Renal cell carcinoma

W.Y. SIOW, S.K.H. YIP, L.G. NG, P.H. TAN *,W.S. CHENG and K.T. FOO
Department of Urology and *Department of Pathology , Singapore General Hospital, Singapore

In developed countries, there has been increased incidental detection of renal cell carcinoma (RCC). The incidence, pathological stage and survival of incidentally detected carcinoma in a developing country in Asia where, from 1990 to 1998, 165 renal cell carcinomas were identified. The clinical presentation, diagnostic-imaging modality employed, pathological staging and patient survival was reviewed. Incidental renal cancers included those that were diagnosed through health screening or detected incidentally through imaging studies for other conditions. The survival between these incidentally detected lesions and their symptomatic counterparts (suspected group) was compared. Sixty-four patients (39%) had their tumours detected incidentally, including 39 who were entirely asymptomatic and 25 who presented with non-specific symptoms, not initially suggestive of RCC. For the entire group, computed tomography provided the definitive diagnosis in 81% of cases. The incidental detection group had significantly smaller size of tumour (5.9 cm c.f. 7.6 cm), lower stage and lower histological grading. In particular, 78% of patients with incidental RCC had stage I or II diseases (TNM stage classification), compared with 57% of patients with suspected tumour (p<0.05; Chi-square test). The disease free survival was significantly better for those with incidental detection (86% c.f. 66% at last follow up; p< 0.05; log-rank test) over a mean follow up period of 33 months (range 1 -91). Regression analysis showed that stage of disease was the only independent variable predictive of clinical outcome. In conclusion, that significant numbers of RCC were detected incidentally. These tumours were of a lower clinical pathological stage and had a better prognosis.

Advances in imaging technology and the increased application of radiological modalities such as ultrasonography (US) and computed tomography (CT) for health screening and evaluation of other conditions, particularly intra-abdominal pathologies, have led to an increased number of early renal cell carcinomas (RCC) being diagnosed incidentally in developed countries.^1-3 However, whether these incidental tumours are of lower stage and grade than their symptomatic counterparts and, thus, are more amenable to curative surgery and better survival remains controversial.^4-9

To the best of our knowledge, there are no reports on asymptomatic renal cancer and their implication on survival in a developing country in Asia. Hence, we studied the incidence of incidental renal carcinoma treated in Singapore General Hospital, a tertiary referral centre with a draining population of 3 million. In particular, the clinical profiles and prognosis of incidental RCC were studied.

A retrospective review was conducted between the year 1990-1998. The cancer registry registered 213 patients with the diagnosis of RCC. One hundred sixty-five cases with complete records of clinical presentation and diagnostic imaging, in addition to the clinical pathological staging information, were included in the study. The majority (153) of patients had histological confirmation; while a small proportion (12) of patients with metastatic disease on presentation was diagnosed solely on imaging.

Patients were divided into two groups according to whether the tumour detection was an incidental finding or whether it was suspected. Patients with incidental tumours included those who were entirely asymptomatic and had abnormalities detected through health screening, as well as those who presented with non-specific symptoms not initially suggestive of RCC. On the contrary, patients presenting with specific symptoms, including haematuria, loin pain and loin masses were assigned as the suspected tumour group. Patients who were being investigated for obvious metastatic diseases upon first presentation were also included in the suspected group.

The range of diagnostic modalities included intravenous urogram (IVU), US, CT, angiography, magnetic resonance imaging and fine needle aspiration/biopsy. Definitive diagnostic modality referred to the modality based on which a RCC was strongly suspected, where further imaging studies were not required to establish the diagnosis. Further imaging was performed, however, to aid an accurate pre-operative staging.

The histology was studied with particular reference to the local extent of the tumour. Pathological staging was based on the TNM classification while the Fuhrman nuclear grades were used for grading. Differences in pathological findings and outcome between the two groups were compared. The recently implemented electronic medical record systems enable us to have cross-reference to all hospital contacts for individual patients. Their visit to the National Cancer Centre and other disciplines was captured on line, in addition to the follow up information obtained from visits to the Department of Urology. Last follow-up status, as at April 2000, were obtained, presence of recurrent disease at last follow-up and death were computed as significant events in the disease-free survival analysis.

A total of 64 cases (39%) of RCC were detected incidentally in a total of 165 cases diagnosed over the period 1990-1998. Of these, 39 (21%) were entirely asymptomatic, with their tumours diagnosed after further work-up for abnormal findings noted during health screening. The other 25 patients presented initially with non-specific symptoms. While imaging studies were performed for other working diagnoses, the presence of renal cancers was revealed. In the suspected group (n = 101), haematuria was the most common presenting symptom (27%). The presenting symptoms for the entire group are shown in Table 1.

Group	Symptoms	Frequency	%
Incidental	Asymptomatic	39	23.6
	Non-specific :Abdominal pain	16	9.7
	Hypertension	3	1.8
	Lower urinary tract symptoms	3	1.8
	Renal impairment	2	1.2
	Urinary tract infection	1	0.8
Suspected	Specific: Haematuria	44	26.7
	Loin pain	20	12.1
	Loin mass	3	1.6
	More than one primary symptom	15	9.1
	Systematic constitutional symptoms	16	9.7
	Metastatic disease	3	1.8

The initial imaging modality employed varied according to the presenting 'symptom'. For those who presented for health screening, US was the most commonly used first line imaging modality detecting the abnormal lesion (29/39, 74%). For those who presented with haematuria, IVU was the most commonly employed first line investigation (20/44, 45%). Irrespective of the initial investigation, CT was the imaging modality that was most frequently diagnostic and conclusive (82%; Table 2). Nephrectomy was performed in 153 patients, of whom, 18 was nephron-sparing surgery (12 from the incidental group and 6 from the suspected group).

Modality	Frequency	%
Computed tomography	133	80.6
Ultrasonography	19	11.5
Angiography	9	5.5
Intravenous urography	2	1.2
Fine needle aspiration	2	1.2

Tumour size was available in 155 of the 165 patients. The mean size of incidental RCC was 5.9 cm (SD, 4.1) as compared to 7.6 cm (SD=3.5) for suspected RCC. This difference is statistically significant (p=0.011, t-test). Incidental RCC was associated with a lower clinicopathological stage than suspected RCC. Organ confined disease i.e. TNM classification stage I and II disease was found in 78% of patients with incidental tumours, as compared with 57% of patients with suspected RCC. This difference is statistically significant (p= 0.001, Chi square test). The incidental cancers also had a significantly lower histological grading (Table 3).

Tumour characteristics	Incidental (%)	Suspected (%)
Stage I	14	4
II	64	49
III	19	26
IV	3	22
Fuhrman nuclear grade 1	26	13
2	58	51
3	14	31
4	2	5

The follow-up period for the patients ranged from 1 month to 91 months (mean = 33 months). The mean follow-up for the incidental group and the symptomatic group were 36 (range 1-90) and 31 (range 1-91) months, respectively. Of the 64 patients with incidental RCC, 55 (86%) were alive with no evidence of disease at last follow up. Of the 101 patients with suspected tumours, 67 (66%) were alive with no evidence of disease (Table 4). This table breaks down the cases by the year of diagnosis, indicating that the increase in the total number of renal tumours diagnosed affects both incidental and symptomatic cases. Figure 1 showed the disease-free survival curves of the two groups. The difference is statistically significant (p < 0.05; log-rank test). Regression analysis revealed that tumour stage is the only independent variable predicting development of disease recurrence and death, while incidental detection and tumour grade are not.

Table 4: Year of diagnosis and development of recurrence or death at last follow-up

Year of diagnosis	Disease status	Incidental	Symptomatic	Total
94 & before	NED	11	10	21
	Recurrence/death	-	10	10
95	NED	5	16	21
	Recurrence/death	1	2	3
96	NED	8	11	19
	Recurrence/death	1	7	8
97	NED	17	17	34
	Recurrence/death	4	5	9
98	NED	14	13	27
	Recurrence/death	3	10	13

The incidence of incidental RCC has been studied in a number of retrospective series as well as population surveys, and it ranged from 15 to 61%.^1-9 A number of authors were able to document an increasing trend of such detection over their study period (Table 5). Such a wide range of incidence in the literature could be related to differences in definition of incidental detection in various studies, as well as the referral pattern and health screening policies in different countries.

Table 5: Studies reporting detection of incidental renal cell carcinoma and survival

Author, year, country	Study	Year	No.	%	Remarks
Konnack (85), US	Case series	80-84	46	48	Increasing trend c.f. 61-73 (13%)
Thompson (88), US	Case series	46-85	212	18	Increasing trend
Smith (89), US	Case series	82-85	122	25	Increasing trend c.f. 74-77 (5%)
Mevorach (92), US	Case series	76-87	235	29	Similar stage; no improved survival
Rodriguez-Rubio (96), Spain	Case series	79-93	157	35	Lower stage; improved survival
Sweeney (96), UK	Case series	72-92	189	15	Lower stage; improved survival
Gudbjartsson(96), Iceland	Population	71-90	406	15	Lower stage; no survival benefit after correcting for stage
Jayson (98), US	Case series	89-93	131	61	Similar stage
Bretheau (98) France	Population	92-94	1486	40	Smaller size

Apparently, most of the major series included asymptomatic patients and patients with non-specific symptoms into the incidental detection group. Using similar criteria, the present series showed an incidence of 39%. While this is a retrospective series from a tertiary referral center, this institute is one of the few major hospitals serving the entire population. Significant portion of patients was treated here when compared with the data from the National Cancer Registry.¹¹ Thus, we believe that our results portray the national incidence of this cancer.

While we had an actual increase in the incidence of RCC, the increase could largely be attributed to the use of US and CT for health care screening as well as investigation of other unrelated symptoms.¹¹ In a large-scale health-screening programme in Japan, US was used primarily and led to detection of a significant number of renal lesions.¹² From the experience of Ozen et al,¹³ the majority of these incidentally detected renal masses turned out to be renal cell carcinoma. In fact, subsequently, a nationwide survey in Japan confirmed that there was an obvious trend towards increased numbers of renal cell carcinoma detected incidentally. Thus, US should be considered the primary imaging modality compared with IVU.¹⁴

Even for patients presenting with gross haematuria, the diagnostic pathway can be enhanced by providing a day case haematuria diagnostic service combining the use of US and flexible cystoscopy.^14,15 This will minimize the delay in radiological appointments and facilitate the arrangement of definitive imaging. CT remained the modality of choice to accurately delineate the extent of the lesion, the stage of the disease and to exclude gross lymph node or metastatic dis-ease. This was shown in our series where CT was the definitive imaging in over 80% of cases, irrespective of the initial modality that detected the lesion in the first instance.¹⁴

There is disagreement in the literature regarding the implication on survival for these incidentally detected lesions. Incidentally detected lesions should be of a lower stage and have better survival. However, there were also interesting series which reported no significant difference in stage or in survival.^4-7 The multivariate analysis by Gudbjartsson et al8 demonstrated that there was no difference in survival with incidental detection, after correction for stage and grade between the incidental and the suspected group.⁷ Similarly, in our series, the difference in survival derived from the difference in pathological stage. Thus, it is clear that the survival benefits are due to the detection at an earlier stage.

To date, radical nephrectomy remains the mainstay of treatment for RCC, incidental or otherwise, in the presence of an anatomically and functionally normal contra-lateral kidney. However, with the favourable pathology and improved prognosis seen with incidental RCC, nephron- sparing surgery may become the preferred treatment of choice.^16-18 Currently, we exercise strict selection for such procedures, namely, that the tumour is of early stage, small size, easily accessible and there is absence of multi-focal lesions. But with increased detection of such small incidental lesions, we are performing nephron-sparing surgery more frequently.

In conclusion, incidental detection of renal cell carcinoma occurred in up to 39% of cases. The tumours were smaller and of lower pathological stage and grade than the symptomatic counterparts. A better survival outcome was noted.

We thank Ms J Lee for her assistance in statistical analysis and Dr. P Moorthy for his assistance in the preparation of the manuscript.

Correspondence: Dr Sidney KH Yip, Department of Urology, Singapore General Hospital, Outram Road, Singapore 169608

Renal cell carcinoma: incidental detection and pathological staging