Outcomes of Salvage Proton Therapy After Primary Cryosurgery or High-Intensity Focused Ultrasound for Recurrent Localized Prostate Cancer

Abstract

Purpose

To report long-term outcomes for patients treated with salvage proton therapy (PT) following primary cryosurgery (CS) or high-intensity focused ultrasound (HIFU) for recurrent localized prostate cancer.

Materials and Methods

We assessed 31 patients who underwent salvage PT at our institution between 2007 and 2017 for biopsy-confirmed, localized prostate cancer recurrence after CS (n = 21) or HIFU (n = 10). Patients had no evidence of nodal or distant metastases. Median PT dose was 74 CGE (range, 70.2-82) delivered at 1.8 to 2 CGE/fraction. Androgen deprivation therapy was administered to 12 patients for a median of 7 months (range, 6-24). Freedom from biochemical failure was defined per the Phoenix criteria. Freedom from regional failure and distant failure defined as the absence of radiologic or clinical evidence of metastatic disease in pelvic lymph nodes and distant to pelvic lymph nodes, respectively. Toxicities were graded using CTCAEV5. Patient-reported outcomes were assessed using the International Prostate Symptom Score and Expanded Prostate Index Composite. Kaplan-Meier estimates were used to calculate survival and oncologic outcomes.

Results

Median follow-up was 8.0 years (range, 1.9-13.5). The 5- and 10-year Freedom from biochemical failure rates were 72.4% and 64.7%, respectively. Overall survival at 5- and 10-year was 86.0% and 71.1%, respectively. At 10-years, freedom from regional failure was 73.7% and failure and distant failure was 96.8%. One late grade 3 hematuria occurred (3.2%), and 2 transient late grade 2 rectal hemorrhages occurred (6.4%). Median Expanded Prostate Index Composite and International Prostate Symptom Score scores remained stable from pre-PT through 5 years after treatment.

Conclusion

Our results demonstrate salvage PT as a safe and effective option following CS or HIFU with low toxicity and minimal impact on reported quality of life. PT outcomes are comparable with recent prostate cancer ablation salvage reports. Future studies should include multi-center collaborations to investigate elective pelvic nodal radiation to optimize treatment strategies for recurrent prostate cancer.

Introduction

An estimated 313 780 men were diagnosed with prostate cancer in 2025; of these, over 200 000 cases were likely localized to the prostate.¹ Most men choose the National Comprehensive Cancer Network (NCCN)-supported treatment regimens of external beam radiotherapy (EBRT), radical prostatectomy (RP), or brachytherapy (BT).2, 3, 4, 5, 6 Cryosurgery (CS) and high-intensity focused ultrasound (HIFU) are emerging primary prostate cancer treatments.7, 8, 9 With each treatment option, including CS and HIFU, there is a risk for local recurrence. EBRT can be used to salvage these recurrences.10, 11

CS, also known as cryotherapy or cryoablation, is a technique used to treat localized prostate cancer by destroying cancerous tissue with extreme cold.12, 13 During the procedure, thin needles are guided into the prostate using transrectal ultrasound, and a specialized gas is circulated through them to freeze the targeted cells. A warming catheter is typically used concurrently to protect the urethra from freezing injury. In a randomized controlled trial comparing CS to EBRT for primary prostate cancer treatment, CS was found to be non-inferior to Conventional 2-Dimensional Radiotherapy at 36 months for the primary endpoint of disease progression with 23.9% progression.¹⁴ The NCCN guidelines list whole gland CS as an experimental local, primary therapy option.⁶

HIFU uses focused, high-frequency sound waves, typically delivered transrectally under real-time monitoring, to generate thermal energy.15, 16 This energy is concentrated on small focal points within the targeted tissue in short pulses. The resulting heat induces coagulation necrosis, causing irreversible damage to the malignant lesion. In a retrospective analysis of 1002 patients treated with HIFU, intermediate-risk prostate cancer patients experienced 63% to 78% freedom from biochemical recurrence at 5-years.¹⁶ The NCCN guidelines point to the lack of a long-term randomized controlled trial comparing HIFU with either RP or EBRT and do not recommend HIFU as a primary treatment option outside of clinical trials.⁶

Salvage therapy options are needed for patients experiencing recurrence after primary CS or HIFU. Options include repeat focal therapy, salvage RP, or radiation therapy (RT).¹⁷ However, salvage treatments can be associated with increased morbidity compared to primary therapies.17, 18 Proton therapy (PT), a form of EBRT, offers potential dosimetric advantages by minimizing dose to surrounding normal tissues, which may be particularly beneficial in the re-treatment setting.¹⁹

A previous report from our institution demonstrated that salvage PT is safe and effective for after CS or HIFU.¹⁰ This study aims to report updated, long-term oncologic outcomes, toxicity profiles, and patient-reported quality of life (QOL) for patients treated with salvage PT following primary CS or HIFU for recurrent localized prostate cancer at our institution.

Materials and methods

We retrospectively analyzed data for 31 patients treated with salvage PT at the University of Florida Health Proton Therapy Institute between 2007 and 2017. Eligible patients were enrolled on an IRB-approved outcomes-tracking protocol (IRB201801009). Patients had biopsy-confirmed, localized prostate cancer recurrence after primary whole-gland CS (n = 21) or HIFU (n = 10). All patients had no evidence of nodal or distant metastatic disease based on standard workup at the time of salvage treatment, which included magnetic resonance imaging +/- pelvis computed tomography (CT), bone scan, chest X-ray, and lab tests [prostate-specific antigen (PSA), testosterone, and alkaline phosphatase]. Patients treated with focal ablation were excluded. No patients received SpaceOAR gel.

Patients underwent CT simulation with immobilization using a rectal balloon and instructions for a full bladder. Treatment planning followed Radiation Therapy Oncology Group (RTOG) contouring guidelines. The median PT dose delivered to the prostate bed was 74 Cobalt Gray Equivalent (CGE) (range, 70.2-82), typically delivered in 1.8-2.0 CGE fractions. The proximal 2 cm of the seminal vesicles were included in the clinical target volume for patients with a Gleason score of 3+4 = 7 or higher. Four patients received elective pelvic nodal irradiation using intensity-modulated radiation therapy (IMRT) up to 45 Gy determined by physician preference, followed by a PT boost. Daily image guidance utilized implanted fiducial markers with orthogonal kV x-rays and cone-beam CT. Androgen deprivation therapy (ADT) was administered to 12 patients (primarily higher-risk) for a median duration of 7 months (range, 6-24).

Follow-up occurred at 6 months, 1 year, 2 years, and then annually post-PT. Toxicities were graded by physicians according to the Common Terminology Criteria for Adverse Events Version 5.0 (retrospectively applied for consistency). Patient-reported outcomes were collected prospectively using the International Prostate Symptom Score (IPSS) and Expanded Prostate Index Composite (EPIC)-26 questionnaires at baseline and subsequent follow-up visits. Freedom from biochemical failure (FFBF) was defined using the Phoenix criteria (nadir PSA + 2 ng/mL). Freedom from regional failure (FFRF) was defined as the absence of pelvic nodal disease, and freedom from distant failure (FFDF) as the absence of disease beyond pelvic nodes, confirmed radiologically or clinically. Overall survival (OS), FFBF, FFRF, and FFDF were calculated from the end of PT using the Kaplan-Meier method.

Results

Thirty-one patients were included. Twenty-one had received primary CS (68%) while 10 had received HIFU (32%). The median age at salvage PT was 71.1 years. Median follow-up was 8.0 years (range, 1.9-13.5). Baseline characteristics, including Gleason score and pre-treatment PSA were recorded post-CS or HIFU. Those, along with treatment details like RT dose and ADT use, are summarized in Table 1. Most patients had a low pre-treatment PSA of ≤10 ng/mL (n = 25), received 74 GyRBE (n = 25), and on histology had a Gleason score of 4+3 (7b) or higher (n = 20). Twelve patients received adjuvant salvage ADT for a median of 7 months (range, 6-24).

Table 1.

Patient demographic, diagnostic, and treatment characteristics (N = 31).

Characteristic	Number or other value
Primary CS	21
Primary HIFU	10
Age (median range)	71.1 y (53.7-85.4)
Diabetes	3
A-bloc or 5FA reductase inhibitor	12
Prior TURP	0
PSA
≤10 >10 and ≤20 >20	25 6 0
Gleason score
6 3+4 = 7 (7a) 4+3 = 7 (7b) ≥8	6 5 11 9
RT dose
70.2 74-78 79.2-82	1 28 2
Pelvic nodal RT	4
ADT	12
Duration of ADT (median range)	7 mo (6-24)
Follow-up (median range)	8.0 y (1.9-13.5)

Abbreviations: CS, cryosurgery; HIFU, high-intensity focused ultrasound; TURP, transurethral resection of the prostate; PSA, prostate-specific antigen; RT, radiation therapy; ADT, androgen-deprivation therapy

Kaplan-Meier estimates for oncologic outcomes are shown in Figure 1 and Table 2. The 5- and 10-year OS rates were 86.0% and 71.1%, respectively. The 5- and 10-year FFBF rates were 72.4% and 64.7%, respectively. The 5- and 10-year freedom from regional failure (FFRF) rates were 83.3% and 73.7%, respectively. In aggregate, 12 patients experienced a biochemical failure, and 8 of these patients were later found to have a nodal failure as shown in Supplemental Table 1. One patient experienced a distant failure approximately 1 year after treatment completion. No patients died because of radiation or prostate cancer complications.

Figure 1.

Kaplan-Meier plots of estimated patient survival and oncologic outcomes through 10 years.

Table 2.

OS and oncological outcomes at 5 and 10 years after treatment completion.

	5-year (95% CI)	10-year (95% CI)
OS	86.0% (68.0-94.7)	71.1% (48.9-86.3)
FFDF	96.8% (85.6-99.8)	96.8% (85.6-99.8)
FFRF	83.3% (67.9-94.1)	73.7% (54.9-89.3)
FFBF	72.4% (55.2-87.3)	64.7% (46.8-82.0)

Abbreviations: CI, confidence interval; OS, overall survival; FFDF, freedom from distant failure; FFRF, freedom from regional failure; FFBF, freedom from biochemical failure

Acute toxicities were minimal. Late gastrointestinal (GI) toxicities grade ≥2 and genitourinary (GU) toxicities grade ≥3 are summarized in Table 3. Two patients (6.4%) experienced late grade 2 rectal hemorrhage requiring formalin application and Anusol suppositories, respectively. One patient (3.2%) experienced late grade 3 hematuria requiring transfusion. These toxicities resolved after treatment. Additionally, 2 patients developed second primary malignancies: melanoma on the chest and cutaneous squamous cell carcinomas on the head and arm, respectively. These malignancies occurred outside the treatment field, and University of Florida Health Proton Therapy Institute physicians evaluated them to be unrelated to PT.

Table 3.

Late GI and GU toxicities.

Category	Event	Grade	Intervention	Time after tx (years)	Incidence (%)
GI 2+					6.4 (2/31)
	Rectal Rectal	2 2	Formalin Anusol suppositories	2 1
GU 3+					3.2 (1/31)
	Hematuria	3	Transfusion	1.5

Abbreviations: GI, gastrointestinal; GU, genitourinary; tx, treatment

Median scores for EPIC urinary, bowel, and sexual domains, as well as IPSS scores, remained relatively stable from baseline through 5 years of follow-up (Figure 2). Bowel scores were high, indicating good function, during baseline evaluation at 96.4/100 for 30 patients responding and remained stable through 5 years when 10 patients reported at 98.2/100. Similarly, the median urinary scores were 95.5 at baseline and 95.2 at 5-year. Median EPIC sexual domain scores were 25.3 at baseline and 24.3 at 5 years. Finally, median IPSS scores, in which a score of 0 indicates no symptoms, were 6.5/35 at baseline and 6.0 at 5 years, indicating most patients were experiencing mild symptoms at those time points. Comparison of baseline characteristics of 5 years responders and 5-year non-responders revealed that 5-year responders had significantly higher urinary and bowel scores, while EPIC sexual domain scores and IPSS scores were similar.

Figure 2.

Bowel (A), urinary (B), sexual (C), and IPSS (D) QOL outcomes reported by patients at pre-tx, 2 yr, and 5 yr, Abbreviations: IPSS, International Prostate Symptom Score; QOL, quality of life; tx, treatment; yr, years.

Discussion

This study provides long-term outcome data for salvage PT following primary CS or HIFU, demonstrating durable biochemical control and survival with minimal toxicity. The 10-year FFBF rate of 64.7%, the OS rate of 71.1%, only a single distant failure, and no treatment- or disease-related mortality in this cohort are encouraging, especially considering the salvage setting. These results build upon our previous report, which showed a 3-year FFBF of 77% in a subset of these patients.¹⁰

A recent meta-analysis of salvage EBRT after ablative therapy provides valuable context.²⁰ Twenty-one studies were included with 435 (75.8%) patients treated with primary HIFU and 157 (24.2%) treated with primary CS. The median age was 68.4 years, and the median follow-up was 36.8 months. A median salvage RT dose of 73 Gy (range 35-79.2) was delivered in fractions of 1.8 to 7 Gy. The highlighted finding was a higher rate of treatment toxicity in patients who received primary HIFU than those who received primary CS before salvage EBRT. For example, 13 of 319 patients in the primary HIFU group experienced a late grade 3 to 4 GU toxicity, while only one of 186 patients in the primary CS group experienced a late grade 3 to 4 toxicity. The pooled biochemical failure rate was 0.20 with no significant subgroup difference between primary HIFU and primary CS groups.

In comparison, the present study has a significantly longer follow-up than any of the previously published cohorts: 96 months compared to the median follow-up of 36.8 (range, 23-76) of studies included in the meta-analysis. We found a late grade ≥3 GU toxicity rate of 3.2%, while the meta-analysis found a rate of 2%. Additionally, no men in our cohort experienced a grade ≥3 GI toxicity compared to about 1% in the meta-analysis. Since 2017, rectal spacers have been used, which may further reduce GI toxicity. Table 4 compares our study to recent publications in more detail. For instance, Nathan et al reported 5 (5%) late grade 3 GU toxicities, 3 (3%) late grade 3 GI toxicities, and one rectourethral fistula requiring surgical repair (grade 4) in 100 patients who underwent salvage IMRT (97%) or 3DCRT (3%) after ablation.¹¹

Table 4.

Literature review of salvage outcomes.

Study	Pts	Primary tx	Salvage tx	Median RT dose (Gy)	Pelvic lymph nodes	ADT	Median FU (mo)	FFBR (%)	GI and GU Toxicities
Lichlyter et al, Current series	31	CS (n = 21) HIFU (n = 10)	PT	74 (range, 70.2-82)	4	12	96 (range, 23-162)	5-y: 72.4 10-y: 64.7	GI: 2 late grade 2 GU: 1 late grade 3
Baude et al25 2025	56	HIFU	VMAT	78-80/2 fx (n = 46) 60/3 fx (n = 10)	21	27	19.5 (range, 5.5-47)	100a	GI: 4 acute grade 2, 1 late grade 2 GU: 1 late grade 3
Perennec et al26 2024	10	CS	IMRT	74-78/(2-2.2 fx	4	7	60 (range, 28-119)	NR, 8/10 no BCP	GI: 1 late grade 3 GU: 1 late grade 3
Nathan et al11 2022	100	HIFU (n = 92) CS (n = 5) EP (n = 3)	RP	N/A	N/A	0	16.5 (IQR: 10-30)	3-y: 69	1 post-op hematomab
	100	HIFU (n = 94) CS (n = 4) EP (2)	IMRT and 3DCRTc	74-78/2 fx (n = 97) 60-64/2 fx (n = 3)	NR	90	37 (IQR: 18.5-64)	3-y: 88.9	GI: 3 late grade 3, 1 rectoethral fistula GU: 5 late grade 3
Cafuta et al27 2022	13	HIFU	IMRT (n = 12) BT (n = 1)	73.6	7	6	76 (IQR: 55-96)	76.9	GI: 1 grade 2 GU: 1 grade 3
Mesci et al28 2022	14	HIFU	IMRT or VMAT	71 (range, 70-74)/2 fx	NR	7	54	5-y: 36	None
Di Lalla et al25 2022	12	HIFU	IMRT	76-78/2 fx (n = 11) 66/3 fx (n = 1)	NR	9	46	83.3	None
Jiang et al29 2021	11	CS	IMRT	68.4-81.0/1.8 fx)	NR	4	23 (range, 3-64)	NR, 10/11 BCP	GI: 1 late grade 2 GU: 1 late grade 3

Abbreviations: pts, patients; tx, treatment; RT, radiation therapy; Gy, Gray; yr, year(s); ADT, androgen deprivation therapy; FU, follow-up; mo, months; fx, fractionation; FFBR, freedom from biochemical failure; BCP, biochemical progression; GI, gastrointestinal; GU, genitourinary; CS, cryosurgery; HIFU, high-intensity focused ultrasound; PT, proton therapy; VMAT, volumetric modulated arc therapy; IMRT, intensity-modulated radiation therapy; BT, brachytherapy; RP, radical prostatectomy; EP, electroporation; IQR, interquartile range; N/A, not applicable; NR, nor reported.

^aBCP defined as nadir +2.

^bClaven-Dindo grade 3.

^cPatients were treated from 2010 to 2020; patients treated with radiation in 2010 and 2011 received 3DCRT while patients treated in 2012-onward received IMRT.

Figure 1 and Table 2 demonstrate that most of the patients who experienced biochemical failure also experienced regional nodal failure by 10 years of follow-up. There is no level 1 evidence supporting the role of elective pelvic nodal RT in this population. However, the recent POP-RT study showed that elective nodal RT reduced the risk for biochemical recurrence among patients with localized, high-risk prostate cancer receiving primary EBRT.²¹ Similarly, in the phase 3 NRG/RTOG 0924 trial, Roach et al. found that elective pelvic nodal radiation reduced the risk for biochemical failure at 10 years among patients with intermediate-risk prostate cancer treated with IMRT when compared to the prostate alone.²² Lastly, the NRG/RTOG 0534 SPPORT randomized trial showed that the addition of elective pelvic nodal RT among patients with biochemical recurrence of prostate cancer after RP improved freedom from biochemical recurrence compared to prostate bed RT alone.²³ Given the results of this study and the benefits demonstrated with elective pelvic nodal RT in other settings, it is strongly suggested to consider elective pelvic nodal RT in the salvage setting after primary CS or HIFU.

Encouragingly, median EPIC bowel and urinary scores, as well as IPSS, showed no statistically or clinically significant changes from baseline through 5 years post-PT. Sexual function scores were low pre-salvage treatment and remained so, suggesting that prior CS or HIFU could have permanently impacted sexual health before PT began. The urinary and bowel scores were impacted by survivorship bias, with the baseline scores of the 5-year responders being significantly higher than the 5-year non-responders.

Notably, the favorable toxicity and QOL outcomes of this study suggest that dose escalation above 74 Gy in fractions of 2 could be considered in this patient population. The results of Baude et al and Di Lalla et al support this option.24, 25 The former treated most of 56 patients with 78 to 80 Gy in fractions of 2 and followed them for a median of 19.5 months, and the latter treated most of 12 patients with 76 to 78 Gy in fractions of 2 and followed them for a median of 46 months. Neither found any grade ≥3 GI or GU toxicities. However, for patients whose recurrences were due to subclinical nodal disease at the time of treatment, escalating the dose to the prostate would not be expected to improve outcomes.

This study has limitations, including its retrospective nature, relatively small sample size (n = 31), and heterogeneity in primary treatments and use of ADT. Furthermore, advancements including PSMA-PET imaging, pencil beam scanning, and SpaceOAR were not standard during the treatment era, potentially impacting patient selection, staging, or outcomes and limiting generalizability compared with current practice. Despite these limitations, our findings suggest salvage PT is a viable and effective treatment option offering good long-term disease control with low rates of severe toxicity and stable urinary/bowel QOL.

A follow-up prospective series, especially a multi-center collaboration, may better reveal how a patient in this post-CS or HIFU population with newly diagnosed recurrent prostate cancer would fare if they received PT in the current era. Such a study should carefully track which patients receive pelvic nodal RT and ADT and which patients experience local failure versus regional failure. Participating physicians could consider a lower threshold for treating pelvic nodes. Ideally, a sufficient number of patients treated with PT, IMRT, and VMAT would be included, allowing for subgroup analyses.

Ethics

The study protocol was approved by the University of Florida Institutional Review Board (IRB201801009). All participants provided informed consent to participate in this study.

Funding

This study received no funding.

CRediT authorship contribution statement

Daniel A. Lichlyter: Conceptualization, Investigation, Data Curation, Formal Analysis, Methodology, Writing – Original Draft, Writing – Review and Editing. Nancy P. Mendenhall: Project Administration, Supervision, Visualization, Methodology, Resources, Validation, Writing – Original Draft, Writing – Review and Editing. Miscia Fortna: Investigation, Resources, Writing – Review and Editing. R. Charles Nichols: Writing – Review and Editing. Curtis M. Bryant: Supervision, Validation, Writing – Original Draft, Writing – Review and Editing.

Data Sharing Statement

The authors agree to share anonymized data upon reasonable request by researchers.

Declaration of Conflicts of Interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: NPM is the Editor-in-Chief of IJPT. WMM is the Advisory Editor for IJPT. CMB is an Associate Editor for IJPR. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

Supplementary material