By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Dr Ronald J. Cohen, Uropath Pty Ltd, 47 Oxford Close, West Leederville, WA 6007, Australia. e-mail: email@example.com
Study Type – Prognosis (case series)
Level of Evidence 4
What's known on the subject? and What does the study add?
Biopsy of incidentally detected small renal masses is a safe and accurate method of determining benign from malignant lesions. Renal biopsy identifies over one quarter of cases as benign; however, one in five biopsy procedures results in insufficient material for a diagnosis.
The study contributes an additional 268 renal biopsy cases to the literature and further assesses the accuracy of biopsy and its ability to reduce the number of patients undergoing radical surgery for benign lesions. The results of lesions managed conservatively with follow-up are also presented. Furthermore, the study identifies an improved diagnostic trend for needle core biopsies over the period.
• To further validate the safety and diagnostic accuracy of percutaneous core biopsy in small renal masses ([SRMs]≤4 cm) in response to the rising prevalence of renal ‘incidentalomas’.
• To determine the value of percutaneous core biopsy in its ability to influence the choice of intervention or surveillance for the management of SRMs.
PATIENTS AND METHODS
• We collected data on the incidence of benign, malignant and non-diagnostic samples from 268 SRM (clinical T1a) biopsies performed at our institution between 1998 and 2009.
• The diagnostic accuracy of biopsy in small renal lesions was examined in cases proceeding to nephrectomy.
• Follow-up on the remaining non-surgical cases was performed.
• SRMs (≤4 cm, clinical T1a) constituted 59% of all renal lesions diagnosed; 80% of these biopsies were diagnostic and 20% were non-diagnostic.
• Of the diagnostic samples, 74% were malignant and 26% were benign.
• Thirty-three percent (17 of the initial 52) non-diagnostic samples underwent re-biopsy, with 94% yielding a histological diagnosis on repeat sampling.
• Fifty-eight percent of our diagnostic masses underwent nephrectomy, returning a renal biopsy accuracy rate of 100% for identifying lesions as benign or malignant.
• Of the 98 observed masses, 49% were benign, 28% malignant and 23% were non-diagnostic. In 63% of these cases, size remained stable in the follow-up period.
• Percutaneous core biopsy continues to provide an accurate and safe tool for preoperative tissue diagnosis of SRMs and should be offered to patients before considering surgical intervention.
• With regard to both benign and malignant T1a renal lesions, in select patient groups, those electing to undergo observation may expect favourable short- to medium-term outcomes.
The widespread use of cross-sectional imaging has led to increased detection of incidental renal tumours, particularly small renal masses (SRMs) ≤4 cm in diameter [1,2]. In contemporary surgical series 19–26% of SRMs were proven benign [3,4] indicating that preoperative diagnosis has the potential to avoid a substantial number of major surgical procedures and/or ablative therapies. The accuracy of percutaneous renal core biopsy for SRMs is an ongoing issue of debate [1,5], with several recent publications investigating large series of SRMs with promising diagnostic biopsy rates of 84–97% [6–9].
Our group has previously published renal biopsy results for 222 renal masses ≤5 cm in diameter that were accrued between 1999 and 2007 [10,11]. We have extended our previous study to December 2009 and present our biopsy data for 250 clinical T1a SRMs, ≤4 cm in maximum diameter, so as to make our findings directly comparable with the current literature . In the present paper, we attempt to reconfirm the accuracy of renal needle core biopsy that will assist clinical decision-making in the management of incidental SRMs.
MATERIALS AND METHODS
We identified 427 sets of renal core biopsies at Uropath, a specialist urological pathology practice in Western Australia, undertaken for renal mass diagnosis between February 1999 and December 2009. The clinical records of the 23 referring urologists were then accessed to identify all renal core biopsies of incidentally detected, asymptomatic lesions with a maximum diameter ≤4 cm and follow-up data collected on these patients. The final group consisted of 268 sets of renal core biopsies from 250 incidentally discovered SRMs in 250 patients. Results for 158 of these patients accrued before December 2007 have previously been published among our findings for 222 SRMs <5 cm [10,11].
All patients had been referred by their treating urologist to an interventional radiologist familiar with renal core biopsy. All cases were initially diagnosed by ultrasonography or CT. Cases diagnosed by ultrasonography were then subjected to pre- and post-contrast CT and these results were reviewed by the attending urologist and consultant radiologist. All lesions appeared predominantly solid, although some had a minor cystic component, with no visible fat and all were considered suspicious for malignancy on imaging. Renal core biopsy was performed under ultrasonography or CT guidance using local anaesthesia and an 18-G core biopsy gun, with a mean (median; range) of 2.7 (2.0; 1–10) cores taken per lesion. All biopsy samples were reviewed by one specialist pathologist (R.J.C.) with a particular interest in renal tumours.
Biopsy cores were designated ‘diagnostic’ and separated into histological tumour types, or were designated ‘non-diagnostic’ if there was (i) insufficient material for analysis, or if the sample contained only (ii) normal renal parenchyma, (iii) fibro-fatty connective tissue, (iv) necrotic tissue or blood clot, or (v) inflammatory/fibrotic tissue.
The biopsy results (malignant vs benign, tumour type and, where assigned, Fuhrman grade) were correlated with definitive pathology findings for patients who went on to undergo radical or partial nephrectomy, while the medical records were reviewed for patients who were managed non-surgically to obtain information about their renal mass after biopsy. From these results, we determined the number of patients electing to undergo radiofrequency ablation as the primary treatment for their tumours. We also identified those electing observation with their subsequent clinical and radiographic follow-up.
Comparisons of renal masses and follow-up characteristics between patient groups were performed using non-parametric methods, with the chi-squared test for categorical data and the Mann–Whitney U-test for continuous data. Univariate and multivariate logistic regression modelling was used to evaluate covariates as predictors of a diagnostic biopsy. All P values were two-sided and a P value of <0.05 was considered to indicate statistical significance. Statistical analysis was performed using SPSS® version 17.0.
Analysis of the 268 sets of renal core biopsies from 250 incidentally discovered asymptomatic SRMs found that the median (range) patient age was 64 (22–88) years, and the median (range) lesion size was 2.5 (0.9–4) cm. Of the 268 biopsies, 214 (80%) were diagnostic, including 16 from patients where the initial biopsy was non-informative. One initially non-diagnostic case that underwent repeat biopsy yielded another non-informative sample and the patient elected to have no further procedures. Therefore, on re-biopsy of all 17 patients with initially non-informative biopsies, diagnosis was obtained in 16 cases, with one patient requiring two repeat procedures for a diagnosis. Of the 214 diagnostic samples 74% (158) were malignant and 26% (56) were benign (Fig. 1). Of the 122 malignant RCCs, where histological grade was indicated, 55% of tumours were Fuhrman grade I, 39% grade II and 6% grade III/IV (Table 1).
Table 1. Clinical Fuhrman Grade assignment for RCCs, where allocated, and correlation with surgical specimens
No. of biopsies (%)
No. of nephrectomies (%)
Grade assigned at biopsy
Fuhrman Grade I
Fuhrman Grade II
Fuhrman Grade III/IV
Of the 72 cases with grade assigned who underwent nephrectomy
Of the 268 biopsies, 54 (20%) were designated non-diagnostic because they contained only normal renal parenchyma (29 cases), inflammation and/or fibrosis (nine cases), fat or fibro-fatty connective tissue (nine cases), insufficient material for analysis (five cases) or necrotic tissue and/or blood clot (two cases). These 54 biopsy sets originated from 52 patients, of whom four were lost to follow-up, 17 underwent repeat biopsy with 16 of those receiving a diagnosis of malignant tumour (10 patients) and benign tumour (six patients). Nine non-diagnostic cases underwent nephrectomy without repeat biopsy, of which eight were found to be malignant and one was found to be benign (Fig. 1). The remaining 23 patients underwent observation.
Of the 214 patients with diagnostic biopsies, 125 (58%) proceeded to radical or partial nephrectomy and 120 had histological information available (the additional five nephrectomy cases were performed outside our institution and final pathology could not be located). In all 120 cases, this confirmed the biopsy diagnosis as malignant (114 cases) or benign (six cases), giving a malignant vs benign biopsy accuracy rate of 100%. For 98% (117) of these 120 lesions, the histological type was also correctly diagnosed on biopsy. Of the remaining three cases, two diagnosed as clear-cell RCC were subsequently found to be a solid variant of papillary RCC and an unclassifiable RCC, respectively, while one diagnosed as papillary RCC was found to be an Xp11 translocation tumour. Fuhrman grade was accurately assigned in 50/72 (69%) biopsy specimens when compared with nephrectomy, upgraded in 17 (24%) and downgraded in five (7% [Table 1]).
Of 116 diagnostic cases where type of surgery was indicated, 53 (46%) of the procedures were nephron-sparing partial nephrectomies. The frequency of partial vs radical nephrectomy rose from 36% (19/53) up to and including 2006, to 54% (34/63) from 2007 to 2011 (P= 0.05). Two patients with a benign mass diagnosed on biopsy elected to undergo radical surgery.
Overall, 39% (98/250) of SRMs had no surgical intervention and instead were followed. The final outcomes for these masses, where follow-up was available, is shown in Table 2. There were a total of 89 diagnostic masses that did not undergo nephrectomy; 50 benign and 39 malignant masses. Of the 50 benign masses, two were lost to follow-up and the remaining 48 were followed with regular ultrasonography for a median (range) of 19 (1–108) months. Two patients (both with oncocytomas that remained stable in size) developed macroscopic haematuria but no patient required interventional treatment during the follow-up period. All 48 patients in this benign group were considered to have been spared nephrectomy to date based on preoperative biopsy findings.
Table 2. Follow-up data for 98 patients with SRMs who did not undergo surgical intervention
No. of cases
Median tumour diameter, mm
Size stable, n (%)
Size increase, n (%)
Median growth rate, mm/year
Size decrease, n (%)
Resolved, n (%)
New symptoms, n (%)
Metastases, n (%)
Nephrectomy, n (%)
Radio-frequency ablation, n (%)
Of the 39 malignant masses, seven of these patients were lost to follow-up, no follow-up was collected for one patient diagnosed with renal metastases of non-renal tumours (metastatic melanoma, 35 mm), and four patients (two with clear-cell RCC, two with papillary RCC) underwent radiofrequency ablation as primary treatment of their lesions. The remaining 27 patients with no immediate active intervention were followed with regular ultrasonography for a median (range) of 17.0 (6–43) months. Radiofrequency ablation therapy was indicated in an additional three of these patients as tumour growth rates exceeded 3.6 mm/year. None of these 27 diagnostic patients developed metastatic disease during the follow-up period.
Age was a significant factor for these malignant cases electing to undergo surveillance over surgical intervention. The median patient age was 79 years compared with 61 years in the same group that instead underwent nephrectomy (P< 0.001). Tumour size however, did not vary significantly between the two groups (median 0.26 cm and 0.27 cm, respectively).
The SRMs that underwent observation in the absence of surgery displayed no significant differences in the proportion of lesions that experienced growth (29% benign vs 30% malignant), the median growth rate (0.37 cm/year benign vs 0.47 cm/year malignant) or onset of symptoms between benign and malignant lesions. Larger sample sizes may further validate these results.
Of the 27 non-diagnostic cases that did not undergo surgical intervention, four patients were lost to follow-up, and the remaining 23 were followed with regular ultrasonography for a median (range) of 26 (6–84) months. Two patients (both with masses that remained stable in size) developed symptoms of loin or abdominal pain, but no patient in this group developed metastatic disease or required treatment during the follow-up period. Masses yielding non-diagnostic biopsies were significantly more likely to resolve and less likely to increase in size compared with masses with diagnostic biopsies. This may reflect mass size; non-diagnostic biopsies tended to be smaller at the time of biopsy, with a median diameter of 2.0 cm vs 2.6 cm for diagnostic masses (P= 0.008). It may also reflect a non-neoplastic aetiology in some cases.
Non-diagnostic cases, as a group, were followed up for longer than diagnostic cases. This was owing to a higher proportion of non-diagnostic procedures performed in the early years of the study (61% from 1999 to 2006). In the diagnostic group, the majority (73%) of biopsies were performed in the final 3 years of the study, resulting in a shorter follow-up.
Multivariate logistic regression analysis demonstrating independent predictors of a diagnostic renal biopsy are shown in Table 3. Diagnostic biopsies were significantly associated with a larger mass diameter, more biopsy tissue available for pathological analysis, and a biopsy during the later years of the study. Age was not a significant predictive variable on univariate analysis (P= 0.231) and renal core number was significant on univariate analysis (P= 0.006, odds ratio 1.242), but was eliminated in the multivariate model owing to the predictive power of tissue volume (core number by sample length).
Table 3. Multivariate logistic regression analysis for predictors of a diagnostic SRM biopsy
Change in AUC value
OR, odds ratio; AUC, area under the curve.
Diameter of mass, mm
Amount of tissue, mm
Year of surgery
Over the course of the study, the number of biopsy cores increased minimally (Table 4); however, clinicians were able to sample a significantly larger volume of tissue per core. The increasing diagnostic rate is reflective of this; from 65% in the years after 1999 to 86% by 2009.
Table 4. Time trends in percutaneous needle biopsy of SRMs
Mass diameter, mm
No. needle cores taken
Amount of biopsy tissue, mm
Diagnostic biopsies, n (%)
Finally, there were only two significant clinical complications noted from the 268 biopsy procedures (morbidity rate of 0.7%). One patient required hospital admission and transfusion of two units of blood owing to post-biopsy bleeding but recovered without sequelae and went on to undergo elective surgery for clear-cell RCC. In a second case, a 2-cm intrarenal haematoma rather than the tumour was mistakenly excised at laparoscopic partial nephrectomy. The patient subsequently underwent radical nephrectomy to remove clear-cell RCC. No instances of tumour implantation along the needle track were noted in surgical specimens or during the follow-up of the malignant tumours that were followed with radiological imaging.
Incidental renal mass lesions account for up to two thirds of all clinically diagnosed renal tumours and, despite apparently earlier diagnosis and treatment, the mortality from RCC remains essentially unchanged [2,12]. In lock-step with the increasing numbers of renal ‘incidentalomas’, benign tumours are more frequently encountered . This poses a significant clinical dilemma as to whether we offer radical intervention and who is likely to benefit from such action.
The present study investigates the application of percutaneous needle core biopsy in an attempt to temporize the use of radical curative surgery. It is clear that in specialist hands, core biopsy provides dependable, accurate results and clearly defines the significant subset of incidentally detected benign mass lesions . In RCC, biopsy further provides a measure of accuracy with regards Fuhrman grade (with a concordance of 69%) allowing more tailored decision-making [6,7]. Immediate complication rates of <1% and the inability to demonstrate the hypothetical risk of tumour seeding in a sizable number of consecutive cases supports its continued and more widespread use [13,14].
The one factor that continually plagues the protagonists of renal biopsy is the unacceptable proportion of cases where non-diagnostic material is sampled. Our overall non-diagnostic rate of 20% is clearly reason for concern. However, further analysis indicates the importance of clinical experience, with a non-diagnostic rate of 35% at the start of the study, and current figures show this rate now approaching a more acceptable 14%. Part of this non-diagnostic rate may be attributable to non-neoplastic aetiology, with no clear diagnostic features recognizable on biopsy core. Needle placement, however, continues to remain a problem, with most cases of clinical T1a lesions yielding diagnostic material on repeat biopsy .
Radical nephrectomy and the loss of half of renal functional reserve poses the significant long-term risk of moderate chronic renal insufficiency (36%; defined as a glomerular filtration rate <45 mL/min/1.73m2) as well as the associated increased risk of cardiovascular pathology [15–17]. Thus, in small lesions there has been a tendency to move towards ‘nephron-sparing’ partial nephrectomy that has led to the more widely accepted surgical intervention, even for benign lesions [16,18]. As the incidence of benign ‘incidentalomas’ approaches 30% it is likely that, even in the best centres, significant morbidity (ranging from 3 to 37%) and mortality will occur [18,19]. Partial nephrectomy with complete tumour clearance and without significant ischaemic and reperfusion damage to the remaining renal parenchyma represents a significant technical challenge [20–22]. Even in the most expert of hands, this is unlikely to be achieved in a reasonable number of cases . The short-term complications and longer-term risks of renal insufficiency as well as significant financial costs cannot in our opinion be justified for small benign tumours. We are unaware of any publication that illustrates any clinical benefit (short- or long-term) for the removal of these benign lesions in the absence of obstruction or excess growth. For this reason, we believe that needle core biopsy performed in a specialist setting can significantly improve the triage of patients with incidental SRMs. Furthermore tailored less rigorous follow-up would be totally appropriate once a benign diagnosis is made on biopsy core, as demonstrated by our 48 surveillance patients that did not undergo intervention as a direct result of their benign renal biopsy.
The fact that lesion growth rates cannot predict malignancy in SRMs is a further argument for biopsy . The 63% of published lesions that experienced growth did so at a mean annual rate of 0.26 cm, regardless of whether they were benign or malignant, and the overall rate of metastases was 1% [16,23,24]. Thus, in select patients where excessive tumour growth rates were not observed, one may assume reasonably safe management through radiological surveillance in biopsy-proven RCC . Yet the real value of biopsy in this setting lies in the ability to identify histologically the indicators of more aggressive disease (high Fuhrman grade or an infiltrative growth pattern) potentially validating treatment in the 20–30% of cases considered to be higher risk .
The benefits of surveillance of malignant SRMs are less clear. Findings of a recent study suggest longer life expectancy in the absence of treatment in patients >74 years . As with the retrospective nature of our study, the literature regarding surveillance of SRMs is sparse and based on a strong selection bias and short follow-up . However, as mortality from RCC has not decreased, surveillance with delayed treatment on progression is fast emerging as a more suitable form of management among the elderly and in those with serious comorbidities [13,23].
The present study did not examine post-surgical cancer-specific survival, renal function or quality of life among the group that underwent surgical intervention. This may be worthy of further investigation given the gap in current data on oncological outcomes .
In conclusion, in the era of incidental SRMs, the inherent risk of chronic renal insufficiency as a product of radical intervention of these lesions often has worse implications than that of disease progression itself. Even partial nephrectomy, essentially offering greater sparing of nephron function while maintaining similar disease control, can have significant quality-of-life ramifications . We re-affirm the significant diagnostic capabilities contemporary renal biopsy offers, in particular the proven ability to identify benign masses where, in most cases, less aggressive management strategies are more suitable than radical surgical intervention.