Meta-Analysis of Shrinkage Mode After Neoadjuvant Chemotherapy for Breast Cancers: Association With Hormonal Receptor

Abstract

Background: Patients with concentric shrinkage mode after neoadjuvant chemotherapy (NAC) is considered to be ideal candidates for breast conserving treatment (BCT). While, what proportion of patients would represent CSM have not been well defined. This study was conducted to pool the rates of concentric shrinkage mode (CSM) in patients undergoing NAC, determine the impact of hormonal receptor on the shrinkage mode after NAC and estimate the rates of the CSM in various subgroups.

Methods: We conducted a systematic review following the guidelines for Meta-Analyses and Systematic reviews for the PRISMA guidelines. We systematically searched the literature about shrinkage mode after NAC from PubMed, Web of Science, Embase, The Cochrane Library, CNKI, Wanfang database published from January 2002 to June 2020 on breast cancer shrinkage mode after NAC and carefully screened the literature by using eligibility criteria: (1) patients with primary breast cancer treated with NAC; (2) publications with available data of shrinkage mode measured by magnetic resonance imaging (MRI), or data of pathology and hormonal receptor. The association between shrinkage mode and hormonal receptor was estimated using Stata 15.1 software.

Results: This analysis included a total of 2434 tumors from 23 papers. The included studies were heterogeneous (I2 = 89.4%, P<0.01). Random effects model was used to estimate the overall rates of CSM: 56.6% [95%CI (50.5%, 62.7%)]. According to the analysis of hormonal receptor, 10 of the paper was included for HR+ (hormone receptor positive) type analysis and the rate of CSM for HR+ type was 45.7% [95%CI (36.4%, 55.0%)]; 9 of the paper was used for HR- type (hormone receptor negative) analysis and the incidence of HR-CSM is 63.1% [95%CI (50.0%, 76.1%)]; with HR+ type as the control, the OR of the HR- CSM rate is 2.32 (1.32, 4.08) folds of HR+ type. From subgroup analyses, the CSM% of luminal A, luminal B, Her2+, and triple negative were 29.7% (16.5%, 42.8%); 47.2% (19.1%, 75.3%); 59.0% (39.7%, 78.3%); 66.2% (52.8%, 79.6%), respectively.

Conclusions: Breast cancer patients undergoing NAC did not get an ideal odds ratio of CSM. The incidence of CSM in breast cancer after NAC is associated with hormonal receptor. Patients with triple-negative breast cancers have the highest rates of CSM after NAC. More care should be taken to select patients with the luminal subtypes for BCT throughout NAC.

Keywords: breast cancer; meta-analysis; neoadjuvant chemotherapy; receptor status; shrinkage mode.

Figure 1

Flow chart of literature screening. CSM, concentric shrinkage mode.

Figure 2

Forest plot of the incidence of CSM after NAC. Pooled estimate incidence of CSM based on 23 papers providing data from 2434 tumors; CSM, concentric shrinkage mode; CI, confidence interval; ES, effect size.

Figure 3

The publication bias of the study was evaluated by funnel chart and Egger’s test. CI, confidence interval; SND, standard normal deviation.

Figure 4

Meta-analysis estimates, given named study is omitted. The results shows that only little variation in the estimates of CSM% is induced by omission of the one selected study. CI, confidence interval.

Figure 5

Association of HR hormonal receptor and CSM%. HR+ CSM% (A), HR- CSM% (B) and the OR value between HR- CSM% and HR- CSM% (C). CSM, concentric shrinkage mode; HR+, hormone receptor positive; HR-,hormone receptor negative; OR, odds ratio; ES, effect size.

Figure 6

Calculation of the subtypes-specific CSM%. (A) 14 of 42 luminal tumors showed CSM (Luminal A); (B) 48 of 95 luminal B tumors showed CSM. (Luminal B); (C) 76 of 153 HER-2 positive tumors showed CSM (HER2+); (D) 166 of 255 triple-negative tumors showed CSM. (triple-negative); CSM, concentric shrinkage mode; ES, effect size.