Percutaneous cryoablation of renal cell carcinoma guided by horizontal open magnetic resonance imaging


Kenta Miki md, Department of Urology, Jikei University School of Medicine, 3-25-8 Nishi-Shinbashi, Minato-ku, Tokyo 105-8461, Japan.


Background:  Cryoablation is a treatment option for some patients with small exophytic lesions of the kidney. The purpose of this study is to determine the feasibility, safety, and intermediate-term treatment outcome of percutaneous cryoablation of renal cell carcinoma guided by horizontal open magnetic resonance imaging (MRI).

Methods:  We prospectively used cryoablation to treat 13 patients with radiographically confirmed enhancing small, solid renal tumors (≤4.8 cm). An argon gas-based cryoablation system was used. One to four cryoprobes with 2 or 3-mm diameters were placed percutaneously into the tumor under local anaesthesia and MRI guidance. Ice ball dimensions were monitored by 2-D MR images. Double freeze-thaw cycles were conducted throughout the procedure. After successful cryoablation, patients were followed on a regular basis to evaluate the treatment’s clinical outcome.

Results:  Median follow up from time of procedure is 35 months (range, 28–42). In all cases the entire procedure was accomplished without significant morbidity or complications. A mild retroperitoneal hematoma, which subsided spontaneously, was noted in one patient. Follow-up dynamic computed tomography (CT) at 3 months after operation confirmed the absence of enhancement in resolved tumor masses for 11 of 13 cases. None of these 11 patients had clinical evidence of recurrent disease at last follow up. The remaining two patients had lesions with some enhanced areas. Subsequent partial nephrectomy histologically confirmed the presence of vital tumor in, respectively, the center and the periphery of the residual masses. One of these patients developed multiple lung and ipsilateral adrenal metastases 13 months after surgical resection.

Conclusions:  Percutaneous cryoablation of small renal cell carcinomas under horizontal open MRI guidance appears to be safe and feasible. An intermediate-term follow up continues to demonstrate efficacy in most patients; however, a few patients experience incomplete ablation with risk of treatment failure. The ideal candidates for this procedure still need to be determined in longer follow up with diligent observation.


Extensive application of modern diagnostic imaging, including abdominal ultrasonography, computed tomography (CT) and/or magnetic resonance imaging (MRI), has led to more frequent serendipitous identification of renal masses in patients without relevant signs or symptoms.1–5 For the last 50 years the standard treatment for renal cell carcinoma has been radical nephrectomy.6 However, the histological characteristics of these small renal masses are typically those associated with slow growth.7–10 Recent studies have shown that nephron-sparing surgery is equally effective in the treatment of small renal masses.7–10

Less invasive therapy options that may better preserve renal function are also being investigated.11–13 In 1996, Delworth et al. described cryoablation of renal cell carcinoma via an open approach.11 Gill et al. performed cryoablation using a laparoscopic approach in a series of 32 cases.12

Percutaneous cryoablation for a small renal mass under horizontal MRI is being developed as a minimally invasive alternative to open or laparoscopic partial nephrectomy.13 MRI allows accurate depiction of the full extent of the ice ball, with excellent contrast between ice and the surrounding tissue. We recently reported the short-term results of our initial experience.14 In the current study, we present the intermediate-term radiographic data from our expanded experience with percutaneous cryoablation.


Entry criteria and patient characteristics

From March 2001 to May 2002, 13 patients underwent percutaneous renal cryoablation at Jikei University Kashiwa Hospital. Mean patient age was 61.9 years (range, 46–73). Tumor size averaged 2.7 cm (range, 2–4.8). One patient was diagnosed with a renal mass in his solitary kidney 7 years after contralateral radical nephrectomy and ipsilateral partial nephrectomy for renal cell carcinoma. The remaining 12 patients had bilateral kidneys at the time of cryoablation. The patient demographic data are listed in Table 1.

Table 1.  Descriptive statistics of the 13 patients who underwent renal cryoablation
  1. Data shown as the mean with the range in parenthesis.

Patients (n)13
Follow-up (months)35 (28–42)
Age (year)61.9 (46–73)
Preoperative mass size (cm) 2.7 (2.0–4.8)
Tumor location (n)
 Anterior 1
 Posterior 2
Tumor position (n)
 Upper pole 4
 Middle 8
 Lower pole 1
Serum creatinine (mg/dL)
 Preoperative 0.70
 Postoperative 0.77
Hospital stay 2.4 (2–7)
Operative blood loss (mL)<10

The study inclusion criteria have been described previously.14 Briefly, all patients had been radiographically diagnosed with a peripherally located circumscribed renal mass that was ≤10 cm in size if solitary or ≤3 cm if multiple; patients were between 20 and 75 years of age with adequate life expectancy. Patients with central renal lesions were not considered for treatment by this technique. Patients were excluded if there was evidence of metastatic disease on any staging study, including laboratory tests, chest radiography, CT, and MRI. All patients were thoroughly counseled regarding the procedures that are considered the standard of care, including radical nephrectomy and partial nephrectomy. After detailed discussion, written informed consent was obtained. This study design was approved by the institutional review board.

Technique and operative procedure

Percutaneous cryosurgery was performed using an argon-based cryoablation system (CRYO-HIT; Galil Medical, Yokneam, Israel) that introduces pressurized argon gas through the probe’s inner lumen. The inner lumen is depressurized through a ball valve at the probe tip into the outer lumen, where the gas is carried up the probe. Depressurization of argon gas results in absorption of surrounding heat and produces a temperature of −185°C. The thawing system uses pressurized helium gas, which functions similarly and may generate temperatures up to 35°C. A 2- or 3-mm reusable probe made of MRI-compatible metal was used for ice ball formation.

The patient was positioned prone on the interventional MRI docking table and advanced into the magnet bore. A flex-4 send and receive coil was positioned around the torso. Axial fast spin-echo images were obtained to localize the kidneys and target the mass. These images were also used to identify the planned probe entry site. The skin site was marked, prepared and draped in normal sterile fashion. After anaesthetizing the skin with 10–15 mL of 1% lidocaine, a 2–3 mm incision was made, through which a cryoprobe 2 or 3 mm in diameter and approximately 15 cm long was inserted with its corresponding 7-F or 10-F pinnacle introducer sheath. Axial fast spin-echo images generated at up to two images per second were used to monitor cryoprobe advancement toward the mass.

The probe was advanced through the center of the mass until the tip was positioned along its distal inner border. For anteriorly located (ventral) tumors the probe was directed through the parenchyma and away from the collecting system, with the angle of approach being calculated to maximize the distance from the collecting system. The cryosystem was activated for 1 min of freezing, at which time imaging was repeated to determine the growth rate of the ice ball in the mass and avoid possible injury to the collecting system. The ice ball that developed was represented by a signal void on T1-weighted images. The ice ball’s size was increased until it completely enveloped the mass and the edge extended 5 mm beyond the margin. If the size or configuration of the mass was such that the maximum ice ball generated failed to obliterate the targeted tissue, the probe was thawed and repositioned using an overlapping technique varied according to lesion size and location. Freezing time was less than 20 min per cycle, mainly dependent on the size of the ice ball. Thawing time was less than 10 min per cycle. Double freeze-thaw cycles were conducted throughout the procedure. The probe was then removed after the second thaw cycle and the access sheath was packed with absorbable gelatin sponge to facilitate hemostasis. A bandage was applied to the incision site. The patient was then removed from the interventional MRI unit and hospitalized overnight. A complete blood count and serum electrolyte tests were performed the morning after the procedure to monitor any hemorrhage or electrolyte disturbance.

Patient evaluation following cryoablation

The follow-up schedule included chest X-ray, dynamic CT, blood urea nitrogen, serum creatinine, and physical examination at 2 and 6 weeks and at 3-month intervals beginning 6 months after operation. Nonenhancement of the cryolesion was considered the primary radiographic hallmark of successful cryoablation. The existence of enhancing areas in the residual tumor mass after 12 months was considered an indication of local failure.

Statistical analysis

Changes in hemoglobin and creatinine levels pre- and postoperatively were examined using the Mann–Whitney U-test with P < 0.05 as significant.


Intraoperative data

The mean surgical time was 156.7 min (range, 110–265). The freeze-thaw cycle took approximately one-third of the surgical time. The freeze-thaw time could not be accurately measured since freezing was continued until a 5-mm margin around the tumor was confirmed. The mean blood loss was estimated at less than 10 mL (Table 1).

Postoperative data

The hospital stay averaged 2.4 days (range, 2–7). The mean preoperative and postoperative serum creatinine level was 0.70 and 0.77 mg/dL, respectively (P > 0.05). The mean preoperative and postoperative serum hematocrit level was 41.1 and 40.9%, respectively (P > 0.05). Median follow up after the procedure was 35 months (range, 28–42).

Complications and morbidity of cryoablation

Percutaneous cryoablation of small renal cell carcinoma was accomplished safely in all 13 cases. No patients required open conversion. No episodes of urinary extravasation, significant gross hematuria, or urinary obstruction occurred. A mild perinephric hematoma, which subsided spontaneously without intervention, was noted in one patient. No other major operative complications were encountered.

Treatment outcome

Most of the cryolesions continued to regress over time (Fig. 1). No residual cryolesion could be identified in three of the 13 patients who completed dynamic CT scanning at 1-year follow up. Serial changes in mean (± standard deviation) cryolesion diameter at 2 and 6 weeks and at 6 and 9 months were, respectively, 4.4 ± 0.7, 3.9 ± 0.6, 2.9 ± 0.3 and 2.6 ± 0.7 cm. Given the baseline tumor diameter of 2.7 cm, this represents a mean change of 176 ± 39.4%, 149 ± 28.0%, 99 ± 26.4%, and 98 ± 32.6%, respectively, at these four intervals. The tentative increase in diameter of the cryolesions immediately following cryoablation is most likely due to edema in the tissue surrounding the tumor.

Figure 1.

Representative case of tumor regression following cryoablation of left renal tumor (arrows). The cryolesion progressively regressed with time following an initial increase in size due to edema (wk, weeks; M, months).

Follow-up dynamic CT 3 or more months postoperatively confirmed no enhancement within resolved tumor masses in 11 of the 13 cases. None of these 11 patients had clinical evidence of recurrent disease at last follow up. At 12 months after cryoablation the remaining two patients still had enhanced areas within their tumors, either in the center (Fig. 2b) or at the peripheral margin (Fig. 3b). Pretreatment size and location of these two tumors were 30 and 25 mm in diameter, located at ventral-medial and dorso-upper aspect of right kidneys, respectively.

Figure 2.

(a) A case of treatment failure following cryoablation. Original tumor, pretreatment image. (b) Centrally enhanced cryolesion of the right kidney is shown in the follow-up dynamic computed tomography 3 months postoperatively (arrow). A large vessel feeding the tumor is also delineated (arrowhead). (c) Histology of the cryolesion in the partial nephrectomy specimen confirmed presence of viable tumor cells.

Figure 3.

(a) A case of treatment failure following cryoablation. Original tumor, pretreatment image. (b) Peripheral rim-like enhancement is shown in the cryolesion of the right kidney in the follow-up dynamic computed tomography 6 months postoperatively. (c) Histology of the cryolesion in the partial nephrectomy specimen confirmed presence of viable tumor cells.

Subsequent partial nephrectomy led to histological confirmation of the presence of vital, clear cell type, G1 ≥ 2 and G2 tumors in the center and the peripheral margin of the residual masses, respectively (Figs 2c,3c). The latter patient developed multiple lung as well as ipsilateral adrenal metastases 13 months after surgical resection. This patient is now under interferon treatment and was still alive at the time this paper was submitted. The other patients continue to have no evidence of disease to date. An intercurrent death due to colon carcinoma was noted in one other patient.


Renal cryoablation using open, percutaneous, or laparoscopic techniques has been employed as a promising alternative to nephron-sparing surgery.4–6 Complete in vivo necrosis of porcine renal parenchyma has been consistently achieved at temperatures of −19.4°C or lower.15 The target lethal temperature of −20°C could be achieved at a depth of 3.1 mm below the leading edge of a 3.2 cm ice ball.16 Optimal outcome thus requires the ice ball to extend well beyond the visible margins of the targeted tumor. Based on these observations, standard technique calls for ice ball formation to extend 5–10 mm beyond the tumor margin to ensure successful cryoablation.

One of the primary criticisms of this technique has been the lack of histologic data regarding the completeness of tumor destruction. Needle biopsies are often used to confirm the elimination of tumor cells. Doceht et al. however, reaffirmed that there is a large degree of inaccuracy in renal biopsy even with open renal sampling in patients undergoing nephrectomy for renal masses.16 A negative postoperative biopsy thus does not necessarily preclude the presence of undetected cancer after cryoablation. Although not required by our original protocol, we attempted an MR-guided biopsy in one case at 12 months following cryoablation; results of this biopsy were negative.

A more serious difficulty is inability to evaluate the margin status of the ablated lesions. We experienced two cases of treatment failure subsequent to incomplete ablation. The cause of failure in one patient was presumed to be an insufficient ice ball margin and poor targeting. This patient subsequently developed distant metastases to the lung and ipsilateral adrenal gland. To our knowledge, however, there have been no reports in the world literature that suggest direct relationship between cryo-procedure and hematogenous metastases. Though the etiology of distant metastases is beyond the scope of this report, caution must be taken for such a possibility. In the other patient the likely cause of incomplete ablation was presence of a large vessel close to the tumor, making it impossible to achieve the required lethal temperature.

Differences in cryotechnique may be worthy of comment. Although other investigators have recommend extension of the ice ball circumferentially to at least 6 mm – preferably 1 cm – beyond the tumor edge,17–19 we elected to set the distance at 5 mm for safety reasons. Since the destructive effects of cryoablation are not limited to the kidney, adjacent organs must be protected from contact with the ice ball. Complete bowel obstruction,17 stricture of the uretero-pelvic junction,20 and pancreatic injury18 have been reported. Other potential complications of renal cryosurgery include post-thaw hemorrhage and urine leakage secondary to caliceal cryoinjury with resulting fistula formation. These safety considerations and the resulting strict limits on ice ball extension may have led to a more significant peripheral sublethal injury zone in our series. Insufficient cryoablation has also been reported by other investigators, however.21,22

Meticulous postoperative follow-up is critical in renal cryoablation. The cryoablated renal tumor may remain radiographically visible at 1 years, as in our cases (Fig. 1). However, definite criteria for tumor recurrence in renal cryolesions and consequent intervention have not been established. Some investigators recommend intensive surveillance, including both CT guided core needle biopsy and radiologic follow up, until the cryolesion is no longer visible.18 The use of MR imaging has been advocated on the basis of its soft-tissue contrast resolution and multiplanar imaging capability.19 Variable MR findings have been noted in the cryolesions. All cryolesions showed a dramatic progressive decrease in size, as also seen by other investigators.19 Any increase in size or change in internal enhancement should be viewed with suspicion. Future studies with longer follow up and more patients with recurrent disease are needed to differentiate normal changes after cryotherapy from tumor recurrence.

Many improvements in cryosurgical equipment have been made in recent years and clinical precedents have been established. The use of MRI to monitor cryoablation offers clear advantages in visualizing both the tumor mass and the formation of the black ice ball. An imaging system designed for the interventional MRI unit (Hitachi Airis; Hitachi Medico, Kashiwa Chiba, Japan) allows the surgeon direct access to the patient for probe placement. This technology, with the patient in a prone position under local anaesthesia, makes the percutaneous approach more practical. The ability of the MRI to produce complete, unobscured images in multiple planes in near real time increases the likelihood that no viable tumor will be left outside the margins of the ice ball.

Although early data from our series as well as others11–13,17 are encouraging, the traditional surgical approach should still be considered the gold standard when feasible. At least for the present, cryoablation should preferably be reserved for elderly patients, especially in imperative cases such as tumor in solitary kidney. Five-year follow-up data will soon be available to confirm the durability of renal cryotherapy. More work is needed to delineate the steps that will ensure cell death at the margins of the cryolesion and to further determine the long-term safety and efficacy of the procedure.