Introduction

Prostate cancer (PC) is very common, being the fifth leading cause of cancer death in males [1]. The prognosis is generally favourable even if the rate of metastasis after the treatment of the primary tumor is relatively high, especially in patients with high-risk cancer. [2]. In patients with metastatic PC, international guidelines recommend androgen deprivation therapy (ADT) alone or in combination with other drugs based on the castration status [3, 4].

In patients with nodal metastases from PC, stereotactic body radiotherapy (SBRT) or single fraction stereotactic body radiosurgery (SRS) were tested with the aim to improve clinical outcomes and to delay next-line systemic treatments (NEST), based on their tolerability and efficacy in achieving prolonged local control [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19].

However, in this setting the evidence is sparse and inhomogeneous. Moreover, although most studies reported data on local control (LC) and/or progression free survival (PFS), specific information on biochemical [8, 9, 11, 15, 18, 19] and clinical response [8, 14], ADT- [7, 11, 18] or NEST-free survival [20] and overall survival are frequently lacking [7, 9, 14].

Furthermore, most available studies included patients with nodal oligorecurrent castration-sensitive PC while only one retrospective study reported on a small series (15 patients) of oligoprogressive castration-resistant nodal PC [21]. Therefore, there is a lack of clear evidence in this setting and in particular in the subgroup of patients with oligoprogressive CP.

Based on this background, the aim of this report is to analyze outcomes and toxicity of patients with nodal oligoprogressive PC enrolled in two dose-escalation phase I trials.

Materials and methods

Study design

This is a post hoc analysis of data from the DESTROY trials [22, 23], two multi-arm phase I studies exploring SBRT and SRS in several cancer settings. In details, the DESTROY-1 trial was a dose escalation study on SBRT delivered with fixed non-coplanar conformal fields or Volumetric Modulated Arc Therapy (VMAT) in patients with primary, oligorecurrent or oligometastatic cancers [22]. The DESTROY-2 trial was based on dose-escalated SRS delivered with VMAT in the same clinical settings [23]. Both trials were approved by the local Ethics Committee and by the Institutional Review Board. All patients signed a written informed consent before treatment. Details on end-points, inclusion–exclusion criteria, treatment planning and delivery, and results of these two trials have been previously reported [22, 23].

Inclusion criteria

In this analysis, we included oligoprogressive castration-resistant PC (mCRPC) patients with a small number (≤ 5) of metastatic sites who underwent SBRT/SRS on nodal metastases. Patients with visceral metastases were excluded. The definition of castration-resistance was based on the European Association of Urology guidelines [24]. Ongoing ADT and type of lesions (synchronous versus metachronous) did not represent exclusion criteria. The gross tumor volume (GTV) was identified by computed tomography (CT) and/or CT-positron emission tomography (PET) and/or magnetic resonance imaging (MRI) and the Clinical Target Volume was defined as the GTV. Organ motion and setup inaccuracies were analyzed to define the planning target volume (PTV) as previously described [22, 23]. Data on age, Eastern Cooperative Oncology Group performance status (ECOG-PS), comorbidities, pre- and post-radiotherapy prostate serum antigen (PSA) levels, systemic therapies, lesions site, fractionation, total prescribed dose, toxicity and outcome were collected.

Toxicity and response evaluation

Acute and late toxicities were evaluated by RTOG and CTCAE 4.03 scales, respectively [22, 23, 25]. The evaluation of biochemical response was carried out four-months after treatment following the Jereczek Fossa et al. proposal: a reduction of the PSA value > 10% compared to the pre-SBRT PSA levels was considered as a response (complete response if > 50%). A stable disease was defined as PSA levels ≤ 10 to ≥ 10% compared to the pre-SBRT value, while a PSA increase > 10% was classified as a biochemical progression [22, 26]. Moreover, in patients with PSA levels > 1 ng/ml, a 11C-choline PET-CT scan was performed and the SBRT/SRS response was classified also according to the PERCIST criteria [27].

Statistical analysis

Patients’ characteristics were reported as frequencies and percentages for categorical variables and as medians and ranges for continuous variables. The Kaplan–Meier method was used to calculate the actuarial outcomes. LC and response were defined on a “per lesion” basis. Moreover, LC was calculated from the date of SBRT/SRS to the date of in-field relapse/progression or to the date of the last follow-up. The NEST-free survival was defined as the time between SBRT/SRS and NEST (second-generation anti-androgens or chemotherapy). Distant progression free survival (DPFS) was calculated from the date of SBRT/SRS to the date of relapse/progression outside SBRT/SRS field or the last seen date. PFS was defined as the time between the date of SBRT/SRS and the date of first progression event (local or distant) or the last follow-up visit for censored patients. Overall survival (OS) was defined as the time from SBRT/SRS to the date of the last follow-up or death. NEST-free survival, DPFS, PFS, and OS were calculated on a “per patient” basis. Univariate and multivariate analysis of factors predicting outcomes was carried out by logistic regression. Differences between subgroups were evaluated by log-rank tests and Cox’s regression model for univariate and multivariate analyses, respectively. Statistical analysis was performed using XLSTAT statistical software (Addinsoft, Paris, France).

Results

Patients’ characteristics

Thirty-eight mCRPC patients with 61 nodal metastases were treated between May 2005 and June 2020 and included in this analysis. All patients had ECOG 0–1 performance status and median age was 73.5 years (range, 62–85). As shown in Table 1, 24 patients (63.2%) had at least one comorbidity, mainly represented by hypertension (47.8%), coronary artery disease (13.5%), diabetes (10.2%), lung disease (10.2%), and hepatopathy (2.6%). All patients were on ADT at the time of SBRT/SRS and were considered as “low burden” patient based on the definition used in the CHAARTED trial [28]. In terms of radiotherapy delivered before SBRT/SRS, 18 patients were previously treated with prophylactic pelvic nodal irradiation, of whom 12 with pelvic nodal relapse occurring after a median interval of 22 months (2–76 months) from the first irradiation. No lesions was previously treated with metastasis-directed therapies.

Table 1 Patients’ characteristics
Full size table

Lesions and treatment details are reported in Table 2. Nodal metastases sites were mainly pelvis (67.2%) and abdomen (24.6%), followed by the thorax (8.2%). Twenty-six patients had one single lesion (68.4%) and received only one SBRT/SRS treatment while concurrent or sequential treatments were performed in 12 patients bearing more than one lesion (total: 35, 9 metachronous and 26 synchronous). Moreover, five patients (13.2%) had been previously irradiated on bone metastases with curative aim. Among these, 2 patients were treated on spine metastases and one on a rib lesion. Another patient treated on 2 metachronous nodal metastases was previously irradiated on 2 bone metastases with SRS in between nodal irradiations. Finally, one patient was previously irradiated on 5 metachronous bone metastases.

Table 2 Characteristics of lesions (N = 61) and treatment’s detail
Full size table

Treatment details

The median PTV was 16 cm3 (range: 1.2–137.0). Thirty-two lesions (52.5%) were treated with SBRT and 29 (47.5%) lesions were treated with SRS. The biologically effective dose (BED)α/β1.5 was calculated as shown in Table 2. The median dose delivered with SRS was 20 Gy (range: 12–24 Gy), with 286 Gy (range: 108–408) median BEDα/β1.5. The most frequent SRS schedule was 20 Gy in single fraction (44.8%). The median SBRT dose was 45 Gy (range: 20–50 Gy) with 315 Gy median BEDα/β1.5 (range: 73–383 Gy). The most frequent SBRT fractionation regimen was 45 Gy in 5 fractions (37.9%) as reported in Table 2.

Response evaluation

In terms of biochemical response, the median PSA before SBRT/SRS was 5.3 ng/ml (0.03–382 ng/ml), while the median PSA at the first evaluation after treatment (4 months) was 2.4 ng/ml (0.001–346 ng/ml). In particular, we recorded 20 (33.0%) complete and 12 (19.0%) partial biochemical responses, 5 (8.0%) stable PSA levels and 19 (32.0%) biochemical progressions, while the PSA level was not available in five cases (8.0%). Moreover, the functional evaluation was performed in 35 lesions whose patient had a detectable PSA (> 1 ng/ml) after SBRT/SRS. We recorded 30 complete responses, 3 partial responses and 2 stable diseases.

Toxicity

Thirty-seven patients (97.4%) experienced none or mild (Grade 0–2) acute toxicity. In details, six grade 1 skin, five grade 1 genitourinary, eight grade 1 gastrointestinal toxicities and one case of grade 2 dysphagia were recorded. In terms of late toxicity, one G2 laryngeal toxicity was recorded 12 months after SRS (24 Gy) in one patient with a nodal lesion in the subclavicular region. He developed a laryngeal stenosis with recurrent laryngeal nerve paralysis due to which he underwent a temporary tracheostomy with subsequent decannulation after two months. Furthermore, one grade 1 skin and one grade 1 gastrointestinal late toxicities were recorded.

Outcome

Median follow-up was 27 months (range: 1–133 months). At last observation (December 2020), seven patients (18.5%) were dead of disease and 31 (81.5%) were alive. Of the latter, 15 patients (39.0%) had out-of-field recurrences while one patient (2.0%) was alive with both in-field and out-of-field relapses. Two-year actuarial LC, NEST-FS, DPFS, PFS, and OS were 94.0, 67.7, 47.2, 47.2, and 90.2%, respectively (Fig. 1, Table 3). Regarding the above mentioned outcomes, no significant differences were found based on age, body mass index, PTV, irradiation technique, BEDα/β1.5, and clinical response (data not shown).

Fig. 1
figure 1

Actuarial clinical outcomes: a local control (LC); b next line sistemic therapy free survival (NEST-FS); c distant progression free survival (DPFS); d progression free survival (PFS); e overall survival (OS)

Full size image
Table 3 Comparison between our series and other studies
Full size table

Discussion

mCRPC patients with nodal oligoprogression during ADT are a highly selected population in which effective local treatments could improve prognosis. Furthermore, sterilizing castration-resistant tumor clones by treating oligoprogressive lesions with SBRT/SRS could delay the systemic treatment shift [29]. The most important result of our study is the prolongation of ADT alone achieved with SBRT/SRS, with approximately two thirds of patients without changes or interruptions of this systemic therapy for at least 3 years. Furthermore, since LC is a necessary condition to achieve this goal, this study confirms the efficacy of SBRT/SRS in patients with mCRPC, the recorded LC rate being close to 90%.

Interestingly, the recorded NEST-free survival (3-year: 67.7%) was comparable to the ADT-free survival in castration-sensitive patients (median: 44 months, Table 3) [30], suggesting that SBRT/SRS could hinder tumor progression also in patients with nodal oligometastases from mCRPC.

In our analysis, based on patients treated with a wide range of BEDs, no significant impact of the delivered dose on LC rates was recorded. This lack of dose–effect may arise from the small sample size or from the relatively high BED (≥ 315 Gy) delivered in most patients. However, this data confirms the lack of clear evidence on the impact of radiation dose in the treatment of lymph node oligometastases [20]. Interestingly, the LC rates recorded in our series seem similar to those recorded in studies [5, 7, 8, 14, 15, 18, 21, 30] on castration-sensitive patients (Table 3). Furthermore, the LC rates registered in the latter analyses were uniformly high, regardless of the percentage of patients undergoing ADT (Table 3). These data suggest that the local effect of SBRT/SRT in PC nodal oligometastases could be independent of the possible impact of concomitant or adjuvant ADT.

Equally noteworthy is the similarity of our results, in terms of PFS, to series including oligorecurrent patients [5, 7, 8, 14, 15, 18, 21, 31, 32] undergoing concomitant or adjuvant ADT (Table 3). This data suggests, also in the setting of mCRPC, the efficacy of metastasis directed therapies in improving tumor control despite the theoretical higher aggressiveness of ADT-refractory patients. Furthermore, the 2-year PFS recorded in our study (47.2%) is similar to that reported by Onal et al. (36.7%) in mCRPC patients with nodal oligoprogression [21]. These figures seem higher than the results recorded in series of oligoprogressive PC bone metastases (median: 10–13.5 months) [31, 32].

Surprisingly enough, the recorded OS rates in nodal oligoprogression from mCRPC, in both the present study and that of Onal et al. [21] (2-year: 90.2–91%), are similar to the figures recorded in castration-sensitive patients (2-year: 67–92%) [7, 14]. This similarity, beyond the intrinsic limits due to the small number and small sample size of the analyzed series, seems to suggest that lymph node oligometastases might have a favourable prognosis regardless of the castration status when treated with SBRT/SRS.

The low recorded incidence and severity of side effects confirm the tolerability of SBRT/SRS in the treatment of lymph node oligometastases [20]. Actually, only one patient required hospitalization due to a temporary laryngeal stenosis resulting from recurrent laryngeal nerve paralysis recorded 12 months after SRS (24 Gy). Peripheral nerve damages were previously reported in patients with lung cancer treated with SBRT. For example, Shultz et al. reported two cases of neuropathy (either vagal or recurrent laryngeal nerves) in a series of 67 non-small cell lung cancers of the upper lobe treated with SBRT. The two subjects had received a moderately higher cumulative dose on the nerves compared to patients not developing vocal fold paresis [33]. Therefore, it should be emphasized that, even if rarely, cases of relevant toxicity are possible especially in the thoracic sites, as reported in a systematic literature review [20].

Our study has inherent limitations related to the sample size and trial design. The relatively small sample size (38 patients), although larger compared to the only study reporting results on SBRT in nodal oligoprogressive PC (15 patients) [21], hampered the identification of significant correlations between analyzed variables and outcomes. Furthermore, our study being a pooled analysis from two dose-escalation trials, total dose and dose/fraction were highly variable. Therefore, an evaluation of the efficacy of a single treatment regimen was not possible. Obviously, studies able to provide clear indications on effective dose/fractionation SBRT/SRS regimens in this setting would be very useful. Finally, the assessment of response was hindered by the exclusion from restaging of patients with PSA < 1 ng/mL after treatment, with a clear bias in the analysis of this outcome.

On the other hand, we analyzed a highly selected subset of metastatic PC patients, with a homogeneous population in terms of disease spread and hormonal status. Hence, a strong point of this study is that it can be considered a very effective model for describing the role of SBRT in mCRPC with few nodal lesions.

In conclusion, our analysis confirms the efficacy and safety of SBRT/SRT in the treatment of mCRPC patients. Despite the use of relatively low doses, the efficacy of this treatment in terms of LC in lymph node metastases is confirmed. Moreover, the analysis suggests the efficacy of metastasis directed therapies in prolonging PFS and NEST-free survival even in castration-resistant patients, with about one third of patients free from disease progression and two third of patients not requiring hormonal switch or discontinuation three years after the treatment. These results justify the design of prospective trials aimed at confirming these preliminary data.