The Use of 68Ga-PET/CT PSMA to Determine Patterns of Disease for Biochemically Recurrent Prostate Cancer Following Primary Radiotherapy

Background - 68Ga-PET/CT PSMA scan is being increasingly used for the staging of biochemically recurrent disease. Early identification of recurrent disease after radiotherapy is important in considering suitability for early salvage therapy to improve prognosis. The aim is to identify patterns of suspected prostate cancer recurrence in relation to post-radiotherapy PSA levels, especially below the accepted Phoenix definition of PSA failure (PSA nadir + 2).

Methods - This was a retrospective single tertiary institution cohort study of consecutive men between July 2014 and June 2018 who received a 68Ga-PSMA PET/CT for elevated PSA levels following radiotherapy as primary treatment of prostate cancer. The primary outcome measure was to determine the relationship between pre-scan PSA and the probability of identifying PSMA-avid disease suggestive of recurrent prostate cancer.

Results - Two hundred and seventy-six patients met criteria for inclusion. The median PSA was 3.60 ng/mL. The overall detection rate for suspected recurrent prostate cancer was 86.3%. Local recurrence was the most common site, occurring in 56.9% (157/276) of men, with isolated local recurrence in 32.6% (90/276). A total of 75.3% (55/73) of men below Phoenix criteria had scans suggestive of recurrent disease, with 52.1% of men having salvageable disease. The regions surrounding the iliac arteries were the most common areas of nodal metastatic disease, with 55.6% of recurrence occurring in the iliac regions.

Conclusions - 68Ga-PSMA PET/CT frequently identifies suspected recurrent disease prior to the accepted Phoenix definition of PSA nadir +2. Prospective outcome studies are required to determine if early identification of local recurrence improves outcomes by increasing the use of salvage local treatments and whether earlier identification of metastatic disease may improve outcomes with prompt initiation of multimodality therapies.


Prostate-specific membrane antigen (PSMA) is a cell surface protein with elevated expression in prostate cancer cells compared with benign prostate tissue, which makes it a favourable target for imaging1. 68Ga-PET/CT PSMA scan is a useful modality in the investigation of recurrent prostate cancer, with sensitivities and specificities for detecting metastatic disease post radical prostatectomy of 33–93% and >99%, respectively2,3. The detection of recurrent disease on 68Ga-PET/CT PSMA also increases with increasing PSA levels4. However there is little data on utilisation in the staging of a rising PSA after radiotherapy as primary treatment for prostate cancer. The Phoenix Consensus defines biochemical recurrence after prostate radiotherapy as any PSA increase 2.0 ug/L over the nadir5. This definition was introduced in 2006 and has not been updated since, despite rapid evolutions in imaging techniques, as well as therapeutic options for both local and distant control. Early identification of recurrent disease after radiotherapy is important in considering suitability for early local salvage, or multimodality systemic therapy to improve prognosis. The aim of this study is to identify patterns of suspected prostate cancer recurrence in relation to postradiotherapy PSA levels, especially below the accepted Phoenix definition of PSA failure.



Ethics approval was obtained from United Care Health Human Research Ethics Committee ref: 2019.02.280. This was a retrospective single tertiary-referral institution cohort study of consecutive men between July 2014 and June 2018 who received a 68Ga-PSMA PET/CT for elevated PSA levels following radiotherapy (either external beam radiotherapy (EBRT) or brachytherapy) for primary treatment of prostate cancer. PSA levels were not available for 27 men who were excluded from analysis. The primary objective was to determine the relationship between pre-scan PSA and the probability of identifying recurrent disease. The secondary objective was to identify the locations for recurrent disease. The final objective was to identify patterns of suspected recurrence below the accepted Phoenix definition of PSA failure. Because of the influence of neoadjuvant androgen deprivation therapy (ADT) on PSA levels in the EBRT group, we used a PSA of ≥2 as a surrogate for biochemical failure. EBRT doses ranged from 60 to 80 Gy (median 78 GY). The low dose rate brachytherapy implants (LDR) were performed with live intra-operative planning and insertion of radioactive I-125 seeds treating to a minimal peripheral dose of 145 Gy. The high dose rate brachytherapy (HDR) treatment combined three fractions of HDR with Ir-192 with three fractions of 7 Gy combined with 46 Gy EBRT.

Imaging Protocol

68Ga-PSMA PET/CT scans were performed on a Phillips Ingenuity camera with 128 slice CT and a GE Discovery MI with 128 slice CT in later scans. The Phillips uses time of flight (TOF) and GE has both TOF and filtered back projection (Qclear). The minimum uptake time was 45–60 min following administration of an average injected dose of 200 MBq Glu-NH-CO-NH-Lys(Ahx)-HBED-CC (PSMAHBED) using 3–4 min frames over 12–13 frames with low dose CT for attenuation correction and anatomic localisation. A diagnostic CT scan with IV contrast was also performed to discriminate radio-urine in the ureters from involved nodes.

Image Analysis

Each 68Ga-PSMA PET/CT was analysed by one of several experienced nuclear physician radiologists. A second read was performed for any equivocal lesions. 68Ga-PSMA PET and CT scans were reported separately, before results were combined. Scans requiring clarification were discussed at the Uro-radiology Multidisciplinary Meeting. For diagnosis of a metastasis, a concordant moderate or intense PSMA uptake in a CT anatomical lesion was required. These included PSMA uptake in a lymph node, bone (sclerotic or lytic) or viscera as identified with CT. For a solitary bone lesion, any discordance between PSMA avidity and CT was further investigated with a dedicated MRI of the bone lesion. Local recurrence was defined as a moderate or intense PSMA avid lesion within the prostate, with an SUV max ≥ 3.

Statistical Analysis

PSA, a continuous variable, was dichotomised (<2 ng/mL, ≥2 ng/mL) and used as the response variable in our analysis. The relationship between PSA and several categorical variables (mostly binary) were analysed using Pearson χ2 test. χ2 tests were considered appropriate as in all cases, individual cell frequencies were >5. Alpha level for significance was set at 0.05.


Two hundred and seventy-six men met criteria for inclusion in analysis. The median age was 75 years (60–94 years). The median PSA was 3.60 ug/L (0.01–83 ug/L). One hundred and twenty-nine men received EBRT. One hundred and forty-seven men received brachytherapy, with 94 men receiving HDR and 30 receiving LDR. The other 23 men received their radiotherapy at a different institution, with exact radiation regimes unknown. The median time between radiotherapy and 68Ga-PSMA PET/CT was 91.5 months.

Detection rate

Two hundred and thirty-nine (86.3%) men showed evidence of suspected recurrent disease on 68Ga-PSMA PET/CT (Table 1), including 73 men (26.3%) with PSA below the Phoenix definition for radiotherapy failure. The highest probability of detecting disease was in men with PSA >2.0, with a detection rate of 90.6% (184/203). Of 8/12 men (66.7%) with PSA <0.5 had PSMA avid areas suggestive of recurrent disease on 68Ga-PSMA PET/CT. Only one man with a PSA <0.5 proceeded to prostate biopsy, which was positive.

Table 1 Detection Rates

Location of avid disease

There were 342 sites of recurrent disease in the 239 men with a positive 68Ga-PSMA PET/CT (Table 2). More than half of all men displayed evidence of local recurrence (56.9%) and was the most common site of recurrent disease for all patients with PSA >0.5. Of the 33 men who proceeded through to biopsy, 28 were confirmed to have histological local recurrence (84.8%). Of the sub-group with PSA <2.0, six men received a prostate biopsy, with four confirmed to have recurrent disease. Local recurrence was more likely in men with a PSA <2, but did not reach statistical significance (p = 0.072). Lymph node metastasis was suspected in 44.4% of men and was the most common site of recurrence with PSA <0.5. There was no difference in detection of lymph node metastasis between PSA < 2 or PSA ≥ 2 (p = 0.332).

Table 2 Location of Avid Disease

Local recurrence only was found in 90/276 (32.6%), including 47/129 (36.4%) of the external beam cohort and 43/141 (30.5%) of the brachytherapy cohort. There was no difference in local recurrence between the external beam or brachytherapy cohort with PSA <2 (p = 0.41) or ≥ 2 (p = 0.40). Of 158/276 (57.2%) men had salvageable disease with local recurrence +/− nodal disease within the boundaries of an extended pelvic lymph node dissection, which includes the obturator, external/internal iliac and common iliac nodes. Of the 73 men below the Phoenix definition for failure, 68Ga-PSMA PET/CT was positive in 55 men (75.3%). Of this group 22/73 (30.1%) had local recurrence only. Of 38/73 (52.1%) had local recurrence and/ or nodal disease within an ePLND template and were thus suitable for local salvage treatment (Fig. 1).

Fig 1 Breakdown of Men with Salvageable and Non Salvageable Disease

Skeletal metastases were identified in 20.9%, while visceral metastasis was relatively rare (2.2%). Biopsies of the bone metastasis were not performed, as there was either concordance with the CT or a subsequent dedicated MRI of the lesion, or alternatively a positive biopsy would not have changed clinical management in men with multiple sites of metastasis.

Location of lymph node metastasis

Of the 122 men with suspected recurrent lymph node metastatic disease on 68Ga-PSMA PET/CT, 282 discrete nodal regions showed evidence of lymph node metastasis (Fig. 2), of which 49 men (39.8%) had node metastasis within the boundary of an ePLND. Of the suspected lymph node metastasis, 55.3% occurred in the common, external and internal iliac regions. A single lymph node metastasis was identified in 56/122 men, of which 38 men (67.9%) had a single node within the field of an ePLND, irrespective of PSA level above or below 2 (p = 0.306). A PSA above or below 2 did not predict the presence of multiple node metastases (p = 0.117). The retroperitoneum was the next most common site of disease (21.7%), with 11.9% of nodal recurrence occurring above the diaphragm.

Fig 2 Location of Lymphatic Metastatic Disease


To our knowledge, this is the largest dataset in the literature on 68Ga-PSMA PET/CT for investigation of men with biochemical failure following primary radiotherapy. Despite dose escalation of 78 Gy with IMRT, 23–44% of men with intermediate or high-risk disease develop biochemical recurrence6. While there have been a few studies investigating the role of 68Ga-PSMA PET/CT in the staging of biochemically recurrent disease after surgery or radiotherapy, most of these studies have either not specified the primary treatment received, or have included small numbers of men who have undergone primary radiotherapy. Our aim was to observe the relationship between pre-scan PSA and the ability of 68Ga-PSMA PET/CT to detect recurrent disease. Of particular interest was the locations of cancer recurrence.

The results indicate that recurrent prostate cancer following primary radiotherapy can be identified at low PSA ranges on 68Ga-PSMA PET/CT, below the Phoenix definition of post radiotherapy/brachytherapy failure. Of the men with PSA <2.0, 75.3% had PSMA-avid evidence of recurrent disease. Despite low PSA levels, only 30.1% of these men had evidence of local-only recurrence suitable for salvage therapy with curative intent. The potential for early salvage treatment below the Phoenix definition of failure increased to 52.1% (38/73) if we include men with lymph node recurrence within the field of salvage radiotherapy or a salvage ePLND. With advances in salvage radiotherapy and surgical techniques, detection of recurrent disease below the Phoenix definition may identify candidates suitable for salvage therapeutic options with curative intent earlier.

Hruby et al.7 recently published findings from 48 men with suspected biochemical recurrence (nadir + 2) following radiotherapy with previous negative conventional imaging (CT or bone scan) who proceeded with 68Ga-PSMA PET/CT. All 48 men were found to have PSMA-avid disease not previously identified by conventional imaging. Fifty-two percent of this cohort had metastatic disease, predominantly to the lymph nodes. The median PSA was 5.1 ug/L, higher than our median of 3.60 ug/L. Unfortunately, areas of lymphatic spread or the relationship between PSA and likelihood of detection of PSMA-avid disease was not clarified. Local recurrence suitable for salvage treatment was only identified in 16.7% of patients. In contrast, in our series, local recurrence only in the entire cohort was identified in 32.6% of men, with 58.0% having recurrent disease amenable to local salvage techniques extending to include the pelvic lymph nodes. Therefore, performing a 68Ga-PSMA PET/CT at lower PSA levels will result in earlier identification of men suitable for salvage with curative intent.

Einspieler et al.8 reported detection rates of 90.7% in their cohort of 118 patients post radiotherapy following either brachytherapy (41 men) or external-beam radiotherapy (77 men). The median pre-scan PSA was 6.4 ug/L and 45 men received ADT within 6 months of the scan8. However unlike our cohort, Einspieler’s study defined biochemical recurrence as per the Phoenix Consensus, with all patients having PSA >2.0. If we compare our subset of patients with PSA >2.0 to Einspieler’s cohort, then our detection rates, (90.6% vs. 90.7%), rates of local recurrence (58.0% vs. 63.5%) and lymph node metastasis (44.2% vs. 42.9%) are similar. Unlike prior studies, we have histological confirmation of 33 men who proceeded with cognitive biopsy of the PSMA-avid area, which identifies recurrent cancer in 28/33 (84.4%) of this group. Our group has previously published the results of histological correlation between pre-operative 68Ga-PSMA PET/CT and the histology of resected lymph nodes at radical prostatectomy, showing a specificity of 93.5%9. Similarly, histological lymph node metastases were also confirmed in 91% of a series of salvage PLND for recurrent PSMA detected oligometastasis avid node metastasis10.

The Phoenix Consensus in 2006 defined PSA failure following primary radiotherapy as any PSA increase 2.0 ug/L higher than the PSA nadir value, regardless of serum concentration of PSA5. With a total PSA >2.0, our results identify 90.6% of men presenting with suspected recurrent disease. However, our findings suggest we are missing the opportunity for early detection of recurrent prostate cancer in a significant portion of men if adhering strictly to the Phoenix criteria. This is notwithstanding the fact 68Ga-PSMA PET/CT has been shown to underestimate the presence of pelvic lymph node metastasis at low post treatment PSA levels9.

It is indisputable that ADT is the most common treatment for men with multiple sites of recurrent disease. The timing of commencement and the regime used remain topics of research11. Men with increasing PSA levels with no symptomatic or clinical evidence of cancer after primary treatment are a therapeutic dilemma with respect to the commencement of ADT, with only 2% of men receiving any form of salvage therapy11,12. The NCCN suggests men with long life expectancy should consider early ADT, regardless of negative imaging and clinical findings of metastatic disease11. ADT is not without considerable negative quality of life issues including metabolic syndrome, hot flushes, impotence and cognitive dysfunction13. Therefore, detection of local recurrence at low PSA by 68Ga-PSMA PET/CT on lesions not seen on standard radiology will provide the option for early local salvage treatments and potentially avoid/delay the introduction of ADT.

Local recurrence after brachytherapy usually occurs at the site of the largest or highest-grade index lesion14,15. With earlier identification of local recurrence, patients have the option of salvage local therapy with radical prostatectomy, brachytherapy or increasingly and focal therapy. Early detection of local recurrence suitable for local salvage focal therapy may decrease the morbidity of a post radiotherapy salvage radical prostatectomy. However, there is excellent 10-year prostate cancer-specific survival following a salvage radical prostatectomy16, with pre-salvage PSA the strongest predictor of progression free, cancer-specific and overall survival17.

Early detection of lymph node metastasis limited to the pelvis also provides the option for salvage pelvic lymph node radiotherapy, or salvage PLND. However, the longterm oncological outcomes of salvage therapies are not robust. If further evaluation with larger scale prospective cohorts identifies a higher probability of men suitable for salvage treatments, then the Phoenix definition may need to be revised. If our results are confirmed by other studies, then evaluation for recurrent prostate cancer after radiotherapy could commence at lower PSA levels above 0.5. While there is little evidence to support a difference in prostate cancer-specific survival between radical prostatectomy and radiotherapy within a decade of treatment, if comparing oncological outcomes of surgery and radiotherapy based on the Phoenix definition, then the oncological outcomes of radiotherapy may be overstated.

The primary limitation in this study is the histological confirmation of recurrent disease. Ideally, regions of 68GaPSMA PET/CT positivity would undergo conformational biopsy to determine the specificity of the imaging. Only 33 men proceeded to prostate biopsy, with a sensitivity of 84.4%. Biopsy was unlikely if metastatic disease was established, or there was a decision not to proceed with local salvage on clinical grounds, as positive biopsy would not change management. Histological correlation of 68GaPSMA PET/CT-avid disease has previously been performed, showing a per-node specificity of 99.5% and positive predictive value of 75% for a PSMA-avid pelvic lymph node9. Unfortunately 68Ga-PSMA PET/CT has a low per-patient and per-node sensitivity below a malignant node diameter of 5 mm9 and therefore some men with a negative 68Ga-PSMA PET/CT may still harbour micrometastatic disease, the oncological outcome of which may not be altered by early salvage local therapy. Also, unlike the high specificity in detecting lymph node metastasis, the specificity of bone metastatic disease on 68Ga-PSMA PET/ CT has not been established. Confirmatory bone biopsy was rarely performed unless discordance between the PSMA avid bone lesion and the CT/MRI existed, and only if the bone biopsy would change clinical management. Another limitation is the lack of data on PSA kinetics, which are a known prognostic factor for local versus metastatic recurrence. Finally, the identification of low volume micrometastatic disease with 68Ga-PSMA PET/CT below the limitations of standard radiology may initiate earlier systemic therapies, but the data showing improvements in long-term oncological outcomes are not available.However, the aim was to identify a relationship between pre-scan PSA and detection of avid-disease on 68Ga-PSMA PET/CT, and this significant cohort size allows for this relationship to be identified.


This study provides evidence of the relationship between pre-scan PSA and the probability of detecting 68Ga-PSMA PET/CT-avid disease following primary radiotherapy for prostate cancer. The study supports the use of 68Ga-PSMA PET/CT in evaluation of recurrent prostate cancer in men with rising PSA levels after radiotherapy or brachytherapy. This detection rate is 90.6% when PSA >2.0. Importantly, this study has identified a significant subset of men with PSMA-avid disease below the standard Phoenix definition of biochemical recurrence. Early identification of local recurrence allows the option of salvage local treatments and timely identification of metastatic disease may improve outcomes with prompt initiation of multimodality therapies. 

Compliance with ethical standards: Conflict of interest - The authors declare that they have no conflict of interest.

Authors: Sheliyan Raveenthiran1, John Yaxley1,2,3,  Troy Gianduzzo4, Boon Kua2, Louise McEwan5, David Wong5, Gail Tsang6, James MacKean6
1. Royal Brisbane and Women’s Hospital, Brisbane, QLD, Australia
2. Wesley Urology Clinic, Brisbane, QLD, Australia
3. Department of Medicine, University of Queensland, Brisbane, QLD, Australia
4. Brisbane Prostate Clinic, Brisbane, QLD, Australia
5. Wesley Medical Imaging, Brisbane, QLD, Australia
6. Genesis Cancer Care, Brisbane, QLD, Australia


1. Ghosh A, Heston WDW. Tumor target prostate specific membrane antigen (PSMA) and its regulation in prostate cancer. J Cell Biochem. 2004;91:528–39.
2. Fitzpatrick C, Lynch O, Marignol L. 68Ga-PSMA-PET/CT has a role in detecting prostate cancer lesions in patients with recurrent disease. Anticancer Res. 2017;37:2753–60.
3. Perera M, Papa N, Roberts M, Williams M, Udovicich C, Vela I, et al. Gallium-68 prostate-specific membrane antigen positron emission tomography in advanced prostate cancer-updated diagnostic utility, sensitivity, specificity, and distribution of prostate-specific membrane antigen-avid lesions: a systematic review and meta-analysis. Eur Urol. 2019. 1016/j.eururo.2019.01.049. pii S0302-2838(19)30095-8 (1873- 7560 (Electronic)).
4. Meredith G, Wong D, Yaxley J, Coughlin G, Thompson L, Kua B, et al. The use of 68Ga-PSMA PET CT in men with biochemical recurrence after definitive treatment of acinar prostate cancer. BJU Int. 2016;118:49–55.
5. Roach M, Hanks G. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOG-ASTRO Phoenix Consensus Conference. Int J Radiat Oncol Biol Phys. 2006;65:965–74.
6. Zumsteg ZS, Spratt DE, Romesser PB, Pei X, Zhang Z, Polkinghorn W, et al. The natural history and predictors of outcome following biochemical relapse in the dose escalation era for prostate cancer patients undergoing definitive external beam radiotherapy. Eur Urol. 2015;67:1009–16.
7. Hruby G, Eade T, Kneebone A, Emmett L, Guo L, Ho B, et al. Delineating biochemical failure with 68Ga-PSMA-PET following definitive external beam radiation treatment for prostate cancer. Radiother Oncol. 2017;122:99–102.
8. Einspieler I, Rauscher I, Duwel C, Kronke M, Rischpler C, Habl G, et al. Detection efficacy of hybrid 68Ga-PSMA Ligand PET/CT in prostate cancer patients with biochemical recurrence after primary radiation therapy defined by phoenix criteria. J Nucl Med. 2017;58:1081–7.
9. Yaxley J, Raveenthiran S, Nouhaud F, Samartunga H, Yaxley A, Coughlin G, et al. Outcomes of primary lymph node staging of prostate cancer with 68Ga-PET/CT PSMA scanning in intermediate and high risk prostate cancer, compared to histological correlation of pelvic lymph node pathology. Can a pre-operative 68Ga-PET/CT PSMA scan replace a PLND for prostate cancer staging? J Urol. 2019;201:815–20.
10. Siriwardana A, Thompson J, van Leeuwen PJ, Doig S, Kalsbeek A, Emmett L, et al. Initial multicentre experience of 68galliumPSMA PET/CT guided robot-assisted salvage lymphadenectomy: acceptable safety profile but oncological benefit appears limited. BJU Int. 2017;120:673–81.
11. James LM, Emmanuel SA, Andrew JA, Anthony VDA, Brian JD, Tanya D, et al. Prostate Cancer, Version 2. 2019, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Cancer Netw. 2019;17:479–505.
12. Tran H, Kwok J, Pickles T, Tyldesley S, Black PC. Underutilization of local salvage therapy after radiation therapy for prostate cancer. Semin Urol Oncol. 2014;32:701–6.
13. Gardiner RA, Green H, Yaxley J, Nicol D, Pakenham K, Mactaggart P, et al. Cognition and hormonal manipulation in prostate cancer. Br J Urol Int. 2000;86(Supplement 3):218.
14. Cellini N, Morganti AG, Mattiucci GC, Valentini V, Leone M, Luzi S, et al. Analysis of intraprostatic failures in patients treated with hormonal therapy and radiotherapy: implications for conformal therapy planning. Int J Radiat Oncol Biol Phys. 2002;53:595–9.
15. Pucar D, Hricak H, Shukla-Dave A, Kuroiwa K, Drobnjak M, Eastham J, et al. Clinically significant prostate cancer local recurrence after radiation therapy occurs at the site of primary tumor: magnetic resonance imaging and step-section pathology evidence. Int J Radiat Oncol Biol Phys. 2007;69:62–9.
16. Chade DC, Shariat SF, Cronin AM, Savage CJ, Karnes RJ, Blute ML, et al. Salvage radical prostatectomy for radiation-recurrent prostate cancer: a multi-institutional collaboration. Eur Urol. 2011;60:205–10.
17. Chade DC, Shariat SF, Cronin AM, Savage CJ, Karnes RJ, Blute ML, et al. Salvage radical prostatectomy for radiation-recurrent prostate cancer: a multi-institutional collaboration. Eur Urol. 2011;60:205–10.

Read a Commentary by Associate Editor of PCAN: Determining Patterns of Disease for Biochemically Recurrent Prostate Cancer by Using 68Ga-PET/CT PSMA - Commentary