess the validity of the 1997 T1/T2 classification change.
The impact of tumor size on survival has been well established.5, 8, 9 Delahunt et al.10 provided an elegant review of the impact of tumor size on the outcome of 116 patients with organ-confined RCC. They reported that two methodologies can be used to evaluate the impact of tumor size, i.e., a step function with finite categorizations and a continuous model to calculate proportional hazards. They explained that the risk of death as a result of RCC is continuous for increasing size if sample sizes are large enough. An apparent lack of significance for any given size cutoff most likely reflects an inadequate sample size rather than a critical biologic difference based on size groupings.
Using a step function model with a large group of patients, we detected a DSS-specific survival difference among patients with 1997 T1N0M0 (Stage I) RCC using a 5.0-cm cutoff. Other investigators have reported a survival difference among patients with 1997 Stage I RCC and proposed the subclassification of T1 tumors into T1a and T1b subgroups. Hafez et al.7 suggested a 4.0-cm cutoff to divide T1 tumors into T1a and T1b subgroups based on their finding that patients with RCC smaller than 4.0 cm have a better DSS compared with patients with tumors larger than 4.0 cm. Krejci et al.8 correlated tumor size with cause-specific survival and proposed a 5.0- cm size cutoff to separate T1a and T1b tumors. Igarashi et al.11 reviewed 333 patients with 1997 T1N0M0 RCC. They proposed a cutoff size of 4.0 cm for dividing T1 tumors into T1a and T1b subclasses because this cutoff maximized the difference in disease-free survival. However, these authors did not compare the survival of patients they would have subgrouped as T1b with a group of patients with T2 disease, which would have more definitively justified subclassification.
Shinohara et al.12 reviewed 224 patients with RCC. At a median follow-up of 67 months, metastases occurred in 20% of patients (i.e., 44 of 224) overall. The metastasis-free survival (MFS) period among patients with T1b disease was less than for patients with T1a and was nearly the same as for patients with Stage T2. The authors recommended similar follow-up regimens for patients with both T1b and T2 disease. Reconsideration of the validity of including all organ-confined tumors up to 7 cm as Stage TI may be more to the point.
In a study more similar to the current study, Zisman et al.4 analyzed the survival of 280 patients with 1997 T1 or T2 disease according to 11 size cutoffs between 1 and 10 cm. In their study, a 4.5-cm cutoff had the greatest predictive value for MFS. They also found that the survival of patients with Stage I RCC tumors larger than their proposed cutoff had a survival equal to that of their patients with 1997 Stage II disease. They concluded that inclusion of all organ-confined tumors up to 7 cm in size as T1 is an error and that subclassification within Stage T1 is unwarranted. Considering all of the above studies, a survival advantage exists for patients with organ-confined tumors smaller than 4.0–5.0 cm.
The effect of changing the T1/T2 cutoff from 7.0 cm to 5.0 cm on stage-specific survival was tested with our data. Seventy-one patients were reclassified from Stage I to Stage II as a result of this hypothetic change. The result of this change on overall stage-specific survival was a higher 5-year DSS rate for patients with Stage I RCC compared with patients with 1997 Stage I RCC (94.6% vs. 89.6%). The survival rate of patients with Stage II RCC using a 5.0-cm cutoff remained unchanged from those with 1997 Stage II RCC (81.1% vs. 82.7%). A 5.0-cm cutoff would have resulted in a greater separation of survival curves for Stage I and Stage II disease. A T1/T2 cutoff between 4 cm and 5 cm imparts greater prognostic power to the TNM staging system.
During the time our data were being analyzed, an international consensus conference recommended subclassification of Stage I RCC. As a result, the 6th edition of the AJCC cancer staging manual has reclassified Stage I into T1a and T1b, defined as tumors measuring 4.0 cm or less and as tumors between 4.1 cm and 7.0 cm, respectively.6
Both radical nephrectomy and nephron-sparing surgery (NSS) are standard treatments for selected patients with localized RCC.7, 13–15 The results of NSS in selected patients with RCC are excellent in regard to high patient survival, low risk of local disease recurrence, and low rates of distant disease recurrence. Several studies have reported excellent results after NSS, but to our knowledge the impact of primary tumor size on tumor recurrence and survival has not been determined definitively. In most series, the upper size limit for elective NSS is 4–5 cm.7 The rationale is related primarily to the risk of residual local or multifocal disease. Lymph node positivity and metastatic disease impart a poor prognosis independent of completeness of primary tumor resection. Therefore, in the choice of local therapy for the primary tumor, tumor size should be considered with regard to the risk that a partial rather than a total nephrectomy would leave tumor behind. All the patients in our retrospective study underwent radical nephrectomy. Therefore, the impact of tumor size on suitability for NSS was evaluated only indirectly. We did find that the risk of nonorgan-confined disease increased as a function of tumor size. As shown in Table 2, a 5-cm cutoff was found to be the best discriminator for this parameter. This may be a relative factor for determining suitability of NSS and the accompanying risk of local tumor spillage and disease recurrence.
Other tumor factors beyond size also affect the prognosis of patients with RCC. In some studies, Furhman nuclear grade was found to be correlated with outcome.16–18 In our study, grade was not available for most patients. This resulted from the long time span under review and the finding that the Fuhrman grading system did not gain widespread acceptance until the mid 1980s. However, where available, small tumor size tended to correlate with low Fuhrman grade (data not shown).
Survival analysis of a large group of patients who underwent radical nephrectomy for RCC suggests that the 1997 TNM cutoff of 7.0 cm used to separate Stage I from Stage II tumors was too high. This finding agrees with changes in the 6th edition of the AJCC cancer staging manual. In our study, a size-related survival difference was found among patients with organ-confined current Stage I disease. A 5.0-cm cutoff best stratified this difference. In the current series, patients with Stage I tumors measuring between 5.1 cm and 7.0 cm were found to have the same survival as patients with Stage II disease. Therefore, subclassification into T1a and T1b as in the new AJCC staging criteria may not be optimal. Furthermore, the 5-cm cutoff was the best stratification for the risk of nonorgan-confined disease. This finding may have an impact on the desirability of NSS in selected patients. The current study findings, as well as those of others, support a size cutoff for Stage I RCC between 4 cm and 5 cm.
- 1SobinLH, WittekindCH, editors. TNM classification of malignant tumours, 5th ed. New York: John Wiley & Sons, 1997.
- 6GreenFL, PageDL, FlemingID, et al., editors. AJCC cancer staging manual, 6th ed. New York: Springer-Verlag, 2002.
- 8Grade and size-specific outcomes for stage T1 renal cell carcinoma after nephron sparing surgery [abstract 651]. J Urol. 2001; 165: 158a., , , et al.
- 12Stage-specific followup protocol after radical nephrectomy for renal cell carcinoma: could we follow the T1b patients by a same followup protocol of T1a patients? [abstract 662]. J Urol. 2001; 165: 161a., , .
- 13Clinical experience with nephron-sparing surgery in the presence of a normal contralateral kidney. Semin Urol Oncol. 1995; 13: 288–291., , .
- 14Outcome of laparoscopic radical and open partial nephrectomy for the sporadic 4 cm or less renal tumor with a normal contralateral kidney. J Urol. 2002; 169: 1356–1359., , , .