Does Concomitant DCIS Affect the Clinical Outcome in Breast Cancer Patients with Invasive Ductal Carcinoma: an Asian Perspective


 BackgroundDuctal carcinoma in situ (DCIS) is an established precursor to invasive ductal carcinoma (IDC) and its coexistence with IDC appear to favour reduced biological aggressiveness. Its prognostic implication and ability to affect clinical outcome has been understudied in Asia. This study aims to explore if concomitant DCIS affects the clinical behavior and outcomes among Asians. MethodsStage I to III breast cancer patients with histologically proven invasive ductal carcinoma, diagnosed and treated in a single institution from 1 June 2004 to 30 June 2014 were included in this study. Statistical analyses were conducted using x2 test, independent T- test, multi-variate logistic regression and Kaplan- Meier test.ResultsA total of 818 patients were identified, including 224 and 594 patients with isolated IDC (No-DCIS) and IDC with coexisting DCIS (IDC-DCIS) respectively. Patients with IDC-DCIS were found to have smaller tumours (median: 22mm, p<=0.01), estrogen receptor positivity (p=0.001), progesterone receptor positivity (p<0.001) and associated with better pathological stage (p=0.001). Patients with No-DCIS were 1.6 times more likely to develop disease progression (95% CI: 1.1- 2.3, p=0.027) and subsequently associated with distant recurrences (20.5% vs 13.6%, p=0.02). The breast cancer specific 5 year overall survival rate for patients with No-DCIS and those with IDC-DCIS was 90.9% (95% CI: 86.2% - 94.5%) and 93.7% (95% CI: 91.4- 95.5%) respectively (p=0.202).ConclusionThe presence of DCIS component in IDC among Asians is associated with favourable tumour biological profile, thereby indicating reduced disease aggressiveness. Our study is the first to report the clinical significance in terms of disease progression and distant recurrences among Asians. Trial registrationThis study had received the approval of the institutional ethics committee prior to its commencement (IRB Ref No: 2019/2884).


Abstract
Background Ductal carcinoma in situ (DCIS) is an established precursor to invasive ductal carcinoma (IDC) and its coexistence with IDC appear to favour reduced biological aggressiveness. Its prognostic implication and ability to affect clinical outcome has been understudied in Asia. This study aims to explore if concomitant DCIS affects the clinical behavior and outcomes among Asians.

Methods
Stage I to III breast cancer patients with histologically proven invasive ductal carcinoma, diagnosed and treated in a single institution from 1 June 2004 to 30 June 2014 were included in this study. Statistical analyses were conducted using x 2 test, independent T-test, multi-variate logistic regression and Kaplan-Meier test.

Conclusion
The presence of DCIS component in IDC among Asians is associated with favourable tumour biological pro le, thereby indicating reduced disease aggressiveness. Our study is the rst to report the clinical signi cance in terms of disease progression and distant recurrences among Asians.

Trial registration
This study had received the approval of the institutional ethics committee prior to its commencement Background Screening mammography has led to a rising detection of early breast cancer, namely ductal carcinoma in situ (DCIS). Though the assumption that invasive cancers are likely to derive from pre-existing DCIS 1,2 , IDC may still evolve de novo in up to 21.4% of cases worldwide. A probable delay in transformation from insitu to invasive form is belived to account for tumours with coexisting DCIS demonstrating lesser biological aggresiveness 3,4 . Recent studies have linked tumours with IDC-DCIS with better clinical features such as smaller and lower grade tumours and lower probability of lymph node invasion.
Nonetheless, it remains controversial if clinical outcomes such as recurrences and overall survival rates are affected 5,6 . The clinical impact of concomitant DCIS in invasive cancers involving Asians has been understudied. Hence, our study aims to address if coexisting DCIS affects the tumour characteristics and clinical outcome such as recurrences, disease progression and overall survival among Asians.

Methods
A retrospective analysis was performed on our prospectively collected breast cancer database with an inclusion period from 1 June 2004 to 30 June 2014. This database comprises of patients who were diagnosed and underwent treatment in a specialized breast unit of a single institution. Only patients with de nitive histo-pathology diagnosis for invasive ductal carcinoma were selected for evaluation. Patients diagnosed with metastatic disease were excluded. Patients with bilateral breast cancer were included as two separate study cases. Figure 1 shows the inclusion and exclusion criteria that resulted in our main study cohort of 818 patients. Quantitative data are shown as median or mean of their values and their variability is expressed as range or standard deviation (SD), as speci ed for each analysis. Qualitative values are shown as absolute values or percentages. Categorical data was presented in frequency and percentage and association between subjects' characteristics and DCIS were tested using Chi-square test.
Numerical data was presented in mean (standard deviation) and association was tested using Independent T-test if normal distribution was ful lled. Otherwise, data was presented in median and interquartile rage (IQR) and Mann-Whitney U test was performed to test for the association between subjects' characteristics and DCIS status. Logistic regression was performed to identify the risk of developing disease progression between DCIS and non-DCIS patients.
Survival analysis and duration to disease free progression between DCIS and non-DCIS were performed using Kaplan-Meier with log rank test to assess for statistical differences.
Propensity scores were calculated using a multiple Logistic Regression with clinical relevant variables: age, overall tumor stage, menopausal status, hormonal status and cerB2 receptor status. Nearestneighbor matching with maximum difference of 5% approach in 1:1 fashion was performed to match the propensity scores values of the DCIS and non-DCIS group.
Statistical signi cance was set at P < 0.05. Analysis was performed using SPSS version 21 (IBM Corp., Armonk, USA). The mean follow up period in our study was 93.7 months (47.2-140.2). The ethnic distribution of our study cohort was consistent with our national demographics. The clinicopathological features and the analysis between the two groups are shown in Table 1  Though majority of tumours in both groups were found to be grade 3 (48.2%), this result was not statistically signi cant (p = 0.171). Similarly, there was no statistical signi cance upon comparison of the presence of lymphovascular invasion between both groups (p = 0.211).

Results
The median size of the invasive tumour of the No-DCIS group was 27mm while the IDC-DCIS group was 22mm (p < 0.001). Patients with IDC-DCIS were associated with lower T stage (p < 0.001) and better overall pathological stage (p = 0.001). Furthermore, patients with IDC-DCIS were likely to express positivity in hormonal receptors (estrogen receptor, p = 0.001; progesterone receptor, p < 0.001) and cerB2 receptor (p < 0.001).
Patients with No-DCIS were 1.6 times more likely to develop disease progression in contrast to IDC-DCIS (OR: 1.6, 95% CI: 1.1-2.6, p = 0.027). Patients with IDC-DCIS recorded a longer survival (14.2 years vs 13.9 years). Log rank test showed that this was not statistically signi cant (p = 0.274). Similarly, patients with IDC-DCIS had an average 12.7 years of disease free progression as compared to those with No-DCIS (12.5 years)(p = 0.600).
The breast cancer speci c 5 year overall survival rate was performed in 163 pairs matched patients with the aid of propensity score matching (PSM) (Fig. 2)and was 92.02% (95% CI 90.9% − 94.7%). Cox proportion hazard ratio showed that patients with IDC-DCIS more likely to progress faster to death (Table 3). Adjustment by age, overall tumour stage, menopausal status, hormonal receptor and cerB2 status showed no signi cant difference in the 5 year overall survival rate (p = 0.608) ( Table 3). This in turn led to the conclusion that this slow progression to the invasive component may then translate into a more favourable clinical prognosis 7,8 . Yet, only a handful of Western studies were able to recognize and associate this to possibly better prognostic features and clinical outcomes. Most yielded insigni cant and inconsistent results 9,10 . The clinical signi cance and role of DCIS among the Asian population remains understudied.
In Singapore, screening mammography is highly subsidized, making it extremely affordable for asymptomatic women. With the increasing awareness for screening mammography, detection of early cancers has also been on the rise. However, currently, the presence of DCIS in invasive cancers has no impact on systemic treatment, which depends solely on the pathological characteristics of the invasive component 11 .
There has yet to be a common consensus between the clinical features and prognostic implications for tumours with concomitant DCIS. With progressive published data among the Westerns, it is widely accepted that patients with IDC-DCIS may present at a younger age with smaller sized tumours with few or no lymph nodal involvement 9,10 . Our results indicated that IDC-DCIS is associated with smaller sized tumours and a lower overall clinical stage, thereby concurring with the hypothesis that tumours with concomitant DCIS were less biologically aggressive.
Likewise, some studies have showed that tumours with concomitant DCIS are likely to express ER, PR and cerB2 positivity as compared to tumours with No-DCIS [13][14][15] . Others disagree 12,15 . Our results showed that IDC-DCIS subjects were likely to express positivity in ER, PR and cerB2 receptors and this was statistically signi cant. While our data suggest that IDC-DCIS cancers may imply a less aggressive phenotype for patients with hormonal receptor or Her2 receptor positive cancers, we do acknowledge that triple negative malignancies may exhibit different biological behavior.
Researchers have since supported the preliminary theory of this slow evolution in deriving the invasive component and thereby resulting in a possibly better clinical prognosis. This study did demonstrate that the group with No-DCIS had higher percentage of distant recurrences compared to the group with IDC-DCIS (20.6% vs 13.5%, p = 0.020). However, we were unable to draw a similar conclusion for local recurrence. This might be attributed to the small total number of patients in our cohort who had developed local recurrences over the surveillance period as a result of better compliance to local radiation therapy prescribed. Our study results also suggest that patients with No-DCIS were 1.6 times more likely to develop disease progression as compared to those with IDC-DCIS, in spite of the adjuvant treatment given. These results add weight to the current speculation that the absence of coexisting DCIS is associated with poorer prognostic features and outcomes among Asian patients, especially in terms of distant recurrences and disease progression. Unfortunately, due to our small sample size of breast cancer related deaths, we were unable to detect a signi cant difference in the breast cancer speci c 5 year overall survival among the 2 groups.
The breast cancer speci c 5 year overall survival rate was analysed with PSM. No signi cant results were found between the group with No-DCIS and IDC-DCIS (92.02%, 95%CI = 90.9% − 94.7%, p = 0.608). PSM is regarded as an advanced statistical technique to minimize any possible confounders in an observational study. It serves to reduce possible treatment assignment bias and mimic randomization 16 . We utilised PSM to assess if concomitant DCIS affects the 5 year breast cancer overall survival rate after adjusting for certain covariates as mentioned above but we were unable demonstrate a more favourable breast cancer speci c 5 year overall survival rate. It might be attributed to the fact that both genomic pro les are highly similar 17,18 .
Lastly, we do recognize the limitations of this study. Being a retrospective analysis, any incomplete data namely, less detailed histology reports in the early years of the 21st century, had to be excluded. Other information such as patients' details may have been missing during the early days of data entry. Other inherent biases associated with retrospective study have to be considered. Secondly, analysis of tumour specimens had been performed in the absence of central pathologic review, hence establishment of details such as presence of isolated tumour cells (ITC) in ltration and micrometastases (mic) in lymph nodes were absent. Thirdly, our sample size was too small to demonstrate a signi cant difference in the 5 year breast cancer speci c survival. Unlike the Western studies, the clinical outcomes among Asians such as recurrences and disease free outcomes as a result from tumours with coexisting DCIS has been understudied 19,20 . This study, with a long median follow up of 94 months, is to add weight to current ndings and further strengthen the belief that coexisting DCIS does lead to better tumour pro le. We are the rst to document the clinical signi cance of concomitant DCIS in the presence of invasive cancers in terms of disease progression and distant recurrences.

Conclusion
With the limited data among Asian population, our study remains the rst to demonstrate improved clinical outcomes in terms of disease progression and distant recurrences. Henceforth, this allows clinicians to better prognosticate and consider vigilant clinical surveillance of patients diagnosed with isolated IDC in remission.

Declarations
Ethics approval and consent to participate This study had received the approval of the institutional ethics committee prior to its commencement (IRB Ref No: 2019/2884). A waiver of consent was permitted.

Consent for publication
Not applicable Availability of data and materials The datasets used during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This study did not receive any funding.
Authors' contributions LWP interpreted the data set and was a major contributor in writing the manuscript. SSS and SCM were heavily involved in data collection and data entry. TPT performed the analysis and interpreted the data set. TSM had edited the manuscript prior to submission. All authors read and approved the nal manuscript Figure 1 Study Cohort Flow Diagram.

Figure 2
Breast Cancer Speci c 5 year Overall Survival before (left) and after (right) PSM.