Real-World Outcomes of CDK4/6 Inhibitors Treatment in Metastatic Breast Cancer in Romania

The introduction in clinical practice of selective cyclin-dependent kinase (CDK) 4/6 inhibitors improves the outcome of patients with hormone receptor (HR)-positive human epidermal growth factor receptor 2 (HER2)-negative metastatic breast cancer (mBC). In Romania, the three available CDK 4/6 inhibitors (Palbociclib, Ribociclib and Ademaciclib) have been approved by the National Agency for Medicines (ANM) in 2019, 2020 and 2021. We conducted a retrospective study from 2019 to 2022 on 107 patients with metastatic breast cancer HR+ that have been treated with CDK 4/6 inhibitors in addition to hormone therapy in the Oncology Department of Colțea Clinical Hospital in Bucharest. The purpose of this study is to calculate the median progression-free survival (PFS) and to compare it with the median PFS from other randomized clinical trials. A key difference from other studies is that our study evaluated both patients with non-visceral mBC and patients with visceral mBC, as these two groups often have different outcomes. A total of 79.4% were postmenopausal patients and 20.6% were premenopausal; 42.1% had different stages at the beginning of disease and 57.9% presented newly metastatic disease. Median PFS was 17 months, unlike randomized clinical trials which reported a median PFS of 25.3 months. The combination of CDK 4/6 inhibitors with endocrine therapy is the golden standard treatment in HR-positive, HER2-negative metastatic breast cancer, bringing a prolongation of survival for these patients. Our results show no major differences compared to randomized clinical trials, despite the smaller patient group. In order to have a picture of the efficacy of the treatment as close as possible to the real-world data, we believe that it would be very useful to have a collaboration between several oncology departments in different institutions to carry out a multi-center study on large groups of patients.


Introduction
Breast cancer is the most common malignancy diagnosis at a global level, accounting for more than 2 million cases each year [1]. It is also the leading cause of cancer death in women worldwide, according to Globocan 2020 [2].
Stage IV breast cancer, either de novo at presentation or secondary after a primary diagnosis, is a heterogeneous disease. Although patients diagnosed with metastatic breast cancer (mBC) are unlikely to be cured, systemic therapies, hormone therapies and targeted therapies have contributed to a significant improvement in overall survival and progressionfree survival over the past decade, especially among those with positive hormone receptor and negative HER growth factor overexpression [3][4][5][6]. A survival of 18 to 36 months has been reported for patients with mBC in different studies, although this range extends from a few months to several years. This gap may vary depending on the availability of effective therapies and the patient's age, performance status, comorbidities, tumor

Treatment Monitoring
In our clinic's practice, we monitor treatment failure, taking into account serial changes in symptoms, physical examination or tumor markers and evidence of disease progression based on serial imaging examinations [24]. The criteria we used to define treatment failure include the following: clinical deterioration during treatment (i.e., increase in diseaseassociated symptoms, intolerable treatment toxicities, decreased performance status), evidence of new metastatic sites, increase in the size of previously documented metastatic lesions and interpretation by individual treating physicians, not using RECIST criteria similar to other clinical trials [13].

Study Limitations
This is an observational, retrospective study. Disease progression was based only on clinical and radiological assessment interpreted by individual clinicians, i.e., not using RECIST criteria similar to clinical trials. This study does not have a control arm consisting only of patients exclusively undergoing hormone therapy. Survival may also be affected by subsequent therapies in patients initially diagnosed with stages I/II/III. Statistical analysis was made using IBM SPSS statistics software version 29.

Results
Out of 107 patients with HR+ HER2-metastatic breast cancer who were treated with CDK4/6 inhibitor along with hormone therapy at Coltea Clinical Hospital Bucharest, 58 patients were treated with Palbociclib, 15 with Abemaciclib, 34 with Ribociclib and the following parameters were studied: age, menopausal status, immunohistochemistry type and outcomes for each category.
As shown in Figure 1, The patients' age ranged from 33 to 87 years, with a strong negative skewness (sk = −0.688) indicating a high proportion of patients over 64 years (CI95%: 61.69-66.44) (sd = 12.38); median value md = 67.0 years; and a quartile range Q1-Q3 (56-74). Approximately 75% of the patients were over 55 years of age. following parameters were studied: age, menopausal status, immunohistochemistry type and outcomes for each category.
As shown in Figure 1, The patients age ranged from 33 to 87 years, with a strong negative skewness (sk = −0.688) indicating a high proportion of patients over 64 years (CI95%: 61.69-66.44) (sd = 12.38); median value md = 67.0 years; and a quartile range Q1-Q3 (56-74). Approximately 75% of the patients were over 55 years of age. From the frequency distribution of the age ranges presented in Table 1, negative skewness is observed, with the proportion of patients over 64 years being about 59%. The age interval with the lowest proportion is 18-49 years (15.0%), and the highest proportion is in the age interval 65-74 years (39.3%).   Table 2 show that the proportion of patients with postmenopausal status is high (79.4%); the difference between the weights of the two status categories is statistically significant p < 0.001. From the frequency distribution of the age ranges presented in Table 1, negative skewness is observed, with the proportion of patients over 64 years being about 59%. The age interval with the lowest proportion is 18-49 years (15.0%), and the highest proportion is in the age interval 65-74 years (39.3%).  Figure 2 and Table 2 show that the proportion of patients with postmenopausal status is high (79.4%); the difference between the weights of the two status categories is statistically significant p < 0.001.    A high Luminal B proportion was observed (57.7%) in Table 3, but not statistically significantly different from the Luminal A weight for a test value of χ 2 (1) = 2.70 and p > 0.05. Out of the total of 107 patients, 3 were initially luminal A then changed to luminal B due to disease mutations or failure of treatment appropriate to the category; thus, they were not included in the statistical analysis, leaving only 104 patients analyzed ( Table 3).
As already well-known, cyclin-dependent kinase (CDK) activities coordinate the cell cycle progression through G1/S phase. However, different CDK activities and cyclins in driving cancer cell cycles are highly heterogeneous and can explain why some metastatic breast cancer patients can initially express high estrogen receptors, during which time they can turn into different expressions of estrogen/progesterone features, thus transforming from Luminal A to Luminal B. [25,26].
Of the total 107 patients, 57.9% (n = 62) patients were diagnosed as stage IV de novo, and the remaining 42.1% (n = 45) patients had stage II or III diagnosis. According to Figure 3, the time elapsed between diagnosis and first metastasis is a random variable that ranged from 1 to 232 months. The distribution shows a strong positive skewness with a high proportion of shorter durations when compared to the mean value (m = 68.0; sd = 58.27); median value md = 57.0 months; and a quartile range Q1-Q3   Figure  3, the time elapsed between diagnosis and first metastasis is a random variable that ranged from 1 to 232 months. The distribution shows a strong positive skewness with a high proportion of shorter durations when compared to the mean value (m = 68.0; sd = 58.27); median value md = 57.0 months; and a quartile range Q1-Q3 (20-103). A low proportion of Abemaciclib inhibitor was observed (14%) according to Figure 4 and Table 4, while the highest proportion was observed for Palbociclib (54.2%), with a statistically significant difference χ 2 (2) = 26.04 and p < 0.001.  A low proportion of Abemaciclib inhibitor was observed (14%) according to Figure 4 and Table 4, while the highest proportion was observed for Palbociclib (54.2%), with a statistically significant difference χ 2 (2) = 26.04 and p < 0.001.  As shown in Table 5 and Figure 5, the most frequently used hormone therapy in our study is Letrozole (54.2%), followed by Fulvestrant (34.6%). Anostrozole was used in only about 11% of cases.   As shown in Table 5 and Figure 5, the most frequently used hormone therapy in our study is Letrozole (54.2%), followed by Fulvestrant (34.6%). Anostrozole was used in only about 11% of cases.
In the case of Fulvestrant, a statistically significant association difference was identified (χ 2 (4) = 14.36 with p < 0.01) with the type of CDK4/6 inhibitor used. Thus, in the case of Abemaciclib and Ribociclib, Letrozole was applied to a greater extent, with 73.3% for Abemaciclib and 67.6% for Ribociclib. In the case of Palbociclib, Fulvestrant (43.1%) and Letrozole (41.4%) were applied in approximately equal proportions.  In the case of Fulvestrant, a statistically significant association di fied (χ 2 (4) = 14.36 with p < 0.01) with the type of CDK4/6 inhibitor use of Abemaciclib and Ribociclib, Letrozole was applied to a greater ext Abemaciclib and 67.6% for Ribociclib. In the case of Palbociclib, Fulv Letrozole (41.4%) were applied in approximately equal proportions. Figure 6 indicates that the distribution of ECOG performance skewed but correct from a medical point of view, with ECOG 0 in 42.1 29.9% cases, ECOG 2 in 18.7 cases and ECOG 3 being observed in only  Figure 6 indicates that the distribution of ECOG performance status is strongly skewed but correct from a medical point of view, with ECOG 0 in 42.1% cases, ECOG 1 in 29.9% cases, ECOG 2 in 18.7 cases and ECOG 3 being observed in only 9.3% cases. In terms of the difference between the ECOG score distributions for the three types of CDK4/6 inhibitor, it was observed that they did not differ significantly (χ 2 (6) = 1.37.36 and p > 0.10).
A high proportion of stage IV (57.9%) was observed in Table 6, followed by stage II (23.4%) and stage III (18.7%). Between patient groups separated according to CDK 4/6 inhibitor type, the difference in disease stage is not statistically significant (χ 2 (4) = 3.34 and p > 0.10).  In terms of the difference between the ECOG score distributions for the three types of CDK4/6 inhibitor, it was observed that they did not differ significantly (χ 2 (6) = 1.37.36 and p > 0.10).
A high proportion of stage IV (57.9%) was observed in Table 6, followed by stage II (23.4%) and stage III (18.7%). Between patient groups separated according to CDK 4/6 inhibitor type, the difference in disease stage is not statistically significant (χ 2 (4) = 3.34 and p > 0.10).  As shown in Table 7   As presented in Table 8, in terms of the number of metastatic sites, the highest proportion of patients had one metastatic site (37.4%) or two metastatic sites (30.8%). As many as 15% of cases had at least four metastatic sites. An ordinal regression analysis was applied using the type of CDK4/6 inhibitor used as a factor. We used Abemaciclib as reference because it had the highest proportion (73.3%) of low numbers of metastatic sites (1-2).
As shown in Table 9, the ordinal logistic regression model does not reach statistical significance (W(2) = 3.51 and p > 0.10), but the odds ratio at the limit of statistical significance (W(1) = 2.89 p < 0.10) is 1.5 times higher for Palbociclib than for Abemaciclib (exp(B) = 1.50; CI95%: 0.94-2.38). In Table 10 and Figure 7, Kaplan-Meier analysis was applied for progression-free survival (PFS), considering the criterion event as lack of progression, resulting in a total PFS of 17.6 months.   As illustrated in Table 11 and Figure 8, the estimated PFS on each CDK4/6 inhibitor was uneven, so patients on Palbociclib (n = 58) had a mean PFS of 22.9 months, those on Ribociclib (n = 34) had a mean PFS of 12.06 months, and those on Abemaciclib (n = 15), 11.7 months. We consider that the inhomogeneous distribution of the number of patients on each of the three arms (patients on Palbociclib, patients on Ribociclib and patients on Abemaciclib) resulted in an inhomogeneous distribution of PFS.  As illustrated in Table 11 and Figure 8, the estimated PFS on each CDK4/6 inhibitor was uneven, so patients on Palbociclib (n = 58) had a mean PFS of 22.9 months, those on Ribociclib (n = 34) had a mean PFS of 12.06 months, and those on Abemaciclib (n = 15), 11.7 months. We consider that the inhomogeneous distribution of the number of patients on each of the three arms (patients on Palbociclib, patients on Ribociclib and patients on Abemaciclib) resulted in an inhomogeneous distribution of PFS.   Table 12 indicates that the death rate is low (22.4%). A binary logistic regression test was applied using CDK4/6 inhibitor type as a factor and Abemaciclib as reference. According to Table 13, the model is not statistically significant and there was no statistically significant difference in the proportion of death depending on the type of CDK4/6 inhibitor used.

Discussion
There is only one ongoing real-world clinical trial investigating the efficacy of CDK4/6 inhibitor in metastatic breast cancer patients in Romania, with expected results in  Table 12 indicates that the death rate is low (22.4%). A binary logistic regression test was applied using CDK4/6 inhibitor type as a factor and Abemaciclib as reference. According to Table 13, the model is not statistically significant and there was no statistically significant difference in the proportion of death depending on the type of CDK4/6 inhibitor used.

Discussion
There is only one ongoing real-world clinical trial investigating the efficacy of CDK4/6 inhibitor in metastatic breast cancer patients in Romania, with expected results in 2024 [27].
An updated search in PubMed of other daily practice analyses of inhibitor CDK4/6 treatment for metastatic breast cancer in Romania revealed just one case report, and no other published data are displayed in databases or registries [28]. Our real-world retrospective study is the first to show efficacy data on this research topic in Romania.
In terms of menopausal status, our group of patients has a high postmenopausal status of 79.4%, which is similar to what was reported in other real-world data studies [14,15].
In other real-world data studies, the patients had an Eastern Cooperative Oncology Group performance status between 0 and 1 (on a 5-point scale, where higher scores reflect higher disability); whereas, in our study, they were between 0 and 3. A proportion of about 72% had ECOG scores of 0 and 1 and only 9.3% of the patients had ECOG 3.
The distribution per CDK4/6 inhibitor was inhomogeneous, with Palbocliclib having the highest proportion of 54.2% and Abemaciclib having the lowest proportion of only 14%, with a statistically significant difference: χ 2 (2) = 26.04 and p < 0.001. According to www.fda.gov, the Food and Drug Administration approved Palbociclib in February 2015 as the first CDK4/6 inhibitor (in combination with Letrozole) for the treatment of HR-positive, HER2-negative advanced breast cancer as an initial endocrine based therapy in postmenopausal women. On 19 February 2016, the FDA approved Palbociclib in combination with Fulvestrant for the treatment of women with hormone receptor (HR)-positive and human epidermal growth factor receptor 2 (HER2)-negative advanced or metastatic breast cancer with disease progression following endocrine therapy [29]. In March 2017, the FDA also approved the second CDK4/6 inhibitor molecule in the treatment of HR + HER2metastatic breast cancer-Ribociclib [30]. Six months later, in September 2017, the FDA approved the third molecule-Abemaciclib [31]. In Romania, the three molecules were approved much later, the first one in 2019 being Palbociclib, followed by Ribociclib in 2020, and Abemaciclib in 2021. The fragmented approval time of the three molecules made the distribution of each CDK4/6 inhibitor uneven [32][33][34][35][36].
The association of CDK 4/6 inhibitor with hormone therapy was in accordance with the literature. In line I of treatment, the PALOMA trials paired Palbociclib with Letrozole, the MONALEESA trials paired Ribociclib with Letrozole and the MONARCH trials paired Abemaciclib with Letrozole. In line II, these studies replaced Letrozole with Fulvestrant [28]. In our study, most patients were treated as per line I, adding as hormone therapy Letrozole (54.2%) + Anastrozole (11%); while line II treatment with Fulvestrant included 34.6% of patients.
In our study both patients who were diagnosed de novo in stage IV and those who received subsequent therapies received CDK4/6 inhibitors. Thus, at the time of diagnosis, we had 57.9% patients in stage IV, 23.4% patients in stage II, and 18.7% patients in stage III. Those in stage II and III at the time of diagnosis of breast cancer received subsequent therapies and, over time, progressed to metastatic stage and were given a CDK 4/6 inhibitor plus endocrine therapy.
In the PALOMA study, 60% of patients had visceral metastases and 40% non-visceral metastases. In our case, 69% (n = 40) of patients on Palbociclib had visceral metastases and 77.6% (n = 45) had non-visceral metastases. In the MONALEESA study, 42% of patients had visceral metastases, while for our Ribociclib arm (n = 34), 61.8% (n = 21) had visceral metastases and 76.5% (n = 26) had non-visceral metastases. The MONARCH study showed 52.7% patients with visceral metastases and 26.2% with non-visceral metastases, but our study on Abemaciclib treatment showed 73.3% (n = 11) of patients with visceral metastases and 53.3% (n = 8) with non-visceral metastases. Our study had a total of 107 patients of which 67.3% (n = 72) had visceral metastases and 73.6% (n = 78) had non-visceral metastases. Our data were similar to the aforementioned clinical trials in terms of the number of metastatic sites, with the highest proportions of patients having one and two metastatic sites (37.4% and 30.8%, respectively). The MONALEESA study divided the number of metastatic sites below three in 42% of cases and above three in 30% of cases.
In terms of progression-free survival in the overall population, the PALOMA study conducted on n = 347 patients treated with Palbociclib in combination with endocrine therapy showed a PFS of 34.9 months. Our study (n = 58), for the same combination of molecules, had a mean PFS of 22.9 months. The MONALEESA study had n = 484 patients treated with Ribociclib in combination with endocrine therapy and showed a PFS of 20.6 months. Our Ribociclib arm of 34 patients had a mean PFS of 12.06 months. The MONARCH study was conducted on n = 267 patients, obtaining a mean PFS of 27.3 months. We had very few patients on Abemaciclib in combination with endocrine therapy (n = 15) and obtained a mean PFS of 11.7 months. We consider that the uneven distribution of the number of patients on each of the three arms-Palbociclib, Ribociclib and Abemaciclib-led to an uneven distribution of PFS, and the much smaller number of patients in this study group compared to the number of patients in the real-world data led to a lower PFS [37][38][39][40][41][42].

Conclusions
The combination of CDK 4/6 inhibitors with endocrine therapy is the gold standard treatment in HR-positive, HER2-negative metastatic breast cancer, effectively prolonging survival for these patients.
Our results show no major differences compared to randomized clinical trials, despite the smaller patient group. In order to acquire a picture of the efficacy of the treatment as close as possible to the real-world data, we believe that it would be very useful to conduct a future collaboration between several oncology departments in different institutions to carry out a multicenter study on large groups of patients.

Institutional Review Board Statement:
The present study is part of the PhD thesis of the first author, for which she received approval from the Ethics Committee of the Colt , ea Clinical Hospital Bucharest where this study was conducted, according to decision 19091/5 October 2021.
Informed Consent Statement: Informed consent was obtained from all subjects involved in this study. Data Availability Statement: Data available on request due to ethical restrictions. The data presented in this study are available on request from the corresponding author and the Coltea Clinical Hospital (secretariat@coltea.ro). The data are not publicly available due to the policy of Coltea Clinical Hospital to have the approval of the Ethics Commitee for each new research study.