This retrospective study found a median PFS of 4.3 months for capecitabine given as monotherapy for HER2 normal MBC. The effect of capecitabine was consistent regardless of the treatment line with longer PFS in ER-positive disease compared to ER-negative disease (HR 0.6, 95% CI, 0.0–0.9, p = 0.006).

The median PFS of 4.3 months is comparable to other studies. A systematic review from 2011 found a PFS of 18 weeks equivalent to 4.2 months [9]. Furthermore, an open-label phase-two trial found a PFS of 4.6 months [10].

We found no significant difference in median PFS when patients were stratified according to the line of treatment. More patients had ER-negative disease in the early lines compared with later lines, addressing the more aggressive nature of the disease. Another potential explanation could be differences in the distribution of the disease burden between the different treatment lines. Consequently, we hypothesize that patients who received capecitabine in early-line settings were more ill than those who received capecitabine in later lines.

Amari et al. investigated capecitabine and time-to-progression (TTP) distributed by lines. Compared with our study, they found a longer median TTP of 8.5 months for patients in the first and fourth-line treatment settings and 6.5 months for patients in the second and third-line treatment settings [5]. Possible explanations could be that Amari et al. did not account for prior lines of endocrine therapy, and patients in their study did not receive CDK4/6i before capecitabine. In contrast, approx. 15% of patients in our study received CDK4/6i before capecitabine initiation. Further, TTP ignores deaths from other causes than breast cancer, which PFS does not.

We found that the distribution of patients with ER-positive disease increased gradually from the first to the fifth line of capecitabine. This might be due to the better treatment options for ER-positive disease, such as endocrine therapy and CDK4/6i before capecitabine. The skewed distribution between the groups regarding ER status may cause the results since PFS is affected by ER status according to different studies [11, 12]. Due to limited sample sizes within each group, this study did not perform PFS calculations for every treatment line stratified by ER-positive and ER-negative.

Hong et al. suggest that ER-positive can be a useful predictive marker for better PFS to second or later line of capecitabine [2]. However, given the retrospective design of their study, as well as our study, we cannot distinguish between prognostic factors and factors predictive for the effect of capecitabine. Thijssen et al. found that patients with ER-positive disease had a significantly longer TTP compared to patients with ER-negative disease [12]. A recent study by Siddiqui et al. enlightened that thymidylate synthase levels were found to be significantly higher in triple-negative breast cancer [13], which can be the rationale for why ER-positive is linked to a longer TTP. However, this needs confirmation in future studies.

The number of patients who did not receive ET before capecitabine in our study was 64/162 (39.5%). Given the number of 41 patients with ER-negative disease, it is considered a relatively high percentage. Additionally, 25 patients received capecitabine as first-line treatment. It should be noted that the 25 patients who received capecitabine in the first line may have developed MBC during their adjuvant ET, which is not accounted for in this study since only treatment from the time of MBC diagnosis was documented. This may overestimate the number of patients who did not receive ET before capecitabine, which is 64/162 (40%) in this study.

Regarding OS, a median of 14.2 months was found, which resembles the OS in Thijssen et al. of 58 weeks (13.3 months) [12]. When stratified into different lines, it seems as if the best line to receive capecitabine is the third line (17.8 months). This is most likely due to an accumulation of patients with ER-positive disease, who generally have a better prognosis, receiving endocrine therapy and/or CDK4/6i in the first or second treatment line and therefore receive capecitabine in third or later line.

A similar tendency is seen when comparing OS for patients with ER-positive and ER-negative disease, where a significant difference was found.

It is a limitation that our findings are based on a single-center retrospective study with a relatively small sample size. We did not account for adjuvant therapy because the report is made from the time of MBC diagnosis; hence, the patients who relapsed during adjuvant endocrine therapy were not considered. Due to the retrospective study design, it cannot be assured that this study is free of detection bias.

A strength of our study is the strict inclusion of patients with HER2 normal MBC receiving capecitabine as monotherapy in any line, regardless of performance status or comorbidities. Further, we managed to reproduce and validate the findings of earlier studies with a larger sample size [2, 9, 12]. To our knowledge, our study is the first study investigating PFS of late-line capecitabine after the introduction of CDK4/6i.