Predictors of Complications After Percutaneous Image-Guided Renal Cryoablation for T1a Renal Cortical Neoplasms

Percutaneous cryoablation

We performed all PCA procedures in a hospital CT suite with the patients under general anesthesia or local anesthesia with conscious sedation. Our technique of PCA has been previously described. ^18,19 All procedures were performed jointly by an interventional radiologist and urologist working together. Our intraoperative technical objective was extension of the ice ball beyond the tumor margin, including a margin of normal tissue.

Oncologic outcomes

In patients with adequate renal function and no allergy for iodinated contrast, CT both with and without intravenous contrast was obtained after cryoprobe removal at the end of the procedure to document technical success and assess potential complications.

We obtained additional follow-up CT or MRI 3–6 months after the initial procedure and annually thereafter. We defined local tumor recurrence as new enhancement in the area of the ablated tumor with initial negative imaging or enlargement in tumor dimension beyond 6 months following the procedure. We considered enhancement at the time of initial imaging or at imaging within 6 months as a residual tumor.

Statistical analyses

The mean and standard deviation were used to describe categorical variables, and the frequency and percentage were used to describe continuous data. We divided patients into two groups based on the occurrence of major peri- and postoperative complications. Comparisons of baseline characteristics, including age at surgery, gender, ethnic group, body mass index, year of surgery, American Society of Anesthesiologists (ASA) score group, tumor greatest dimension (in cm, continuous), preablation biopsy, number of probes used, and nephrometry score (continuous and divided into three complexity groups: low [4–6], moderate [7–9], and high [10–12]), between patients with complications and those without complications were performed using chi-squared test for categorical variables, and the rank sum and Student's t-test for continuous data. Multivariable analysis of the patient and disease characteristics by complication status was done with logistic regression. Given the collinearity between tumor dimension, the number of cryoprobes used, and nephrometry score, only tumor dimension was included in the initial multivariable model. We then analyzed the association of tumor dimension, number of cryoprobes used, and nephrometry score with major complications separately in univariable and multivariable analyses adjusting for patient and disease characteristics. All statistical analyses were two-tailed and performed using Stata 12.0 (StataCorp, College Station, TX). A p value <0.05 was considered statistically significant.

Oncologic outcomes

Overall malignancy rate was 78% with 11% benign rate and 11% of nondiagnostic rate. Histopathologic features of patients with and without complications are described in Table 1. Type of histopathology on preablation biopsy was not predictive of complications (p = 0.65). With a mean follow-up of 24 months (6–64), 11 patients (5.8%) experienced tumor recurrence.

Tumor characteristics (dimension, location, complexity, and R.E.N.A.L. score)

The American Urological Association considers PCA to be a viable treatment option for high-risk surgical patients with T1a (≤4 cm) SRM. ¹ PCA has been shown to achieve good local tumor control with complication rates <10% in SRMs. Local recurrence-free survival in these cases is >90% at 5 years after PCA. ²⁰ SRM with a diameter of ≥3.0 cm, however, has been reported as an independent predictor of complications in patients undergoing PCA and may present challenges for the PCA approach. ^3,21,22 Lehman and colleagues were first to demonstrate tumor dimension of >3.5 cm to be associated with increased complications. ²² In a recent single-center evaluation of perioperative outcomes following PCA, Kim and colleagues reported that larger tumor dimension (>3.0 cm) was the only significant predictor of PCA treatment failure. ³ Increased tumor dimension is consistently associated with complications, such as bleeding, hematoma, and postoperative hematuria. ^23,24 Tumor dimension remains one of the main criteria for patient selection for optimal surgical outcomes. In the current study evaluating just T1a lesions, tumors up to 3 cm had a complication rate <3.5%, while those >3 cm had a 33% recurrence rate. The significant difference in complications based on size should be incorporated into clinical decision-making and planning.

Previous studies have demonstrated the impact of tumor location on complication rates during PCA. Hruby and colleagues. were first to observe and report that centrally located and hilar tumors are more prone for bleeding due to ice ball cracking during the ablation. ²⁵ Subsequent studies have confirmed these findings. ^3,10 Thus, centrally located, endophytic, and hilar tumor should be approached with a great care. ²⁶

Anteriorly located tumors present unique surgical challenges due to location and adjacent organs. However, expertise and utilization of advanced surgical techniques such as hydrodissection for mobilization of surrounding bowel and organs has demonstrated promising results. ²⁷ This is supported by the work of Schmit and colleagues where complication rates did not significantly differ by location, demonstrating rates of 5.5%, and 4.5% for anterior and posterior tumors, respectively (p = 0.074). ²⁸ However, these techniques are carried out in the hands of experts with extensive experience in image-guided ablative procedures at tertiary institutions. As such, the literature has suggested that partially exophytic, posterior, and laterally located tumors are more amenable for PCA. ²⁹

To quantify renal tumor complexity and improve outcome prediction as well as academic reporting, Kutikov and colleagues developed a comprehensive standardized system called R.E.N.A.L. nephrometry. ¹⁷ The system comprises all renal anatomic features into one scoring system by providing a quantitative measure of tumor complexity. A correlation between increasing R.E.N.A.L scores and surgical outcomes following ablative procedures has also been reported. ^28,30,31 Blute and colleagues reported that with each unit increase in R.E.N.A.L. score, patients were 1.5 times more likely to experience a complication. ³¹ In addition, Schmit and colleagues reported mean nephrometry scores of 8.1 ± 2 vs 6.8 ± 1.9 (p < 0.001) for patients in whom major complications did vs did not develop, respectively. Because of the increased risk of major complication and local treatment failure, they suggest avoiding PCA in patients with R.E.N.A.L scores of 10–12. ²⁸ Overall reports published demonstrated that moderate and high complexity tumors, as defined by the R.E.N.A.L. scoring system, will likely experience major complications thus should be avoided. All previous studies evaluating nephrometry score included all tumors above 4 cm, thus our analysis excluding those >4 cm patients, did not show that nephrometry score was predictive of complications. In terms of tumor features, size remains the main factor that should be taken into consideration in surgical planning of patients with SRM.

Patient characteristics

Indications for PCA are similar to indications for other nephron-sparing procedures. Elderly patients with shorter life expectancies or high-risk patients with multiple comorbidities may be good candidates for PCA. In addition, due to its less invasive nature, PCA is preferred for patients with prior renal surgeries, a solitary kidney, multiple tumors, and those with inherited genetic diseases (e.g., von Hippel-Lindau Syndrome, in which patients have a propensity for developing numerous renal tumors over the course of a lifetime). ²⁹ In addition, in the setting of locally recurrent tumors after initial laparoscopic partial nephrectomy or ablative therapy, PCA remains the procedure of choice. Contraindications for PCA include history of coagulopathy or the presence of overlying abdominal organs, such as the spleen or liver, preventing proper probe placement. ³² Although our own data did not reveal these findings, these considerations remain crucial for patient safety and optimal outcomes. Finally, Blute and colleagues demonstrated the importance of skin-to-tumor distance in patients undergoing PCA. ³¹ A longer ablation track, which correlates with patient's higher body mass index, is not a favorable factor for optimal outcomes. With each centimeter increase of skin-to-tumor distance, there was 1.5-fold increase in the likelihood of tumor recurrence. Further studies are necessary to validate these findings.

Collaborative ablative approach and surgical proficiency

It is suggested that PCA procedures will have optimal results both technically and clinically when performed by a urologist and an interventional radiologist. ³³ In 2006, percutaneous ablation procedures involved radiology alone at 58% of institutions, urology alone at 9%, and collaboration between radiologists and urologists at 31%. ³⁴ By 2008, collaboration between a radiologist and urologist occurred in 51% of responding institutions, with urologists alone and radiologists alone performing 35% and 13% of PCA procedures, respectively. ⁷ More recently, Patel and colleagues reported that interdisciplinary collaboration occurred at 42% of all institutions, with urologists regularly being present at the time of ablation (59%) and being responsible for placing needles for ablation (32%). ⁸ In the current study, all procedures were performed as a collaborative effort between urologist and interventional radiologist.

Analysis of our data demonstrated that PCA procedures performed in our early experience before 2008 were associated with significantly higher complications compared with more recent years. There is an obvious learning curve with PCA, as there is with any surgical procedure. Thus, collaborating with interventional radiology is crucial to avoid life-threatening complications and optimize outcomes. To support this, multiple studies have examined practice patterns at academic institutions in the use of ablative technologies for the management of SRMs. ^{7 –9} In 2006, Bandi and colleagues reported that cryoablation was offered by 93% of responding institutions, with cryoablation being more frequently utilized than radiofrequency ablation (79% and 55%, respectively). ⁷ We believe that this is directly related to surgeons comfort with performing the procedure and availability of equipment and interventional radiology colleagues. Moreover, limitations with PCA training lay in the core of resident education. In a study conducted by Elliot and colleagues, ⁹ only 54% of urology residents had ever participated in an ablative procedure during residency training. Of those who had participated, 57% were involved in fewer than five procedures. ⁹ Limited exposure and training can further hinder surgical proficiency. In 2013, George and colleagues incorporated a minimum 2-month interventional radiology rotation for seven urology residents. Residents experienced significant increases in performance proficiency in endovascular technique, ablative therapies, and needle biopsy following the 2-month training. ³⁵ These findings report that expanding urologic residency curriculum to include interventional radiology would be beneficial.

Technology