Incidence and Pattern of Recurrence after Surgical Resection in Organ-Confined Renal Cell Carcinoma

Abstract

Purpose

To evaluate the incidence and pattern of recurrence after surgery in patients with organ-confined renal cell carcinoma (RCC) to establish an appropriate follow-up plan.

Materials and Methods

In this retrospective study, we evaluated data from 2960 patients who underwent radical or partial nephrectomy for stage 1 or 2 RCC. We investigated the location of first recurrence and recurrence-free survival (RFS) by plotting Kaplan–Meier curves and analyzed the associated variables using Cox regression analysis.

Results

During a median follow-up of 59 months, the 10-year RFS rates were 94.5%, 75.0%, and 57.9%, for T1a, T1b, and T2 RCC, respectively. A total of 211 patients experienced recurrence: 67 after 5 years, and 14 after 10 years. The most common sites of the first recurrence were the lungs, bones, and contralateral kidneys. Male sex, older age, higher pathologic T stage, higher nuclear grade, clear-cell RCC, and presence of differentiation were associated with recurrence. Among patients followed up for more than 60 months, higher pathologic T stage and grade, as well as clear cell RCC were predictors of RFS.

Conclusion

Late recurrence after surgery is common in patients with organ-confined RCC, with recurrence occurring even after 10 years. Consequently, long-term follow-up, of perhaps 10 years or more, including imaging studies of the abdomen, lungs, and bone, should be considered for the early detection of recurrence.

Graphical Abstract

INTRODUCTION

Renal cell carcinoma (RCC) is the 14th most common cancer, accounting for approximately 2%–3% of all carcinomas. The incidence of RCC is increasing gradually, possibly due to the widespread use of imaging studies, such as ultrasound and computed tomography (CT).¹ An increase in the use of imaging has also led to an increase in the rate of the early diagnosis of kidney cancer. Since the 5-year survival rate of stage 1 and 2 RCC is approximately 90% and 65%–75%, respectively, the mortality rate due to RCC has gradually decreased since 1990; in 2018, RCC mortality was reported to be 2.9/100000.²

However, approximately 20%–30% of localized RCCs recur within 5 years despite surgical treatment.³,⁴ Organs prone to recurrence include the lungs, bones, liver, lymph nodes, adrenal glands, and brain.⁵ Several studies have reported that delayed recurrence after 5 years is not uncommon.⁵,⁶,⁷,⁸ Additionally, the recurrence site in relapsed RCC is treatable, improving the survival rate when treated.⁹,¹⁰ This suggests that early detection of recurrent lesions is important. Several guidelines recommend mandatory follow-up testing for up to 5 years. However, follow-up observation is not mandatory after 5 years. Particularly in high-risk patients, it is recommended to be performed according to the patient’s preference or the surgeon’s judgment.¹¹,¹² Since patients are lost to follow-up after 5 years and miss out on appropriate treatment due to multiple metastases, appropriate follow-up is necessary even after 5 years.⁸ Therefore, it is crucial to establish a long-term follow-up plan for patients with stage 1 and stage 2 RCC, wherein relatively several patients survive even after 5 years.

In the absence of clear guidelines for long-term follow-up, we aimed to evaluate the incidence and pattern of recurrence after surgery in patients with stage 1 and 2 RCC to establish an appropriate follow-up plan.

MATERIALS AND METHODS

Patient selection

We reviewed the medical records of patients who underwent partial or radical nephrectomy (RN) for pT1 and pT2 RCC between January 2006 and December 2019 at Severance Hospital, Seoul, Korea. The exclusion criteria for patients were as follows: 1) any lymph node metastasis or distant metastasis, 2) diagnosis of cystic RCC, 3) synchronous multiple RCC, 4) patients who were not followed up after surgery, and 5) patients diagnosed with hereditary RCC.

We investigated the sex, age, operative type, operative extent, side of the tumor, tumor size, pathological T stage, pathological type, nuclear grade, and type of differentiation for all patients. Additionally, we investigated whether there was recurrence, as well as the site of the first recurrence and the time until the recurrence.

Follow-up

After surgical resection, imaging studies, such as abdominopelvic CT, magnetic resonance imaging (MRI), chest CT, bone scan, positron emission tomography, and CT, were performed every 3–6 months for the first year, 6–12 months for the next 2 years, and 12–24 months for the next 2 years. When suspicious lesions were identified, imaging tests of the relevant area were performed every 2–3 months. Additionally, a biopsy was performed if necessary. Patients who were followed up regularly for 5 years after nephrectomy underwent imaging studies every 12–24 months. Imaging tests during the follow-up period were performed at the discretion of the surgeons.

Statistical analyses

Kaplan–Meier curves were plotted to depict recurrence-free survival (RFS). Multivariate Cox regression models were constructed to determine the variables associated with RFS. All tests were two-sided, and p<0.05 was considered statistically significant. All statistical analyses were performed using STATA version 15.1 (StataCorp LLC, College Station, TX, USA).

Good clinical practice protocols

This study was approved by the Institutional Review Board of Yongin Severance Hospital (2023-0266-001). This study was performed in accordance with the applicable laws and regulations, good clinical practice, and ethical principles described in the Declaration of Helsinki.

RESULTS

Clinicopathologic characteristics

Table 1 shows the clinical and pathological characteristics of the patients included in this study. Of the 2960 patients, 2022 (68.3%) were male and 938 (31.7%) were female. The median patient age was 56 years. In 1518 (51.3%) patients, the tumor was located on the right side, and in the remaining 1442 (48.7%), it was located on the left side. In total, 1298 (43.9%), 488 (16.5%), and 1174 (39.7%) patients underwent open, laparoscopic, and robotic surgery, respectively. Additionally, 972 (32.8%) patients underwent RN, and 1988 (67.2%) underwent partial nephrectomy (PN). The median tumor size was 2.9 cm. The tumor stage was T1a, T1b, T2a, and T2b in 2137 (72.2%), 611 (20.6%), 180 (6.1%), and 32 (1.1%) patients, respectively. Of the 2492 (84.2%) clear cell RCCs reviewed, 1624 (54.9%) were grades 1 or 2, and 1207 (40.7%) were grades 3 or 4.

Table 1. Clinicopathological Characteristics of Patients (n=2960).

Characteristics		Value
Sex
	Male	2022 (68.3)
	Female	938 (31.7)
Age (yr)		56 (47–64)
Surgical method
	Open	1298 (43.9)
	Laparoscopic	488 (16.5)
	Robotic	1174 (39.7)
Operative extent
	Radical nephrectomy	972 (32.8)
	Partial nephrectomy	1988 (67.2)
Operative side
	Right	1518 (51.3)
	Left	1442 (48.7)
Tumor size (cm)		2.9 (1.9–4.3)
Pathologic T stage
	pT1a	2137 (72.2)
	pT1b	611 (20.6)
	pT2a	180 (6.1)
	pT2b	32 (1.1)
Pathological type
	Clear cell	2492 (84.2)
	Papillary	195 (6.6)
	Chromophobe	223 (7.5)
	Other	50 (1.7)
Nuclear grade
	Grade 1 and 2	1624 (54.9)
	Grade 3 and 4	1207 (40.7)
	Unknown	129 (4.4)
Differentiation
	No	2920 (98.6)
	Sarcomatoid	29 (1.0)
	Other	11 (0.4)
Follow up duration (month)		59 (46–90)

Data are presented as median (interquartile range) or n (%).

Incidence and pattern of recurrence

During a median follow-up period of 59 months, 211 patients experienced recurrence. Among them, 78 (3.7%), 80 (13.1%), and 53 (19.4%) patients had T1a, T1b, and T2 tumors, respectively. The 10-year RFS rates were 94.5%, 75.0%, and 57.9%, respectively. The mean risk of recurrence after 5 years was 0.5% per year for T1a, 3.0% for T1b, and 3.2% for T2 tumors. The mean risk of recurrence after 5 years was 0.5% per year for T1a, 3.0% for T1b, and 3.2% for T2 tumors (Fig. 1).

Fig. 1. Kaplan–Meier curve of recurrence-free survival according to pathological T stage.

Recurrence occurred in 144 patients within 5 years, in 53 patients between 5–10 years, and in 14 patients after 10 years. In the T1a group, 12 patients experienced recurrence after 10 years. A total of 262 organs were involved in the first recurrence, with the lungs, bones, and contralateral kidneys being the most common. Among patients who relapsed after 5 years, 36 had an interval of >18 months between the test at which recurrence was diagnosed and the test performed before that, and 31 had an interval of <18 months. Patients with long intervals between follow-ups had a high rate of multiorgan recurrence (25.8% vs. 6.5%, p=0.028). This seemed to indicate that if the interval between follow-ups was long, there was a possibility that the opportunity for metastasectomy would be less (Table 2).

Table 2. Recurrence According to Pathologic T Stage and First Site of Recurrence; Classification According to the Postoperative Period.

			Total	0–5 years	5–10 years	10 years–
Patients experience recurrence			211	144 (68.2)	53 (25.1)	14 (6.7)
	Pathologic T stage
		pT1a	78	51 (65.4)	15 (19.2)	12 (15.4)
		pT1b	80	52 (65.0)	27 (33.8)	1 (1.2)
		pT2	53	41 (77.3)	11 (20.8)	1 (1.9)
First recurrence site			262	175 (66.8)	66 (25.2)	21 (8.0)
	Lung		81	57 (70.4)	20 (24.7)	4 (4.9)
	Bone		40	32 (80.0)	5 (12.5)	3 (7.5)
	Contralateral kidney		35	19 (54.3)	13 (37.1)	3 (8.6)
	Lymph node		25	18 (72.0)	6 (24.0)	1 (4.0)
	Ipsilateral kidney		17	10 (58.8)	3 (17.6)	4 (23.6)
	Pancreas		11	4 (36.4)	5 (45.5)	2 (18.1)
	Operative site		10	7 (70.0)	3 (30.0)	0 (0)
	Liver		7	5 (71.4)	2 (28.6)	0 (0)
	Brain		7	4 (57.1)	2 (28.6)	1 (14.3)
	Adrenal gland		6	6 (100)	0 (0)	0 (0)
	Other		23	13 (56.5)	7 (30.4)	3 (13.1)

Data are presented as n (%).

Factors associated with RFS

Table 3 shows the factors associated with RFS. RFS was better in female patients [hazard ratio (HR)=0.713, 95% confidence interval (CI)=0.524–0.971, p=0.032] and in patients with non-clear cell RCC (HR=0.447, 95% CI=0.262–0.763, p=0.003). Additionally, older age (HR=1.026, 95% CI=1.014–1.039, p<0.001), higher T stage (HR=3.257, 95% CI=2.374–4.468, p<0.001 for T1b; HR=7.321, 95% CI=5.084–10.543, p<0.001 for pT2), higher nuclear grade (HR=1.748, 95% CI=1.306–2.339, p<0.001), and presence of differentiation (HR=2.099, 95% CI=1.049–4.198, p=0.036) were associated with worse RFS. Factors associated with recurrence after 60 months were similar to those of the entire cohort.

Table 3. Cox Proportional Hazard Model of Factors Associated with Recurrence-Free Survival in All Patients Followed Up for >60 Months.

Variables		Total		Recurrence >60 months
		HR (95% CI)	p value	HR (95% CI)	p value
Sex			0.032		0.922
	Male	1 (Ref)		1 (Ref)
	Female	0.713 (0.524–0.971)		0.974 (0.573–1.655)
Age		1.026 (1.014–1.039)	<0.001	1.029 (1.007–1.052)	0.010
Operative side			0.208		0.752
	Right	1 (Ref)		1 (Ref)
	Left	0.839 (0.638-1.103)		0.924 (0.566–1.509)
Pathologic T stage
	pT1a	1 (Ref)		1 (Ref)
	pT1b	3.257 (2.374–4.468)	<0.001	2.836 (1.655–4.862)	<0.001
	pT2	7.321 (5.084–10.543)	<0.001	5.140 (2.581–10.235)	<0.001
Pathological type			0.003		0.035
	Clear cell	1 (Ref)		1 (Ref)
	Non-clear cell	0.447 (0.262–0.763)		0.285 (0.089–0.913)
Nuclear grade
	Grade 1 and 2	1 (Ref)		1 (Ref)
	Grade 3 and 4	1.748 (1.306–2.339)	<0.001	2.610 (1.554–4.382)	<0.001
	Unknown	0.763 (0.220–2.645)	0.670	4.170 (0.662–26.274)	0.128
Differentiation			0.036
	No	1 (Ref)		1 (Ref)
	Present	2.099 (1.049–4.198)		N/A

HR, hazard ratio; CI, confidence interval.

The factors associated with the first recurrence in the lung, bone, and contralateral kidney are presented in Table 4. Notably, unlike lung and bone metastases, female sex (HR=0.570, 95% CI=0.307–1.056, p=0.074) and age (HR=1.010, 95% CI=0.986–1.034, p=0.419) were not risk factors for recurrence in the contralateral kidney.

Table 4. Cox Proportional Hazards Model of Factors Associated with Recurrence-Free Survival According to the Organ in Which Recurrence Occurred.

Variables		Lung		Bone		Contralateral kidney
		HR (95% CI)	p value	HR (95% CI)	p value	HR (95% CI)	p value
Sex			0.024		0.010		0.059
	Male	1 (Ref)		1 (Ref)		1 (Ref)
	Female	0.556 (0.334–0.927)		0.321 (0.135–0.766)		1.913 (0.975–3.756)
Age		1.044 (1.024–1.066)	<0.001	1.042 (1.013–1.071)	0.004	1.014 (0.985–1.044)	0.360
Operative side			0.344		0.415		0.456
	Right	1 (Ref)		1 (Ref)		1 (Ref)
	Left	0.806 (0.516–1.259)		0.769 (0.409–1.446)		1.290 (0.661–2.520)
Pathologic T stage
	pT1a	1 (Ref)		1 (Ref)		1 (Ref)
	pT1b	7.053 (4.063–12.243)	<0.001	3.582 (1.750–7.332)	<0.001	1.834 (0.842–3.995)	0.127
	pT2	13.785 (7.367–25.796)	<0.001	8.654 (3.876–19.325)	<0.001	3.987 (1.673–9.500)	0.002
Pathological type			<0.009		0.719		0.286
	Clear cell	1 (Ref)		1 (Ref)		1 (Ref)
	Non-clear cell	0.298 (0.120–0.739)		1.172 (0.495–2.776)		0.522 (0.158–1.724)
Nuclear grade
	Grade 1 and 2	1 (Ref)		1 (Ref)		1 (Ref)
	Grade 3 and 4	1.476 (0.914–2.383)	0.111	1.344 (0.695–2.598)	0.379	3.977 (1.828–8.655)	0.001
	Unknown	0.671 (0.079–5.676)	0.714	0.580 (0.069–4.904)	0.617	N/A
Differentiation			<0.001		0.718		0.959
	No	1 (Ref)		1 (Ref)		1 (Ref)
	Present	5.462 (2.409–12.388)		1.453 (0.191–11.044)		0.947 (0.120–7.448)

HR, hazard ratio; CI, confidence interval.

DISCUSSION

Localized RCC recurs within 5 years in approximately 20%–30% of patients, even after curative treatment.³,⁴ Therefore, several guidelines recommend regular follow-ups for up to 5 years. However, the decision to follow-up after 5 years is usually based on the patient’s preference or the attending physician’s judgment. For follow-up beyond 5 years after surgery for RCC, the European Association of Urology (EAU) guidelines recommend CT every 2 years for intermediate- and high-risk patients.¹¹ The National Comprehensive Cancer Network (NCCN) guidelines recommend risk-based follow-up, but do not specify the frequency of follow-up or what tests should be performed.¹³ Conversely, the American Urological Association (AUA) guidelines recommend that abdominal and chest imaging follow-up in high-risk patients should be determined through informed decision-making discussion.¹²

Among the most diverse prognostic factors in RCC, the pathologic stage is one of the strongest and most important prognostic factors; consequently, the higher the stage, the worse the oncologic outcome.¹⁴ Stage 1 RCC has a 5-year survival rate of 90%; however, even small RCC may not always have a good prognosis. Even patients with T1a RCC can experience cancer-specific death after 10 years. Furthermore, Arai, et al.⁸ reported a recurrence rate of 20% at 20 years after surgery for stage 1 RCC, emphasizing the need for long-term follow-up. They also reported that loss to follow-up was more likely to occur in older recurrence-free patients, suggesting that precautions should be taken to ensure that these patients are not lost to follow-up.

Similar to previous studies, our results indicate it would be difficult to conclude that the recurrence rate at 5 years after surgery is so low that it can be ignored in patients with stage 1 and 2 RCC. In stage T1a, the average annualized recurrence rate after 5 years was 0.5%, which was not significantly different from the recurrence rate before 5 years. For T1b and T2, the average annualized recurrence rates after 5 years were approximately 3% and 3.1%, respectively. Although these rates were lower than the recurrence rates within 5 years of surgery, they were not negligible. In our results, the proportion of patients lost to follow-up at around 5 years was quite high (17.5%). Additionally, 10 patients who were lost to follow-up after 5 years without recurrence returned due to confirmed RCC recurrence at another institution.

In previous studies, the most common sites of RCC recurrence were the lungs, bone, liver, lymph nodes, adrenal gland, and brain.¹⁵,¹⁶ In a study by Almdalal, et al.⁵ involving patients enrolled in the National Swedish Kidney Cancer Register, RCCs <7 cm were most likely to recur in the lungs, bones, and treated kidneys. In this study, the recurrence rate in contralateral kidneys was 3.3%. In our study, the most common sites of first recurrence were the lungs, bones, and contralateral kidney.⁵ We found a significantly higher rate of contralateral kidney recurrence in our study than that in previous studies. RCC occurring in the contralateral kidney may be a recurrent or a primary cancer itself. It is also possible that contralateral RCCs are caused by genetic factors. The risk factors associated with recurrence in the contralateral kidney were slightly different from those in the overall patient group and in those with lung or bone metastases. Furthermore, although we excluded patients with definite hereditary RCC, those with undiagnosed hereditary RCC may have been included in our study. However, in situations where all gene mutations cannot be confirmed through tests, such as NGS, there is a need for early detection and treatment, even when recurrence occurs in the kidney. Therefore, we included patients with recurrence in the contralateral and ipsilateral kidneys in our analysis.

In addition to the contralateral kidney, the remaining common metastatic sites including the lungs and bones were identical. Based on these results, it seems necessary to evaluate bone metastasis as well as metastasis to the abdomen and thorax during follow-up. Unfortunately, only a few reliable modalities are available for diagnosing bone metastases. Previous studies have reported that bone scans and PET/CT have limited sensitivity and specificity for diagnosing bone metastases.¹⁶,¹⁷ Therefore, they generally do not recommend diagnosing bone recurrence during follow-up. Consequently, it is necessary to develop testing methods for bone metastasis.

The follow-up interval may be an important decision during follow-up after surgery for RCC. The EAU and AUA guidelines suggest that follow-up examinations should be performed every 2 years between 5–10 years postoperatively. However, we also observed that patients with a follow-up period of >18 months were more likely to have multiple sites of first recurrence compared to those with a follow-up period of <18 months. Thus, the early detection of recurrence is important for improving treatment outcomes. In particular, when recurrence occurs in a surgically resectable area, metastasectomy may improve survival. Therefore, a follow-up period of 2 years may be slightly longer for some patients, and we believe that a shorter surveillance duration is necessary for some patients.

However, it is not efficient to conduct follow-up observations every year for all patients after 5 years. Therefore, it is recommended that the attending physician and patient thoroughly discuss the risk of recurrence before deciding on a follow-up test. Similar to previous studies, sex, age, T stage, clear cell type, tumor grade, and tumor differentiation were factors associated with recurrence in our study. These factors are similar to previously known prognostic factors. In addition to these factors, various studies have shown that various hematological and genetic markers can affect the prognosis.¹⁷,¹⁸,¹⁹,²⁰ However, prognostic prediction models have not been developed to predict delayed recurrence; therefore, they may not be suitable for selecting patients who require long-term follow-up. Consequently, it is necessary to identify prognostic factors for delayed recurrence, either by integrating previously published prognostic factors or by identifying novel factors for predicting delayed recurrence. If we are able to predict the prognosis more accurately, we may be able to develop strategies for longer and more aggressive follow-ups in selected patients at a very high risk of recurrence.

Nonetheless, this study had a few limitations. In our study, we did not analyze the impact of the extent of surgery on recurrence. We initially intended to include whether PN or RN was performed in the Cox-regression analysis to assess their impact on RFS, as they could be associated with oncological outcomes. However, we observed that the larger the tumor size, or the higher the T stage, the more likely the patients were to receive RN. In T1a, 18.1% of patients underwent RN, but this increased to 63.5% in T1b and dramatically to 92.9% in T2. Since patients who underwent RN generally had higher stages, it is natural to expect a higher recurrence rate compared to those who underwent PN. Even if we were to perform Cox regression analysis, it would be impossible to completely adjust for stage differences, leading to a high likelihood of biased results, which could also affect the analysis results of other variables’ impact on RFS. For these reasons, we did not include the extent of surgery in the Cox regression analysis. Furthermore, this study included retrospective data and a single-institution experience. Therefore, large-scale multicenter prospective studies with long-term follow-ups are required to fully characterize the natural history and biology of RCC confined to small organs.

Late recurrence after surgery in organ-confined RCC is common, with recurrence occurring even after 10 years. Long-term follow-up, of perhaps 10 years or more, including imaging studies of the abdomen, lungs, and bone, should be considered for the early detection of recurrence in selected patients. The identification of prognostic factors for patients at a high risk of recurrence requiring long-term and aggressive follow-up is needed.