| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Massachusetts General Hospital, 55 Fruit St., White
270, Boston, MA 02114.
2 Department of Urology, Massachusetts General Hospital, Boston, MA 02114.
Received June 29, 2004;
accepted after revision October 6, 2004.
Address correspondence to D. A. Gervais.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Over 6 years, 100 renal tumors in 85 patients underwent radiofrequency ablation. Images were reviewed to determine the following: effect of initial electrode approach at and parallel to the tumor-kidney interface; effect of collecting system proximity to the tumor and to the zone of ablation; bowel proximity to the tumor and strategies to protect bowel; patterns of residual disease; and approaches at subsequent sessions.
RESULTS. The initial placement of the electrode at and parallel to the tumor-kidney interface did not result in significantly fewer overlapping ablations (p = 0.91) or a lower rate of residual disease (p = 0.86). Direct contiguity of tumor or zone of ablation to the collecting system did not increase the complication rate. However, obscuration of calyces by a central tumor was a significant predictor of collecting system hemorrhage necessitating treatment (p < 0.001) seen in three of nine tumors obscuring calyces. Clinically significant urine leaks were rare (1/100) and related to downstream obstruction. There were no bowel complications despite the fact that 27 of the tumors were within 1 cm of bowel. Protective strategies progressed from reliance on electrode positioning to hydrodissection. Residual patterns were predominantly nodules or crescents, and straight electrodes were commonly used to approach residual disease.
CONCLUSION. Initial electrode position at and parallel to the tumor-kidney interface does not result in less difficult or more successful ablations. Contiguity of tumor or zone of ablation to the collecting system does not increase the risk of complications, but obscuration of calyces does. Bowel complications are rare, and protection with hydrodissection is becoming more common. Residual tumor presents as nodules or crescents of persistent enhancement.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Radiofrequency Ablation
Radiofrequency ablation was performed by one of three radiologists with
experience in radiofrequency ablation as the primary operator in consultation
with the other two radiologists with respect to treatment planning. The choice
of imaging technique, radiofrequency system, and angle of approach of the
electrode was at the discretion of the primary operator. Multiple overlapping
ablations were performed as needed to cover the entire tumor, making sure to
ablate the region of the interface of tumor and normal kidney regardless of
the angle of approach of the electrode. The number of overlapping ablations
performed was based on tumor size and geometry. If residual disease was
detected on imaging after ablation, repeat ablation sessions were scheduled as
needed and as appropriate depending on the clinical status of the patient.
Two monopolar radiofrequency systems were available, and the operators were familiar with both systems. A 200-W generator with internally cooled electrodes (Cool-Tip, Valleylab) and impedance-controlled pulsed current was used in 121 ablation sessions. Each overlapping ablation was 12 min in duration. In 15 ablation sessions, tumors were ablated with a 150-W generator using multitined expandable electrodes ([3-5 cm] Starburst XL, RITA Medical Systems) and a target temperature of 105°C. Tract ablation was not routinely performed, but was performed if the risk of hemorrhage was thought to be elevated such as in cases of central tumors, coagulopathy, or evidence of bleeding during the procedure.
For the Radionics system, 40 tumors were initially ablated with single electrodes and 52 with cluster electrodes. The decision as to which electrode to use for the Radionics system was determined by the attending radiologist on the basis of the following guidelines: All tumors except one that were 1.5 cm or less were treated with a single electrode. The exception was the use of a cluster electrode in a patient whose first tumor required a cluster electrode, but who also had a second small tumor for which we used the same electrode. Tumors between 1.5 and 3.5 cm were treated with single or cluster electrodes at the discretion of the attending radiologist involved in the case. Tumors that were 3.5 cm or larger were treated with a cluster electrode. The exceptions—tumors larger than 3.5 cm for which single electrodes were used—included a long but narrow tumor in which the electrode could be advanced along the long axis of the tumor, a tumor near bowel for which there was concern that the greater pressure needed to insert the cluster electrode would push the tumor so that it would be adjacent to bowel, and our first central tumor.
Effect of the Angle of Approach to the Tumor-Kidney Interface
Based on theoretic considerations, an initial ablation at and parallel to
the interface of an exophytic tumor with the kidney is believed by some to
devascularize the interface, thereby facilitating subsequent overlapping
ablations by eliminating the heat sink effect from adjacent blood vessels
[9]. We reviewed all exophytic
tumors and determined the angle of approach of the initial electrode placement
with respect to the tumor-kidney interface. Angles were recorded as parallel
to and at the tumor-kidney interface, perpendicular, less than 45° from
parallel, greater than 45° from parallel, and parallel but greater than 1
cm away from the tumor-kidney interface. The number of overlapping ablations
in each ablation session and absence of residual tumor after the first
ablation session were recorded. Retrospective review of imaging of the 100
renal tumors in the 85 patients who underwent radiofrequency ablation revealed
68 exophytic renal tumors that were subsequently analyzed in this fashion.
The Effects on the Collecting System
Baseline CT (n = 87 tumors) or MRI (n = 13 tumors)
examinations were reviewed for proximity of the renal cell carcinoma to
calyces, infundibula, ureteropelvic junction, and ureter. Calyces and
infundibula were recorded as being either directly contiguous with the tumor
or not directly contiguous. If the calyces, infundibula, or both were not
opacified where tumor was present in the renal sinus, they were categorized as
obscured by tumor. The nearest proximity of the tumor to the ureteropelvic
junction and to the ureter was measured in centimeters. For tumors not in the
same axial plane as the ureter or ureteropelvic junction on CT, triangulation
of the shortest distance to these structures was performed.
After radiofrequency ablation, the first cross-sectional imaging study in each case was reviewed in the same manner for proximity of the zone of ablation to calyces or infundibula. In addition, collecting system changes detected on CT or MRI when compared with preablation imaging were recorded, as were collecting system complications. Complications were classified as major or minor based on the classification of the Society of Interventional Radiology with major complications requiring treatment or hospitalization and minor complications needing only conservative monitoring. Finally, the results of radiofrequency ablation for tumors abutting or obscuring the collecting system were compared with those for tumors not in continuity with the collecting system.
Bowel Proximity and Considerations for Bowel Displacement
The need to protect bowel from thermal injury during radiofrequency
ablation has recently led to the use of numerous bowel displacement techniques
such as instillation of sterile water, solution of 5% dextrose and water,
CO2, or air or interposition of balloon catheters between tumor and
bowel [10]. However, absolute
distances defining safe zones between bowel and tumor have not been
established empirically. Thus, we reviewed all 100 cases treated with
radiofrequency ablation with respect to tumor proximity to bowel, changes in
bowel proximity during radiofrequency ablation, bowel-related complications,
and evolution of the use of bowel displacement techniques in our practice.
Changes in bowel distance could be measured only for those tumors ablated with CT guidance. Data recorded included changes in proximity from diagnostic imaging to the radiofrequency ablation images, changes during ablation, maneuvers at ablation aimed at displacing bowel or protecting bowel, and postablation bowel-related complications. Because the absolute safe distances are not known, we defined changes that could possibly be clinically significant as changes from one of the following categories into another category: distance decreasing to, at, or within 5 mm of bowel; to, at, or within 1 cm of bowel; or within 1-2 cm of bowel; and increases in distance that moved bowel to a different, presumably safer, category. Given that bowel position can be problematic in close proximity not only to tumor but also to the percutaneous path of the electrode, we also recorded changes in bowel position that blocked or opened the anticipated approach of the needle electrode as clinically significant.
Imaging Follow-Up: Patterns of Residual Disease and Approach to Treating Residual Disease
All patients underwent contrast-enhanced imaging (MRI or CT) before
radiofrequency ablation that served as a baseline comparison for subsequent
imaging after ablation. Patients with a serum creatinine level of 2.0 mg/dL or
less underwent follow-up CT (n = 73), and patients with a serum
creatinine level of greater than 2.0 mg/dL underwent follow-up MRI (n
= 12).
After radiofrequency ablation, CT or MRI without and with contrast material was performed at 1 month, 3 months, and 6 months. Subsequent imaging follow-up depended on the clinical condition of the patient and comorbid conditions, but was generally at 6- to 12-month intervals. The mean duration of follow-up was 2.3 years (range, 3.5-6 years).
Images obtained after radiofrequency ablation were interpreted by consensus of two experienced radiologists. Enhancement of any portion of the tumor was considered residual viable tumor, and absence of enhancement was considered completely ablated tumor [11, 12]. Images were also reviewed for the presence of any new metastatic disease or new renal cell carcinomas. Given that the timing of follow-up imaging studies varies among institutions where radiofrequency ablation is performed and that the optimal timing remains to be determined, we sought to determine the timing of detection of local viable disease at CT and MRI follow-up studies and recorded this information. In addition, in cases with residual disease on the first CT or MR image after ablation, size, location, and morphology of the residual disease were recorded as were the details and results of repeat ablation sessions.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Residual disease was present at the first follow-up CT examination in six (20.7%) of the 29 renal cell carcinomas approached at and parallel to the renal cell carcinoma-kidney interface and in five (22.7%) of 22 of those that were not (p = 0.86). The mean number of overlapping ablations did not differ between groups (parallel approach, 3.5; other approaches, 3.5) (p = 0.91). The analysis was performed in this fashion excluding those tumors that could not be approached parallel to the interface to make the two groups as alike as possible and to exclude the possibility that the tumors with obstructing structures differed in some way that might influence the results. However, the results did not change when a similar analysis that included the 17 tumors with intervening structures precluding a parallel approach was performed: no difference in the frequency of residual disease after the first ablation (p = 0.83) or in the total number of ablations (p = 0.84) was detected. Thus, it appears that initial placement of the electrode at and parallel to the kidney-tumor interface for ablation of exophytic renal tumors does not significantly influence the likelihood of achieving complete necrosis.
Effects of Radiofrequency Ablation on the Urinary Collecting System
On imaging before radiofrequency ablation, 23 tumors were directly adjacent
to a calyx (n = 16), infundibulum (n = 4), or both
(n = 3). An additional nine tumors obscured calyces, and of these
nine tumors, seven also obscured infundibula. On imaging after radiofrequency
ablation, the zone of ablation extended to a calyx in 26 of 100 cases and
obliterated a previously opacified calyx in three cases (Figs.
3A, and
3B). Whether obliteration of
calyces was from thermal destruction of a calyx or destruction of renal tissue
that drained to the calyx could not be determined.
|
|
In 13 of the 26 cases in which the zone of ablation extended to a calyx and in all three cases with calyceal obliteration, the tumor extended to a calyx on preablation imaging. Thus, there were 13 cases in which the zone of ablation extended beyond the tumor to reach a calyx or infundibulum. In these 13 cases, the tumor was between 3 and 10 mm (median, 6 mm) from the calyx before ablation. In the entire cohort of 100 tumors, the mean distance of the renal cell carcinoma from the ureteropelvic junction was 3.0 cm and from the ureter, 3.4 cm. Eight tumors were within 1 cm or less of ureter, and 10 were 1-2 cm from ureter. Eight tumors were within 1 cm of the ureteropelvic junction, and 20 were 1-2 cm from the ureteropelvic junction.
With respect to treatment success, 25 of the 32 tumors adjacent to or obscuring the collecting system underwent complete necrosis at imaging compared with 65 of 68 tumors more remote from the collecting system (p =0.01). Although this difference is significant, the difference is explained almost entirely by the subgroup of tumors obscuring a portion of the collecting system. Of the 23 tumors adjacent to a calyx or infundibulum without obscuration, 20 underwent complete necrosis compared with five of nine tumors that obscured the collecting system. If the tumors obscuring the collecting system are excluded, the 23 tumors adjacent to but not obscuring the collecting system did not differ significantly from the remainder of the tumors with 20 of 23 versus 65 of 68 tumors completely ablated (p = 0.15). Thus, tumor proximity to a calyx or infundibulum alone did not significantly increase the risk of incomplete ablation.
Hemorrhage—The most frequent complication involving the collecting system was hemorrhage. This complication affected the collecting system in four cases (major, n =1; minor, n = 3) and was treated by placement of a ureteral stent in one patient and placement of a bladder catheter in two patients. Stent placement was indicated for ureteral obstruction resulting in anuria in a patient with a solitary kidney, and bladder catheter placement was indicated for bladder clots causing persistent bladder outlet obstruction in two patients. In all three patients who experienced hemorrhage that required treatment, the renal tumor obscured calyces, infundibula, or both on imaging before radiofrequency ablation. Thus, of nine renal cell carcinomas that obscured calyces, infundibula, or both, a bleeding complication into the collecting system requiring treatment developed in 33% (3/9). This was highly statistically significant (p < 0.001) when compared with the lower incidence of hemorrhage in one of the other 91 patients. The fourth patient was asymptomatic and her transient ureteral clot resolved spontaneously, likely from urokinase in urine, after ablation of a renal cell carcinoma that extended to a calyx only. Thus, there was no collecting system hemorrhage requiring treatment in the 23 patients in whom the tumor extended to a calyx or infundibulum or in the 29 patients in whom the zone of ablation extended to the calyx or obliterated the calyx.
Ureteral strictures—The next most common clinically significant complication was development of a ureteral stricture. One proximal ureteral stricture required nephrostomy and stenting for anuria hours after radiofrequency ablation of a renal cell carcinoma 2 mm from the ureter in a solitary kidney. However, two asymptomatic strictures detected incidentally on follow-up imaging were observed without any clinical sequelae. In one patient, a ureteral stricture developed in a lower moiety of a duplex system resulting in asymptomatic and self-contained forniceal rupture within 1 month of radiofrequency ablation of a renal cell carcinoma 4 mm from the ureter. Although the absence of symptoms may have been from the suppressive effects of medications taken by the patient for a heart transplant, no superimposed infection occurred and mild hydronephrosis on MRI with symmetric enhancement of the lower moiety persisted on follow-up imaging up to 2 years later. In another patient, a mild ureteropelvic junction obstruction developed on serial CT over several months after radiofrequency ablation of a renal cell carcinoma 1.4 cm from the ureteropelvic junction with the zone of ablation 1.2 cm from the ureteropelvic junction. Renal parenchymal enhancement remained symmetric, indicating that the mild dilatation did not reflect a clinically significant obstruction. Imaging review showed that the axis of the kidney had shifted slightly from the changes within the tumor and the decrease in tumor size, possibly resulting in kinking at the ureteropelvic junction in this patient with an extrarenal pelvis. Thus, although strictures of the collecting system developed in two of the eight tumors within 1 cm of the ureter or ureteropelvic junction and one of 10 of those renal cell carcinomas 1-2 cm away, the only stricture that resulted in a clinically important obstruction was within 2 mm of the ureter.
Urine leaks—Urine leaks were present in three cases, but only one urine leak became clinically significant in the 100 tumors (1%). The two benign leaks manifested on delayed enhanced CT images as trace contrast material outside the collecting system but within the zone of ablation without urinoma accumulation (Fig. 4). These imaging findings persisted up to 12 months later without clinical sequelae in one patient and resolved by 6 months in the other. One patient required urinoma drainage and placement of a ureteral stent that required conversion to a nephroureteral catheter (Figs. 5A, 5B, 5C, 5D, 5E, and 5F). She had von Hippel-Lindau disease and had both upper and lower pole tumors successfully treated on the same day. The upper pole zone of ablation was the site of the urine leak. However, the tendency to leak may have resulted from the ureteral injury sustained from ablation of the lower pole tumor that was 1 cm from the ureter. Retrograde placement of a ureteral stent showed upper ureter perforation at a site that had been narrowed, likely from spasm after radiofrequency ablation. As a result, the stent was removed, and an antegrade approach was used to place a nephroureteral catheter across the ureteral perforation. Both the ureteral injury and upper pole urine leak healed with these measures.
|
|
|
|
|
|
|
Finally, in one patient, an incidental stone migration from a calyx to the middle ureter occurred during positioning for radiofrequency ablation of a 4.4-cm mixed-type renal cell carcinoma. The ensuing ureteral obstruction was treated with a ureteral stent.
Bowel Proximity and Considerations for Displacement
The mean distance of tumor to bowel at diagnostic imaging was 3.5 cm
(range, 0.10-12 cm). Eighteen tumors were within 5 mm of bowel, and 27 tumors
were within 1 cm. The number of tumors within subsequent centimeter increments
is delineated in Table 1. Of
those tumors within 1 cm or less of bowel, the nearest bowel in decreasing
order of frequency was colon (n = 15), small bowel (nonduodenum,
n = 8), duodenum (n =3), and stomach (n = 1).
Potentially clinically significant changes in tumor proximity to bowel
occurred in 21 tumors between the diagnostic CT examination and the localizing
CT examination at the time of ablation. Among these cases, the nearest bowel
at diagnostic imaging was colon (n =10), small bowel (n =
8), duodenum (n = 2), and stomach (n =1).
|
Decreases in the distance between the bowel and tumor were slightly more common than increases. The distance between bowel and tumor decreased in 12 cases, whereas it increased for nine cases. Changes in category are delineated in Table 2. Notably, in two cases in which small bowel was the closest at diagnostic imaging, the colon was the bowel that moved significantly closer, touching the tumor in one case and at 0.6 cm in another. Trends in bowel changes with position were observed but did not reach significance (p = 0.07). With prone patients, decreases in distance (n = 10) were more common than increases (n = 3). With patients in the decubitus position, increases in distance (n = 5) were more common than decreases (n = 3).
|
Changes in bowel location resulted in changes in the percutaneous puncture site for three tumors. In one case, with the patient prone, the bowel moved into the anticipated posterior approach to an exophytic anterior lower pole tumor, but a more lateral approach was possible. In the second case, with the patient prone, the colon collapsed nearly 1 cm away from the tumor, thereby opening an approach for the electrode to reach the tumor. In the third case, with the patient prone, the hepatic flexure and ascending colon came into contact with tumor, but repositioning to the right side down separated colon and tumor to allow safe ablation.
During the ablation, potentially clinically significant changes in bowel distance occurred in eight patients. However, this change was a decrease in distance in only two cases. In both of these cases, bowel dilatation (stomach, n = 1; colon, n = 1) occurred during the ablation and resulted in the bowel either touching the tumor or being within 2 mm of the tumor. In the remaining six cases, the bowel-to-tumor distance increased because of small space-occupying focal hemorrhage, inflammatory changes, or both (n =4); the electrode pushed the kidney away from bowel as it penetrated the tumor (n = 1); or the colon actually shifted away from the tumor without repositioning the patient (n =1).
Efforts to protect bowel changed over the course of this series. A total of 25 tumors required a protective maneuver after accounting for changes in bowel distance during ablation. The instillation of 50-200 mL of sterile water (Fig. 6) to displace colon was performed for four tumors that started 0.1-0.4 cm from bowel, all later in the series and within the past 2 years. In two of these four cases, the displacement was transient: the water diffused during ablation and thus repeat instillation of water was required.
|
With this approach, 21 tumors within 1 cm of bowel were ablated, 10 of which were within 5 mm. This approach was effective: 20 of these 21 tumors were completely ablated. The single failure was an 8.9-cm tumor. Nine of the 21 tumors required more than one ablation session for residual viable tumor. All the tumors treated in this fashion were larger than 3 cm, and these results compare favorably with the other tumors in this size category for which 18 of 26 tumors required more than one session for residual tumor. No recurrences developed in the group of patients with tumor near bowel. The single tumor recurrence developed in a patient with tumor 6 cm from bowel.
There were no complications related to bowel perforation or inflammation. CT performed after radiofrequency ablation showed the nearest bowel to be normal in all cases.
Patterns of Residual Disease and Timing of Detection and Approach to Repeat Ablation
Thirty tumors in 29 patients showed residual enhancement interpreted as
viable tumor on the first follow-up CT (n =26) or MRI (n =
3) examination after the first radiofrequency ablation visit. In 26 of these
30 tumors, the residual enhancement was definitively diagnosed on the 1-month
scan. In three tumors, the enhancement was present and detected on the 1-month
CT scan, but was considered indeterminate because of the possibility that it
represented volume-averaging artifact with normal enhancing renal parenchyma
at the cephalad or caudal aspects of the tumor. By the 3-month CT scans, the
enhancement was clearly shown to reflect tumor enhancement, and these three
tumors were ablated again. In the final patient, indeterminate tumor
enhancement was present on the 6-month scan but was interpreted as
volume-averaging artifact; residual disease was definitively diagnosed at 14
months after the initial session when a second ablation session was
performed.
Among these tumors were 25 of 92 tumors (1.3-8.9 cm) first treated with internally cooled electrodes and five of eight tumors (3.6-7.5 cm) first treated with multitined expandable electrodes with an overall mean of 3.4 cm. The 30 tumors showed a total of 46 discrete foci of enhancement (Figs. 7A, 7B, 8A, and 8B) ranging in morphology from nodules (3 cm, n = 18) and masses (> 3 cm, n =7) to crescents or half moons (n = 20) with a single finger-shaped focus. The mean tumor diameters did not differ significantly between these two groups (p = 0.11). The distribution of viable tumor and morphology of enhancement by electrode type are detailed in Table 3. The hooked array system tended to leave peripheral nodules and masses rather than crescents. The straight electrodes left crescents and nodules or masses.
|
|
|
|
|
Twenty-three of the 30 tumors were successfully ablated with one or more additional sessions, 18 with straight needle electrodes and five with expandable electrodes; two patients await repeat ablation. Straight needle electrodes were preferred for the repeat visits because electrode geometry was thought to be more favorable to the crescent-shaped areas and because straight needle electrodes were thought to be easier to use for small nodules, especially when multiple nodules were present.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
First, in performing radiofrequency ablation of exophytic tumors, the electrode approach at and parallel to the tumor-kidney interface has been postulated to confer advantages by devascularizing the interface and limiting blood flow to the tumor during subsequent overlapping ablations. Limited or absent blood flow allows larger zones of ablation by limiting tissue cooling that results from perfusion [9]. Our results, however, showed no significant difference in the number of overlapping ablations performed or in the rate of residual disease with the parallel approach compared with other approaches. Other factors likely play a dominant role in determining these performance and outcome measures. These results do not imply that this angle of approach should not be used, but rather that other angles of approach can be effective and that a radiologist need not be limited to a single angle when considering patient positioning and approach to the tumors. Moreover, for those patients in whom the parallel approach is precluded by intervening organs or osseous structures, one has no reason to expect a more difficult ablation or a higher risk of residual disease. A major limitation of this aspect of our study is the retrospective nature. Future study could address this question from a different perspective such as whether the timing of the individual ablations could be reduced to 6 min, for example, after ablation of the tumor-kidney interface.
Next, we have shown that clinically significant injuries to the urinary collecting system are rare. Equally important, we have shown that tumors that abut a calyx, but do not obscure it, are just as likely as tumors more distant from the collecting system to undergo complete ablation. Effects of radiofrequency ablation on the collecting system based on proximity to the tumor and to the thermal lesion have not previously been reported in detail. Although there has been some concern in animal work with respect to the risk of possible urine leak and urinoma formation after radiofrequency ablation with injury to the collecting system [13, 14], we have shown that extension of the zone of ablation to a calyx appears safe. Given that clinically significant urine leaks did not occur except as a result of downstream obstruction, direct thermal injury to the collecting system appears more likely to result in stricture formation, not perforation. Strictures of the ureter, ureteropelvic junction, or both are most likely within 1 month of radiofrequency ablation. One limitation of our study is the relatively small number of patients whose tumors or zones of ablation were contiguous to a calyx or infundibulum or within 1 cm of the ureter or ureteropelvic junction. Thus, larger series or the experiences of other investigators might shed additional light on this matter.
With respect to other collecting system complications, the 33% risk of bleeding in those renal cell carcinomas that obscured calyces on CT or MRI offers a new perspective on central tumors. Although obscuration of calyces alone would not preclude radiofrequency ablation, this highly significant increased risk of bleeding into the collecting system will, in the future, alert the interventional radiologist to expect this complication more often in this subset of patients. Our findings with respect to ureteral or ureteropelvic junction strictures provide opportunities for further study. In particular, we did not use prophylactic strategies to protect the ureter from thermal injury. Anecdotal cases of prophylactic ureteral stent placement with or without cold perfusate exist, and this practice could be evaluated to determine whether stent placement before radiofrequency ablation of renal cell carcinomas within 5-10 mm of the ureter is beneficial in protecting against stricture formation.
Third, we have confirmed that bowel complications after radiofrequency ablation are exceedingly rare given the absence of any perforations in ours and other series [1-7], but the potential for catastrophic consequences, as experience with liver tumor ablation has shown, requires constant vigilance [15, 16]. Increasing awareness of bowel displacement techniques and their application in selected cases may decrease the risk, but this hypothesis is difficult to prove scientifically given the low incidence of clinically significant bowel injury. Indeed, we safely ablated 21 tumors within 1 cm of the bowel without bowel displacement techniques using caution in electrode positioning. Despite the fact that we cannot prove or disprove a significant benefit of bowel displacement techniques, awareness of bowel proximity to tumor, potential changes in distance from diagnostic imaging to ablation, and changes during ablation as shown in our study remain important. In our series, the colon was not only the most common structure to be closest to tumor but also the most likely to change positions from diagnostic imaging to ablation and during ablation. Hydrodissection will likely remain an important tool in our arsenal for tumors close to bowel. However, as shown by two of our four cases, the tumor-to-bowel distance may decrease during ablation as water dissects through tissues, and repeat instillations may be needed for long procedures.
Optimization of the timing of imaging after radiofrequency ablation remains an open question. Controversy remains about whether an immediate (i.e., same or next day) CT or MRI study is needed and what the ideal intervals for subsequent follow-up are [1-7]. In the interest of detecting and treating residual tumor promptly, all authors of studies about imaging-guided radiofrequency ablation of renal cell carcinoma have reported initial follow-up imaging at short intervals, usually within 1 month or less after the ablation. Given that most of the residual viable tumor in our study was evident at 1-3 months (22/23 tumors or 96%), this practice is undoubtedly worthwhile.
One additional finding that we report is the difficulty in some cases to determine whether subtle enhancement at the caudal or cephalad aspect of the zone of ablation is residual viable tumor versus enhancement of normal renal parenchyma. Short-term interval follow-up imaging was of paramount importance in definitively diagnosing these tumors and ensuring prompt repeat ablation for complete necrosis of the renal cell carcinoma. Thus, we think that frequent imaging in the early postablation period is justified. Given the absence of any new enhancement detected beyond the 6-month follow-up scan, longer intervals between imaging studies are probably justified once a patient shows no residual enhancement over a 1-year period.
Finally, our descriptions of patterns of residual viable tumor could prove useful in planning repeat radiofrequency ablations in these patients. Residual enhancement after the initial radiofrequency ablation of renal cell carcinoma is often completely treated with a second ablation session. When ablation is incomplete, hooked array systems tend to leave a residual pattern of multiple peripheral nodules, whereas straight needle electrodes leave crescent-shaped edges in many cases but can also leave nodules or masses. Both geometric patterns are as expected based on the electrode geometry, and both can be treated with straight or multitined expandable needle electrodes.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. D. Michaelson, O. Iliopoulos, D. F. McDermott, F. J. McGovern, M. G. Harisinghani, and E. Oliva Case 17-2008 -- A 63-Year-Old Man with Metastatic Renal-Cell Carcinoma N. Engl. J. Med., May 29, 2008; 358(22): 2389 - 2396. [Full Text] [PDF]
|
|
|
|
 |
|
 G. E. Wile, J. R. Leyendecker, K. A. Krehbiel, R. B. Dyer, and R. J. Zagoria CT and MR Imaging after Imaging-guided Thermal Ablation of Renal Neoplasms RadioGraphics, March 1, 2007; 27(2): 325 - 339. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kawamoto, S. Permpongkosol, D. A. Bluemke, E. K. Fishman, and S. B. Solomon Sequential Changes after Radiofrequency Ablation and Cryoablation of Renal Neoplasms: Role of CT and MR Imaging RadioGraphics, March 1, 2007; 27(2): 343 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. P. Ryan Nonsurgical Approaches to Colorectal Cancer Oncologist, October 1, 2006; 11(9): 999 - 1002. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. C. Lucey Radiofrequency ablation: the future is now. Am. J. Roentgenol., May 1, 2006; 186(5 Suppl): S237 - S240. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
