Functional Outcomes after Local Salvage Therapies for Radiation-Recurrent Prostate Cancer Patients: A Systematic Review

Simple Summary Local salvage therapies are offered to prostate cancer (PCa) patients with recurrent disease following primary radiation therapy with initial curative intent. Favorable oncologic outcome is the primary aim of salvage therapy, but many patients may be left with treatment-related adverse consequences, potentially affecting their quality of living. The aim of this review was to evaluate the rates and severity of various functional outcomes after salvage therapy in patients with radiation recurrent PCa. Local salvage therapies are associated with impaired urinary and sexual functions depending on the specific salvage strategy. However, accurate estimation of the likelihood of these sequalae may be predictable based on the high pre-salvage general and disease specific health status. Despite these adverse consequences and impaired quality of life, oncologic advantage of local salvage treatment post radiation recurrence prostate cancer seems justified in general, but shared decision working with an informed patient is essential. This paper serves as a discussion platform for this process. Abstract Purpose: To assess the rate and severity of functional outcomes after salvage therapy for radiation recurrent prostate cancer. Methods: This systematic review of the MEDLINE/PubMed database yielded 35 studies, evaluating salvage radical prostatectomy (RP), brachytherapy (BT), high-intensity focal ultrasound (HIFU) and cryotherapy (CT) after failure of primary radiation therapy. Data on pre- and post-salvage rates and severity of functional outcomes (urinary incontinence, erectile dysfunction, and lower urinary tract symptoms) were collected from each study. Results: The rates of severe urinary incontinence ranged from 28–88%, 4.5–42%, 0–6.5%, 2.4–8% post salvage RP, HIFU, CT and BT, respectively. The rates of erectile dysfunction were relatively high reaching as much as 90%, 94.6%, 100%, 62% following RP, HIFU, CT and BT, respectively. Nonetheless, the high pre-salvage rates of ED preclude accurate estimation of the effect of salvage therapy. There was an increase in the median IPSS following salvage HIFU, BT and CT ranging from 2.5–3.4, 3.5–12, and 2, respectively. Extended follow-up showed a return-to-baseline IPSS in a salvage BT study. The reported data suffer from selection, reporting, publication and period of study biases, making inter-study comparisons inappropriate. Conclusions: local salvage therapies for radiation recurrent PCa affect continence, lower urinary tract symptoms and sexual functions. The use of local salvage therapies may be warranted in the setting of local disease control, but each individual decision must be made with the informed patient in a shared decision working process.


Introduction
The prevalence of prostate cancer (PCa) has been steadily increasing over the last few decades. The Global Burden of Disease study demonstrated a 3.7-fold rise in PCa incident cases between 1990 and 2015. This, together with the relatively modest PCa mortality and aging population, has generated a large pool of PCa survivors [1]. Despite the recent advances in PCa diagnosis and treatment, the burden of PCa remains significant, and PCa still constitutes the fifth most common cause of cancer mortality in the male population worldwide [2]. In addition, the economic burden of PCa management, in the United States alone, is considered the highest rising costs among all malignancies [3]. Furthermore, patients with PCa may suffer from disease-and treatment-related adverse events. [3].
There are several standard and widely used treatment options for patients with nonmetastatic PCa, including external beam radiotherapy (EBRT), brachytherapy (BT), and radical prostatectomy (RP) [4]. Elderly patients with several comorbidities and limited life expectancy are less likely to undergo RP and consequently, with a proportionally increasing probability, receive radiation therapy (RT) [5,6]. Many other patients opt for RT because of other reasons, including the desire to avoid adverse events of RP [7,8]. The use of escalated dose radiotherapy is taking over conventional control-dose radiation as the results of randomized trials have shown that dose escalation has superior benefit in terms of biochemical recurrence (BCR) rates, local disease progression and distant metastasis [9,10]. Despite curative intent, a significant proportion of these patients ultimately experience PCa recurrence. Biochemical recurrence is estimated to affect 32-57% in control-dose RT and 16-43% in escalated dose RT after 10-years follow-up [10,11]. BCR is, in fact, a robust surrogate of clinical disease recurrence, including local recurrence and distant metastasis [12][13][14].
The rapid advancement in diagnostic imaging has improved the sensitivity of tumor staging and detecting small metastatic deposits, decreasing the number of patients diagnosed with isolated local recurrence [15]. Diagnosing local recurrence can be challenging due to radiation-induced changes in the prostate, compromising MRI findings, and histopathological evaluation [16,17].
Nonetheless, when local recurrence is detected, patients may benefit from local salvage therapies, including salvage RT, cryoablation, high-intensity focal ultrasound (HIFU), and RP [12]. However, salvage therapies are underutilized due to the limited high-quality data, unclear survival benefits, and treatment-related adverse effects [18][19][20]. These therapies might also be under-studied or simply not reported owing to the adverse consequences and lack of consensus on treatment algorithm. Androgen deprivation therapy (ADT) is, therefore, the most commonly utilized therapy in patients with radiation recurrent PCa [13], and the decision to pursue additional salvage interventions should balance the oncologic efficacy, adverse effects, the expertise as well as patient wishes and life expectancy [12].
Contemporary evidence on functional outcomes after local salvage therapies in patients with radiation recurrent PCa is unclear. We aimed to explore the functional outcomes and quality of life after local salvage therapies in patients with radiation recurrent PCa, and to create an evidence catalogue serving as a framework for research and shared decision making.

Evidence Acquisition
This review followed the Preferred Reporting in Systematic Reviews and Meta-Analyses (PRISMA) protocol [21]. We conducted a systematic literature search using the PubMed-Medline database in July 2020, including articles published in the English language from January 2005 to June 2020. The search strategy included the following keywords in isolation or combination: "radio recurrent prostate cancer", "radiation relapse in prostate cancer", "radiation failure in prostate cancer", "local salvage therapy", "local therapy", "salvage therapy [MeSH]", "salvage cryoablation", "salvage cryotherapy", "salvage cryosurgery", "salvage high-intensity focal ultrasound", "salvage re-irradiation", "salvage brachytherapy", "salvage surgery", "salvage prostatectomy" AND "functional outcomes", "morbidity [MeSH]", "adverse effects", "patient-reported outcomes", "clinical outcomes" and "quality of life [MeSH]". The search was carried out by two authors independently, and any dispute was resolved by referring to a third author who acted as an arbiter.
The Population (P), Intervention (I), Comparator (C), Outcomes (O), and Study design (S) (PICOS) approach was used to specify the eligibility criteria. We considered a study eligible if it reported on PCa patients who were treated with primary radiation therapy (EBRT, BT, or both) and subsequently developed biochemical or clinical recurrence (P). Additionally, eligible studies should include data on patients treated by any local salvage therapy with or without ADT (I), and data on the treatment-related functional outcomes such as urinary function, sexual function, or quality of life (O) in prospective or retrospective studies (S). A comparator group (C) in each study was not necessary as an inclusion criterion since the outcomes of interest can be retrieved from case series or cohort studies and these outcomes can be compared across different studies.
We excluded review articles, case reports, articles reporting on ≤39 patients, comments, editorials, and conference abstracts. After duplicate removal, careful inspection of the remaining articles' titles and abstracts was undertaken to rule out non-relevant articles. Studies involving multiple-modality primary therapies (except ADT) without separate analyses for radiation therapy as primary treatment were excluded. We also excluded studies reporting on oncological but not functional outcomes. An additional search in the references of all included studies was performed to screen for any articles that might have been missed in the primary search. All included articles were obtained as full-text articles for vigilant evaluation.
Risk of bias assessment was performed using the National Institutes of Health (NIH) quality assessment tool for observational and cross-sectional studies [25]. This tool comprises 14 questions on each included study, addressing the quality of the study, the included cohort, ascertainment of exposure and outcomes of interest and follow-up data. A median follow-up period of 24 months post salvage therapy was defined as "adequate" for each included study. A point of "1" was given if the study fulfilled the information needed in the question. If not, a "0" point was given. If the question does not apply to the study or information not necessarily provided, a "N/A" score was given. As a total score, the summation of all the "1" scores was undertaken for each study.

Study Selection
The primary search identified a total of 3209 articles while searching through additional sources yielded another 113 articles. After duplicates were removed, 683 articles remained, which were evaluated by reading the title and abstract of each. A subsequent full-text evaluation resulted in 35 articles to be included in evidence synthesis. Figure 1 shows the PRISMA flow chart and study selection process.

Urinary Incontinence
UI as a functional outcome of post-surgical salvage therapy was measured using the number of daily pads required following surgery. The reported UI rates at 12 months postsurgery ranged from 48% to 85% [26,27,40,41]. The rate of severe UI was more than 23% in all studies reporting on salvage surgery, with Mohler et al. and Seabra et al. reporting rates of 85% and 72%, respectively [26,27].
In patients who underwent cryotherapy after RT, the reported UI rates were between 3.2% and 52%, but two studies reported an UI rate of 0%: de Castro Abreu et al. in the focal cryotherapy group and Clarke et al. [28,37,38,47,[49][50][51][52]58,59]. In addition, the reported rates of severe UI were less than 7%. Table 2 shows data on continence rates among included studies.    : mild UI was defined as any UI up to but not exceeding two pads per day, grade 1 (UI with coughing or sneezing) and 2 (UI with running or picking an object from the floor) Ingleman-Sundberg UI, grade 1 (occasional UI, no pads needed) CTCAE v4.0/v4.3, and grade 1 (occasional, no pads needed) and 2 (pads needed, not interfering with ADL) CTCAE v3.0. £: severe UI was defined as any UI exceeding two pads per day, UI requiring instrumentation or surgery, grade 3 (UI with walking) Ingleman-Sundberg UI, grade 2 (spontaneous UI, no intervention needed) and 3 (spontaneous, intervention needed) CTCAE v4.0/v4.3, and grade 3 (UI interfering with ADL) and 4 (surgery is needed) CTCAE v3.0 § : no further details on pads or grades provided. ₡ : Tan (a) used focal CT, while Tan (b) used total CT *: ≥1 pad with no other specifications **: the numbers reported do not match those of our table (e.g., <3 pads and ≥3 pads); the lower limit underestimates UI and the upper limit overestimates UI, so "less than" and "at least" were added accordingly. +: "Mild to moderate" was [  : Tan (a) used focal CT, while Tan (b) used total CT *: ≥1 pad with no other specifications **: the numbers reported do not match those of our table (e.g., <3 pads and ≥3 pads); the lower limit underestimates UI and the upper limit overestimates UI, so "less than" and "at least" were added accordingly. +: "Mild to moderate" was put as (mild UI), "severe" was put as (severe UI), [n] = number of patients assessed for given outcome, UCLA-PCI = University of California, Los Angeles Prostate Cancer Index.  [29,30,43,54,55]. Three studies reported severe UI rates of 8%, 4.6%, and 2.4%, respectively [29,30,43].
Van Son et al. reported a median IIEF score of 11.0 preoperatively, and of 7.0 at 1 month and 3.0 at 36 months follow-up in patients who received BT [29]. Yamada et al. reported a decrease in the median IIEF score from 2.0 to 1.5 among 42 patients at 36 months follow-up [30].
Only one study reported IIEF scores in patients subjected to cryotherapy, in which Bomers et al. demonstrated a decrease in mean IIEF score from 11.7 to 9.0 in 44 patients [58]. Table 3 shows detailed information about ED in patients who underwent salvage interventions.

Urinary Obstruction
The International Prostate Symptoms Score (IPSS) was used as a tool to assess the lower urinary tract symptoms in patients who underwent several salvage interventions. In patients who underwent salvage HIFU, Hostiou et al. and Ahmed et al. reported an increase in mean (median) IPSS from 5.6 [4] to 8.1 (7.4) and from 8.3 [7] to 11.6 (9.5) in 50 and 46 patients, respectively [35,44]. Baco et al. reported an increase in mean IPSS from 7.1 to 8.6 in 47 patients [33]. In

Risk of Bias Assessment
Other than Berge et al. [46], none of the included studies used a comparator group. Thus, the question on blinding the outcome assessors did not apply to the included studies and is thus given a "N/A" score (Question 12). The mean (median) bias scores of the studies were 8.6 [9], respectively. Most included studies showed an "intermediate" risk of bias, as 76% of the studies had a score 8-10 (Table 4). *: a median postoperative follow-up period of 24 months was considered adequate.

Discussion
The treatment of radiation recurrent PCa represents a challenge given the lack of consensus on patients' selection, heterogenous efficacy of local salvage modalities, and their variable toxicity profiles. It is often the last opportunity for local disease control and prevention of local and distant progression with its sequalae. The current review reports a relatively high prevalence of adverse functional outcomes following local salvage treatment for radiation recurrent PCa. It is worth noting that the pre-salvage rates of urinary and sexual dysfunction were generally high, probably owing to the effect of primary treatment. It is difficult to accurately determine the effect of salvage modalities on functional outcomes due to the significant heterogeneity among included studies, as well as the lack of standardized reporting methods and tools. Compared to primary RP, salvage RP is associated with a higher risk of complications such as ED, anastomotic stricture, urinary retention, urinary fistula, abscess, and rectal injury [61]. We observed high UI rates post salvage surgery, ranging from 48% to 85%, with severe UI rates exceeding those of mild UI. In addition, the rates of ED are high post salvage surgery. With longer follow-up, however, UI and ED rates declined. Moreover, the series were mostly older with significant changes in surgical technique and post-operative follow-up. Nonetheless, the relatively high preoperative rates preclude an accurate estimation of the effect of surgery, and limited follow-up of the studies hamper any long-term conclusions.
We found that ED rates are high, with at least half of men undergoing a local salvage treatment reporting ED. Interestingly, there was a high rate of ED before salvage therapy, with as much as 70% reporting ED. In a previous systematic review, Chade et al. found that 50-91% of patients had ED prior to salvage RP and 80-100% reported ED following salvage surgery [19]. In addition, urinary continence ranged between 21% and 90% after salvage surgery on longer follow up [19]. The variability of the findings may be mainly due to patient selection and surgical technique; in more recent series, these complications appear to be much less common, due to the progress in surgical technique and patient selection [62].
In patients who underwent salvage cryotherapy, we found that most of the included studies reported modest UI rates, mostly being mild. Two studies reported UI rates of 0% [38,51]. In a retrospective study of 143 patients who underwent cryotherapy after RT failure, Cespedes et al. reported high long-term UI rates of 28% at least 12 months after salvage cryotherapy, with 8-40% of patients reporting persistent rectal pain and 4% needing to undergo surgical procedures for the management of treatment-associated complications [63]. Notably, the introduction of cryotherapy has significantly decreased complications such as UI, fistulae, obstruction and ED [64]. For instance, in a recent study comprising only 14 patients, Boissier et al. reported de novo UI and ED in one patient, respectively [64]. Ideally, salvage cryotherapy has been suggested to be considered only for patients with comorbidities, a life expectancy of at least 10 years, an initial clinical stage of T1/T2, initial ISUP grade ≤ 2/3, a pre-salvage PSA-DT ≥ 16 months and a pre-salvage PSA ≤ 10 ng/mL [65].
High dose rate and low dose rateBT have been shown to be effective treatment options for histologically proven local PCa recurrence after RT with moderate Grade 3-4 gastrointestinal and genitourinary toxicity profiles ranging from 2.7-20% and 3-47%, respectively [66]. Overall, most salvage BT studies reported a 2.4% to 26.7% UI rate exceeding 1 pad daily. A meta-regression analysis comparing functional outcomes of surgical and non-surgical salvage modalities in radiation recurrent PCa found that both salvage BT and cryotherapy had significantly better results in terms of continence than salvage RP [67]. A subgroup analysis of patients who underwent non-surgical salvage modalities revealed that salvage BT and cryotherapy were comparable in UI rates. At the same time both were significantly better than salvage HIFU [67].
In terms of erectile function, the included studies illustrate the high rates of ED in this patient population, as shown by the decrease in IIEF scores compared to before therapy. However, a high-level of heterogeneity was noticeable between surgical and non-surgical salvage modalities indicating substantial differences between studies.
Salvage HIFU has recently emerged as an alternative thermal ablation option for radiation recurrent PCa [65]. In this review, most salvage HIFU studies reported continence rates above 50%. A meta-analysis showed that salvage HIFU did not demonstrate significantly better continence rates than salvage RP [67]. In general, the median followup did not exceed 24 months, a relatively adequate follow up period to report changes in outcomes [36,45]. Currently, there is a lack of high-quality data precluding robust recommendations regarding the indications for salvage HIFU.
We found considerable variability of functional outcomes after salvage therapy of radiation recurrent PCa. Proper patient selection and thorough consideration of the oncologic outcomes are critical factors in patient counselling and decision-making to achieve durable cancer control with the best possible quality of life.
In this review, we believe that the limitations of our work mainly stem from the heterogeneity of included studies, which even precluded a proper quantitative analysis. This review was based on case series and small cohort studies that lack a comparator group. The lack of extended follow-up duration was an issue only for some studies with outcomes as ED, whose rates might significantly change with time. Furthermore, this review focused on pertinent urologic functional outcomes, without analyzing other therapy-related adverse consequences such as GI toxicity. The significant heterogeneity of reporting outcomes and lack of pre-salvage rates of key outcomes also added to the problem. Therefore, it is difficult to draw solid conclusions with a high level of evidence. Indeed, prospective and comparative studies between different salvage modalities with long follow-up duration are needed to generate reliable evidence and validate long-term functional outcomes.

Conclusions
Local salvage therapies after radiation recurrent PCa are associated with impaired urinary and sexual functions. Accurate estimation of the impact of these therapies is precluded by the preoperative morbidity associated with primaryRT. Despite these adverse consequences, the oncologic advantage may justify the use of local salvage therapy post radiation failure in select informed patients who benefit from a balanced shared decisionmaking process.