Paul L. Nguyen, MD ^{1 *}, Anthony V. D'Amico, MD, PhD ^{. Cancer 2007;110:1417}

Among men who experience prostate-specific antigen (PSA) failure after external beam radiation or brachytherapy (RT), many will harbor occult micrometastases; however, a significant minority will have a true local-only failure and, thus, potentially may benefit from a salvage local therapy. Those most likely to have a local-only failure initially have low-risk disease (PSA < 10 ng/mL, Gleason score 6, clinical T1c or T2a tumor status), pretreatment PSA velocity < 2.0 ng/mL per year at the time of initial presentation, interval to PSA failure > 3 years, PSA doubling time > 12 months, negative bone scan and pelvic imaging, and positive rebiopsy. In addition, men with presalvage PSA levels > 10 ng/mL, presalvage T3/T4 disease, or presalvage Gleason scores 7 on a rebiopsy sample without significant RT effects are unlikely to be cured by salvage local therapy.

Based on a review of all series of post-RT salvage prostatectomy, cryosurgery, and brachytherapy published in English since 1990, morbidity can be substantial. Although urinary incontinence appeared to be greater after salvage prostatectomy (41%) or cryosurgery (36%) than after brachytherapy (6%), patients who received salvage brachytherapy faced a 17% risk of grade 3 or 4 genitourinary complications and a fistula risk that averaged 3.4% across all series. From this review, the authors concluded that prospective randomized studies are needed to determine the relative efficacy of the 3 major local salvage modalities and that additional research is needed to identify factors associated with an increased risk of significant complications to improve patient selection and to augment the benefit/risk ratio associated with attempts to cure local-only recurrences after radiation therapy.

After either external beam radiation or interstitial brachytherapy for clinically localized prostate cancer, between 20% and 50% of men may experience prostate-specific antigen (PSA) failure. Many of these men will harbor a component of micrometastatic disease at the time of PSA failure, but a significant minority will have a true local-only recurrence and potentially can be cured with a salvage local therapy. Currently, there is no consensus regarding the optimal management of patients who are believed to have a local-only recurrence after prostate radiotherapy (RT). In this review, we describe the selection criteria for men who are most likely to benefit from a salvage local therapy, and we summarize the published literature on the oncologic outcome and toxicity of the currently available treatment approaches.

The greatest challenge in selecting a man with PSA failure after RT for further local therapy is determining whether the rising PSA represents distant disease, local disease, or both. Because the morbidity of a local salvage therapy can be substantial, it is crucial to identify which patients are most likely to have local-only disease and, thus, have the greatest chance of benefiting from the treatment.

The workup for a man who is being considered for a local salvage should include a bone scan to rule out distant metastases and a pelvic magnetic resonance imaging (MRI) study or computed tomography scan to rule out regional lymph node disease. However, because PSA failure precedes clinically evident metastases by a median of 8 years after prostatectomy and at least 7 years after radiation, these studies are unlikely to have a high yield. A more sensitive test for detecting clinically occult lymph node metastases in the de novo setting is high-resolution MRI used in conjunction with lymphotropic superparamagnetic nanoparticles, as described by Harisinghani  which reportedly had 90% sensitivity for pathologic lymph node metastases compared with the 35% sensitivity of MRI alone. This technology is not available yet for clinical use and is undergoing further study, but it may play a role in the work-up of these men in the future.

A transrectal prostate needle biopsy should be obtained to demonstrate local disease before local salvage is considered. It should be reviewed by a pathologist who is familiar with radiation effects on the prostate, because radiation changes can include the same features that are consistent with high-grade disease (eg, necrosis), thereby leading to an inaccurate Gleason score or a false-positive result. Therefore, some experts have recommended that a Gleason pattern should be assigned in this setting only if radiation effect is not present in the biopsy sample. The timing of the biopsy also can affect the test characteristics, because approximately 30% of positive biopsies that are obtained 12 months after radiation will convert to negative status by 24 to 30 months. Like in the de novo setting, there are resulting sampling errors and false-negative results, and Crook et al. have reported that 19% of patients who have a negative posttreatment biopsy later had a positive rebiopsy.

Men who present initially with clinical characteristics that place them at high risk for developing distant metastases are unlikely to have local-only disease at the time of PSA failure. Specifically, men with an initial presenting Gleason score 8, a clinical tumor category T2c, or a PSA level > 20 ng/mL have a shorter time to PSA failure, have a higher rate of developing distant metastases, and have a 14-fold increase in prostate cancer-specific mortality compared with patients who present initially with low-risk clinical features. Ideally, candidates for salvage local therapy are those who present initially with low-risk clinical features (PSA < 10 ng/mL, biopsy Gleason score 6, T1c or T2a tumor status), because their PSA failure is less likely to be caused by occult micrometastases that were missed at the time of initial therapy.

Pretreatment PSA kinetics also have emerged as important predictors of outcome after definitive therapy. Specifically, a pretreatment PSA velocity > 2.0 ng/mL per year in the 18 months prior to diagnosis is associated with both prostate cancer-specific mortality and all-cause mortality after radiation therapy or radical prostatectomy (RP). In 1 RT study, men with a PSA velocity > 2.0 ng/mL per year versus < 2.0 ng/mL per year had a 24-fold greater risk of death from prostate cancer, and this cutoff point of approximately 2.0 ng/mL per year has been confirmed by others. Therefore, patients with a pretreatment PSA velocity > 2.0 ng/mL per year at the time of initial presentation are suboptimal candidates for local salvage therapy at the time of PSA failure, because they are more likely to have failed from occult micrometastases.

From a large series of patients who underwent RP, Pound et al. observed that a time to PSA failure < 2 years was associated with a significantly increased risk of developing distant metastases. An update of that series by Freedland et al. indicated that, with longer follow-up, a cutoff point of 3 years was a better differentiator. The concept that a shorter interval to PSA failure is associated with a higher risk of distant metastases also applies to the postradiation setting. Zagars and Pollack reported that men who developed metastases after RT experienced PSA failure at a median of 9 months compared with a median time to failure of 18.4 months for men who failed but did not develop metastases.Those authors suggested that a cutoff point of 1 year was an important predictor of distant metastatic risk. Taking the most conservative approach, the optimal candidate for salvage local therapy should have an interval > 3 years between RT and PSA failure.

PSA doubling time (PSA-DT) at the time of PSA failure is an important predictor of the time to subsequent distant metastases. Specifically, Zagars and Pollack observed that a PSA-DT < 8 months was associated with a 54% actuarial rate of distant metastases at 7 years compared with only 7% for men who had a PSA-DT > 8 months. Similarly, Zelefsky et al. observed a 3-year actuarial rate of distant metastases of 49% for men who had a PSA-DT < 3 months compared with a rate of only 7% for men who had a PSA-DT > 12 months. The outcome of patients who have a PSA-DT < 3 months is particularly poor, and work by D'Amico et al. and Freedland et al. have demonstrated a very strong association between a PSA-DT < 3 months and prostate cancer-specific mortality, with a median interval of 6 years between PSA failure and death. Therefore, the optimal candidate for salvage local therapy should have a PSA-DT > 8 months and, ideally, > 12 months.

Several studies of salvage cryotherapy have indicated consistently that a PSA level > 10 ng/mL at the time of salvage is associated with a worse outcome. For example, for patients who have a presalvage PSA level < 10 ng/mL versus 10 ng/mL, Izawa et al. reported a 5-year disease-free survival (DFS) of 57% versus 23%, and de la Taille et al. reported a 9-month PSA failure-free survival rate of 86% versus 42%, respectively.In salvage prostatectomy series, a PSA level 10 ng/mL has been associated with a worse pathologic stage and a significantly higher rate of subsequent PSA failure, which makes patients who have PSA levels 10 ng/mL suboptimal candidates for salvage local therapy.

RP is the local salvage modality with the longest history, and current series have included patients who were treated in the 1960s.

The 5-year biochemical DFS (bDFS) rate typically ranges from 50% to 60% in most salvage RP series. The largest series is from the Mayo Clinic, which dates back to 1967 and contained 199 patients, including 138 men who underwent RP. At the time of salvage RP, PSA levels were < 10 ng/mL in 78% of these patients, and pathologic (p)T2 disease was identified in 39%. After a median follow-up of 92 months, the 5-year estimate of bDFS was 63% when PSA failure was defined as a PSA level > 0.4 ng/mL. The next largest series is from Memorial Sloan-Kettering Cancer Center (MSKCC)/Baylor Medical Center and included 100 patients who were followed for a median of 60 months. Fifty-nine percent of those patients had a PSA level < 10 ng/mL, 35% had pT2 disease, and 55% were PSA failure-free at 5 years when PSA failure was defined as a PSA level > 0.2 ng/mL. In both of these series, men who had PSA levels < 10 ng/mL at the time of salvage RP had significantly lower estimates of PSA failure compared with men who had PSA levels 10 ng/mL.

Additional analysis of factors that were predictive of outcome was performed by Cheng et al. on a subset of 86 patients from the Mayo Clinic series. Those authors observed that a preoperative PSA level < 10 ng/mL (P = .02), a prostatectomy Gleason score of 7 or of 8 to 10 with < 7 as the baseline (P = .0005), and DNA aneuploidy or tetraploidy with diploid as the baseline (P = .02) each was associated with cancer-specific survival; although, on multivariable analysis, only the associations for Gleason score and DNA ploidy were significant.

Oncologic outcome was worse for patients who underwent cystoprostatectomy (CP) compared with patients who underwent RP, with a 5-year bDFS rate of 19% versus 63%, respectively, in the Mayo Clinic series and 30% versus 50%, respectively, in the Wayne State series. Because CP typically was reserved for men with bulkier or locally advanced disease, including bladder neck invasion, these results most likely reflect differences in patient selection rather than differences in the efficacy of the procedure. However, these poor results suggest that the chance of curing a man who has a locally advanced recurrence with a local salvage therapy is very low and calls into question the utility of salvage CP as monotherapy given its substantial morbidity.

In the de novo setting, 3 randomized trials have demonstrated that short-course neoadjuvant androgen-deprivation therapy (ADT) decreases the rate of positive surgical margins, but this has not translated into improved biochemical outcomes. In the salvage setting, there are no randomized studies, and retrospective data from the Mayo Clinic and the Netherlands Cancer Institute series showed no reduction in the positive margin rate or improvement in outcome with neoadjuvant ADT, although it is possible that the selection of men with more advanced disease to receive ADT obscured any potential benefit.Conversely, some indirect evidence in favor of neoadjuvant ADT comes from the Gainesville series, in which the 24 men who received neoadjuvant ADT had a 21% positive margin rate, which compares favorably to the 33% rate in the Mayo Clinic series and the 35% rate in the University of Southern California series. In addition, those 24 men had an 80% 5-year bDFS rate, which is the highest reported in the salvage RP literature. However, caution must be used when making comparisons across series in which the data are retrospective, because imbalances in both known and unknown prognostic factors may exist. Therefore, in the absence of randomized data or consistent results from retrospective series, we cannot recommend the routine use of neoadjuvant ADT prior to salvage RP.

The potential for cure with salvage RP must be balanced against the risks of substantial potential toxicities. These risks are greater in the salvage RP setting than in the de novo setting because of radiation changes in the operative field that may cause fibrosis, merging of tissue planes used for dissection, and poorer wound healing.

Perioperative mortality is very low, with only a single treatment-related death reported from the combined series of 531 men. Urinary incontinence has ranged from 0% to 67% in various series, with a weighted average across all series of 41% for patients who underwent RP. Bladder neck stricture generally is reported in approximately 20% to 30% of patients (weighted average, 24%) and often leads to dysuria, moderate pelvic pain, and the subsequent need for dilation. The average rate of rectal injury was 4.7% across all series.

In the MSKCC/Baylor series, complication rates appeared to be improving over time, suggesting that there is a learning curve for this procedure. Specifically, men who were treated from 1984 to 1992 had a 15% rectal injury rate compared with 2% for men who were treated after 1992 (P = .01). The overall rate of grade 2 through grade 4 complications was reduced from 33% to 13% (P = .02) before versus after 1993, respectively. However, neither the rates of incontinence (43% vs 32%, respectively) nor the rates of anastomotic strictures (28% vs 32%, respectively) changed before versus after 1993.

Although it was attempted first and abandoned in the 1960s, prostate cryosurgery came to be adopted increasingly in the early 1990s with the introduction of a transrectal ultrasound-guided technique described by Onik et al. that allowed for improved visualization and monitoring of freezing. Moreover, the introduction of urethral warming catheters has decreased the risk of urethral sloughing and injury. Consequently, by the late 1990s, salvage cryotherapy had emerged as an alternative to salvage prostatectomy.

It is apparent from the data that it is difficult to derive a single number for PSA control after salvage cryosurgery, because the series used varying definitions of PSA failure and had varied lengths of median follow-up. The most extensively studied cohort is the University of Texas M. D. Anderson Cancer Center (MD Anderson) series, in which 131 men who were treated from 1992 to 1995 had a 57-month median follow-up and experienced a 5-year PSA failure-free survival rate of 40% with failure defined as a PSA level that exceeded the nadir value by 2 ng/mL. The next largest series was from Western Ontario and reported on 125 men who had a crude PSA failure free-survival rate of 34% after a median follow-up of 19 months, although failure most likely was detected earlier than in the MD Anderson series, because any PSA level > 0.5 ng/mL was scored as a failure. In both series, the factors associated with a significant decrease in PSA control were a PSA level > 10 ng/mL before cryosurgery, clinical category T3 or T4 disease (prior to cryosurgery in the Ontario series and prior to RT in the MD Anderson series), and Gleason score (> 8 before RT in the Ontario series and 8 before cryosurgery in the MD Anderson series). In addition, a reanalysis of the MD Anderson series has suggested that a PSA DT < 16 months before cryosurgery also was associated with a worse PSA outcome.

It appears that, in general, the PSA failure-free survival rate for cryosurgery is somewhat lower than the 50% to 60% 5-year control rate reported in the RP series; however, because of uncertainties regarding potential imbalances in prognostic factors among the patients in the cohorts, no conclusions concerning the relative oncologic control can be drawn. However, a potential explanation for the apparently lower PSA control rates with cryosurgery compared with RP is the possibility of incomplete ablation of the prostate gland, as evidenced by the 50% prevalence of viable glands at 24 months in the Western Ontario series and 59% prevalence of residual normal or atypical glands or cancer seen in the MD Anderson series. Izawa et al. reported that an increased number of freeze-thaw cycles (P = .002) and a greater number of probes (P = .04) were associated with improved glandular ablation, although more aggressive ablation also has been associated with increased morbidity, as described below.

Table  lists the complication rates of all salvage cryosurgery series published in the English literature since 1990. Similar to the salvage RP series, perioperative mortality is extremely low, and only 1 death (0.2%) has been reported in 510 total patients from 9 series. Also like the RP series, it has been demonstrated that morbidity for salvage cryotherapy is more severe than when cryotherapy is performed on a nonirradiated prostate gland.

Urinary incontinence was reported in 72% of patients in the MD Anderson series compared with only 20% of patients in the Ontario series; however, follow-up was shorter (19 months) in the latter series, and late complications may increase with time. In addition, the MD Anderson results are patient-reported, which may be more sensitive than physician assessments, and used a strict definition of incontinence, which included any urinary dribbling, even if a man did not require any pads. Incontinence was reported in as few as 4.3% of patients and in as many as 96% of patients in various series, with a weighted average of 36% for all series. Data from the Ontario series suggests that men with a prior transurethral resection of the prostate in addition to radiation are at particularly high risk of developing incontinence after salvage cryosurgery. Perotte et al. implicated the lack of a urethral warming catheter as another important risk factor for incontinence.

Urethral sloughing was observed in up to 27% of patients and averaged 11% across all series. In 1 series, the incidence of sloughing was significantly worse for patients who were treated without versus with a urethral warming catheter (24% vs 4%, respectively.

Bladder neck stricture or retention occurred in 17% of patients across all series, and perineal pain occurred in 36%. It is noteworthy that perineal pain occurred in 44% of patients in the MD Anderson series and was so severe that it interfered with normal daily activities in 38%. It also is important to note that the presence of pain was associated with the lack of a warming catheter (70% vs 34%

Although the development of fistulas was reported in up to 11% of patients in 1 series, it was an uncommon event and occurred at a rate of 2.6% over all series. Chin et al. observed that all 4 of the patients who developed rectal fistulas in their series had either T3b or T4a disease and suggested that patients with T3/T4 disease are particularly poor candidates for salvage cryosurgery, because they are unlikely to be cured, and they appear.

Follow-up from the MD Anderson series indicated that 33% of patients were satisfied overall with their cryosurgery, and that overall satisfaction was lower for those with perineal pain and higher for those with controlled disease. The authors of that study concluded that, even with the use of a urethral warming catheter, the frequency of morbid events is high, and salvage cryosurgery does not appear to have any advantages over salvage prostatectomy in terms of morbidity or quality of life.

Prostate brachytherapy was described first by Pasteau et al. in 1911; although, until recently, it was considered a relatively untested salvage modality compared with the more robust experiences with salvage RP and cryosurgery. However, since 2003, 5 series that included 124 patients have been reported, nearly doubling the total number of patients reported since 1990 to 251.

Like in the salvage cryotherapy series, the overall PSA outcome with salvage brachytherapy is difficult to determine both because of varying definitions of PSA failure and because of the different lengths of median follow-up. The largest series from Grado et al. included 49 patients with a median follow-up of 64 months and reported a relatively low 5-year failure-free survival rate of 34% using 2 rises above the nadir as the definition of failure. The Arizona Oncology series (n = 17 patients) published by Beyer and the Mount Sinai series (n = 30 patients) published by Lo et al. also had mature follow-up at 62 months and 59 months, respectively, and reported 5-year PSA control rates (American Society for Therapeutic Radiology and Oncology [ASTRO] 1997 consensus definition) of 53% and 57%, respectively. Early PSA results from Lee et al. at the University of California-San Francisco (UCSF) have been encouraging: Those authors reported an 89% PSA control rate at 2 years using high-dose-rate brachytherapy. In addition, early data from the Uro-Radiology Prostate Institute indicated an 87% crude PSA control rate at 30 months. These encouraging early results potentially may be related to the significant number of men with PSA levels < 10 ng/mL at salvage (100% and 97%, respectively). However, both series also used the ASTRO 1997 consensus definition in which failure is not determined until there have been 3 rises above the nadir; then, failure is backdated to the midpoint between the nadir and the first rise. This can create an artifact of a time lag in detecting failures and has the effect of underestimating failure when follow-up is short. Consequently, given the relatively short follow-up in those series (19 months and 30 months, respectively), PSA control may have been overestimated, and further follow-up will be needed to determine whether these early results are maintained. In the Dana Farber series, which used a nadir +2 ng/mL definition for failure that was not backdated, the 4-year PSA control rate was 70%, which was similar to the 75% rate reported at the Mayo Clinic by Wong et al. using the ASTRO 1997 definition of failure but with a 44-month median follow-up.

None of the published series reported any presalvage factors that had a statistically significant association with PSA outcome, although all 4 failures in the Uro-Radiology Prostate Institute series had both a presalvage PSA level > 10 ng/mL and presalvage Gleason scores 7. Similarly, investigators in the Arizona Oncology series reported that the 5-year PSA failure-free survival rate was worse numerically for patients with presalvage PSA levels 10 ng/mL versus < 10 ng/mL (25% vs 67%, respectively; P = .12) and for patients with presalvage Gleason scores 7 versus 6 (30% vs 83%, respectively; P = .15).

Overall, the rate PSA control achieved with salvage brachytherapy appears to be comparable to what has been reported for salvage RP and somewhat better than what has been reported for salvage cryosurgery. However, as discussed above, it is not possible to make definitive statements regarding relative efficacy without a prospective randomized trial because of the different definitions of PSA failure, the unknown impact of patient selection, the associated known and unknown prognostic factors on PSA outcome.

The incidence of morbidity and the complications in salvage brachytherapy series published in English since 1990 are listed No perioperative mortality has been reported in any of the series. Although it was not studied formally, based on first principles and clinical common sense, reirradiating a patient who already has ongoing RT proctopathy or cystitis from the prior RT would be contraindicated because of the very high risk of rectal or bladder injury.

Incontinence was reported by from 0% to 31% of men, with a prevalence of 6% among all men in the combined series. Grade 3 or 4 GI toxicity ranged from 0% to 24% and was reported in 5.6% of all men in the combined series. Grade 3 or 4 GU toxicity was reported in 0% to 47% of patients, with an overall rate in the combined series of 17%. The most serious complications were prostatic-rectal fistulas, which accounted for the majority of complications that required surgical intervention. Specifically, the rate for developing of a fistula ranged from 0% to 12% and averaged 3.4% across all men reported from all series.

Given the significant decrement in quality of life imposed by grade 3 or greater toxicity, Nguyen et al. examined the Dana Farber Cancer Institute/Brigham and Women's Hospital series for factors that were associated significantly with grade 3 or 4 GI or GU toxicity and observed that an interval < 4.5 years between initial radiation and salvage radiation was associated significantly with the time to grade 3 or 4 toxicity (hazards ratio [HR], 12.03; 95% confidence interval [95% CI], 1.44-100; P = .02). Similarly, an interval < 4.5 years between radiation courses also was associated with a shorter time to colostomy and urostomy because of a fistula (HR, 24.6; 95% CI, 1.1-529; P = .04).

Although the rectal injury rate of < 5% is similar across all 3 major forms of local salvage, the 6% incontinence rate for salvage brachytherapy appears to be lower than the average of 36% that was estimated for salvage cryotherapy and the 41% average that was estimated for salvage RP. The frequency of fistulas and of grade 3 or 4 toxicity is not very high; however, when they occur, they can have a significant impact on quality of life. Therefore, research is needed to identify pretreatment factors that are associated with the occurrence of grade 3 and 4 complications after all types of salvage local therapy.

In addition to the 3 main salvage therapies discussed above, there is interest in alternative therapies aimed at treating local radiation failures with minimal morbidity, although data are limited. In 2004, Gelet et al. reported on 71 men who received salvage high-intensity focused ultrasonography (HIFU) to the whole gland after the failure of external beam radiation. Currently, HIFU is not practiced in the U.S., but the 30-month rate of freedom from biochemical or histologic failure in that European study was 38%. Morbidity included 7% grade 3 incontinence, 17% bladder neck stenosis, and 6% rectal fistula, although follow-up was only 14.8 months and may have underestimated late toxicity. In addition, Master et al. of UCSF have described ferromagnetic thermal ablation of local recurrence in a series of 14 men, and there has been a planning study on salvage interstitial microwave thermal therapy after brachytherapy as well as a case report of external beam radiation after local failure on brachytherapy. Currently, each of these modalities requires more study before they can be compared with other salvage therapies that have more data available.

After critically reviewing the literature, based on the retrospective nature of the comparisons, it is not possible to ascertain whether PSA outcome is best after salvage prostatectomy, cryosurgery, or brachytherapy. Only a prospective randomized trial can make definitive statements regarding comparative outcomes, and such studies are needed.

The men who are most likely to have local-only disease at the time of PSA failure and, thus, who are most likely to benefit from salvage local therapy can be selected based on factors identified in the literature, including initial low-risk disease (PSA < 10 ng/mL, Gleason score 6, clinical T1c or T2a tumor status), pretreatment PSA velocity < 2.0 ng/mL per year at the time of initial presentation, interval to PSA failure > 3 years, PSA-DT > 12 months, negative bone scan and pelvic imaging studies, and positive rebiopsy.

In addition, prognostic factors have emerged that are associated with lower rates of PSA recurrence by 5 years after any of the salvage local therapies, and these include a presalvage PSA level < 10 ng/mL and a presalvage Gleason score < 7 read from a biopsy specimen without significant RT effect. Also, there is evidence from the salvage prostatectomy and cryotherapy series to suggest that patients with bulky or locally advanced clinical T3/T4 disease at the time of salvage have a very low likelihood of being cured by a salvage local therapy. Patients who meet the above criteria likely will have relatively low-volume and relatively indolent disease and, thus, also should have a life expectancy > 10 years to have a reasonable chance of gaining a benefit from salvage therapy that will justify the toxicity of therapy.

Complications after any salvage local therapy in the post-RT setting are significant; therefore, these techniques should be performed only on study or by a physician in a center of excellence with significant prior experience. Patient selection may help reduce morbidity by avoiding salvage modalities in patients with pre-existing bladder and/or bowel toxicity from prior RT. Additional efforts currently are underway to reduce the long-term morbidities of these therapies, and there is evidence to suggest that morbidity outcomes have been improving over time. Conversely, most of the series that we examined in this review reported on patients who received RT in the conventional or lower dose era (66-70 gray [Gy]); and, because many patients now are moving to higher radiation doses for their primary therapy (75.6-79.2 Gy), it is unknown whether morbidity from a salvage local therapy will be more severe for the patients who receive initial high-dose radiation.

Although urinary incontinence appears to be greater after salvage RP (41%) or cryosurgery (36%) than after salvage brachytherapy (6%), patients who receive salvage brachytherapy face a 17% risk of grade 3 or 4 genitourinary complications and a fistula risk that ranges from 0% to 11% and averages 3.4% across all series. Additional research is needed to identify the factors associated with an increased risk of significant complications after salvage local therapies to help improve patient selection and to augment the benefit/risk ratio associated with attempts to cure local-only recurrences after radiation therapy.