Clinical significance of IDC‐P as predictive factor after intensity‐modulated radiation therapy

Abstract The clinical significance of intraductal carcinoma of the prostate (IDC‐P) in men with nonmetastatic prostate cancer (PCa) treated with high‐dose external‐beam radiation therapy remains unclear. The aim of this study was to evaluate the impact of IDC‐P in men who received intensity‐modulated radiation therapy (IMRT) for nonmetastatic PCa. All patients with high‐risk (H‐R) and very high–risk (VH‐R) PCa who received IMRT between September 2000 and December 2013 at our institution were analyzed retrospectively. We re‐reviewed biopsy cores for the presence of IDC‐P. Treatment consisted of IMRT (median: 78 Gy at 2 Gy per fraction) plus 6‐month neoadjuvant hormonal therapy (HT). In total, 154 consecutive patients with H‐R and VH‐R PCa were analyzed. Intraductal carcinoma of the prostate was present in 27.9% (n = 43). The median follow‐up period was 8.4 years. The 10‐year PCa‐specific survival, biochemical failure (BF), clinical failure, and castration‐resistant PCa rates were 90.0%, 47.8%, 27.5%, and 24.5% in patients with IDC‐P, and 96.6%, 32.6%, 10.8%, and 7.0% in those without IDC‐P, respectively (p = 0.12, 0.04, 0.0031, and 0.012, respectively). In multivariable analysis, IDC‐P was not identified as an independent predictive factor for BF (p = 0.26). The presence of IDC‐P was correlated with a significantly higher incidence of disease progression in men with H‐R and VH‐R PCa who received IMRT, although it was not identified as an independent predictive factor for BF. Further investigations are needed to determine the significance of IDC‐P as an independent predictive factor for survival outcomes.


| INTRODUC TI ON
Prostate cancer (PCa) shows heterogenous pathological features.
Intraductal carcinoma of the prostate (IDC-P) has been reported as an adverse pathological form of PCa, characterized as retrograde extension of PCa cells to pre-existing prostatic ducts. 1 Currently, the presence of IDC-P is listed as an adverse prognostic factor independent of the Gleason grading system in the current guideline of the European Association of Urology. 2 Reports on the prognostic impact of IDC-P among men with nonmetastatic PCa who received definitive external-beam radiation therapy (EBRT) are limited, because most previous studies on nonmetastatic PCa were conducted in the setting of radical prostatectomy (RP). [3][4][5] To our knowledge, no prospective study has been conducted, and only two retrospective studies on definitive EBRT have been reported, 6,7 mainly involving patients with intermediate-risk (IR) PCa or those with high-risk (H-R) PCa treated with a suboptimal dose (≤70 Gy with conventional fractionation). The clinical significance of IDC-P in men with H-R and very high-risk (VH-R) PCa who received the current standard high-dose EBRT remains unclear. Therefore, we aimed to assess the impact of IDC-P in men who received intensity-modulated radiation therapy (IMRT) for H-R and VH-R PCa.

| MATERIAL S AND ME THODS
This study followed the tenets of the Helsinki Declaration, with approval from the institutional ethical review board (approval number: R1898). Written informed consent was obtained from all patients.

| Patients
We retrospectively reviewed our institutional radiotherapy database, and searched for eligible patients. The eligibility criteria were as follows: (1) nonmetastatic PCa categorized into H-R or VH-R groups at the initial diagnosis according to National Comprehensive Cancer Network risk classification ver. 2. 2019, 8 specifically, PCa with at least one of the following risk features: ≥T3a, initial prostate-specific antigen (iPSA) > 20 ng/ml, or Gleason score sum ≥8 at the initial pathological evaluation; (2) treated with conventional fractionated IMRT between September 2000 and December 2013 at our institution; (3) underwent systematic prostate needle biopsy with ≥6 cores at our institution.
Patients whose biopsy had been performed at other institutions were excluded from this analysis because their histological specimens could not be re-reviewed. Patients with castration-resistant PCa (CRPC) at the initiation of IMRT were also excluded.

| Re-review of pathological specimens and pathological evaluation
Hematoxylin and eosin (HE)-stained slides of all available cores from prostate needle biopsy were systematically re-reviewed by a dedicated genitourinary pathologist (T.T.) who was blinded to the patients' information and clinical outcomes.

| Treatment
We previously reported the details of our institutional treatment protocol. 10,11 In brief, treatments consisted of short-term neoadjuvant hormonal therapy (NA-HT) and IMRT. For HT, all patients received short-term NA-HT, which basically comprised 6 months of combined androgen blockade. No adjuvant HT (A-HT) was applied.

| Patient follow-up and salvage treatment for recurrence
After the completion of IMRT, patients were followed every 1-4 months during the first 2 years and every 4-6 months thereafter. Instead of the application of long-term A-HT, patients who developed recurrence after IMRT were treated with salvage HT (S-HT) in the early phase after recurrence (basically PSA >4.0 ng/ml). No additional radiographic study after IMRT was required, unless an increase in the PSA level or symptoms suggesting clinical failure (CF) were observed. Before initiating salvage therapy, computed tomography and bone scintigraphy were conducted. No androgen receptor axis-targeted agents or chemotherapies were used in a castrationsensitive state.

| Statistical analyses
Chi-square analysis or the Mann-Whitney U test was used to compare the characteristics of patients and treatment with versus without IDC-P. Chi-square analysis was used to compare the initial pattern of CF between patients with versus without IDC-P.
Time zero was defined as the date of IMRT initiation. The Kaplan-Meier method was used to estimate overall survival (OS) and PCa-specific survival (PCSS), and the cumulative incidence method accounting for death without each event being a competing risk was used to calculate biochemical failure (BF), CF, and CRPC. Patients who were lost to follow-up with best supportive care due to disease progression were categorized as "died from PCa" at the time of the final visit. Biochemical failure was defined as the earliest timing of a PSA increase of >2.0 ng/ml above the nadir according to the Phoenix definition, 12 CF, or the initiation of HT due to disease progression.
Clinical failure was defined as recurrent disease confirmed via radiographic studies. The definition of CRPC was as follows: PSA increase >2.0 ng/ml above the nadir with testosterone <50 ng/dl or during HT (except off-period of intermittent HT), change in the contents of S-HT due to disease progression, or CF during salvage therapy.
The Cox proportional hazard model or Fine and Gray's regression model was used to estimate the hazard ratio (HR) in univariate analysis (UVA). Multivariable analysis (MVA) using Fine and Gray's regression model was performed to evaluate the impact of the presence of IDC-P on BF. Along with the presence of IDC-P, the following reported predictive factors were included as covariates: GG (continuous), iPSA (>20 vs. ≤20 ng/ml), clinical T stage (T3-4 vs. T1-2), rate of positive core (continuous), and age at the initiation of IMRT (continuous). 13 Due to the small number of events of PCa-specific mortality, CF, or CRPC, MVA was conducted only for BF. Prior to MVA for BF, we performed

| Patient and treatment characteristics
We identified 172 consecutive patients who met the eligibility criteria. Among them, five patients were downgraded to unfavorable IR or lower after the re-review of histological specimens, IMRT was discontinued in two (at 38 and 66 Gy due to the patient's request or a poor general condition), and bone metastasis at the initial diagnosis was retrospectively detected in one patient. In addition, 10 patients received A-HT after the completion of IMRT. These 18 patients were excluded, and the remaining 154 patients were included in the analysis. %. When stratified by the IDC-P status, patients with IDC-P had a significantly advanced stage or pathological features ( Table 1). Of note, all patients with IDC-P had disease with ≥GG 4 PCa (p < 0.001).
Patients' characteristics and comparisons are summarized in Table 1.  Table 1.

| Oncological outcomes and association with the presence of IDC-P
The median follow-up period was 8.4 (IQR: 6.3-11.5) years. There were nine deaths due to PCa with a median period of 10.0 (IQR:  Figure 1A,B. Overall survival and PCSS rates are summarized in Table 2. During follow-up, 33.8% (n = 52), 13.6% (n = 21), and 11.0%

| Oncological outcomes among ISUP GG 4-5 prostate cancer
As all patients with IDC-P had GG 4-5 disease, we additionally in-

| DISCUSS ION
In the current study, we retrospectively evaluated the impact of IDC-P in men who received IMRT for H-R and VH-R PCa. The presence of IDC-P was correlated with a significantly higher rate of disease progression (BF, CF, and CRPC), although it was not detected as an independent predictive factor for BF in MVA. To the best of our knowledge, this is the first study to evaluate the clinical significance of IDC-P among patients with H-R and VH-R PCa treated with IMRT.
It has been reported that IDC-P represents an advanced stage and pathological features of a tumor, and is correlated with a higher risk of disease progression and poor survival outcome. 3,[15][16][17][18][19][20][21] According to a population-based study of men undergoing RP, PCa patients with IDC-P showed a higher Gleason score, advanced T stage, lymph node metastases, and positive surgical margins. 15 Subsequently, IDC-P was correlated with a threefold increase in   Intraductal carcinoma of the prostate and/or cribriform morphology in RP specimens was correlated with adverse pathological findings, whereas those detected in biopsy specimens were not correlated with an adverse pathology. They concluded that this conflicting result was likely due to the low sensitivity of biopsy. In the current study, IDC-P was detected in 27.9% of all patients and 41.3% of GG4-5 patients, respectively. As our pathological evaluation was based solely on needle biopsy specimens, this may have influenced our results.
Despite the limitations discussed above, our findings are novel in that IDC-P was significantly correlated with worse CRPC progression, which is considered a more appropriate surrogate for survival outcomes than BF. 23 In the current study, a more than threefold increase of HR in CRPC progression (HR: 3.38, 95% CI: 1.31-8.74, p = 0.012) was observed in the patients with IDC-P in UVA. Our observation was consistent with results in previous reports for metastatic PCa. 21,24,25 According to a retrospective analysis of metastatic PCa with GS 8-10 docetaxel for hormone-sensitive metastatic PCa patients yielded better oncologic outcomes in several recent randomized controlled trials. 26 Although speculative, the use of those agents in combination with EBRT may be a promising therapeutic strategy against PCa with IDC-P. Our findings should be further investigated from the aspects of both diagnostic significance and treatment strategy for patients with IDC-P, especially in the setting of prospective trials with larger cohorts.
Our study had several limitations, including its retrospective nature and analysis of only a small cohort. Firstly, the pathological review was performed solely using HE-stained slides of needle biopsy specimens, and immunohistochemical staining to detect IDC-P was not applied. However, according to the current consensus, immunohistochemical staining is not essential for diagnosis of IDC-P. 27 Therefore, our results can be applied to current daily clinical practice.
Secondly, we performed MVA only for BF due to a small number of events. Lastly, as discussed above, our patients only received shortterm NA-HT (median: 6.8 months) because we designed the treatment protocol before the combination of long-term A-HT for unfavorable PCa was established as the standard of care. Our results may not be applicable to patients treated with the current standard long-term HT.
However, instead of the application of long-term A-HT, patients were salvaged early after BF (PSA >4.0 ng/mL). This early salvage method was shown to be a viable alternative to long-term A-HT among locally advanced PCa patients who received definitive EBRT in a phase III randomized controlled trial: nonmetastatic CRPC-free survival rate at 5 years, 84.8% in the long-term A-HT group versus 82.8% in the early S-HT group (p = 0.5619). 28 Therefore, our results are consequently considered appropriate for assessing CRPC rates and survival outcomes, although those regarding BF may be partly limited by the lack of long-term A-HT, as discussed above. For these reasons, our findings regarding the clinical significance of IDC-P are not conclusive but merely hypothesis generating. Nevertheless, we believe that our results provide baseline data on the clinical significance of IDC-P among H-R and VH-R PCa patients treated with IMRT, as well as a focal point for further research in this area. Given the paucity of evidence based on prospective trials assessing the significance of IDC-P, these findings are of particular importance.
In conclusion, this study showed that the presence of IDC-P was correlated with a significantly higher incidence of disease progression in men with H-R and VH-R PCa who received IMRT, although it was not detected as an independent predictive factor. Further investigations are warranted to confirm our findings.

ACK N OWLED G M ENT
This work was partly supported by JSPS KAKENHI Grant No.21 K06933 and Fujiwara memorial foundation. We thank Professor F I G U R E 4 Cumulative incidence curves of biochemical failure (A), clinical failure (B), and castration-resistant prostate cancer (C) in grade group 4-5 prostate cancer after intensity-modulated radiation therapy (IMRT) stratified by the presence of intraductal carcinoma of the prostate (IDC-P) Satoshi Morita for his advice regarding statistics. We also thank Dr.
Ryo Ashida for his assistance with data management.

D I SCLOS U R E
The authors have no conflict of interest. This document has been checked by a professional medical editor, who is a native speaker of English.