Prognostic value of intraductal carcinoma subtypes and postoperative radiotherapy for localized prostate cancer

Abstract

Background

Intraductal carcinoma of the prostate cancer (IDC-P), as a specific pathological type in prostate cancer which usually implies a poor prognosis. IDC-P morphology can be divided into two subtypes: Pattern 1, sieve like or loose cribriform structures; Pattern 2, solid or dense cribriform structures. The purpose of the study is to identify the impact of IDC-P and its subtypes on the prognosis of patients undergoing post-operative radiotherapy (PORT) after radical prostatectomy (RP) due to localized prostate cancer(PCa).

Methods

We performed a retrospective study of patients with localized PCa treated by RP followed by PORT or not. Patients with localized PCa who underwent RP from August 2013 to December 2020 were included in this study. Inclusion criteria: post-operative PSA dropped to less than 0.1 ng/ml after RP, had at least 1 poor prognostic risk factor (including high Gleason’s grouping; positive surgical margins; seminal vesicle invasion; extraprostatic extension; and lympho-vascular invasion), and were eligible for adjuvant radiotherapy.; In this study, patients who underwent salvage radiotherapy after RP due to biochemical recurrence (two consecutive PSA > 0.2 ng/ml) were also included, but not patients with persistent postoperative PSA > 0.1 ng/ml. Exclusion criteria: patients using other types of therapy prior to biochemical recurrence. Screening cases with pathological results of intraductal carcinoma, subtyping was completed by a pathologist, grouped by intraductal carcinoma (+/-; pattern 1/ 2) and treatment regimen (RP + PORT / RP only), Kaplan-Meier curves were plotted based on the time to biochemical recurrence-free and overall survival of the patients, and Cox regression analyses were performed. Finally, based on the results of Cox regression analysis, we initially predicted the probability of biochemical recurrence and death of the patients by plotting the nomogram.

Results

A total of 139 patients were included in this study with a median follow-up of 61.5 months. K-M curves showed that patients with “IDC-P (+) RP only” had the worst prognosis; patients with IDC-P could have a survival benefit after receiving PORT; whereas patients with non-intraductal carcinoma had a better prognosis than the above patients with or without PORT. In addition, patients with IDC-P(+) pattern 2 were more likely to experience biochemical recurrence and death. Multivariate Cox regression analysis showed that pattern 2 was a risk factor for biochemical recurrence and death. Other BCR-related risk factors in the research: Gleason grading group 5 (HR = 3.343, 95% CI: 1.616–6.916, P = 0.001), PM (HR = 2.124, 95% CI: 1.044–4.320,P = 0.038) and PORT (HR = 0.266, 95%CI: 0.109–0.647, P = 0.004). Other OS-related risk factors in the research: Grading group 5 (HR = 3.642, 95%CI:1.475–8.991, P = 0.005), SVI (HR = 2.522, 95% CI: 1.118–5.691, P = 0.026) and PORT (HR = 0.319, 95%CI: 0.107–0.949, P = 0.040).

Conclusion

Patients suffering from localized prostate cancer with IDC-P(+), especially IDC-P pattern 2, are more susceptible to biochemical recurrence and death after radical prostatectomy. While postoperative radiotherapy can alleviate the negative prognostic impact from IDC-P. It is implied that IDC-P can also be an indicator to be considered in PORT decision making to some extent.

Introduction

Prostate cancer (PCa) is the most common tumor of the male genitourinary system. The incidence of PCa in China is increasing annually, especially in large cities [1]. Remarkable progress were obtained in the clinical treatment of PCa in recent years. Local treatment like radical prostatectomy (RP) is available for clinically localized PCa, but about one third of patients who undergo RP still progress to localized advanced prostate cancer or metastatic prostate cancer within 10 years after initial diagnosis [2]. To minimize recurrence and prolong patients’ survival time as much as possible, clinicians will administer post-operative radiotherapy (PORT) scheme to some patients. Among them, adjuvant radiotherapy is recommended for patients depending on whether patients have poor prognostic factors, including positive surgical margins (PM), lympho-vascular invasion (LVI), extraprostatic extension (EPE), high Gleason grading group (GG 4–5) and seminal vesicle invasion (SVI) after RP [3–5]. The patients can be either treated with salvage radiotherapy, depending on whether they have persistent measurable PSA or biochemical recurrence(BCR) after surgery, provided that the patients have excluded distant metastases [6].

Intraductal carcinoma of the prostate (IDC-P) is a kind of pathological type with unique features and clinically significance. According to the 2014 International Society of Urological Pathology (ISUP) meeting, IDC-P is considered to be the retrograde spread of adenocarcinoma within the prostate duct, usually associated with high-grade adenocarcinoma [7–9]. Its detection rate in low-risk localized prostate cancer was 2.1%, compared to 23.1%, 36.7% and 56.0% for intermediate, high-risk and metastatic prostate cancers respectively [8]. A increasing number of studies have found that IDC-P is usually associated with shorter time to PSA progression, shorter overall survival time and poorer clinical outcomes [10–12]. Interestingly, several studies have found that the prognosis of patients with IDC-P still remains conflicting and it has been suggested that this may be related to the different subtypes of IDC-P [13]. According to the Epstein criteria and the 2016 WHO classification, IDC-P morphology can be divided into two types: Pattern 1—sieve like or loose cribriform structures; Pattern 2—solid or dense cribriform structures [14, 15]. It has been documented that IDCP pattern 2 appears to be more aggressive and malignant than IDC-P pattern 1 in locally advanced prostate cancer and metastatic hormone-sensitive prostate cancer (mHSPC) [13, 16]. Recently, Wang et al. confirmed it in a study of castration resistant prostate cancer (CRPC) patients, finding that patients with IDC-P pattern 2 treated with either abiraterone or docetaxel had a poorer outcome compared to pattern 1 [17].

The negative effect of IDC-P on prognosis in patients with clinically localized prostate cancer has been reported in some research but still requires a mass of data to support this point, and it is unclear whether there are differences in clinical outcomes between the two groups of patients with different subtypes of IDC-P. Besides, it remains uncertain whether the negative prognostic impact of IDC-P can be mitigated by PORT as well. Therefore, the main purpose of this study is to investigate the prognostic value of IDC-P and its subtypes on patients receiving RP. In addition, whether PORT provides a survival benefit for patients with IDC-P is another major theme of this study. The results of this study will provide a data basis for clinicians when dealing with the need for further radiotherapy from patients after RP, and to some extent guide clinicians in making adjustments on the frequency for future follow-up of patients.

Materials and methods

Study population

There are totally 341 patients who underwent laparoscopic radical prostatectomy for localized prostate cancer between January 2013 and September 2020 were included in this study. Our study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and was approved by the Institutional Review Board of Beijing Chaoyang Hospital, Capital Medical University (NO.: 2022-Ke-55), which waived the requirement of informed consent for this retrospective analysis.

The including criteria is as followed: (1) Patients with PORT: (1) Indications for adjuvant radiotherapy: postoperative PSA drop to less than 0.1 ng/ml after RP with at least 1 poor prognostic risk factor (including high Gleason’s grouping; positive surgical margins; seminal vesicle invasion; extraprostatic extension; and lympho-vascular invasion). (2) Indications for salvage radiotherapy: patients who received salvage radiotherapy after RP due to biochemical recurrence(two consecutive PSA > 0.2 ng/ml), it is worth mentioning that patients with PSA persistence after RP or PORT (persistent PSA > 0.1 ng/ml) were not included in this study. (2) Patients without PORT: patients with at least 1 poor prognostic factor but not receiving PORT served as the control group.

Exclusion criteria: patients treated with any other types of therapy before the appearance of biochemical recurrence, including patients treated with adjuvant endocrine therapy or chemotherapy following RP and PORT due to the PSA not dropping below 0.1 ng/ml. The reason is as follows: (1) If the PSA level of a patient was greater than 0.1 ng/ml after RP, the patient might have received postoperative adjuvant endocrine therapy alone or salvage radiotherapy, and adjuvant endocrine therapy would have had an impact on the patient’s prognosis. Consequently, it is unfortunate that we did not include patients who had a persistent PSA after RP in the study because we could not separate two treatment above at the time of the medical history search. (2) When a patient’s PSA level does not fall below 0.1 ng/ml after receiving PORT, this may be due to the presence of distant metastases or poor radiosensitivity, which may be followed with other forms of treatment which have an impact on prognosis as well.

The main information recorded were Gleason Grading Group (GG), positive margins (PM), lymphovascular vascular invasion (LVI), extraprostatic entension (EPE), seminal vesicle infiltration (SVI), IDC-P and its subtypes. Other information collected mainly included PORT modality (adjuvant radiotherapy/salvage radiotherapy), whether hormone therapy was combined with PORT and follow-up results about BCR and OS.

Histopathological review

Since IDC-P is usually a localized lesion that may be difficult to be detected, and previous pathology slides have been reviewed by different pathologists in almost 10 years, which may cause some bias in the diagnosis. Therefore, all of the postoperative pathological slides of 139 patients were re-evaluated by an experienced pathologist according to the American Joint Committee on Cancer’s Prostate Staging System(AJCC) [18, 19]. The diagnosis of IDC-P and its subtypes is based on Epstein’s criteria, in which the integrity of the basal layer of the prostatic ducts is the most important observation, and immunohistochemical staining may be required if necessary [15]. According to whether the tumor cells in the ducts account for 50% of the total ductal area, the pathological subtypes can be divided into loose cribriform(Pattern 1, < 50%) and dense cribriform (Pattern 2, ≥ 50%) [9]. Other evaluation items include PM, EPE, SVI, LVI, GG and so on. The definition of ISUP grading was according to the International Society of Urological Pathology 2014 grade groups: GG1 = GS ≤ 6, GG2 = GS3 + 4, GG3 = GS4 + 3, GG4 = GS3 + 5/4 + 4/5 + 3, GG5 = 5 + 5/5 + 4/4 + 5 [9] .

Radiotherapy protocol

Patients were given adjuvant radiotherapy or salvage radiotherapy: (1) Adjuvant radiotherapy was scheduled for patients with high risk factors within 1 year after RP to prevent biochemical recurrence when surgery-related side effects have resolved or stabilized. The indications includes positive surgical margins, lympho-vascular invasion, extraprostatic extension, high Gleason grading group and seminal vesicle invasion [4, 5]. (2) Salvage radiotherapy was given to the tumor bed ± pelvis for pitents with biochemical recurrence after RP, after exclusion of distant metastasis is [6].

The detailed protocol is IMRT(intensity modulated radiotherapy, IMRT) external radiotherapy technique, in which 1.8–2 Gy/time, 37–45 f, total radiation dose of 64–72 Gy is given to the tumor bed while ensuring the safety of normal tissues. Radiotherapy range should include tumour bed ± pelvic radiotherapy. The tumour bed should include the anastomosis, the bladder neck and the rectovesical gap. The specific range is as follows: from the vas deferens stump to 8–12 mm below the anastomosis of bladder urethral or at the level of the upper edge of bulb of penis, with the upper border generally within 3–4 cm above the pubic symphysis. If the histological results indicate seminal vesicle gland was invaded, the seminal vesicle gland area should be covered.

Follow-up protocol and endpoint definition

Patients after RP were followed up regularly according to the guideline recommendations, the details of which were to retest serum PSA at least 2 times for the first 3 months after surgery, and if it dropped below 0.1ng/ml, to test it every 3 months for 1 year and if there was no progression it changed to every 6 months after 1 year. Diagnostic Rectal Examination (DRE) may be performed once when the clinician suspects recurrence and then followed by MRI, and if PSA stays below 0.2ng/ml, DRE can be suspended [20].

Study endpoints included biochemical recurrence and death. patient’s PSA can generally drop to under 0.2ng/ml after RP, if the PSA returned to above 0.2ng/ml level with an increasing trend in two consecutive follow-ups, it was defined as PSA recurrence or BCR after RP. If the PSA failed to fall below 0.2ng/ml, it was called PSA persistence. Biochemical-free survival time (BFS) was defined as the time between the date of RP and the presence of BCR. Overall survival time(OST) was defined as the time duration between the date of RP and patients death. Based on whether the patients received PORT and the IDC-P status, the population included in the study was divided into four groups(as shown in the following pictures and tables), and Kaplan-Meier curves (K-M curves) were plotted according to the follow-up results to compare BFS and OST.

Data statistic

Clinical and pathological data were analyzed with SPSS v23.0 and Graphpad prism version-8.0.2 was required to plot K-M curves. Univariate methods included Pearson’s Chi-square test, Welch’s T-test and Mann-Whitney U test. Survival analyses were tested with the K-M curves and Cox regression analysis.

Variate candidates were analyzed using univariate and multivariate Cox regression, and then established the models for predicting the BCR-free rate and overall survival rate. The details are selecting the variables from those with P-values < 0.05 in multivariate Cox regression analysis. Subsequently, nomograms were constructed as graphic representations of the prediction models for BCR and OS using “rms” package. To test the efficacy of each model, receiver operating characteristic (ROC) curve analysis was performed using “pROC” package, the area under the curve (AUC) and 95% confidence interval (CI) were calculated as well. Sangerbox (http://sangerbox.com/) were utilized for constructing graph.

Result

Patients’ characteristic

From January 2013 to September 2020, there were 341 patients who underwent radical prostatectomy, of whom 139 localized prostate cancer patients with complete clinical and pathological data who had at least 1 poor prognostic factor were included in the study. Among them 46 patients who received RP + PORT and 93 patients who underwent RP only. The median follow-up time in this study was 61.5(Interquartile range, 27.1–87.4) months, with 54.6(IQR: 23.7–72.9) months in the RP + PORT group and 64.0(IQR: 28.0-93.6) months in the RP only group and there is no difference between two groups (P = 0.110) (Table 1).

Table 1.

Clinical characteristics of patients with at least 1 poor prognostic factor after RP

	Total (n = 139)	Radiotherapy (n = 46)	Without Radiotherapy (n = 93)	P
Mean Age(Std Dev)	67.0(6.6)	66.8(6.4)	67.1(6.7)	0.841
Gleason grading group(%)*				< 0.0001
≤ 3	70(50.4%)	10(21.7%)	60(64.5%)
4	19(13.6%)	8(19.6%)	11(11.8%)
5	50(36.0%)	28(60.9%)	22(23.7%)
Mean Pre-operative PSA(ng/ml)(Std Dev)*	37.1(81.6)	65.1(127.3)	22.4(32.4)	0.004
Positive margins(%)*	48(34.5%)	35(41.3%)	32(31.2%)	< 0.001
Lymphovascular invasion(%)	32(23.0%)	7(15.2%)	25(26.9%)	0.124
Extraprostatic extension(%)	38(27.3%)	14(30.4%)	24(25.8%)	0.565
Seminal vesicle invasion(%)	34(24.5%)	15(32.6%)	19(20.4%)	0.116
Concomitant ADT (combined with radiotherapy)	—	15(32.6%)	—	NA
IDC-P(%)*	47	18	29	0.004
Pattern 1	9	0(0%)	9(31.%)
Pattern 2	38	18(100%)	20(68.9%)
Follow-up months(IQR)	61.5(27.1–87.4)	54.6 (23.7–72.9)	67.0 (28.0–93.6)	0.110
Events
BCR(%)	38(27.3%)	10(21.7%)	28(32.2%)	0.298
Death(%)	27(20.1%)	8(17.4%)	19(20.4%)	0.670
Mean Pre-PORT PSA(ng/ml)	3.1(6.8)	3.1(6.8)	—	NA
PORT modality
Adjuvant (%)	23	23(50%)	—	NA
Salvage (%)	22	22(47.8%)	—	NA
Missing(%)	1	1(2.2%)	—	NA

Clinical characteristics of patients undergo radical prostatectomy with at least 1 poor prognosis factor; Std Dev Standard deviation, ADT Androgen deprivation therapy, IDC-P Intraductal carcinoma of the prostate, IQR Interquartile range, BCR Biochemical recurrence, PORT Post-operative radiotherapy;

There were 18 cases (39.1%) diagnosed with IDC-P in the RP + PORT group, all of which were IDC-P pattern 2(100%), with 2 cases of mixed type recorded as Pattern 2 IDC-P and no pure IDC-P pattern 1 detected. In total, 29 (31.2%) cases IDC-P were detected in the RP only group, of which 20 (68.9%) cases were Pattern 2, 9 (31.1%) were Pattern 1 and 6 cases with mixed type were recorded as pattern 2. In the PORT group, the detection rate of Pattern 2 IDC-P was higher (P = 0.004). In addition, the RP + PORT group had a higher Gleason grading group, pre-operative PSA level and a higher detection rate of positive surgical margins (all P < 0.01), with no significant differences found in the analysis of the remaining variables including LVI, EPE and SVI. Table 1 demonstrates the characteristics information between the two groups.

The effect of IDC-P on BCR and OS after RP

Among the patients enrolled in this study, patient with IDC-P(+) was more likely to develop biochemical recurrence and among them who received RP only suffered the worst prognosis (Fig. 1a). They progressed to BCR more rapidly than the other three groups with 41.4% rate of BCR-free survival at 5 years after RP and the median BCR-free time being only 41.0 months. Fortunately, the negative effect can be ameliorated by RP + PORT, with the rate of BCR at 5 years after RP incresed to 66.8% for “IDC-P(+) RP + PORT” and the median BCR-free time extend to 72.2 months, which was statistically longer than the “IDC-P(+) RP only” (logRank P = 0.0410). The prognosis of patients with IDC-P(-) was better than the IDC-P(+), even patients who did not receive PORT still had a BCR-free survival of 87.5% at 5 years postoperatively (median BCR-free time 95.2 months), while the “IDC-P (-) RP + PORT” could maintain to 79.3% as well (median BCR-free time not reached), much higher than the “IDC-P(+) RP only” (logRank P < 0.0001). These results suggest that IDC-P(+) has a strong negative impact on BCR-free survival.

Fig. 1.

The effect of IDC-P and PORT on the prognosis of patients after RP. Kaplan-Meier Curve for survival analysis of BCR and OS in patients with at least 1 poor prognosis factor after RP: (a) the median BCR-free time of “IDC-P(+) RP only” vs. “IDC-P(+) RP + PORT” : 41.0 vs. 72.2 months (logRank P = 0.0410); “IDC-P(-) RP only” vs. “IDC-P(-) RP + PORT” : 95.2 months vs. median not reached, which was significantly better than that of patients with IDC-P(+) RP only(logRank P < 0.0001); (b) the median overall survival time of “IDC-P(+) RP only” vs. “IDC-P(+) RP + PORT” : 70.1 vs. 90.0 months (logRank P = 0.1436) ; “IDC-P(-) RP only” vs. “IDC-P(+) RP only” : 70.1 months vs. median not reached, logRank P = 0.0021)

In the follow-up regarding OS we found it similar to the outcome of BCR (Fig. 1b). Patients with IDC-P(+) have higher mortality, which IDC-P(+) patients who receive RP only had the worst prognosis as low as 58.0% overall survival rate at 5 years postoperatively, while those who received RP + PORT had higher overall survival rate of 78.6%(median overall survival time 70.1 months vs. 90.0 months). However, there was no significant differences between the two groups (logRank P = 0.1436) (Fig. 1b). We considered that this result may be owing to that we didn’t distinguish the two subtypes of IDC-P. When we eliminated the Pattern 1 cases, the results showed that PORT improved the prognosis of patients with IDC-P Pattern 2 (median overall survival time 55.0 months vs. 90.0 months, logRank P = 0.0288). Besides, the prognosis of “IDC-P(-) RP only” was much better than “IDC-P(+) RP only” as the K-M curves showed(median overall survival time 70.1 months vs. not reached, logRank P = 0.0021).

Effect of different subtypes of IDC-P on BCR and OS

IDC-P can be divided into different subtypes based on whether the tumor cells in the duct occupy 50% of the duct cross-sectional area.(Fig. 2) Since there were no patients with “pure” IDC-P pattern 1 in the RP + PORT group, the population in the section finally was divided into 3 groups in the survival analysis of different subtypes of IDC-P including “pattern 1 RP only” “pattern 2 RP only” and “pattern 2 RP + PORT”.

Fig. 2.

Different subtypes of IDC-P according to the proportion of tumor cells in the duct. HE staining for different subtypes of IDC-P: (a) Pattern 1 (loose cribriform): tumour/ductal area < 50%; (b) Pattern 2 (dense cribriform): tumour/ductal area > 50%

Follow-up results showed that patients with IDC-P pattern 2 are more susceptible to BCR (Fig. 3a), with patients who received RP only having a biochemical recurrence-free rate of 15.6% compared to 63.1% treated with RP + PORT at 5 years after RP, suggesting that RP + PORT could significantly improved biochemical recurrence-free survival rate in patiens with Pattern 2 (median BCR-free survival time: Pattern 2 RP only vs. Pattern 2 RP + PORT, 33.0 months vs. 72.2 months, logRank P = 0.0037). The prognosis of Pattern 1 patients seemed much more encouraging since no biochemical recurrence in the RP only group at 5 years (median BCR-free survival time: Pattern 2 RP only vs. Pattern 1 RP only, 33.0 months vs. median not reached, logRank P = 0.0011).

Fig. 3.

The effect of different subtypes of IDC-P on BCR and OS in patients after RP. Imapct of different IDC-P subtypes on BCR and OS in patients after RP: (a) the median BCR-free time of “Pattern 2 RP only” vs. “Pattern 2 RP + PORT”: 33.0 vs. 72.2 months (logRank P = 0.0037); “Pattern 2 RP only” vs. “Pattern 1 RP only”: 33 months vs. median not reached (logRank P = 0.0011); (b) the median overall survival time of “Pattern 2 RP only” vs. “Pattern 2 RP + PORT” : 46.2. vs. 90.5 months (logRank P = 0.0265); “Pattern 2 RP only” vs. “Pattern 1 RP only” : 46.2 vs. 91.4 months (logRank P = 0.0070)

Similar results were obtained in OS as well, where the overall survival rate at 5 years for patients with IDC-P pattern 2 who recieved RP only dropped to 40.7%, while for those who received RP + PORT it increased to 76.3% (median overall survival time: Pattern 2 RP only vs. Pattern 2 RP + PORT, 46.2 months vs. 90.5 months, logRank P = 0.0265). IDC-P pattern 1 also had a better overall survival rate than pattern 2, with no deaths at 5 years postoperatively and only one death during the entire follow-up period in the study (median overall survival time: Pattern 2 RP only vs. Pattern 1 RP only,46.2 months vs. 91.4 months, logRank P = 0.0070).(Fig. 3b).

Cox regression analysis

Of the 139 patients included in the study, a total of 38 patients experienced BCR and 27 patients died during the follow-up. Cox regression analyses for BCR and OS are presented respectively (Table 2). Variables including IDC-P and its subtypes, age, preoperative PSA, GG, PM, LVI, EPE, SVI, and PORT were included in this analysis. The results of the multivariate analysis suggested that the risk factors affecting biochemical recurrence and death of patients varied. The results showed that Gleason grading group 5 (HR = 3.343, 95% CI: 1.616–6.916, P = 0.001), PM (HR = 2.124, 95% CI: 1.044–4.320,P = 0.038) and IDC-P pattern 2 (HR = 2.908, 95% CI:1.462–5.784, P = 0.002)were the risk factors of BCR, while PORT (HR = 0.266, 95%CI: 0.109–0.647, P = 0.004) being its protective factor.

Table 2.

Cox regression for prognosis of patients after RP

(a)
Risk Factors	Biochemical Recurrence
	Univariate Analysis			Multivariate Analysis
	HR(95% CI)	P		HR(95% CI)				P
Age(years)	0.989 (0.942–1.038)	0.646
PSA(ng/ml)	0.996 (0.985–1.007)	0.466
Grading group*		0.005						0.005
4 vs. ≤ 3	1.633 (0.596–4.472)	0.340		2.152 (0.726–6.385)				0.167
5 vs. ≤ 3	3.142 (1.568–6.296)	0.001		3.343 (1.616–6.916)				0.001
Positive Margins*	2.249 (1.162–4.355)	0.016		2.124 (1.044–4.320)				0.038
Lymphovascular Invasion	0.961 (0.465–1.985)	0.915
Extraprostatic extension	1.069 (0.537–2.127)	0.849
Seminal vesicle invasion	1.571 (0.828–2.981)	0.167
IDC-P*		0.007						0.007
Pattern 1 vs. IDC-P (-)	1.038 (0.241–4.463)	0.961		0.760 (0.171–3.369)				0.718
Pattern 2 vs. IDC-P (-)	2.842 (1.469–5.499)	0.002		2.908 (1.462–5.784)				0.002
PORT*	0.368 (0.154–0.882)	0.025		0.266(0.109–0.647)				0.004
(b)
Age(years)	1.029 (0.972–1.090)		0.328
PSA(ng/ml)	0.999 (0.993–1.004)		0.663
Gleason grading group*			0.004					0.015
4 vs. ≤ 3	2.708(0.883-8.300)		0.081		1.606 (0.497–5.187)			0.428
5 vs. ≤ 3	4.467 (1.860-10.727)		0.001		3.642 (1.475–8.991)			0.005
Positive Margins	1.823 (0.848–3.955)		0.123
Lymphovascular Invasion	1.681 (0.722–3.914)		0.229
Extraprostatic extension	0.802 (0.339-1.900)		0.616
Seminal vesicle invasion*	2.826 (1.291–6.188)		0.009		2.522(1.118–5.691)			0.026
IDC-P*			0.033					0.026
Pattern 1 vs. IDC-P (-)	1.269 (0.688–5.585)		0.753		0.758 (0.166–3.457)			0.720
Pattern 2 vs. IDC-P (-)	2.859 (1.292–6.327)		0.010		2.968(1.283–6.865)			0.011
PORT*	0.368 (0.127–1.067)		0.066		0.319(0.107–0.949)			0.040

IDC-P Intraductal carcinoma of the prostate, PORT Post-operative radiotherapy, CI Confidence interval.

Similarly, Grading group 5 (HR = 3.642, 95%CI:1.475–8.991, P = 0.005), IDC-P pattern 2 (HR = 2.968, 95%CI: 1.283–6.865, P = 0.011) were the risk factors of OS, while PORT (HR = 0.319, 95%CI: 0.107–0.949, P = 0.040) was a protective factor. Distinct from BCR, the variable SVI emerged as a risk factor for OS after RP as well(HR = 2.522, 95% CI: 1.118–5.691, P = 0.026). All of these results suggest an adverse impact of IDC-P, especially pattern 2, on the prognosis of localized prostate cancer. It is worth mentioning that IDC-P pattern 1 has significantly less adverse prognostic effects compared to pattern 2.

Nomogram development and validation

Nearly half of the patients in this study have not experienced the endpoint event of BCR or death. Combined with the median follow-up time of 61.5 (IQR 27.1–87.4) months in the study, it was proposed to construct nomograms using the above filtered variables GG, PM, SVI, IDC-P and PORT by Cox regression analysis to predict BCR-free survival rate and OS rate at 3 and 5 years after RP respectively. The AUC of the prediction mode for BCR-free rate in 3-years and 5-years was 0.78 (95%CI: 0.90 − 0.67 ) and 0.81 (95%CI: 0.91 − 0.70), 0.73 (95%CI: 0.98 − 0.48) and 0.77 (95%CI: 0.92 − 0.61) for OS rate in 3-years and 5-years, suggesting a valid prediction model for BCR-free survival or OS.(Fig. 4).

Fig. 4.

Nomogram constructed for patients after RP. Figure 4. Nomograms were constructed based on risk factors screened by Cox regression and ROC curves were plotted: (a) Nomogram for predicting the probability of BCR at 3 and 5 years after surgery (AUC at the 3 years-blue and 5 years-red: 0.78 and 0.81); (b) Nomogram for predicting the probability of overall survival rate at 3 and 5 years after surgery (AUC at the 3 years-blue and 5 years-red: 0.73 and 0.77)

Discussion

To date, the most popular and effective treatment for localized prostate cancer is still radical prostatectomy, but for some aggressive PCa, BCR is nearly inevitable after surgery. Postoperative combined radiotherapy is an effective therapy to prevent the progression of prostate cancer. In recent years, several studies have found that IDC-P detected either by prostate puncture biopsy specimens or RP specimens suggests a poor prognosis [21–24]. However, there is very few research on whether postoperative radiotherapy can improve the prognosis of IDC-P patients. In this study, we used K-M curves and Cox regression analysis to identify the factors affecting patients’ prognosis after radical prostatectomy and construct a prognostic model, which confirmed the negative effect of IDC-P, especially pattern 2, on the prognosis of prostate cancer patients, meanwhile revealed the protective effect of postoperative radiotherapy. The nomogram will allow us to predict the future biochemical recurrence-free survival and overall survival rate of patients in a more accurate and individualized way, which will provide important clues of the subsequent therapeutic regime for these patients.

The definition of IDC-P was initiated by McNeal et al. in 1996 and this unique pathological subtype was subsequently found to be significantly associated with risk factors that clearly predict poor survival outcomes such as high-grade invasive prostate cancer, high Gleason score, large tumour volume, extraprostatic invasion and lymph-vascular invasion [22]. In 2006, Guo and Epstein proposed that IDC-P is characterized by the presence of solid cribriform structures or loose cribriform/micro-papillary structures with marked nuclear heterogeneity within the ducts, provided that the basal cells are intact [14]. In terms of molecular mechanism studies, a recent study has found decreased PSMA expression and Elevated GLUT-1 expression among IDC-P patients, suggesting that the poor prognosis of IDC-P patients may be related to glucose metabolism [25]. With increased awareness and emphasis on IDC-P, its detection rate is significantly higher than before and a growing number of studies have confirmed that IDC-P (+) is an independent risk factor for BCR and shorter OS time after surgery [12, 21, 26–28]. Recently, a review of current studies related to IDC-P has been conducted around investigators, clinicians and pathologists with an MDT perspective, reflecting the uniqueness of IDC-P compared to PAC, which provides a foundation for subsequent precision treatment of PCa patients [29]. In addition to the potential molecular mechanisms mentioned above, whether IDC-P should be included in the Gleason scoring system is a hot topic as well. Postoperative radiotherapy, including adjuvant and salvage radiotherapy, has been proven to slow disease progression. And traditional external radiotherapy combined with brachytherapy may further improve the patient’s prognosis [30]. Vincent Q. Trinh et al. found that the combination of adjuvant radiotherapy in patients with IDC-P could extend the BCR-free time of patients [13], but there are still few studies on the benefit of PORT for patients with IDC-P. Our results demonstrate that BCR-free time and overall survival time of patients with IDC-P(+) is markedly less than patients with IDC-P(-), which is consistent with previous studies. Among this proportion of patients with IDC-P(+), those who received PORT had a significantly better prognosis, suggesting that PORT may diminish the negative impact of IDC-P(+) on prognosis. Combined with the results of the Cox regression analysis, IDC-P has the potential to become the next important indicator for radiotherapy in the future, just like other pathological parameters such as “positive margins”, “seminal vesicle invasion” and so on. The frequency of PSA follow-up might also be increased appropriately for patients with IDC-P (+),so that recurrence can be detected and treated as early as possible.

The prognosis of prostate cancer in patients with different histological subtypes of IDC-P has been found to be different, patients with solid/dense cribriform IDC-P (Pattern 2 in this study) have markedly rapid disease progression than micropapillary/loose cribriform IDC-P (Pattern 1 in this study), and are more likely to develop biochemical recurrence and castration-resistant [23]. Furthermore, Wang et al. found that in patients with metastatic CRPC, IDC-P pattern 2 was not responsive to either abiraterone or docetaxel, which severely shorten patients’ progression-free survival [17]. Our study found that IDC-P pattern 2 was a risk factor for biochemical recurrence and death in patients with localized prostate cancer, while the prognosis of patients with IDC-P pattern 1 did not appear to be different from those with IDC-P (-), which is similar with previous studies. It is certain that for patients with IDC-P pattern 2, the K-M curve showed that PORT does extend the biochemical-recurrence free time and prolong disease process as well as the overall survival time. Due to the small sample size in this study and the fact that none of the patients with IDC-P pattern 1 received post-operative radiotherapy, it is not clear whether these patients could benefit from radiotherapy in terms of BCR and OS.

Recently, there have been an increasing number of studies combining different risk factors to construct a Nomogram model for predicting the prognosis of prostate cancer. Compared to other conventional prediction methods, the Nomogram provides individualized prognostic prediction method by integrating multiple risk factors and plotting them at a certain scale, which allows a more direct demonstration of the impact that various risk factors may play on the prognosis of the disease. Zhao Jinge et al. selected IDC-P detected in prostate needle biopsy as the main object of their study, and construct a nomogram to predict the probability of biochemical recurrence after surgery for high-risk prostate cancer, they found that the inclusion of IDC-P significantly increased the C-index and improved the predictive efficacy of the model [16]. In another study, the author integrated the Gleason Score, IDC-P status, Baseline PSA level into nomogram and successfully predict the incidence of CRPC in 12, 24, 36 months and overall survival in patients with distant metastases [31]. Although IDC-P plays an important role in prostate cancer, there are no other studies that have considered IDC-P in prognostic models except for the studies mentioned above. We used Cox regression to identify prognostic factors associated with biochemical recurrence and death. Then, combined with the commonly used clinical pathological indicators, IDC-P and its subtypes were also included in the prediction model as well, which initially predicted the biochemical recurrence and survival outcomes of patients suffering from localized prostate cancer with at least one poor prognostic factor after radical surgery. In the upcoming studies, in addition to expand the sample size, we will also explore the role played by IDC-P and its subtypes in different stages of prostate cancer respectively and develop a more accurate predictive models to provide a data base for individualized patient therapy.

There are some limitations in our study. Firstly, this is a retrospective study with a limited sample size, especially for patients with IDC-P pattern (1) All nine patients were distributed in the group that did not receive PORT, which made it impossible to compare the survival rates of patients with IDC-P pattern 1 who received PORT. Besides, the relative proportions between pattern 1 and pattern 2 were different from previous literature, as almost all IDC-P(+) patients in this study being pattern (2) Secondly, patients with adjuvant radiotherapy and salvage radiotherapy were enrolled in the study and the therapeutic effects between two modalities can not be distinguished. In addition, the part concerning the OS survival analysis of this study was also biased by the fact that we could no longer check the adjuvant endocrine therapy regimen of some patients due to the large time span of the patients included in this study. Finally, some pathological and serological items which are important indicators for clinical decision making were not included in the Nomogram as they were not statistically significant in Cox regression analysis results in this study, which may have led to a bias in the predicted outcome compared to the reality.

Conclusions

After radical prostatectomy, patients diagnosed localized prostate cancer with IDC-P(+), especially IDC-P pattern 2(dense cribriform), are more susceptible to biochemical recurrence, their overall survival may be compromised compared to those with the IDC-P (-). Postoperative radiotherapy can significantly alleviate the negative effect by extending the biochemical recurrence free time and overall survival time.

Abbreviations

IDC-P

Intraductal carcinoma of the prostate cancer

RP

Prostatectomy

PCa

Prostate cancer

PORT

Post-operative radiotherapy

PM

Positive surgical margins

LVI

Lympho-vascular invasion

EPE

Extraprostatic extension

GG

Gleason grading group

SVI

Seminal vesicle invasion

BCR

Biochemical recurrence

ISUP

International Society of Urological Pathology

mHSPC

Metastatic hormone-sensitive prostate cancer

CRPC

Castration resistant prostate cancer

BFS

Biochemical-free survival time

OST

Overall survival time

CI

Confidence interval

Author contributions

Cao Fang wrote the main manuscript text and prepared Figs. 1, 2, 3 and 4 as well as Table 1, and 2; Li qing reviewed all the pathology slides and provided diagnostic reports. All authors reviewed the manuscript.

Funding

This study was supported by the National Natural Science Foundation of China (NSFC 82170783).

Data availability

The clinical and follow-up data used to support the findings of this study are restricted by the Institutional Review Board of Beijing Chaoyang Hospital, Capital Medical University, in order to protect the patient privacy. Data are available from Yinong Niu (E-mail: niuyinong@mail.ccmu.edu.cn) for researchers who meet the criteria for access to confidential data.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Beijing Chaoyang Hospital, Capital Medical University (NO: 2022-Ke-55), which waived the requirement of informed consent for this retrospective analysis.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.