- To present the oncological and functional outcomes of salvage focal (SFC) and salvage total (STC) cryoablation for recurrent prostate cancer (PCa) after failed primary radiotherapy.
The management of locally recurrent prostate cancer (PCa) after primary radiation therapy (RT) is controversial, with the majority of patients undergoing androgen deprivation therapy (ADT), which may further jeopardize their quality of life and increase the risk of skeletal fracture and cardiovascular mortality [1, 2]. Salvage treatment options for curative intent include radical prostatectomy (RP), cryoablation, high-intensity focused ultrasonography (HIFU) or brachytherapy [3, 4]. Since salvage RP is associated with significant morbidity, salvage cryoablation has emerged as an option that maintains cancer control whilst minimizing treatment side effects [3-5]. Salvage cryoablation for the whole prostate gland, has been performed with acceptable long-term oncological outcomes, but significant side effects persist, including incontinence (13%), urinary retention (2%) and recto-urethral fistula (1–3.4%) [4, 6-9]. This has prompted a technical modification to ensure tissue preservation, namely focal salvage therapy to decrease treatment-related morbidity further [6, 10, 11].
We recently reported our median 3.7-year follow-up experience of primary focal cryoablation for clinically unilateral PCa . We obtained encouraging functional results whilst achieving similar oncological outcomes to those of a matched-pair RP cohort . In the present study, we report the oncological and functional outcomes of our 7-year experience with salvage focal cryoablation (SFC) and salvage total cryoablation (STC) for biopsy-proven locally recurrent PCa after primary RT.
After obtaining institutional review board approval and informed consent from the patients, we retrospectively analysed the records of 50 consecutive patients who underwent SFC (n = 25) or STC (n = 25) from March 2003 to August 2010. At entry, all patients underwent systematic and lesion-targeted TRUS-guided biopsies plus Doppler studies, by the same experienced radiologist (D.B.), to confirm locally recurrent PCa and to stage accurately. Each hypoechoic lesion (HEL) on TRUS and biopsy-proven cancer location were mapped, recorded and schematically represented on a worksheet (Fig. 1). Dimensions, location and Doppler grading of increased blood flow within the HEL were documented. When extraprostatic extension was suspected, staging targeted biopsies of the involved neurovascular bundles or seminal vesicles was performed. All patients with pre-salvage therapy PSA >10 ng/mL or a biopsy Gleason score of ≥7 underwent metastatic evaluation with bone scintigraphy and CT or MRI of the abdomen and pelvis.
Patients with biopsy-proven unilateral PCa underwent SFC whilst patients with biopsy-proven bilateral PCa underwent STC. Any ADT that had been prescribed by an outside institution was discontinued at study entry, therefore, during the entire follow-up period, no patient received ADT before recurrence was established.
Salvage focal cryoablation was defined as hemi-ablation of a single lobe of the prostate, whilst STC was defined as ablation of the whole prostate gland (Fig. 2). Cryoablation was performed under TRUS-guidance (Type EUB-6500, Hitachi Medical Systems America, Inc. Tarrytown, NJ, USA), using a freehand technique (without the use of a grid template), with a third generation, argon/helium gas-based cryomachine (Endocare, HeathTonics Inc., Austin, TX, USA). Five thermocouple sensors were transperineally inserted into five critical points: the right and left neurovascular bundles, Denonvillier's fascia, the rhabdo-sphincter and the prostate apex (Fig. 3). A double freeze–thaw cycle with a urethral warming device was used.
Each freeze cycle took ∼8–10 min to achieve a lethal freezing temperature of −40 °C in the thermocouple sensors of the neurovascular bundle, apex and Denonvillier's fascia adjacent to the cancer. In cases of biopsy-proven extraprostatic extension (stage T3a), the cryolesion (iceball) was postero-laterally extended to cover the area. In cases of biopsy-proven involvement of the seminal vesicle (stage T3b), a cryoprobe was placed within the seminal vesicle for two additional freeze–thaw cycles (Fig. 4). All patients were discharged on the day of the procedure with a Foley catheter that was self-removed 3–4 days after SFC or 10–14 days after STC.
Follow-up protocol included PSA testing and TRUS imaging with Doppler at 3, 6 and 12 months after treatment and annually thereafter. Follow-up biopsies were offered at 6, 12, 24 and 60 months or when indicated. Potency was defined as erections adequate for penetration in over half of attempts (score ≥3 in question 2 of the International Index of Erectile Function) . Continence was defined as the use of no pads. Biochemical failure (BF) was defined as PSA nadir plus 2 ng/mL (Phoenix criteria) [6, 10, 11, 13].
The statistical software package SAS, Version 9.2 (SAS Institute Inc., Cary, NC, USA) was used for all the analyses. Pearson's chi-squared or Fisher's exact test were used to examine the association between categorical demographic and clinical variables. Wilcoxon's rank-sum test was used to test differences in not normally distributed continuous variables. Kaplan–Meier plots were used to estimate the probabilities of BF-free survival (BFFS). All P values reported are two-sided and a P value <0.05 was considered to indicate statistical significance.
Demographic and baseline characteristics are shown in Table 1. The median (range) age, time interval between primary RT and salvage therapy, PSA and Gleason score for SFC were 71 (59–81) years; 8.3 (3–15) years, 2.8 (0.1–8.2) ng/mL and 7 (6–8), respectively, whilst those for STC were 73 (57–83) years, 6.3 (2–13) years, 3.9 (0.1–12) ng/mL and 7 (6–9), respectively. ADT was prescribed at outside institutions after RT and before salvage cryotherapy in nine patients (36%) undergoing SFC, and in seven patients (28%) undergoing STC. All salvage cryoablations were successfully completed without intra-operative complications. The type of primary RT received did not affect the feasibility of the procedures.
|No. of patients||25||25||–|
|Median (range) age at RT, years||61 (51–74)||66 (54–76)||0.02|
|Median (range) PSA before RT, ng/mL||7 (2.8–19.8)||6 (1.8–23.8)||0.23|
|Median (range) time to PSA nadir after RT, months||20 (1–72)||20 (3–48)||0.67|
|Radiation type, n (%)||–|
|EBRT||11 (44)||11 (44)|
|Proton beam||8 (32)||5 (20)|
|Brachytheraphy||5 (20)||7 (28)|
|Brachytheraphy +EBRT||1 (4)||2 (8)|
|Median (range) Gleason score before RT||7 (3–7)||6 (5–8)||0.59|
|Gleason distribution before RT, n (%)||0.30|
|≤6||12 (48)||16 (64)|
|7||12 (48)||7 (28)|
|8||0 (0)||1 (4)|
|Unknown||1 (4)||1 (4)|
|Clinical stage before RT, n (%)||0.20|
|T1c||12 (48)||8 (32)|
|T2a||2 (8)||3 (12)|
|T2b||2 (8)||5 (20)|
|T2c||0 (0)||3 (12)|
|T3a||0 (0)||0 (0)|
|T3b||1 (4)||0 (0)|
|Unknown||8 (32)||6 (24)|
|Time from RT to salvage cryoablation, years||8.3 (3–15)||6.3 (2–13)||0.13|
|ADT after RT and before cryoablation*, n (%)||9 (36)||7 (28)||0.76|
|Type of ADT after RT and before cryoablation*, n (%)||–|
|Leuprolide||2 (8)||0 (0)|
|Goserelin||0 (0)||1 (4)|
|Bicalutamide||0 (0)||1 (4)|
|Leuprolide + bicalutamide||7 (28)||5 (20)|
|Duration of ADT after RT and before cryoablation*, n (%)||–|
|≤1 year||3 (12)||7 (28)|
|≤2 years||4 (16)||0 (0)|
|Intermittent||2 (8)||0 (0)|
|Median (range) salvage cryoablation age, years||71 (59–81)||73 (57–83)||0.14|
|Median (range) pre-cryoablation PSA, ng/mL||2.8 (0.1–8.2)||3.9 (0.1–12)||0.84|
|Median (range) pre-cryoablation Gleason score||7 (6–8)||7 (6–9)||0.69|
|Median (range) pre-cryoablation Gleason distribution, n (%)||0.21|
|≤6||5 (20)||7 (28)|
|7||14 (56)||8 (32)|
|8||6 (24)||8 (32)|
|9||0 (0)||2 (8)|
|Pre-cryoablation TRUS + biopsy stage, n (%)||0.06|
|T1c||0 (0)||1 (4)|
|T2a||13 (52)||14 (56)|
|T2b||7 (28)||2 (8)|
|T2c||0 (0)||4 (16)|
|T3a||3 (12)||1 (4)|
|T3b||2 (8)||3 (12)|
|Median (range) pre-cryoablation prostate volume (TRUS), cm3||24 (16–45)||30 (13–55)||0.01|
The median (range) follow-up was 31 (4–90) months for SFC and 53 (12–92) months for STC. Two patients in the STC group were lost to follow-up. BF occurred in 32% of patients (8/25) for SFC, and 12% (3/23) for STC, respectively. In the patients without BF (SFC, n = 17 and STC, n = 20), the mean post-cryoablation PSA decreased by 86% for SFC (from 3.3 to 0.2 ng/mL), and 90% for STC (from 3.7 to 0.1 ng/mL [Fig. 5]) within the first year and remained stable, during the entire follow-up period, ranging from 0.2 to 0.6 ng/mL for SFC and from 0.1 to 0.2 ng/mL for STC (Fig. 6). The estimated 5-year BFFS was 54.4% for SFC and 86.5% for STC (Fig. 7).
No patient died during follow-up. One patient in the STC group developed metastases, despite having a negative follow-up biopsy with a PSA of 7.3 ng/mL after STC. The detailed clinical course of this patient was as follows: initial diagnostic biopsy was Gleason score 7, clinical stage T2a, and PSA 6.4 ng/mL, treated with combined external beam RT (EBRT) and brachytherapy. PSA failure occurred 8 years after RT (PSA of 0.24 ng/mL), and was initially managed with bicalutamide. The patient was subsequently referred for salvage therapy and TRUS showed bilateral suspicious HEL (15 and 10 mm in diameter), with Gleason 6 disease identified on targeted biopsies. After a negative metastatic evaluation the patient underwent STC; however, we speculate that the use of bicalutamide may have maskedthe metastatic findings on CT and bone scan performed before STC.
A TRUS-guided follow-up biopsy was performed in 48% of patients (12/25) in the SFC group and in 28% (7/25) in the STC group. Patients who deferred follow-up biopsy had no evidence of clinical progression after treatment, with a consistently low PSA as well as disappearance/shrinking of the biopsy-proven HEL on post-treatment TRUS (Fig. 8). In the SFC group, no patient had biopsy-proven cancer on the treated side; however, two patients had positive biopsies on the untreated side, resulting in the need for repeat SFC (n = 1) or ADT (n = 1). One patient in the STC group had biopsy-proven cancer and was subsequently treated with ADT. New onset urinary incontinence occurred in three patients in the STC group (13%), whereas no patient in the SFC group developed incontinence (P = 0.10). In the SFC group, 29% of the patients (2/7) retained postoperative potency but, by contrast, none of the four potent patients in the STC group recovered their potency after treatment, despite the use of a phosphodiesterase -5 inhibitor (P = 0.48). One (4%) patient in the STC group developed a recto-urethral fistula, but none occurred in the SFC group (P = 0.48).
The cumulative mortality from PCa among men who failed primary RT may reach 11, 20 and 42% at 5, 10 and 15 years, respectively . Recent reports show that men with low- to intermediate-risk PCa may derive no survival benefits from primary PCa local treatment ; however, such expectant management has not been assessed in the setting of radio-recurrent disease. Salvage treatment, including focal cryoablation, should be considered because tissue-preservation modifications may minimize treatment-related morbidity [4, 12].
The index lesion was defined as the biopsy core with higher Gleason score, followed by the larger actual cancer core length (mm) ; however, since the interpretation of the Gleason score grading after definitive RT and ADT is limited , the index lesion definition after RT may be misleading. In the present study, treatment was attempted for all the known cancer (index and secondary foci). Depending on the results of the pretreatment systematic and targeted biopsies, hemi- or total cryoablation was performed for unilateral or bilateral PCa, respectively.
The key to successful SFC is appropriate patient selection, especially in accurately determining the location of the recurrent cancer. We believe that image-guided biopsy-proven localization of disease is critical. In our study all men received a preoperative TRUS-guided biopsy to identify unilateral or bilateral disease and thus select the appropriate treatment option (SFC vs STC). As the exclusion of metastatic disease is necessary, preoperative imaging was performed in all patients with potential risks. As yet there is no consensus as to the PSA threshold necessary to trigger a metastatic evaluation. Investigators suggest that a PSA >10 ng/mL at the time of salvage could predict a worse outcome after local salvage surgical therapy . In the present series, metastatic evaluation was performed in 16 (64%) and 17 (68%) patients in the SFC and STC groups, respectively. It should be noted that, as we experienced one case of metastasis after STC (in a patient who received bicalutamide at the time of their PSA failure), the use of any hormonal agent could potentially result in misinterpretation of the metastatic evaluation.
A major criticism of ‘primary’ focal therapy for PCa is based on the tenet that PCa is a multifocal disease. By contrast, in the ‘salvage’ setting after radiation failure, PCamay have different biological properties from those seen in the treatment-naïve prostate. Interestingly, Arrayeh et al.  and also Pucar et al.  both reported that clinically significant locally recurrent PCa after RT is more likely to occur at the same site as the primary dominant tumour, thus ‘suggesting supplementary focal therapy aimed at enhancing local tumor control would be a rational addition to management’. Huang et al.  analysed step-section specimens of 46 salvage RPs and found that only 28% (13/46) had multifocal disease, whilst 72% had a single, unilateral focus. By contrast, Wise et al.  reported that in the treatment-naïve prostate, 83% of cancers were multifocal, and only 17% (82/486) had a single focus in one lobe. Furthermore, the study by Huang et al. reports a median total cancer volume of 0.68 cm3 compared with a median total cancer volume of 3.43 cm3 (2.78 cm3 for the index tumour and 0.18 cm3 for secondary tumours) in the study by Wise et al. These indicate that in the ‘salvage’ setting after RT, recurrent cancer has less volume, and is more likely to be a single, unilateral focus; therefore, it is more amenable to targeted focal therapy than is ‘primary, treatment-naïve’ PCa.
To our knowledge, there is no study comparing the freehand technique with the use of grid template-guided cryoablation. The freehand cryoprobe delivery technique requires significant TRUS-imaging expertise to understand the three-dimensional surgical field; however, without the limitations imposed by the grid template, the freehand technique has a number of advantages which include: (i) more options for the point of cryoprobe skin insertion; (ii) greater degree of freedom in angulating the cryoprobe during insertion (thus permitting probe insertion parallel to the rectal wall to avoid injury); and (iii) easier adjustment of the distance between two cryoprobes or between a cryoprobe and an anatomical landmark (such as the contour of the visible cancer, prostate margin, urethra or rectal wall). As such, we believe these advantages of freehand manoeuverability facilitate the precise targeting of the biopsy-proven cancer, and may result in better oncologicalal control and reduced morbidity.
There is a lack of consensus with regard to the definition of BF after salvage cryoablation. In the present study, we used the Phoenix definition as applied to post-RT recurrence [6, 10, 11, 13]. Since hemi-ablation therapy preserves one lobe of the prostate, the postoperative PSA value after focal therapy might remain higher than after whole-gland therapy, requiring the development of a new definition. We believe that the hypothesized theory of an index lesion is worth discussing in the salvage setting . Since the index lesion may be responsible for 80% of the total cancer volume in the treatment-naïve prostate , secondary cancer lesions on multifocal PCa are not significant predictors of PSA relapse . Given the majority (72%) of disease recurrence in irradiated prostates is as a single index cancer , focal ablative therapy potentially reduces the postoperative PSA by ≥80% if the index lesion is properly targeted and ablated. In our study, the mean postoperative PSA decreased by 86% in patients without BF after SFC and by 90% after STC, suggesting that the index lesion was adequately ablated. In the five cases with an advanced index lesion of pT3b disease, in which the seminal vesicle was treated, there has been no biochemical recurrence with a mean (range) follow-up of 9.2 (4–24) months. Clearly, however, further follow-up is necessary for these high-risk patients with seminal vesicle involvement who probably had a risk of micrometastasis.
Several authors report BFFS after STC using various criteria including Phoenix (Table 2) [6, 7, 10, 11, 26]. In the present study, we report a 5-year BFFS rate for STC of 86% using the Phoenix criteria. By contrast, to our knowledge, there are only two studies that reported outcomes after partial salvage cryoablation. Eisenberg and Shinohara  reported BFFS rates (Phoenix criteria) of 89, 79 and 79% at 1, 2 and 3 years, respectively . Wenske et al.  reported 5- and 10-year disease-free survival rates after SFC (Phoenix criteria) of 47 and 42%, respectively, but in their study 20% (n = 11) of the patients had cryoablation instead of RT as primary treatment . Ahmed et al.  reported their pilot study with salvage focal HIFU in 39 patients with a median follow-up of 17 months. The actual 1- and 2-year BFFS rates (Phoenix criteria) were 69 and 49%, respectively. Hormone therapy after salvage focal HIFU was applied in 16 (41%) patients and follow-up biopsy was performed in nine (23%) patients. We report a 5-year BFFS rate of 54% for SFC (Phoenix criteria), no patient received adjuvant ADT and a follow-up biopsy was performed in 38% (19/50) of the patients (Table 2).
|Authors||Year of publication||Cryoablation period||N||Cryoablation type||Median (range) follow-up time, months||BF criteria||5 years BFFS rate, %||Follow-up biopsy rate, %||Incontinence rate, %||Potency preservation rate, %||Rectal fistula rate|
|Williams et al. ||2010||1995–2004||176||STC||7.46 (1–14)1,2||Phoenix3||47||95||NA||NA||NA|
|Bahn et al. ||2003||1993–2001||59||STC||82.3 (NA)|| |
PSA ≥0.5 ng/mL
PSA ≥1 ng/mL
|Ismail et al. ||2007||2000–2005||100||STC||33.5 (12–79)2||ASTRO||594||NA||13||24||1%|
|Pisters et al. 5||2008||NA||279||STC||21.6 (24.9)6|| |
|Eisenberg and Shinohara ||2008||2004–2007||19||SFC||18 (6–33)|| |
|52||1 out of 19||2 out of 5||0%|
|Wenske et al. ||2012||1999–2011||328||STC (N = 273)||49 (2–204)||Phoenix|| |
|SFC (N = 44)8||37 (2–173)||Phoenix|| |
|Ahmed et al. ||2012|| |
|39||SFC||17 (10–29)||Phoenix|| |
|Present study||2012||2003–2010||50||STC (N = 25)||53 (12–92)||Phoenix||86.5||28||13||0 out of 4||4%|
|SFC (N = 25)||31 (4–90)||Phoenix||54.4||48||0||2 out of 7||0%|
Indications for treatment were different between STC vs SFC (the STC group included more advanced cases with biopsy-proven bilateral cancer while the SFC group included cases with unilateral cancer). Subsequently, there were also differences in areas of the prostate ablated (whole gland vs unilateral), which directly affected postoperative PSA levels. Furthermore, patients undergoing STC showed a trend towards higher clinical stage (P = 0.06) as well as more patients with biopsy Gleason pattern 8 and 9 than those in the SFC group. In addition, the follow-up time was longer for STC, therefore, it was not possible to compare oncological outcomes directly between the two groups with regard to postoperative PSA levels and biochemical failure. It should be noted that since SFC left more untreated tissues than did STC, the postoperative PSA might remain higher even after achieving therapeutic success (Fig. 7).
Treatment-related morbidity for SFC was less than that for STC, although the difference was not statistically significant. In fact, Wenske et al.  reported an overall complication rate of 15% (42/273) for STC and 7% (4/55) for SFC. In the present study, no patients in the SFC group reported new onset urinary incontinence, compared with three patients in the STC group (13%). Incontinence after SFC was reported in one of 19 patients by Eisenberg et al.  and in none of 55 patients (defined as the need for a male sling or artificial urinary sphincter) by Wenske et al. ; however, after STC it ranges from 0 to 13% [6, 8, 9, 25]. After RT a limited number of patients remain sexually potent . This is reflected in our study where only 28% (7/25) of patients in the SFC group and 15% (4/25) of those in the STC group were sexually potent before their salvage treatment. Two of the seven patients in the SFC group and no patient in the STC group regained potency postoperatively, which is consistent with the literature [9, 10]. A further study with increased patient numbers undergoing SFC is currently under way to confirm the statistical benefits in erectile function outcomes when compared with STC. Recto-urethral fistula occurred in one (4%) patient in the STC group. The rate of recto-urethral fistula after STC is reported to be between 1 and 3.4% [6, 8, 9, 26] and 0 and 5.5% after SFC [6, 10].
The main limitations of our study are the small number of patients, its retrospective design and the small number (38%) of patients with follow-up biopsy. Although we strongly recommended patients undergo follow-up biopsy, we found that a favourable clinical course with no evidence of disease progression (such as a consistently low PSA or disappearance of the targeted biopsy-proven lesion) might affect the patients' decision to undergo or to defer the follow-up biopsy. The ideal indication/protocol to perform follow-up biopsy, accounting for the balance between patient benefit and harm, has yet to be determined. In the near future, evolving imaging that includes multiparametric MRI, multiparametric ultrasonography, and ultrasonography/MRI fusion may further improve patient selection and enhance oncological and morbidity outcomes [28, 29].
In conclusion, SFC and STC are feasible and safe with acceptable mid-term oncological outcomes. For carefully selected patients SFC is an option that could be associated with lower treatment-related morbidity compared with STC. Although longer follow-up and more patient numbers are needed, our initial oncological and functional outcomes of SFC and STC are encouraging.
There was no financial support or relationship that may pose conflict of interest. Authors' financial disclosures: Mihir M. Desai is paid consultant for Baxter and Hansen Medical. Inderbir S. Gill is paid consultant for Hansen Medical. The remaining authors have nothing to disclose.
salvage focal cryoablation
salvage total cryoablation
androgen deprivation therapy
high-intensity focused ultrasonography
biochemical failure-free survival
external beam radiotherapy