The increasing importance of histologic grading in tailoring adjuvant systemic therapy in 30,843 breast cancer patients

The key-objectives of this study were to evaluate the importance of grading in tailoring adjuvant systemic therapy within the past and current Dutch breast cancer guideline, by identifying specific groups of patients for whom the systemic therapy indication depends on histologic grade. Second, we evaluated guideline adherence to the past Dutch breast cancer guideline, and we evaluated which factors, among which histologic grade, may have influenced this.

Reports of patients without a primary tumor (n = 2104) were excluded. Furthermore, reports from patients who received neoadjuvant treatment (n = 5829) were excluded, because treatment for these patients is based upon biopsies, whereas biomarkers like grade, and receptor-status may alter after systemic treatment [16‐21], and as such may differ from the initial biopsy. For synchronous IBC, defined as ipsilateral IBC within six months after the first IBC resection-specimen report, only the first report was included, as this tumor is usually the index-tumor on which treatment decisions are based. Remaining patients with > 1 pathology report (n = 1495), which concerned bilateral tumors in the majority (> 90%), were excluded, as it could not be determined which tumor was considered the index-tumor. Lastly, we excluded reports with any missing data (which concerned missing values on grade, ER/PR-, and HER2-receptor status) (eFigure 1).

To evaluate the role of grading in tailoring adjuvant systemic therapy in daily practice, the pathology-specific data were linked to treatment data from the Netherlands Cancer Registry (NCR) adding information on adjuvant systemic treatment and information on N-status. In the final step of linking, < 5% of patients (1454) could not be matched, leaving a total of 30,843 IBC patients for data analysis (eFigure 1).

This study was approved by the scientific and privacy committee of PALGA and the monitory board of the NCR.

Grade was assessed according to the modified Bloom and Richardson guideline, which combines three sub-scores (nuclear pleomorphism, tubule formation and mitotic count), resulting in a total score and derived grade [22, 23]. The combined ER/PR receptor-status was considered positive when ≥ 10% of tumor cells showed nuclear staining by immunohistochemistry (IHC) for either or both ER and PR, which is the cutoff used in the Netherlands [12]. Based on guideline cutoffs, tumor size was categorized (≤ 1 cm, 1.1-2 cm, > 2 cm). For age, we used different cutoffs based on the guideline indication, as this was deemed clinically most relevant. For GEPs, we only presented numbers on 70-GS, as 21-GS was barely used in the Netherlands between 2013 and 2016 (< 1%) [24].

According to the guideline that was valid during the study period (2013–2016) [25], four patient groups had an aCT-indication. First this included all N + patients < 70 years (subgroup 1). Second this concerned N0 patients < 70 years with unfavorable characteristics, which were defined as age < 35 years (except grade I tumors ≤ 1 cm) (subgroup 2), or, within patients aged 35–70, grade II-III tumors of 1.1-2 cm (subgroup 3), and all tumors > 2 cm (subgroup 4).

When specifically focusing on grade, it can be concluded that the aCT-indication according to the 2013–2016 guideline specifically depended on grade (i.e. grade was the determinative factor) in patients < 35 years with a N0 tumor ≤ 1 cm, and in patients aged 35–70 with a N0 tumor between 1.1 and 2 cm, as aCT was indicated when their tumor was graded as ≥ grade II.

To evaluate whether accurate grading by pathologists remains relevant, we then evaluated grading within the current Dutch breast cancer guideline, which was published in 2019 [11]. Currently, aCT is indicated in N + patients < 70 years (except grade I tumors < 2 cm) (subgroup 1), and patients < 70 years N0/N0(i +)/N1(mi) with unfavorable characteristics. These unfavorable characteristics are defined as age < 35 years with a grade I tumor > 2 cm, or a grade II/III tumor > 1 cm (subgroup 2), age 35–70 with a grade I tumor > 3 cm, a grade II tumor > 2 cm, or a grade III tumor > 1 cm (subgroup 3). Lastly, aCT is indicated in all HER2-overexpressing tumors (subgroup 4).

Thus the aCT indication within the current guideline depends on grade in three specific patient subgroups. First this concerns patients < 35 years, N0/N0(i +)/N1(mi) with a HER2-negative tumor between 1.1 and 2 cm. Second, this concerns patients aged 35–70, N1-3, and a hormone-positive, HER2-negative tumor ≤ 2 cm. Lastly, this concerns patients aged 35–70, N0/N0(i +)/N1(mi) with a hormone receptor positive, HER2-negative tumor between 1.1 and 3 cm [11].

In addition, within the current guideline, grading also plays a role in tailoring (aET). Overall, aET is indicated in patients of all ages with an ER- and/or PR-positive tumor receptor status and N1-status (including (N1(mi)), while in case of an N0-status (including N0(i +)), aET is only indicated in grade I tumors > 2 cm, and ≥ grade II tumors > 1 cm. Thus, for N0/N0(i +) patients with a hormone-positive tumor between 1.1 and 2 cm, the aET-indication is determined by grade.

Patient-, tumor- and treatment characteristics were summarized using counts and proportions, and means and standard deviations.

First, we identified patients for whom the aCT-indication, when strictly applying the 2013–2016 guideline, depended on grade. We then performed the same analyses for the 2019 guideline, and we also identified patients for whom the aET-indication would depend on grade, as this was added in the 2019 guideline. Second, we calculated which proportion of patients with an aCT-indication according to the 2013–2016 guideline actually received aCT. Third, the relation of grade and other clinicopathologic variables (age, ER/PR- and HER2-receptor status, tumor size) and 70-GS use with aCT-administration was evaluated by multivariate logistic regression for the overall population with a 2013–2016 guideline aCT-indication (i.e. all four subgroups). Adjusted odds ratios (AORs) and 95% confidence intervals (CIs) were provided for aCT versus no aCT. Grade I was taken into account as reference category for grade. Lastly, to identify factors that may have influenced guideline adherence with regard to aCT, within the four subgroups with a guideline aCT-indication, we compared clinicopathologic variables (grade, ER/PR- and HER2-receptor status, age, tumor size) and 70-GS-use between patients who did and did not receive aCT, using counts and proportions, which were tested by means of a χ²-test. For age, a cutoff of ≥ 60 years was used, as a considerable group of patients with an aCT indication was aged between 60 and 70 years, which may have influenced guideline adherence.

We identified 7744 (25.1%) patients for whom grade would be the determinative factor for their aCT-indication, when strictly applying the guideline that was valid during the study period (2013–2016) (eTable 1). When applying the guideline that was published in 2019 [11], we identified an increase in the percentage of patients for whom the aCT-indication would depend on grade, being 35.2% (n = 10,869) (eTable 2). In addition, new in the current guideline, the indication for aET would depend on grade in 9,173 (29.7%) patients (eTable 3).

Applying the Dutch breast cancer guideline that was valid between 2013 and 2016, we identified 14,954 patients (48.5%) with an aCT-indication (Table 2). Of this group, only 9,010 patients (60.3%) actually received aCT. As 9,950 patients actually received aCT, 940 patients (9.4%) apparently did so without having a guideline-indication. Administration of aCT decreased slightly between 2013 and 2015, while a considerable decrease of 12.8% was observed in 2016 (Fig. 1).

In general, for all patients with an aCT-indication, aCT-administration was significantly more apparent in patients with tumors of higher grade [adjusted OR (AOR) grade II: 1.4 (95%CI 1.3–1.7), AOR grade III: 4.7 (95%CI 4.1–5.4)], HER2-receptor-positive tumors [AOR 4.9 (95%CI 4.1–5.8)], and tumors of larger size [AOR per cm increase: 1.4 (95%CI 1.3–1.4)]. In contrast, patients with hormone-receptor-positive tumors (AOR: 0.4 (95%CI 0.4–0.5), an age ≥ 60 years (AOR: 0.3 (95% CI 0.3–0.3), or patients with a 70-GS performed (irrespective of 70-GS result-category) (AOR: 0.4 (95%CI 0.4–0.4) had significantly lower odds of receiving aCT (Table 3).

Guideline adherence was most apparent for patients < 35 years with an aCT-indication (89.1%), while it was lowest for the group of N0-patients aged 35–69 with tumors of 1.1-2 cm size (45.4%) (Table 4). Notably, within the latter group, only 20.9% of patients with grade II tumors received aCT, whereas 78.8% of patients with grade III tumors received aCT (p = 0.000). In addition, 70-GS-use was highest within this group (30.2%) and aCT was only administered in 34.6% of these patients.

As to the role of the 70-GS, this was used in 2,950 patients with a guideline-aCT-indication (19.7%), and only about a third of these patients received aCT (Table 4). We also observed an only marginal increase in the use of 70-GS in the years between 2013 and 2015, followed by a considerable increase of 12.4% in 2016 (Fig. 1).

This study with nationwide data on 30,843 Dutch primary breast cancer patients illustrates the increasing importance of histologic grading with regard to tailoring adjuvant systemic therapy, as the indications for aCT and aET depend on grade in roughly a third of IBC patients. This is due to the large proportion of patients who fall into the category in which histologic grade finally determines the need for adjuvant therapy when national guidelines are followed. This stresses the need for optimizing grading by pathologists, to diminish the risk of worse patient outcome due to non-optimal treatment decisions.