Prognostic factors of brain metastasis and survival among HER2-positive metastatic breast ﻿cancer patients: a systematic literature review

Background Patients with breast cancer who overexpress the human epidermal growth factor receptor 2 (HER2) and subsequently develop brain metastasis (BM) typically experience poor quality of life and low survival. We conducted a comprehensive literature review to identify prognostic factors for BM and predictors of survival after developing BM, and the effects of therapies with different mechanisms of action among patients with HER2+ breast cancer (BC). Methods A prespecified search strategy was used to identify research studies investigating BM in patients with HER2+ BC published in English during January 1, 2009–to June 25, 2021. Articles were screened using a two-phase process, and data from selected articles were extracted. Results We identified 25 published articles including 4097 patients with HER2+ BC and BM. Prognostic factors associated with shorter time to BM diagnosis after initial BC diagnosis included younger age, hormone receptor negative status, larger tumor size or higher tumor grade, and lack of treatment with anti-HER2 therapy. Factors predictive of longer survival after BM included having fewer brain lesions (< 3 or a single lesion) and receipt of any treatment after BM, including radiosurgery, neurosurgery and/or systemic therapy. Patients receiving combination trastuzumab and lapatinib therapy or trastuzumab and pertuzumab therapy had the longest median survival compared with other therapies assessed in this review. Conclusions More research is needed to better understand risk factors for BM and survival after BM in the context of HER2+ BC, as well as the assessment of new anti-HER2 therapy regimens that may provide additional therapeutic options for BM in these patients. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-021-08708-5.

We conducted a literature review to assess the epidemiology of patients with HER2+ metastatic or advanced BC with BM by describing prognostic factors for developing BM and factors predictive of survival among patients with BM. Additionally, differences in survival and time to progression by HER2-targeting treatment drug classes were explored based on the drug mechanism of action.

Search design
Electronic searches were conducted in PubMed and Embase. A predefined search strategy (Online Resource 1 and Online Resource 2) was used to identify research studies investigating BM in patients with HER2+ BC. The search was restricted to studies published in English. The original search was restricted to studies published during the period of January 1, 2009 to July 30, 2019. However, the protocol was amended to expand the search to June 25, 2021 to capture the most recent published literature. Titles and abstracts identified from the electronic databases were exported to an Excel (Microsoft Corporation; Redmond, Washington) file for screening. Systematic literature reviews and meta-analyses relevant to the study objectives were not themselves included within the scope of this literature review, but the bibliographies were reviewed to identify potential additional publications.

Screening and extraction
Articles were screened in a two-level process. In Level 1 screening, one researcher reviewed the titles and abstracts of the identified articles according to the literature review inclusion and exclusion criteria ( Table 1) ▪ PK/PD of treatments BM brain metastasis; CDK4/6 cyclin-dependent kinase 4 and 6; HER2 human epidermal growth factor receptor 2; PD pharmacodynamics; PK pharmacokinetics a Literature reviews and meta-analyses were not be included in the review but were used to identify primary studies not previously identified and selected articles for further review. In Level 2 screening, the full text of articles selected at Level 1 were reviewed by one researcher using the same set of inclusion and exclusion criteria. If there was any uncertainty about the inclusion of articles, the Level 2 reviewer discussed the article with a second researcher to confirm, by consensus, whether the article met the study inclusion and exclusion criteria. Data were extracted according to prespecified data fields using the full-text articles, including study population, country, observation period, sample size, and select sample characteristics. A checklist was used to assess the quality of the studies.

Results
The number of studies included and excluded at each stage of screening was documented in a PRISMA diagram ( Fig. 1) [14]. The original search strategy yielded 232 records for Level 1 screening and the expanded search added 162 records for Level 1 screening, for a total of 394 records retrieved. The bibliographies of 8 systematic reviews/meta-analyses were reviewed and yielded one additional study to be included for full-text screening for a total of 138 articles that were included in the Level 2 full-text screening. A total of 25articles met the inclusion criteria described in Table 1 and were selected for data extraction.
Prognostic factors for developing brain metastasis Table 2 presents information on time to first BM diagnosis (TTBM) from initial HER2+ BC diagnosis and prognostic factors associated with a shorter TTBM. Across the 20 studies that reported information on TTBM, the shortest reported median TTBM was 10.8 months [31] and the longest was 76.2 months [26]. Among the 10 studies that reported on prognostic factors for BM diagnosis [15, 17, 18, 21-23, 26, 29-31], the most commonly assessed prognostic factors included age, HR status, receipt of anti-HER2 therapy, and tumor grade.
In three studies, age was not found to be associated with TTBM [21,22,29], while two studies reported an association between younger age at BC diagnosis (aged ≤40 or < 50 years in Maurer et al. [26] and Brufsky et al. [31], respectively) and shorter TTBM [26,31]. Hormone receptor status was assessed as a prognostic factor for TTBM in eight studies; four studies reported no association between HR status and TTBM [21,22,26,29], while four other studies reported that patients with HRstatus had a shorter TTBM compared with those who were HR+ [15,23,30,31]. Among the eight studies that assessed trastuzumab use and TTBM, five reported no association between receipt of trastuzumab and TTBM [17,22,26,29,30], while three reported that patients who received trastuzumab had a longer TTBM compared with those who did not [18,23,31]. Two studies reported an association between higher tumor grade and shorter TTBM [21,23], while two studies reported no association with tumor grade but did report an association between larger tumor size and shorter TTBM [26,29].
Other treatment-related factors and TTBM were also assessed. In one study, longer TTBM was reported among those receiving treatment with endocrine therapy versus no endocrine therapy in the palliative care setting [30]. Conversely, another study reported that patients receiving adjuvant endocrine treatment had a shorter TTBM [26]. Longer TTBM was reported among patients who had surgery for their BC and among those treated with (neo) adjuvant anthracyclines + taxanes [26].
Other reported prognostic factors for shorter TTBM included higher disease burden (i.e., ≥ 2 metastatic sites) [31], higher HER2 levels (measured by the HERmark® Breast Cancer Assay) [21], distant relapse in ≤2 years [22], first distant relapse at a visceral site [23], and nodal involvement [26]. One study reported on genetic factors of TTBM, but the reported associations in the discovery set were not observed in the validation set [23].
Predictors of survival after brain metastasis among patients with HER2+ breast cancer Table 3 presents median time to death (i.e., survival) after BM diagnosis and predictive factors associated with shorter survival after BM diagnosis. Across the 23 studies that reported information on median survival after BM, the shortest reported median survival was 5.2 months [34] and the longest was 28 months [21].
The most commonly assessed predictive factors for shorter survival after BM diagnosis included age, HR status, number of brain lesions, receipt of anti-HER2 therapy, and receipt of any systemic therapy. Four studies reported no association between age at BM diagnosis and survival [15,27,31,35], three reported shorter survival among older patients [18,24,33], and one reported shorter survival among younger patients [28]. The nine studies that observed HR status reported no association between HR status and survival after BM diagnosis [15,24,25,27,28,[31][32][33]36]. Six studies reported that the presence of a higher number brain lesions compared with fewer brain lesions was associated with shorter survival after BM diagnosis [18,24,25,27,28,33]. Other reported predictors of shorter survival after BM diagnosis included the presence of neurologic symptoms [24,26,27], tumor grade 3, two or more extracranial metastatic sites [33], central nervous system disease at BC diagnosis [31], uncontrolled extracranial metastases [27,28], visceral metastases, and Radiation Therapy Oncology Group recursive partitioning analysis prognostic class 3 versus class 1/2 [39].
While specific regimens were not assessed in this literature review, data were included from studies that did not assign treatment and for which any anti-HER2 treatment was captured within typical clinical practice. All 13 studies that assessed treatment with anti-HER2 therapy and survival after BM diagnosis reported an association between anti-HER2 therapy and survival. Twelve studies reported that patients who received anti-HER2 therapy after their BM diagnosis had a longer survival compared with patients who did not receive anti-HER2 therapy [17-20, 24, 25, 27, 28, 30, 31, 36, 39]. Four studies [18,25,33,36] found that patients receiving both trastuzumab and lapatinib after their BM diagnosis had longer      Includes chemotherapy, endocrine therapy, and HER2-targeted therapy survival than those receiving either agent alone or no anti-HER2 therapy. One study found that patients receiving trastuzumab and pertuzumab after their BM diagnosis had longer survival than those receiving other HER2-targeted therapies or no HER2-targeted therapy [19]. Four studies [18,31,35,39] assessed non-HER2targeted therapies and survival after BM diagnosis and reported that shorter survival was associated with no chemotherapy versus chemotherapy [18,31], no hormonal therapy versus hormonal therapy [18], WBRT alone versus any systemic therapy after WBRT [39], and no systemic therapy versus any systemic therapy [35]. In addition, three studies reported that patients receiving surgery or stereotactic radiosurgery had longer survival than those not receiving these treatments [24,31,33], while one study reported no association [35]. One study reported no association between the total dose of radiotherapy and survival after BM diagnosis [28].
Treatment mechanism of action and outcomes after brain metastasis HER2-targeted monoclonal antibodies Table 4 presents information on anti-HER2 treatment type by mechanism of action, and disease progression and survival after BM. Four studies evaluated survival among patients receiving a HER2-targeted monoclonal antibody (i.e., trastuzumab) after BM diagnosis and reported that overall survival (OS) was longer in patients who received trastuzumab after local therapy compared with patients who did not receive trastuzumab [17,30,31,39]. In Ahn et al. [17], Berghoff et al. [30], and Brufsky et al. [31], the difference in median OS after BM between trastuzumab users and nonusers was 7.8 months, 10 months, and 13.8 months, respectively. Niwinska et al. [39] reported that among patients with HR+ status, trastuzumab users had a 11-month longer median OS compared with nonusers (P < 0.001), and among patients with HR-status, trastuzumab users had a 6-month longer median OS compared with nonusers (P = 0.004). One study evaluated survival among patients receiving HER2-targeted monoclonal antibody combination therapy (i.e., trastuzumab + pertuzumab) and reported that OS was longer in patients who received trastuzumab + pertuzumab (44 months) compared to those who received other HER2-targeted therapy (17 months) or no HER2-targeted therapy (3 months) [19].

HER2-targeted non-monoclonal antibodies
One study evaluated survival among patients receiving a HER2-targeted non-monoclonal antibody (i.e., pyrotinib) with or without surgery/radiation, and reported that OS was longer in those with combination HER2-targeted non-monoclonal antibody and surgery/radiation [32].

Tyrosine kinase inhibitor combination therapies
Four studies evaluated the association between patients receiving lapatinib in combination with either trastuzumab or another anti-HER2 therapy, not otherwise specified [25,27,33,36]. Braccini et al. [36] and Kaplan et al. [33] reported that patients receiving both lapatinib and trastuzumab had longer median OS (25.7 months and 23.6 months, respectively) after BM compared with those receiving either lapatinib alone or trastuzumab alone (9.6 months [P < 0.001] and 14.6 months [P = 0.023], respectively). Hayashi et al. [25] reported longer OS after BM in patients receiving lapatinib and trastuzumab compared with patients receiving only one of the two targeted therapies or no HER2-targeting therapy (P < 0.001). Similarly, Morikawa et al. [27] reported that patients receiving lapatinib in combination with another anti-HER2 therapy, not otherwise specified, had lower mortality compared with those not receiving the combination therapy (adjusted hazard ratio = 0.26; 95% confidence interval [CI], 0.13-0.52).

Any anti-HER2 therapy
The use of anti-HER2 therapy, not otherwise specified, after BM diagnosis was associated with an increase in OS in all studies except one [26]. Among the seven studies that reported longer survival in patients treated with anti-HER2 therapy after BM diagnosis [18-20, 24, 28, 33, 36], the median OS among those receiving anti-HER2 therapy ranged from 15.2 to 44 months compared with the median OS among those not receiving anti-HER2 therapy, which ranged from 3 to 13.8 months. Among patients receiving anti-HER2 therapy after BM diagnosis, median OS ranged from 11.8 [36] to 17.5 months [20] longer after their BM diagnosis compared with those not receiving anti-HER2 therapy. Conversely, Maurer et al. [26] reported no association between anti-HER2 therapy after BM diagnosis and OS.

Quality of studies
Quality assessment of included studies was conducted using the Good Research for Comparative Effectiveness (GRACE) checklist [40,41]. This 11-item scale contains 6 items related to quality of data and 5 items related to methodology. For each question, the quality is assessed based on "fit for purpose", and the quality is considered sufficient if the data or information provided per item is adequate for study purposes. The quality assessment revealed that all studies were eligible to be included in this review, even though 5 (20%) studies have some limitations in the scientific methods (See Table 5).

Discussion
This literature review included 25 published articles that assessed a total of 4097patients with HER2+ BC with BM. Among these patients, prognostic factors of BM development and predictive factors of survival after BM diagnosis were assessed. Prognostic factors associated with shorter TTBM among patients with HER2+ BC included younger age at BC diagnosis, HR-versus HR+ status [15,23,30,31], no receipt of trastuzumab versus receipt of trastuzumab [18,23,31], and higher tumor grade versus lower grades [21,23]. While these associations were observed across multiple studies, six studies found no association with TTBM and these prognostic factors [17,21,22,26,29,30]. Some studies reported longer TTBM in patients receiving trastuzumab or anti-HER2 therapy, not otherwise specified, while some studies found no association. Overall survival after BM diagnosis was shorter in patients with a higher number of lesions, was unaffected by HR status, and was variably affected by age at diagnosis. Treatment-related factors predictive of longer survival after BM included receipt of any systemic therapy.
Studies that assessed treatment with anti-HER2 therapy after BM diagnosis and survival (n = 13) reported that patients who received anti-HER2 therapy had longer survival after BM compared with patients who did not receive anti-HER2 therapy. Among studies that assessed survival differences between trastuzumab-based and lapatinib-based therapies [18,25,33,36], patients receiving both trastuzumab and lapatinib after BM diagnosis had longer survival than those receiving either agent alone or no anti-HER2 therapy. One study found that patients receiving both trastuzumab and pertuzumab therapy after BM diagnosis had longer survival than those receiving other HER2-targeted therapy or no HER2-targeted therapy [19]. While trastuzumab has limited capability to cross the intact BBB, elevated concentrations of trastuzumab in the cerebrospinal fluid have been demonstrated when the BBB is impaired via radiotherapy and/or meningeal carcinomatosis [42]. As described in this review, trastuzumab in combination with lapatinib, which has been shown to cross the BBB in the BM setting [43], is favorable for survival in patients with BM. Lapatinib-based therapies may be an alternative therapeutic option for patients with BM and trastuzumab resistance [44]. The studies in this review also reported that survival after BM was improved with other therapies, including surgery or radiosurgery [24,31,33] and other systemic therapies [18,31,39]; however, the improvements in survival in patients receiving anti-HER2 therapy superseded survival in those receiving one of these other therapies alone [24,28,39].
Contemporary patients with HER2+ BC have better OS compared with patients with HER2+ status who received treatment for BC 20 years ago [45]. As more patients with HER2+ BC are living longer, with more opportunity to develop distant metastases, new  [46]. A recent randomized controlled trial found the addition of tucatinib compared with placebo to trastuzumab and capecitabine regimens improved progression-free survival and OS [47]. Trastuzumab deruxtecan, an antibody-drug conjugate, demonstrated encouraging antitumor activity during a phase 2 trial that evaluated patients with HER2+ metastatic BC, including 24 patients with BM [48]. Neratinib, a pan-HER TKI, showed efficacy in combination with capecitabine for treatment of refractory HER2+ BC with BM [49]. Treatment with anti-HER2 therapies, including current investigational and newly approved therapies, may provide additional options for patients with HER2+ BC with BM. This review includes some limitations, such as the search was limited to the past 10 years and to Englishlanguage articles only. Only studies that assessed patients with BM are included in this review. Clinical trials often exclude patients with any BM or enroll only patients with clinically stable BM [50], limiting the ability to comprehensively assess the predictors of survival in patients with BM.

Conclusion
In this literature review, we describe the epidemiology of patients with HER2+ metastatic or advanced BC with BM, including prognostic factors for developing BM, a -Data attributes of exposure, outcomes (recording, objective measurement, validation) and important covariates/co-founders were assessed using 6 item checklist b -Scientific methods were assessed using 5 item checklist and focused on the following areas -New initiators of treatment -Concurrent comparators -Control of covariates/confounders/effect modifier -Control of immortal time bias -Analyses to evaluate the potential for bias for biased assessment factors predictive of survival among patients with BM, and differences in survival and time to progression by HER2-targeting drug class based on drug mechanism of action. Prognostic factors associated with shorter TTBM included younger age, HR-status, no receipt of trastuzumab or anti-HER2 therapy, higher tumor grade, and larger tumor size. Predictors of longer OS after BM included receipt of anti-HER2 therapy or any systemic therapy, and the presence of fewer brain lesions (< 3 or a single lesion). Trastuzumab and lapatinib combination therapy after BM diagnosis was associated with longer OS after BM compared with other treatments assessed in this review. More research is needed to better understand risk factors for BM and treatments that may improve outcomes.