Are There Parameters that Predict a Nondiagnostic Biopsy Outcome Taken During Laparoscopic-Assisted Cryoablation of Small Renal Tumors?

Abstract

Background and Purpose:

The histopathologic diagnosis of a small renal mass (SRM) that is managed with cryoablation relies on preoperative or intraoperative biopsies. Because a considerable number of these SRMs are benign, accurate diagnosis has prognostic and follow-up implications. The main problem in SRMs is the high rate of nondiagnostic biopsies. Our purpose was to assess whether certain tumor and biopsy characteristics are correlated with a diagnostic biopsy outcome.

Patients and Methods:

One hundred tumors that were smaller than 4.5 cm in 94 patients were managed with laparoscopic cryoablation. After dissection of the perirenal fat and identification of the tumor by intra-abdominal ultrasonography, one or more biopsies were obtained before freezing. Using the Student t/Mann Whitney U test, the following parameters were evaluated for predicting biopsy outcome: Tumor size, location, and exophytic part of the tumor, size of the biopsy needle, the number of biopsies taken, and presence of nonenhancing areas compatible with necrosis inside the tumors. Correlations among parameters were assessed using a Spearman correlation or Kruskal-Wallis test.

Results:

Twenty-two (22%) biopsies were nondiagnostic and consisted of normal kidney tissue, connective tissue, fat, fibrosis, necrosis, and/or blood. There were no significant differences in parameters between the diagnostic and nondiagnostic group. There was a positive correlation between tumor size and number of biopsies (P=0.029) and between the presence of nonenhancing areas and both size (P<0.001) and the number of biopsies taken (P<0.001).

Conclusion:

No statistical significant correlation was found between biopsy outcome and tumor or biopsy characteristics. More biopsies were taken in larger tumors, and larger tumors contained more nonenhancing areas that were suspect for necrosis.

Introduction

R enal cryosurgery is a treatment option for small renal masses (SRMs) in which the tumor is ablated in situ by lethal freezing injury. Consequently, there is no surgical specimen for pathologic examination and, therefore, the histopathologic diagnosis relies exclusively on biopsies. A large study has shown that 23% of operated renal tumors that were smaller than 4 cm turn out to be benign, and even when cancer is present in these tumors, a considerable percentage is of low grade.¹ Preoperative diagnosis can be obtained by performing a percutaneous core biopsy or fine needle aspiration; however, concerns about accuracy have prevented their routine use.^2
–4 Of percutaneous core biopsies, up to 21% are nondiagnostic,⁴ and reports on intraoperative biopsies taken before cryoablation show a similar range of nondiagnostic biopsy outcomes.^5

–8 Nondiagnostic biopsies are usually because of either an insufficient amount of tissue or erroneous sampling of necrotic/fibrotic areas or normal kidney tissue.⁴

Because determination of treatment success after cryoablation is mainly based on cross-sectional imaging, obtaining a histopathologic diagnosis should be warranted to withhold patients with a benign tumor from an intense follow-up. Moreover, it has been demonstrated that a vast amount of nondiagnostic biopsies appear to be false negative in SRMs.⁹ This study evaluated whether tumor and biopsy characteristics can predict biopsy outcome in a series of patients who were treated with laparoscopic renal cryoablation in our center.

Patients and Methods

We identified consecutive patients who were treated with laparoscopic cryoablation between July 2004 and May 2010 for one or more solid enhancing SRMs and who underwent an intraoperative biopsy before freezing. The cryoablation was performed using a third-generation argon-based system with 1.47 mm (17-gauge) cryoprobes (type SeedNet^® or IceRods,^® Seednet Gold System,™ Galil, Tel Aviv, Israel).

After mobilization of the kidney and dissection of the peritumoral fat, real-time laparoscopic ultrasonography (US) was used to identify and measure the size of the tumor. Percutaneous biopsies were taken under laparoscopic vision using a 16- or 18-gauge core biopsy system (18G Topnotch™ and TruPath,™ Boston Scientific, Natick, MA, and 16-gauge Quick-Core^® Biopsy Needle, Cook Medical, Denmark) without the use of a guiding sheath. To ensure an accurate position of the needle, the tumor was first penetrated for 1 or 2 mm before the biopsy gun was fired. The number of biopsies taken was determined by the number of shots needed to obtain a proper core of tissue as identified visually by the surgeon. All biopsies were fixed in formalin, embedded in paraffin, and stained using hematoxylin and eosin. All biopsies were viewed and there were multidisciplinary discussions by specialized genitourinary pathologists.

Pathology reports were retrieved from the hospital computer system, and results were marked in a database as either diagnostic or nondiagnostic. Biopsies were considered diagnostic when a definitive diagnosis of benign or malignant tumor was possible. A biopsy was considered non-diagnostic if no tumor cells were present in the biopsy tissue, or insufficient tumor cells were present to differentiate between a benign and malignant tumour. The histopathologic diagnosis was determined according to the World Health Organization 2004 classification.¹⁰ If necessary, additional immunohistochemical staining was performed at the discretion of the pathologist.

The following tumor and biopsy characteristics were evaluated for their influence on biopsy outcome:

1. Tumor size; the maximum diameter as measured on preoperative CT or MRI.

2. The exophytic part of the tumor; the percentage of tumor protruding outside the surface of the kidney as measured on CT or MRI.

3. Tumor location; the location of the tumor epicenter was recorded as upper, interpolar, or lower pole of the kidney, assessed on the preoperative CT or MRI.

4. The presence of nonenhancing areas compatible with necrosis inside the tumors on preoperative CT or MRI (categorized as present of absent).

5. The number of biopsies taken (ie, number of shots, see above).

6. The size of the biopsy needle (16- or 18-gauge).

7. The quality of the biopsies (after preparation for pathologic assessment).

To assess the quality of the biopsies, the slides used for microscopic investigation by the pathologist were retrieved from our clinics pathologic archive and scored by two investigators (KB and MH) according to the following standards (also see Fig. 1):

• Good: One or more compact cores >10 mm in length.

• Intermediate: One or multiple fragments between 5 and 10 mm.

• Poor: Multiple or solitary separate fragments smaller than 5 mm.

FIG. 1.

A solid tumor (A) with a good quality biopsy specimen consisting of a compact core (C). Below, a tumor with nonenhancing areas suggestive for necrosis (B) with a poor quality biopsy specimen consisting of multiple short fragments (D).

To evaluate differences in tumor and biopsy characteristics between diagnostic and nondiagnostic biopsy outcomes, we used the Student t test or in case of a skewed distribution, the Mann Whitney U test for numerical parameters and the chi-square test for the categorical parameters.

To analyze correlations between individual parameters, we performed a Spearman correlation for numerical and ordinal parameters and the Kruskal-Wallis test for numerical vs multicategorical parameters. For all analyses, the level of significance was set at 5%.

Results

A total number of 100 tumors in 94 patients were identified. The mean age of the 57 (61.3%) men and 36 women was 67 years (range 38–91 y). In five patients, multiple tumors were treated. Characteristics of the tumors and the biopsy results are shown in Table 1. Our series showed 22 (22%) nondiagnostic biopsies. In 47% of tumors, immunohistochemistry was used to differentiate benign from malignant tumors and/or to further distinguish subtypes of renal-cell carcinma (RCC).

Table 1.

Characteristics of Patients and Tumors

Patients	94
Tumors	100
Patients with multiple tumors	5
Side of tumor L/R	49/51
Diagnostic biopsy outcome	78 (78%)
Renal–cell carcinoma	56 (72%)
Oncocytoma	20 (26%)
Angiomyolipoma	2 (3%)
Nondiagnostic biopsy outcome	22 (22%)
Normal renal tissue	7 (32%)
Fibrous tissue or sclerosis	8 (36%)
Infection (chronic)	3 (14%)
Fat tissue	2 (9%)
Necrosis or blood	2 (9%)
Mean tumor size (SD) (cm)	2.58 (0.77)
No. of tumors 1.0–1.9 cm	23 (23%)
No. of tumors 2.0–2.9 cm	38 (38%)
No. of tumors 3.0–3.9 cm	36 (36%)
No. of tumors 4.0–4.5 cm	3 (3%)
Exophytic part of tumors
No of tumors 0–24 % exo	18 (18%)
No of tumors 25–49 % exo	14 (14%)
No of tumors 50–74 % exo	54 (54%)
No of tumors 75–99 % exo	14 (14%)
Tumor location
Upper pole	33 (33%)
Interpolar	41 (41%)
Lower pole	26 (26%)
Mean no. of biopsies taken (SD)	2.68 (1.00)
1	4 (4%)
2	47 (47%)
3	30 (30%)
4	9 (9%)
5	4 (4%)
(missing data)	6 (6%)
Biopsy needle size 16G/18G	17/83

SD=standard deviation; G=gauge.

Table 2 shows comparison of different variables (radiologic and procedural) between the groups of diagnostic and nondiagnostic biopsy results. There was no statistically significant difference between the groups of diagnostic and nondiagnostic biopsies. There was a positive correlation between tumor size and the number of biopsies (Spearman r=0.191, P=0.029), between tumor size and exophytic part of the tumor (Spearman r=0.209, P=0.018) and between the size of the biopsy needle and the quality of the biopsy (Spearman r=0.342, P=<0.001). Furthermore, central nonenhancing regions were significantly more present in larger tumors, with a mean tumor size of 2.11 cm for tumors without nonenhancing intratumoral areas and 2.88 cm for tumors with this feature (Student t test, P<0.001) and more biopsies were taken when nonenhancing areas were present in the tumor (Spearman r=0.227, P<0.001).

Table 2.

Differences Between Diagnostic and Nondiagnostic Biopsies for Different Parameters

Parameter	Diagnostic biopsies (n=78)	Nondiagnostic biopsies (n=22)	P value
Mean tumor size (SD)	2.61 (0.76) cm	2.52 (0.81) cm	0.64 (t test)
Median exophytic part (IQR)	50% (20–60)	50% (30–60)	0.85 (MWU test)
Median number of biopsies (IQR)	2 (2–3)	3 (2–3)	0.39 (MWU test)
Tumor location
Upper pole (n=33)	26 (79%)	7 (21%)	0.77
Interpolar (n=41)	33 (80%)	8 (20%)	(chi square)
Lower pole (n=26)	19 (73%)	7 (27%)
Size of biopsy needle
16-gauge (n=17)	13 (76%)	4 (24%)	0.87
18-gauge (n=83)	65 (78%)	18 (22%)	(chi square)
Quality of biopsies^*
Good (n=22)	15 (68%)	7 (32%)	0.32
Intermediate (n=30)	25 (83%)	5 (17%)	(chi square)
Poor (n=42)	32 (76%)	10 (24%)
NA (n=6)	-	-
Nonenhancing regions within tumor
Present (n=58)	47 (81%)	11 (19%)	0.36
Absent (n=39)	28 (72%)	11 (28%)	(chi square)
NA (n=3)	-	-

For six biopsies, the original specimen was not available, and therefore the quality could not be assessed.

SD=standard deviation; IQR=interquartile range; MWU=Mann-Whitney U.

Discussion

In our series on laparoscopic cryoablation of SRMs, 22% of intraoperatively taken biopsies were nondiagnostic. This is within the range reported in other studies on renal cryoablation, although figures vary widely with 6% to 32% for laparoscopic cryoablation and 11% to 23% for percutaneous cryoablation.^{6,7,11

–18} The wide range can be caused by different reasons, such as tumor characteristics, procedural details, variation in the definition of a diagnostic biopsy, intraobserver variability,^19,20 and inherent pathologic diagnostic difficulties in RCC.

In the present study, a diagnostic biopsy was strictly defined by tissue that would explain a renal mass on an imaging study, such as carcinoma or benign renal tumors. In some studies, biopsies consisting of normal renal tissue are also categorized as a diagnostic biopsy. We strongly believe, however, that the presence of normal renal tissue usually implies a sampling error, because theoretically renal masses are not composed of normal renal tissue. In other studies, the presence of fibrosis and necrotic tissue is regarded as a sign of tumor and categorized as a diagnostic biopsy. We also regarded this as nondiagnostic because of the absence of tumor cells and therefore the inability to determine a definitive diagnosis.

When comparing the ranges of nondiagnostic biopsies between laparoscopic and percutaneous cryoablation series, it is remarkable that these numbers differ so little; one would expect that the laparoscopic vision would be an advantage in obtaining a biopsy and therefore the nondiagnostic biopsy rate would be lower. We experienced, however, that with the biopsy needle suspending freely in the abdominal or retroperitoneal cavity during laparoscopy, it is difficult to aim the exact location of the biopsy, something that is not encountered when the biopsy is taken percutaneously. Perhaps the use of a guiding sheath during laparoscopy could be helpful and improve biopsy results.

In the present study, none of the studied parameters on radiologic tumor characteristics, on biopsy needle size, number of biopsies, or quality of biopsies was found to influence the percentage of nondiagnostic biopsies significantly.

The size of the tumor (all tumours in our study were ≤4.5 cm) did not influence biopsy outcome. Wunderlich and colleagues,²¹ however, did find that tumor size affected biopsy outcome when including larger sized tumors. In their series, 250 ex-vivo biopsies were taken from 50 tumors (range 2–20 cm), one central and four peripheral biopsies from each tumor. For the 30 tumors that were smaller than 4 cm, 16.7% of the central biopsies and 25% of the peripheral biopsies consisted of necrosis and/or fibrosis as a result of which the tumor biology (benign or malignant) was indefinable. For the 20 tumors that were larger than 4 cm, this was the case for 30% of the central biopsies and 32.8% of peripheral biopsies. In our study, the percentage of biopsies (all our biopsies were taken centrally) that contained mostly necrosis and/or fibrosis was 10% (Table 1).

We found that significantly more biopsies were taken in larger tumors. In our setting, the number of biopsies depended on the macroscopic appearance of the obtained biopsy assessed by the surgeon (ie, the biopsy gun was fired until a proper core of tissue was yielded). Furthermore, there was a positive correlation (P=<0.001) between tumor size and the presence of intratumoral nonenhancing areas on imaging suggestive for necrosis. This is compatible with the conclusion of Wunderlich and coworkers²¹ that larger tumors result in a higher amount of nondiagnostic biopsies, especially when taken centrally from the tumor, as was done in our series.

After dissection of the perirenal fat, an exophytic tumor protruding from the kidney surface is more easily recognizable than a predominantly intrarenal tumor. We expected that under laparoscopic vision, it might be easier to take a diagnostic biopsy from a more exophytic tumor than from a less exophytic tumor. This is not demonstrated by our results, however. A possible explanation is the use of intra-abdominal US; in this way, even scarcely exophytic tumors can also easily be identified intraoperatively, provided that they are not isoechoic.²² Furthermore, we found tumor size to be correlated with a larger exophytic part of the tumor, which is expected, because large tumors are more likely to deform the kidney surface.

Laparoscopically, the upper pole is usually more difficult to reach with a biopsy needle than the interpolar area and lower pole—the reason we included this parameter in our study. Nonetheless, our analysis shows that this did not significantly influence the biopsy outcome.

Breda and associates²³ compared the accuracy of three different sized biopsy needles in a prospective, ex-vivo study. Their study shows that 100% of the biopsies obtained with the 14- and 18-gauge biopsy needles were diagnostic, while this number decreased to 84% with the 20-gauge needle. The histologic accuracy of the 14-, 18-, and 20-gauge needles was 92%, 97%, and 81%, respectively. They therefore advise to use an 18-gauge biopsy needle as a minimum size. This was the case in our center where intraoperative biopsies were performed with an 18-gauge needle. The proven safety of larger needles²⁴ together with an interim analysis of our results prompted us to use a 16-gauge needle in an attempt to improve our diagnostic rate. Therefore we were able to assess potential advantages of a larger caliber needle in an in-vivo setting. Needle size did not turn out in a statistically significant increase in the rate of diagnostic biopsies, but larger needle size positively influenced the quality of the biopsies. In concordance with this, Hruby and colleagues²⁵ found that larger caliber needles (12–14 gauge vs 16–20 gauge) resulted in improved tissue specimens in terms of width and number of glomeruli and vessels captured in a porcine model.

To our knowledge, there is no standardized way to score the quality of a biopsy. We based our scoring system on the length and number of (fragmented) cores after preparation for pathologic assessment. According to our definition, a total of 42 biopsies were of poor quality (32 [76%] of the diagnostic biopsies and 10 [24%] of the nondiagnostic biopsies). Apparently, the macroscopic appearance of the biopsies says little about the microscopic appearance and the ability of the pathologist to make the diagnosis.

The main limitation of this ad hoc study is the retrospective collection of some of the variables and the lack of randomization. Consequently the number of biopsies, and the needle size reflect clinical practice and, as mentioned, the attempt of improving diagnostic results. The sample size might seem insufficient to show any statistically significant difference in an observational study; however, indications to perform laparoscopic renal cryoablation are limited. A randomized controlled trial shall definitively answer the question, although a big sample would be necessary to adjust for the number of core biopsies related to tumor size and biopsy needle caliber. A certain percentage of nondiagnostic biopsies is ultimately unavoidable and inherent to the sampling of SRMs.

Efforts should be made to improve the diagnostic rate in ablation of SRMs, because decreasing the nondiagnostic rate has significant impact on follow-up, health care costs, and patient apprehension.

Conclusion

In our setting, 22% of intraoperative biopsies are nondiagnostic during laparoscopic cryoablation. Radiologic tumor characteristics (tumor size, exophytic part, tumor location, or nonenhancing parts in the tumor) or the number of biopsies taken, the size of the biopsy needle, or the quality of the biopsies did not predict the biopsy outcome.

Footnotes

Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Frank

, Blute

, Cheville

et al. Solid renal tumors: An analysis of pathological features related to tumor size. J Urol, 2003; 170:2217–2220.

Smith

. Complications of percutaneous abdominal fine-needle biopsy. Review. Radiology, 1991; 178:253–258.

Terrone

, Volpe

. The role of pathology for clinical decision-making in renal cell carcinoma is increasing. Eur Urol, 2007; 51:1166–1170.

Volpe

, Kachura

, Geddie

et al. Techniques, safety and accuracy of sampling of renal tumors by fine needle aspiration and core biopsy. J Urol, 2007; 178:379–386.

Gill

, Remer

, Hasan

et al. Renal cryoablation: Outcome at 3 years. J Urol, 2005; 173:1903–1907.

Davol

, Fulmer

, Rukstalis

. Long-term results of cryoablation for renal cancer and complex renal masses. Urology, 2006; 68,suppl 1:2–6.

Schwartz

, Rewcastle

, Powell

et al. Cryoablation of small peripheral renal masses: A retrospective analysis. Urology, 2006; 68,suppl 1:14–18.

Cestari

, Guazzoni

, dell'Acqua

et al. Laparoscopic cryoablation of solid renal masses: Intermediate term followup. J Urol, 2004; 172:1267–1270.

Laguna

, Kümmerlin

, Rioja

, de la Rosette

. Biopsy of a renal mass: Where are we now? Curr Opin Urol, 2009; 19:447–453.

10.

Lopez-Beltran

, Scarpelli

, Montironi

, Kirkall

. 2004 WHO classification of the renal tumors of the adults. Eur Urol, 2006; 49:798–805.

11.

Permpongkosol

, Link

, Solomon

, Kavoussi

. Results of computerized tomography guided percutaneous ablation of renal masses with nondiagnostic pre-ablation pathological findings. J Urol, 2006; 176:463–467.

12.

Silverman

, Tuncali

, vanSonnenberg

et al. Renal tumors: MR imaging-guided percutaneous cryotherapy—initial experience in 23 patients. Radiology, 2005; 236:716–724.

13.

Gupta

, Allaf

, Kavoussi

et al. Computerized tomography guided percutaneous renal cryoablation with the patient under conscious sedation: Initial clinical experience. J Urol, 2006; 175:447–453.

14.

Desai

, Aron

, Gill

. Laparoscopic partial nephrectomy versus laparoscopic cryoablation for the small renal tumor. Urology, 2005; 66,suppl 5:23–28.

15.

Laguna

, Beemster

, Kumar

et al. Perioperative morbidity of laparoscopic cryoablation of small renal masses with ultrathin probes: A European multicentre experience. Eur Urol, 2009; 56:355–361.

16.

Lawatsch

, Langenstroer

, Byrd

et al. Intermediate results of laparoscopic cryoablation in 59 patients at the Medical College of Wisconsin. J Urol, 2006; 175:1225–1229.

17.

Bandi

, Wen

, Hedican

et al. Cryoablation of small renal masses: Assessment of the outcome at one institution. BJU Int, 2007; 100:98–801.

18.

Atwell

, Farrell

, Leibovich

et al. Percutaneous renal cryoablation: Experience treating 115 tumors. J Urol, 2008; 179:2136–2141.

19.

Kümmerlin

, ten Kate

, Smedts

et al. Core biopsies of renal tumors: A study on diagnostic accuracy, interobserver, and intraobserver variability. Eur Urol, 2008; 53:1219–1225.

20.

Kümmerlin

, Smedts

, ten Kate

et al. Cytological punctures in the diagnosis of renal tumours: A study on accuracy and reproducibility. Eur Urol, 2009; 55:187–195.

21.

Wunderlich

, Hindermann

, Al Mustafa

et al. The accuracy of 250 fine needle biopsies of renal tumors. J Urol, 2005; 174:44–46.

22.

Reichelt

, Gajda

, Chyhrai

et al. Ultrasound-guided biopsy of homogenous solid renal masses. Eur Urol, 2007; 52:1421–1426.

23.

Breda

, Treat

, Haft-Candell

et al. Comparison of accuracy of 14-, 18- and 20-G needles in ex-vivo renal mass biopsy: A prospective, blinded study. BJU Int, 2010; 105:940–945.

24.

Wood

, Khan

, McGovern

et al. Imaging guided biopsy of renal masses: Indications, accuracy and impact on clinical management. J Urol, 1999; 161:1470–1474.

25.

Hruby

, Lehman

, Mitchell

et al. Optimizing renal cortical neoplasm tissue sampling through a modified biopsy technique: Laboratory experience and initial clinical experience. Urology, 2007; 70:431–434.