Development and assessment of a novel core biopsy-based prediction model for pathological complete response to neoadjuvant chemotherapy in breast cancer

doi:10.21203/rs.3.rs-2063499/v1

Download PDF

Research Article

Development and assessment of a novel core biopsy-based prediction model for pathological complete response to neoadjuvant chemotherapy in breast cancer

https://doi.org/10.21203/rs.3.rs-2063499/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose

Pathological complete response (pCR), the goal of NAC, is considered a surrogate for favorable outcomes in breast cancer (BC) patients administrated neoadjuvant chemotherapy (NAC). This study aimed to develop and assess a novel nomogram model for predicting the probability of pCR based on the core biopsy.

Methods

It was a retrospective study involving 769 BC patients administrated NAC between January 2012 and December 2017. After converting continuous variables to categorical variables, variables entering the model were sequentially identified by univariate analysis, multicollinearity test, and binary logistic regression analysis, and then a nomogram model was developed. The performance of the model was assessed concerning its discrimination, accuracy, and clinical utility. Internal validation was performed by the Bootstrap method.

Results

The optimal predictive threshold for estrogen receptor (ER), progesterone receptor (PR), Ki67, and p53 were 22.5% (95%CI: 0.603-0.718), 6.5% (95%CI: 0.578-0.693), 32.5% (95%CI: 0.610-0.731), and 37.5% (95%CI: 0.586-0.716), respectively (all P <0.001). Five variables were selected to develop the model, clinical T staging (cT), clinical nodal status (cN), ER status, Ki67 status, and p53 status (all P ≤0.001). The nomogram showed good discrimination with the area under the curve (AUC) of 0.804 (95% CI: 0.756-0.853; P <0.001) and good calibration. Decision curve analysis (DCA) showed that the model had practical clinical value.

Conclusion

This study constructed a novel nomogram model based on cT, cN, ER status, Ki67 status, and p53 status, which could be applied to personalize the prediction of pCR in BC patients treated with NAC.

breast cancer

neoadjuvant chemotherapy

pathological complete response

prediction model

nomogram

According to the latest global cancer data, breast cancer (BC) has become the most commonly diagnosed female malignant tumor and the fifth leading cause of cancer mortality [1]. Core biopsy, regarded as the gold standard for tissue acquisition for pathological assessment when considering malignancy, is widely used in BC patients. According to core biopsy, biomarkers including estrogen receptor (ER), progesterone receptor (PR), Ki67, p53 protein, and human epidermal growth factor receptor 2 (HER2) expression could be measured by immunohistochemistry (IHC) or fluorescence in situ hybridization (FISH). Molecular profiling is routinely applied to pre-treatment core biopsy to assist in pre-operative management decisions, which is a vital reference when deciding whether to proceed with neoadjuvant chemotherapy (NAC) treatment [2].

NAC, the effective pre-operation treatment, is widely applied in patients with locally advanced BC, or the molecular subtype of tumors is associated with a high likelihood of response [3]. The primary aim of NAC is to downstage tumor staging and convert inoperable tumors to resectable [4]. Furthermore, it may provide crucial prognostic information based on response to NAC [5].

Pathological complete response (pCR) has been considered a reliable predictive factor since achieving pCR is related to favorable survival outcomes irrespective of BC molecular subtypes in patients who received NAC [6]. However, due to the heterogeneity of BC, patients' response to NAC is differential, which makes it possible for patients with analogous clinical TNM staging to have diverse outcomes [7]. Therefore, discrimination against these heterogeneities and accurate prediction of response to NAC based on data available in the early pre-treatment period is meaningful; it may inform the therapeutic options, thereby impacting the treatment selection.

A vital unresolved clinical issue to consider when selecting NAC is the possibility of future pCR. Today, some predictors of response to NAC have been reported in the previous literature, including clinical, pathological, radiographic, and biological parameters [8–11]. Based on these parameters, quite a few prediction models have been built [12, 13]; however, the prognostic biomarkers involved in most models are not accessible to every patient for economic reasons or cumbersome steps. These reasons limited the assessment of which of the NAC-treated patients would have a good response. To fill this research gap, we developed and assessed a novel prediction model based on the core biopsy at the time of diagnosis. In our setting, the well-established predictive factors incorporate clinical T staging (cT), clinical nodal status (cN), ER status, Ki67 status, and p53 status. These factors, identified as prognostic factors, were discussed in previous studies [14–16]. To improve the accuracy of the prediction model, we found the predictive cut-off value of pre-treatment ER, Ki67, and p53 and combined them with other clinical predictive factors. To our knowledge, this is the first pCR prediction model in BC patients who received NAC treatment that involves the expression level of pre-treatment p53 protein with a specific value. This novel model may provide a convenient and economical method to predict the response to NAC.

Study subjects

It was a retrospective analysis of female patients with core biopsy confirmed invasive BC at our institution between January 2012 and December 2017. This study was approved by the ethics committee of First Affiliated Hospital of Chongqing Medical University (ID: No. 2020–59) and according to the Declaration of Helsinki; as a result of a retrospective observational study design, the hospital waived the request for informed consent. Female patients ≥ 18 years old who underwent NAC and subsequent surgery were incorporated into this study. Patients were excluded if (1) received < 4 cycles of NAC; (2) received non-anthracycline- and taxane-based chemotherapy; (3) HER2 status was unknown; (4) no specific value of ER, PR, Ki67, and p53 was available or unknown on core biopsy. Finally, a total of 769 eligible patients were included in this study (Fig. 1A).

Pathologic assessment

All specimens from core biopsies were fixed with formalin solution within 2 hours of isolation and then delivered to the department of pathology, Chongqing Medical University. All were processed in the pathology department by the same standards according to immunohistochemistry-related procedures. ER, PR, HER2, Ki67, and p53 were measured by IHC or FISH. The value of ER, PR, Ki67, and p53 was identified as the percentage of nuclear staining positive cells in the invasive front of the tumor (0%-100%). If the proportion of positive cells was ⩽10%, it was considered consistent; if it was > 10%, the result needed re-assessment (unblinded) to reach a consensus. The final percentage of positive cells was the average value assessed by two experienced pathologists. HER2-positive was defined as either a score of 3 + by IHC or an over 2.2-fold growth in FISH. The pathological response of tumors to NAC was assessed by the Miller-Payne scoring system [17]. pCR was regarded as ypT0/Tis ypN0, which means no residual malignant cells in any excision of breast tissues and lymph nodes after completion of NAC treatment [18].

Clinical assessment

The clinical nodal status is assessed by clinical diagnostic imaging (ultrasonography and magnetic resonance imaging).

Statistical method

The IBM SPSS 26.0 software and the RStudio 1.4.1 software were adopted to process data.

Convert continuous variables to categorical variables

Receiver operating characteristic (ROC) analysis was performed to measure the cut-off value of ER, PR, Ki67, and p53 indications to convert continuous variables to categorical variables.

Variable description

The Kolmogorov-Smirnov goodness-of-fit test was utilized to assess the normality of continuous variables. Those that conformed to an abnormal distribution used the Mann-Whitney U test, described as a median and interquartile range. Categorical variables were described as frequency and percentage and assessed by the chi-square test.

Variable selection

First, independent variables were included in the univariate analysis; those variables with p < 0.05 were involved in the multicollinearity test to examine whether there was multicollinearity amongst the variables. The variance inflation factor (VIF) > 5 was regarded as strong collinearity. The selected variables were added to the binary logistic regression analysis with a forward LR method to obtain the independent predictive factors for pCR in the NAC setting.

Nomogram development and assessment

The variables selected in the binary logistic regression model were utilized as the final predictive factors to construct a nomogram predicting the likelihood of pCR. We assessed the model in the following aspects: discrimination, accuracy, and clinical utility.

The area under the curve (AUC) was utilized to evaluate the discriminatory capability of this nomogram model; in addition, the sensitivity and specificity of the cut-off value of the model were computed.

The internal validation of the nomogram model was conducted using the Bootstrap method; after 1,000 repeated sampling from the original data, the calibration curve was plotted. Because of the Bootstrap method with 1,000 internal repeated sampling, it was not necessary to divide the data into training and validation sets. The Hosmer-Leme show test was utilized to evaluate the consistency of the model; P > 0.05 was regarded as that the model prediction is in good agreement.

Decision curve analysis (DCA), which estimated the net benefit of the prediction model according to the discrepancy between the number of true and false predicted positive outcomes [19], was utilized to judge the models' clinical utility and predictive value; additionally, it could differentiate between patients with a high and low probability of occurrence. Considering it was the first pCR prediction model in BC involving the expression level of pre-treatment p53 protein with a specific value, the DCA of the model with p53 and the model without p53 were performed.

The flow chart of the statistical process was shown in Fig. 1B.

Convert continuous variables to categorical variables

ROC curve analysis was utilized to calculate the optimal cut-off value for pre-treatment ER, PR, Ki67, and p53 indications to convert continuous variables to categorical variables. Our results showed that the optimal cut-off values for ER, PR, Ki67, and p53 were 22.5% (95%CI: 0.603–0.718), 6.5% (95%CI: 0.578–0.693), 32.5% (95%CI: 0.610–0.731), and 37.5% (95%CI: 0.586–0.716), respectively (all P < 0.001), as detailed in Fig. 2. Then, continuous variables were transformed into categorical variables based on their cut-off values.

Patient baseline characteristics and univariate analysis

In total, 769 eligible subjects were enrolled in the analysis. 82 patients (10.7%, 82/769) achieved pCR, while 687 patients (89.3%, 687/769) did not achieve. Univariate analysis of predictive factors between the pCR group and non-pCR group showed statistically significant distinctions (P < 0.05) in terms of the six factors of cT, cN, ER status, PR status, Ki67 status, and p53 status, while there were no statistically significant distinctions between the two groups between age at diagnosis, menopausal status, HER2 status, and chemotherapy cycles. Details are shown in Table 1. Then the six statistically significant predictors derived above were conducted multicollinearity test. Since the VIF of these predictors was all < 5, there was no covariance. Then the above predictive factors were assigned values for binary logistic regression analysis, and the results of the multicollinearity test and assignments were summarized in Table 2.

Table 1

Baseline clinicopathological characteristics and univariate analysis.
Predictive factors	Comparison of predictive factors between the two groups [M (P25, P75)]
Predictive factors	All subjects (n = 769) n (%)	pCR group (n = 82) n (%)	Non-pCR group (n = 687) n (%)	Z/x²	P
Age at diagnosis, y Menopausal status Pre-menopause Post-menopause cT cT1 cT2 cT3 + cT4 cN Negative Positive ER status (%) <22.5 ≥ 22.5 PR status (%) <6.5 ≥ 6.5 HER2 status Negative Positive Ki67 status (%) < 32.5 ≥ 32.5 p53 status (%) < 37.5 ≥ 37.5 Chemotherapy cycles 4 5–8	48.0 (43.0–56.0) 468 (60.9%) 301 (39.1%) 48 (6.2%) 537 (69.8%) 184 (23.9%) 303 (39.4%) 466 (60.6%) 361 (46.9%) 408 (53.1%) 432 (56.2%) 337 (43.8%) 437 (56.8%) 332 (43.2%) 552 (71.8%) 217 (28.2%) 426 (55.4%) 343 (44.6%) 705 (91.7%) 64 (8.3%)	47.0 (41.8–54.0) 51 (62.2%) 31 (37.8%) 13 (15.9%) 56 (68.3%) 13 (15.9%) 54 (65.9%) 28 (34.1%) 62 (75.6%) 20 (24.4%) 64 (78.0%) 18 (22.0%) 44 (53.7%) 38 (46.3%) 40 (48.8%) 42 (51.2%) 26 (31.7%) 56 (68.3%) 74 (90.2%) 8 (9.8%)	48.0 (43.0–56.0) 417 (60.7%) 270 (39.3%) 35 (5.1%) 481 (70.0%) 171 (24.9%) 249 (36.2%) 438 (63.8%) 299 (43.5%) 388 (56.5%) 368 (53.6%) 319 (46.4%) 393 (57.2%) 294 (42.8%) 512 (74.5%) 175 (25.5%) 400 (58.2%) 287 (41.8%) 631 (91.8%) 56 (8.2%)	-1.225 0.069 16.117 26.898 30.283 17.836 0.376 23.974 20.847 0.247	0.221 0.812 < 0.001 < 0.001 < 0.001 < 0.001 0.557 < 0.001 < 0.001 0.619
pCR pathological complete response, cT clinical T staging, cN clinical nodal status, ER estrogen receptor, PR progesterone receptor, HER2 human epidermal growth factor receptor 2.

Table 2

Multicollinearity test of predictive factors and the approach to assigning values.
Factors	Tolerance	VIF	Assignment
cT cN ER status PR status Ki67 status p53 status	0.986 0.964 0.489 0.499 0.938 0.948	1.014 1.037 2.046 2.004 1.067 1.054	“cT1” = 1, “cT2” = 2, “cT3” = 3 “Negative” = 0, “Positive” = 1 “<22.5%” = 0, “≥22.5%” = 1 “<6.5%” = 0, “≥6.5%” = 1 “<32.5%” = 0, “≥32.5%” = 1 “<37.5%” = 0, “≥37.5%” = 1
VIF variance inflation factor, cT clinical T staging, cN clinical nodal status, ER estrogen receptor, PR progesterone receptor.

Binary logistic regression analysis

The binary logistic regression analysis was conducted with the six statistically significant predictive factors derived above (cT, cN, ER status, PR status, Ki67 status, and p53 status) as the independent variables and whether achieving the pCR (“no” = 0, “yes” = 1) as the dependent variable. Finally, cT, cN, ER status, Ki67 status, and p53 status (all P ≤ 0.001) were independent predictors of pCR; these results were detailed in Table 3.

Table 3

Results of binary logistic regression analysis.
Predictive factors	B	SE	Wals	P	OR (95% CI)
cT cT1 cT2 cT3 + cT4 cN: neg vs pos ER status: <22.5% vs ≥ 22.5% Ki67 status: <32.5% vs ≥ 32.5% p53 status: <37.5% vs ≥ 37.5% Constants	-1.550 -1.963 1.045 1.237 -0.851 -0.917 -0.994	0.403 0.483 0.262 0.288 0.258 0.267 0.466	14.802 16.526 15.900 18.498 10.861 11.847 4.541	< 0.001 < 0.001 < 0.001 < 0.001 0.001 0.001 0.033	Ref. 0.212 (0.096–0.468) 0.140 (0.054–0.362) 2.843 (1.701–4.751) 3.446 (1.961–6.056) 0.427 (0.257–0.708) 0.400 (0.237–0.674) 0.370
SE standard error, OR odds ratio, CI confidence interval, cT clinical T staging, cN clinical nodal status, ER estrogen receptor.

Develop and assess the nomogram

We constructed the nomogram shown in Fig. 3A based on the predictive factors derived from the above binary logistic regression analysis. The scores corresponding to five indicators (cT, cN, ER status, Ki67 status, and p53 status) of each patient in the nomogram were summed to calculate the final score, and the probability of achieving pCR for that patient could be derived.

As shown in Fig. 3B, the ROC curve was utilized to evaluate the discrimination of the nomogram model. The ROC analysis showed AUC = 0.804 (95% CI: 0.756–0.853; P < 0.001), the cut-off value was 0.110, and sensitivity and specificity of the value were 73.2% and 74.7%, respectively.

The internal validation of the nomogram model was conducted by the Bootstrap method, and the calibration curve was plotted after 1,000 repeated samplings from the original data (Fig. 3C). Since the Hosmer-Lemeshow goodness-of-fit test showed χ2 = 7.089, P = 0.527, suggesting that the model has good prediction accuracy.

The DCA of the model with p53 and the model without p53 were carried out, considering it was the first pCR prediction model in BC involving the expression level of pre-treatment p53 protein with a specific value; details were shown in Fig. 4. The DCA, with the threshold probability as the horizontal coordinate and the net benefit rate as the vertical coordinate, was utilized to judge the models' clinical utility and predictive value. The greater the net clinical benefit obtained using this prediction model when nearer the DCA curve is to the upper right corner. The graph shows that the net benefit of the nomogram model with p53 is higher than that without p53.

Figure 5 exhibited the clinical impact curves of the nomogram model. The blue curve (Number high risk with event) was the number of true positives at each threshold probability, while the red curve (Number high risk) represented the number of subjects classified as positive by the prediction model at each threshold probability.

Since achieving pCR is related to better survival outcomes (event-free survival [HR = 0.40, P < 0.001] and overall survival [HR = 0.32, P < 0.001]) in the NAC setting, it has been regarded as a dependable predictor [6]. Hence, the prediction of pCR in the early pre-treatment period is of great significance. However, the predictive biomarkers in other literature are not accessible to every patient for economic reasons or cumbersome steps. To fill this research gap and achieve the maximum utilization of resources, this study selected simple and easy-to-access core biopsy and clinical information of patients as predictive factors and built a prediction model for pCR in the NAC setting. It helps to predict chemotherapy response at the time of diagnosis and makes it possible for clinicians to intervene early in some high-risk patients. The selection of variables and the conditions of the model development are described in the ensuing paragraphs.

Clinical tumor staging

Clinical tumor staging plays a crucial role in chemotherapy response. Livingston-Rosanoff et al. reported a retrospective study that included 38,864 patients who underwent NAC treatment and subsequent surgery for a solitary lesion varying from cT1 to cT3, which revealed that cT3 tumors have a lower probability of achieving pCR irrespective of molecular subtypes [20], which is consistent with our study. The possible explanation for this finding is that larger tumors have a higher chance of revealing heterogeneity of elevation, which may affect the sensitivity of chemotherapy [21]. Many prediction models involved clinical tumor staging (tumor size) as a predictive factor [22–24], which indicates it is a reliable factor for predicting pCR.

Clinical nodal status

In the last decade, the administration of systemic treatment in patients with node-positive disease has switched from the adjuvant to the neoadjuvant setting. According to previous studies, 20–42% of the firstly node-positive patients finally achieve pCR of the axillary lymph nodes [25]. Our results showed that pre-treatment clinical nodal status was associated with chemotherapy response; in other words, there was a greater possibility of pCR in patients with clinically node-negative disease, which is consistent with the previous study [26]. It reported the low probability of pathologic nodal positivity in patients with clinical node-negative and breast pCR disease [26], highlighting the crucial role of clinically node-negative in achieving pCR. Meanwhile, it indicated that clinical assessment by clinical diagnostic imaging played a vital role in pCR prediction. An accurate clinical assessment technique for recognizing patients with node-negative disease would be beneficial for BC care. Some models incorporated pre-treatment lymph node status for predicting axillary pCR for node-positive BC [27, 28], indicating the vital role of nodal status in pCR prediction.

ER status

ER, a vital factor that defines tumor subtype, has extensively been identified as a feature that affects the response to NAC [29]. Previous studies reported that ER-negative subtypes such as HER2-enriched and triple-negative BC were more likely to achieve pCR and favorable long-term outcomes [29, 30], which is consistent with our results. By analyzing pre-treatment ER as a continuous variable, we could divide patients into ER-positive and ER-negative or ER-high and ER-low diseases. Further, we found the cut-off value of pre-treatment status rather than simply dividing it into ER-positive and ER-negative, which could be explained by the fact that ER-low disease and ER-negative disease have similar biological behavior. Weisman et al. [31] found that ER-low malignancies had a semblable pathologic response to NAC treatment as ER-negative diseases, demonstrating the above point. It will separate the ER-positive patients into different subgroups with different probabilities of achieving pCR. Nevertheless, the cut-off value remains controversial. Few previous studies assessed ER status quantitatively, one of which reported a cut-off value of 30% when distinctions in responses could be seen among patients with ER < 30% and those with ER > 30% diseases [32], which is consistent with our results showing that the cut-off value of ER was 22.5%. Also, previous literature reported that the threshold of 80% best predicted the relation to pCR [33].

Ki67 status

Ki-67 is a biomarker of cell proliferation used to evaluate the invasiveness of the tumor; except for the G0 phase, the expression of Ki67 exists in all the cell cycle phases [34]. Ki-67 has been assessed in several studies for its predictive role in the NAC setting, but its cut-off value of it remains controversial. However, a large-scale meta-analysis incorporated 44 studies reported that pre-treatment high-Ki-67 was related to elevated pCR rates in BC patients who received NAC using distinct cut-off values of Ki-67 [35]. Our study found a threshold of Ki-67 was 32.5%, which is in the range from 15–50%, as previous literature reported [35].

p53 status

p53 protein, coded by the TP53 gene, the most frequently mutated gene in BC, plays a crucial role in metabolism, apoptosis, DNA repair, and cellular sensitivity to chemotherapy [36]. Numerous BC patients who will accept NAC treatment have cancers harboring TP53 mutations. Many studies have tried to identify the role of that mutation in pathological response, which showed that compared with wild-type counterparts, tumors with TP53 mutations have a statistically higher probability of pCR in BC [14, 37]. Replaced by simply dividing p53 status into positive and negative, the cut-off of 37.5% best predicted the correlation with pCR in our cohort, and patients with p53 ≥ 37.5% were more likely to have a pCR. Limited studies reported the threshold value of p53 in BC; Lee et al. [38] found that a threshold of 10% for p53 was a predictive factor of survival outcome. However, there were no studies on the cut-off value of p53 in the NAC setting. To our knowledge, this is the first study to explore the role of pre-treatment p53 in the NAC prediction model; moreover, the cut-off predictive value of p53 was found.

Prediction model

In summary, in our study, the model for predicting pCR had five variables, cT, cN, ER status, Ki67 status, and p53 status, respectively. Notably, we visualize the prediction model and present it as a nomogram. The ROC analysis and Hosmer-Lemeshow goodness-of-fit test indicated that the model had good discrimination and calibration; AUC = 0.804 (95% CI: 0.756–0.853; P < 0.001), sensitivity, and specificity of 73.2% and 74.7%, respectively, Hosmer-Lemeshow goodness-of-fit test χ2 = 7.089, P = 0.527 > 0.05. In clinical application, the probability of pCR can be judged by the total score obtained by adding the scores of each risk factor. For all we know, this is the first pCR prediction model that includes the expression level of pre-treatment p53 status. The optimal model is the model with the maximum net benefit for any given probability threshold. The DCA shows that outcome prediction using the model with p53 has a greater net benefit than the model without p53.

Limitation

The major limitation of this study is that the prediction model was developed based on the Chinese population and may not apply to other ethnic groups. Additionally, internal validation was conducted in the single data set because of the limited sample size, and no external validation group was established to validate this model. It is hoped that external validation will be conducted in further studies by conducting multicenter studies.

The model developed in this study incorporated five variables, cT, cN, ER status, Ki67 status, and p53 status, which had good predictive power for pCR. The model is simple, and the selected predictors are all easy-to-access variables from core biopsy and clinical information, which makes it easy to utilize and can be widely applied to BC patients who received NAC for identifying groups that need early intervention.

Author contributions

Conception and design: All authors; Acquisition of the data: AL, JC, CL, YJ, YW, YD, LJ, HL, and YP; Data analysis and interpretation: AL, JC, CL, and YJ; Drafting or revising of the manuscript: AL, JC, CL, YJ, and SL; Review and editing: AL, JC, CL, and SL; The guarantors of the integrity of the entire study: SL and AL; Final approval of manuscript: All authors.

Founding

This study is supported by the Key Research and Development Project of Chongqing’s Technology Innovation and Application Development Special Big Health Field (Grant no. CSTC2021jscx-gksb-N0027).

Data availability

The original datasets used and/or analyzed during the present study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest The authors declare that the research was conducted with no commercial or financial relationships that could be construed as a potential conflict of interest.

Ethics approval Approval was obtained from the ethics committee of First Affiliated Hospital of Chongqing Medical University (ID: No. 2020–59). Due to having a retrospective observational design, the hospital waived the requirement for informed consent.

Consent to participate As this study had a retrospective observational design and the data are anonymous, the requirement for informed consent was waived by the hospital.

Consent for publication Not applicable.

Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209–49.
Slostad JA, Yun NK, Schad AE, Warrior S, Fogg LF, Rao R. Concordance of breast cancer biomarker testing in core needle biopsy and surgical specimens: A single institution experience. Cancer Med. 2022.
Burstein HJ, Curigliano G, Thürlimann B, Weber WP, Poortmans P, Regan MM, et al. Customizing local and systemic therapies for women with early breast cancer: the St. Gallen International Consensus Guidelines for treatment of early breast cancer 2021. Ann Oncol. 2021;32(10):1216–35.
Gralow JR, Burstein HJ, Wood W, Hortobagyi GN, Gianni L, von Minckwitz G, et al. Preoperative therapy in invasive breast cancer: pathologic assessment and systemic therapy issues in operable disease. J Clin Oncol. 2008;26(5):814–9.
Cortazar P, Zhang L, Untch M, Mehta K, Costantino JP, Wolmark N, et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. Lancet. 2014;384(9938):164–72.
Bonnefoi H, Litière S, Piccart M, MacGrogan G, Fumoleau P, Brain E, et al. Pathological complete response after neoadjuvant chemotherapy is an independent predictive factor irrespective of simplified breast cancer intrinsic subtypes: a landmark and two-step approach analyses from the EORTC 10994/BIG 1 – 00 phase III trial. Ann Oncol. 2014;25(6):1128–36.
Wang-Lopez Q, Chalabi N, Abrial C, Radosevic-Robin N, Durando X, Mouret-Reynier MA, et al. Can pathologic complete response (pCR) be used as a surrogate marker of survival after neoadjuvant therapy for breast cancer? Crit Rev Oncol Hematol. 2015;95(1):88–104.
Piñeros M, Parkin DM, Ward K, Chokunonga E, Ervik M, Farrugia H, et al. Essential TNM: a registry tool to reduce gaps in cancer staging information. Lancet Oncol. 2019;20(2):e103-e11.
Kwa M, Makris A, Esteva FJ. Clinical utility of gene-expression signatures in early stage breast cancer. Nat Rev Clin Oncol. 2017;14(10):595–610.
Stanton SE, Adams S, Disis ML. Variation in the Incidence and Magnitude of Tumor-Infiltrating Lymphocytes in Breast Cancer Subtypes: A Systematic Review. JAMA Oncol. 2016;2(10):1354–60.
Conti A, Duggento A, Indovina I, Guerrisi M, Toschi N. Radiomics in breast cancer classification and prediction. Semin Cancer Biol. 2021;72:238–50.
Phung MT, Tin Tin S, Elwood JM. Prognostic models for breast cancer: a systematic review. BMC Cancer. 2019;19(1):230.
Wu C, Jarrett AM, Zhou Z, Elshafeey N, Adrada BE, Candelaria RP, et al. MRI-based digital models forecast patient-specific treatment responses to neoadjuvant chemotherapy in triple-negative breast cancer. Cancer Res. 2022.
Shahbandi A, Nguyen HD, Jackson JG. TP53 Mutations and Outcomes in Breast Cancer: Reading beyond the Headlines. Trends Cancer. 2020;6(2):98–110.
Nicolini A, Ferrari P, Duffy MJ. Prognostic and predictive biomarkers in breast cancer: Past, present and future. Semin Cancer Biol. 2018;52(Pt 1):56–73.
Fitzgibbons PL, Page DL, Weaver D, Thor AD, Allred DC, Clark GM, et al. Prognostic factors in breast cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med. 2000;124(7):966–78.
Ogston KN, Miller ID, Payne S, Hutcheon AW, Sarkar TK, Smith I, et al. A new histological grading system to assess response of breast cancers to primary chemotherapy: prognostic significance and survival. Breast. 2003;12(5):320–7.
Mazouni C, Peintinger F, Wan-Kau S, Andre F, Gonzalez-Angulo AM, Symmans WF, et al. Residual ductal carcinoma in situ in patients with complete eradication of invasive breast cancer after neoadjuvant chemotherapy does not adversely affect patient outcome. J Clin Oncol. 2007;25(19):2650–5.
Vickers AJ, Holland F. Decision curve analysis to evaluate the clinical benefit of prediction models. Spine J. 2021;21(10):1643–8.
Livingston-Rosanoff D, Schumacher J, Vande Walle K, Stankowski-Drengler T, Greenberg CC, Neuman H, et al. Does Tumor Size Predict Response to Neoadjuvant Chemotherapy in the Modern Era of Biologically Driven Treatment? A Nationwide Study of US Breast Cancer Patients. Clin Breast Cancer. 2019;19(6):e741-e7.
Derouane F, van Marcke C, Berlière M, Gerday A, Fellah L, Leconte I, et al. Predictive Biomarkers of Response to Neoadjuvant Chemotherapy in Breast Cancer: Current and Future Perspectives for Precision Medicine. Cancers (Basel). 2022;14(16).
Du S, Gao S, Zhao R, Liu H, Wang Y, Qi X, et al. Contrast-free MRI quantitative parameters for early prediction of pathological response to neoadjuvant chemotherapy in breast cancer. Eur Radiol. 2022;32(8):5759–72.
Ye P, Duan H, Zhao Z, Fang S. A Practical Predictive Model Based on Ultrasound Imaging and Clinical Indices for Estimation of Response to Neoadjuvant Chemotherapy in Patients with Breast Cancer. Cancer Manag Res. 2021;13:7783–93.
Keam B, Im SA, Park S, Nam BH, Han SW, Oh DY, et al. Nomogram predicting clinical outcomes in breast cancer patients treated with neoadjuvant chemotherapy. J Cancer Res Clin Oncol. 2011;137(9):1301–8.
Schipper RJ, Moossdorff M, Beets-Tan RGH, Smidt ML, Lobbes MBI. Noninvasive nodal restaging in clinically node positive breast cancer patients after neoadjuvant systemic therapy: a systematic review. Eur J Radiol. 2015;84(1):41–7.
Barron AU, Hoskin TL, Day CN, Hwang ES, Kuerer HM, Boughey JC. Association of Low Nodal Positivity Rate Among Patients With ERBB2-Positive or Triple-Negative Breast Cancer and Breast Pathologic Complete Response to Neoadjuvant Chemotherapy. JAMA Surg. 2018;153(12):1120–6.
Zhu J, Jiao D, Yan M, Chen X, Wang C, Lu Z, et al. Establishment and Verification of a Predictive Model for Node Pathological Complete Response After Neoadjuvant Chemotherapy for Initial Node Positive Early Breast Cancer. Front Oncol. 2021;11:675070.
Kim R, Chang JM, Lee HB, Lee SH, Kim SY, Kim ES, et al. Predicting Axillary Response to Neoadjuvant Chemotherapy: Breast MRI and US in Patients with Node-Positive Breast Cancer. Radiology. 2019;293(1):49–57.
Haque W, Verma V, Hatch S, Suzanne Klimberg V, Brian Butler E, Teh BS. Response rates and pathologic complete response by breast cancer molecular subtype following neoadjuvant chemotherapy. Breast Cancer Res Treat. 2018;170(3):559–67.
Shen G, Zhao F, Huo X, Ren D, Du F, Zheng F, et al. Meta-Analysis of HER2-Enriched Subtype Predicting the Pathological Complete Response Within HER2-Positive Breast Cancer in Patients Who Received Neoadjuvant Treatment. Front Oncol. 2021;11:632357.
Weisman P, Ospina-Romero M, Yu Q, Wisinski K, Xu J. HER2-positive/ER-low breast carcinoma shows a response to neoadjuvant chemotherapy similar to that of HER2-positive/ER-negative breast carcinoma. Pathol Res Pract. 2022;238:154087.
Osako T, Nishimura R, Okumura Y, Toyozumi Y, Arima N. Predictive significance of the proportion of ER-positive or PgR-positive tumor cells in response to neoadjuvant chemotherapy for operable HER2-negative breast cancer. Exp Ther Med. 2012;3(1):66–71.
Raphael J, Gandhi S, Li N, Lu FI, Trudeau M. The role of quantitative estrogen receptor status in predicting tumor response at surgery in breast cancer patients treated with neoadjuvant chemotherapy. Breast Cancer Res Treat. 2017;164(2):285–94.
Li L, Han D, Wang X, Wang Q, Tian J, Yao J, et al. Prognostic values of Ki-67 in neoadjuvant setting for breast cancer: a systematic review and meta-analysis. Future Oncol. 2017;13(11):1021–34.
Chen X, He C, Han D, Zhou M, Wang Q, Tian J, et al. The predictive value of Ki-67 before neoadjuvant chemotherapy for breast cancer: a systematic review and meta-analysis. Future Oncol. 2017;13(9):843–57.
Kastenhuber ER, Lowe SW. Putting p53 in Context. Cell. 2017;170(6):1062–78.
Chen MB, Zhu YQ, Xu JY, Wang LQ, Liu CY, Ji ZY, et al. Value of TP53 status for predicting response to neoadjuvant chemotherapy in breast cancer: a meta-analysis. PLoS One. 2012;7(6):e39655.
Lee SK, Bae SY, Lee JH, Lee HC, Yi H, Kil WH, et al. Distinguishing Low-Risk Luminal A Breast Cancer Subtypes with Ki-67 and p53 Is More Predictive of Long-Term Survival. PLoS One. 2015;10(8):e0124658.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Development and assessment of a novel core biopsy-based prediction model for pathological complete response to neoadjuvant chemotherapy in breast cancer

Status:

Version 1

Abstract

Figures

Introduction

Patients And Methods

Study subjects

Pathologic assessment

Clinical assessment

Statistical method

Convert continuous variables to categorical variables

Variable description

Variable selection

Nomogram development and assessment

Results

Convert continuous variables to categorical variables

Patient baseline characteristics and univariate analysis

Binary logistic regression analysis

Develop and assess the nomogram

Discussion

Clinical tumor staging

Clinical nodal status

ER status

Ki67 status

p53 status

Prediction model

Limitation

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1