In September 2020, pharmacological counselling provided by the Clinical Pharmacology Unit of the National Cancer Institute CRO Aviano was opened to support medical oncologists in decisions making for first- or second-line CDKis (palbociclib, abemaciclib, and ribociclib) and endocrine therapy (letrozole) for HR-positive/HER2-negative MBC patients. Our counselling service was already available for other KI used in oncology as imatinib, sorafenib, regorafenib, sunitinib, lenvatinib, and three PARP inhibitors: olaparib, niraparib, and rucaparib. Written informed consent was obtained for pharmacogenetic and TDM analyses and the publication of the here presented reports. Any potentially identifying information was omitted. Data concerning age, disease, stage, and molecular profiling, treatment regimen and setting, drug dose, adverse drug reactions, and coadministered treatments were retrieved from the electronic medical record upon patients’ reporting by the prescribing medical oncologist. Toxicities were retrospectively collected through clinical records revision and graded according to Common Terminology Criteria for Adverse Events (CTCAE) Version 5.0.

The pharmacology laboratory is undergoing the UNI EN ISO-15189 accreditation program and is certified according to EMQN (www.emqn.org) and SKML (www.skml.nl) proficiency testing schemes for pharmacogenetic and TDM routine diagnostics, respectively.

Candidate genes were selected based on a literature search (PubMed-MEDLINE) focusing on those encoding for proteins involved in CDKis ADME (Roncato et al., 2020). Considering that CDKis are often administered in association with letrozole, SLCO1B1*5/*15/*17 was also genotyped (Gregory et al., 2017). Patients were genotyped for CYP3A4 (*1B, rs2740574; *1G rs2242480; *3, rs4986910; *20, rs67666821; *22, rs35599367; *26, rs1381053638); CYP3A5 (*3, rs776746; *6, rs10264272; *7, rs41303343); SLCO1B1 *5/*15/*17 (rs4149056); ABCB1 (1236C > T, rs1128503; 3435C > T, rs1045642; 2677G > T/A, rs2032582); and ABCG2 (421C > A, rs2231142). The pharmacogenetic analysis was performed by SNPline PCR Genotyping System platform employing Kompetitive allele-specific PCR (KASP) assays (LGC Genomics, Hoddesdon, United Kingdom) according to manufacturer’s instructions (Biosearch Technologies, 2021). Regarding ABCB1 2677G > T/A the tri-allelic discrimination was assessed using Pyrosequencing technology by PyroMark Q48 (Qiagen, Hilden, Germany). Primer sequences and genotyping details are available upon request. Positive and negative control samples were included in each analysis.

Plasma was obtained by centrifugation at 2,450 g for 10 min at 4°C of whole blood EDTA tubes and stored at −80°C until analysis. Patients’ samples were analyzed with a newly developed LC-MS/MS method as previously reported (Posocco et al., 2020). Drug concentration was usually evaluated at specific time points, which allowed the evaluation of C_min or C_max at steady state. Patients were asked to have their last drug intake 24 h (C_min), or 1–4 h (C_max of ribociclib) before the sampling time. Last administration (self-reported) and sampling times were also recorded. According to the literature (Verheijen et al., 2017), the average exposure of the approved efficacious dose was used as a proxy of target C_min and will be referred to as “target C_min” in this article. In more detail, patients’ concentrations were compared with the reported population mean C_min of 61 ng/ml (US Food and Drug Administration, 2017b) for palbociclib at the standard dose of 125 mg/day and with the reported mean C_min of 732 ng/ml (FDA, 2021b) for ribociclib at the standard dose of 600 mg/day. The mean steady-state population C_max for ribociclib is 2,237 ng/ml. For letrozole, a C_min target value has already been proposed at 85.6 ng/ml by dedicated exposure-efficacy studies (FDA, 2021a).

Potential DDIs were identified using Lexicomp (UpToDate, 2021), Drug Interactions Checker on Drugs.com (Drugs.com, 2021), Flockhart Interaction Table (Flockhart et al., 2021), a summary of each coadministered product characteristics (EMA) (EMA, 2021b, EMA, 2021a) and Medscape (Medscape, 2021). All potential DDIs were analyzed and classified based on their clinical impact as moderate (pharmacological effects must be controlled) or severe (drug combination should be avoided).

Descriptive statistics were used to present and analyze the TDM data. The mean population C_min value reported in the literature for patients treated with the standard dose was used to evaluate the TDM target, and the ±20% intervals were calculated. This interval was based on acceptable variability of analytical data according to incurred sample reanalysis criteria (FDA, 2018; EMA, 2022). Patients who fell outside this range were considered to be potentially under or overexposed to the drugs. For palbociclib, concentrations within 49–73 ng/ml and for ribociclib C_min within 586–878 ng/ml were considered to be within the range, regardless of the dose patients received.

At the age of 54, Case I was diagnosed with breast cancer and, after neoadjuvant chemotherapy, underwent a radical mastectomy followed by adjuvant endocrine therapy with tamoxifen for 5 years and with letrozole for another year. Five years later, after a diagnosis of nodal recurrence, letrozole was reintroduced as first-line therapy. After 2 years, because of bone and nodal progression, second-line therapy with fulvestrant and palbociclib was started. In the most recent period, she started venlafaxine treatment due to a moderate depression status in addition to duloxetine, already prescribed to treat a mood disorder, and a mild bone progressive disease (PD) was reported at the subsequent positron emission tomography (PET) scan. The woman was referred to pharmacological counselling to better characterize the potential impact of DDIs between palbociclib and venlafaxine on the outcome of a CDKi treatment. By the time pharmacological counselling was required, palbociclib dosing had already been reduced (75 mg/day instead of 125 mg/day, 3 weeks on/1 week off), due to previous recurrent hematological toxicity (neutropenia and thrombocytopenia).

A blood sample was collected at steady state to assess the concentration of palbociclib in plasma 24 h after the last intake (C_min). As specified earlier, the patient was receiving a daily dose of 75 mg at the time of blood sampling. The reported average C_min in patients treated at the standard dose of 125 mg/day palbociclib is 61 ng/ml. The measured C_min, which was equal to 27.3 ng/ml was consistent with the reduced dose administered (75 mg/day). When analyzing the case from a pharmacogenetic point of view, no defective genetic variants affecting the metabolism or transport of palbociclib (in the genes CYP3A4, CYP3A5, ABCB1, and ABCG2) were highlighted.

The analysis of potential DDIs highlighted only an increased risk of serotonin syndrome/serotonin toxicity development due to coadministration of serotonin/norepinephrine reuptake inhibitors. Although this information is not strictly useful for the purposes of the requested pharmacological counselling, a potentially damaging DDI was highlighted further supporting the utility of intensified pharmacological care in this setting.

Possibly a prolonged underexposure to the active drug, among other pathological factors, could have affected treatment efficacy. It is also likely that the toxicity experienced by the patient was not related to overexposure to the drug but more likely to high sensitivity to the drug’s toxic effect related to other causes. The patient presented a baseline absolute neutrophil count (ANC) of 3.03 × 10³/mm³ that, according to previous reports, could predispose to a higher risk of neutropenia (Iwata et al., 2021). Based on our analysis, a switch to abemaciclib could have been considered due to a lower incidence of treatment-associated hematologic toxicity neutropenia (Sledge et al., 2020).

The bone lesion for which the patient was referred was treated with locoregional radiotherapy, and the patient continued treatment with fulvestrant and palbociclib. Subsequently, after 5 months, another mild bone PD was noted on the PET scan and locoregional radiotherapy was again performed and treatment continued. Finally, a systemic PD was registered 1 month later and second-line therapy with capecitabine was started. No more blood draws were collected for C_min quantification of palbociclib and letrozole.

Key takeaway: The pharmacological evaluation excluded an interaction between treatment with venlafaxine or duloxetine and palbociclib outcome, avoiding the necessity to modify the anti-depressive treatment. The counselling also ruled out the presence of pharmacogenetic variants and overexposure to palbociclib as a reason for the observed hematologic toxicity and highlighted the presence of a baseline ANC median value below 3.60 (× 10³/mm³) as a risk factor for it. The dose reduction put the patient at risk for sub-optimal exposure to the drug. Switching to a compound less associated with bone marrow suppression, such as abemaciclib, could have been a valuable strategy to overcome recurrent hematologic toxicity. Unfortunately, at the time of writing switch between CDKis was considered an off-label intervention.

Case II was diagnosed with breast cancer when she was 60 and at that time underwent conservative breast surgery, followed by 5-year adjuvant endocrine therapy with letrozole. Recently, a computed tomography (CT) scan has highlighted distant metastasis with lungs, lymph nodes, and bone involvement. First-line endocrine therapy was therefore started with palbociclib 125 mg/day (3 weeks on/1 week off) and letrozole 2.5 mg/day with the occurrence of recurrent neutropenia grade 3 for which the dose was reduced to 100 mg/day and the patient referred to the pharmacological counselling. Plasma concentration at steady state of both palbociclib and letrozole was determined at 22 h after the last drug assumption. The analysis showed: 1) a plasma concentration of 85.2 ng/ml of palbociclib treated at 125 mg/day (approximately 30% higher than the mean C_min of 61 ng/ml reported in the literature for treatment at 125 mg/day); 2) a plasma concentration of 93.1 ng/ml for letrozole, in line with the desired threshold (85.6 ng/ml). The pharmacogenetic analysis highlighted an ABCB1 haplotype with the homozygous presence of ABCB1 rs1128503; ABCB1 rs1045642 and ABCB1 rs2032582 resulting in P-gp protein low function/expression (Salama et al., 2006) that could be compatible with increased drug exposure, further corroborated by the heterozygous presence of ABCG2 421C > A (Morisaki et al., 2005). The patient also presents a SLCO1B1*5/*5 genotype associated with a poor function phenotype (Cpicpgx, 2015).

From our analysis no additional risk factors were present and our suggestion was to monitor treatment at a reduced dose. The advice was followed and palbociclib C_min was found to be 62.3 ng/ml, after a dose reduction to 100 mg/day, in line with the population target C_min. Regardless, the patient developed again up to grade 4 neutropenia. The patient presented a baseline ANC of 3.14 × 10³/mm³.

Key takeaway: The pharmacological evaluation highlighted the presence of three risk factors for the development of neutropenia: 1) overexposure to 125 mg/day palbociclib; 2) baseline ANC median value below 3.60 (×10³/mm³); 3) low function ABCB1 haplotype.

The pharmacological evaluation also highlighted the importance of TDM to exclude a causal link between overexosure and recurrent neutropenia, as this adverse reaction was still observed after reducing palbociclib dose. In fact, at 100 mg/day of palbociclib, systemic exposure was within the C_min target range with persisting toxicity.

Case III was diagnosed with breast cancer when she was 57 and underwent conservative breast surgery, followed by adjuvant chemotherapy and adjuvant endocrine therapy with tamoxifen. Sixteen years later, lung metastases were detected, and first-line endocrine therapy was started with ribociclib (600 mg/day, 3 weeks on/1 week off) and letrozole 2.5 mg/day. The treatment was well tolerated but after 2 months a persistent grade 2 skin rash was observed. After 1 year of intermitting treatment at full dosage with several treatment suspensions (up to 2 months), the patient was referred for pharmacological counselling.

Two samples were collected at steady state at 24 h, and an hour and a half from the last drug intake to evaluate ribociclib and letrozole plasma levels. The samples allowed an accurate assessment of the drug C_min and a hypothetical estimation of ribociclib C_max (reached between 1 and 4 h after drug assumption). C_min resulted to be 1,100 ng/ml for ribociclib and 70.3 ng/ml for letrozole, while the estimated C_max values were 2020 ng/ml and 94.1 ng/ml, respectively. Ribociclib concentration at 24 h after the last dose far exceeded the reported target population C_min of 732 ng/ml (US Food and Drug Administration, 2017a), suggesting a potential role in the development of skin toxicity, while letrozole C_min appeared slightly lower than the target population C_min of 85.6 ng/ml.

Neither the pharmacogenetic analysis, focusing on the search for defective polymorphisms in CYP3A4, CYP3A5, SLCO1B1, ABCB1, and ABCG2, nor the analysis of DDIs explained the patient’s overexposure to ribociclib. Because the drug’s package insert does not suggest a treatment strategy for the occurrence of skin toxicity, treatment with ribociclib was continued by switching from letrozole to fulvestrant and, a few months later, reducing the dose first to 400 mg and then to 200 mg. The patient developed less severe skin reactions afterward. However, after 4 months of treatment with ribociclib at a reduced dosage, a disease progression was reported to CT scan and second-line treatment with capecitabine was initiated.

Within the pharmacological counselling, TDM could have guided an earlier dose reduction to a more tolerated dosage, ensuring an adequate plasma exposure and avoiding an intermittent therapy that could have compromised treatment efficacy.

Key takeaway: The pharmacological evalutation highlighted an overexposure to ribociclib administered according to the standard regimen (600 mg/day) as a possible risk factor for the cutaneous toxicity and ruled out pharmacogenetic variants as a potential cause. TDM-guided early dose reduction to a more tolerated dose could have ensured adequate plasma exposure. Unfortunately, the counselling service was not made available until the patient had been treated for more than a year and recurrent episodes of toxicity had occurred.

Case IV concerns a 38-year-old woman diagnosed with early breast cancer who, after neoadjuvant chemotherapy underwent radical mastectomy and adjuvant endocrine therapy with tamoxifen and a Luteinizing Hormone-Releasing Hormone analogue (LHRHa) for 5 years. Seven years later, after the diagnosis of tumor relapse with nodal metastases, first-line therapy with letrozole and ribociclib was started. The patient was referred to pharmacological counselling after several episodes of persistent skin dryness, rash, and swallowing difficulty, a low-grade allergic reaction probably associated with ribociclib. Two episodes of grade 3 neutropenia were also recorded throughout the treatment despite a baseline ANC of 7.21 × 10³/mm³. In the last 14 months, the patient was treated with ribociclib 600 mg/day (according to 3 weeks on/1 week off schedule) and letrozole 2.5 mg/day.

To evaluate ribociclib and letrozole plasma exposure at steady-state, a blood sample was collected approximately 23.5 hours after the last drug intake. Patient’s C_min resulted in being 1717.6 ng/ml for ribociclib and 181.9 ng/ml for letrozole. These concentrations largely exceeded the target population C_min values reported in the literature (i.e., 732 ng/ml for ribociclib and 85.6 ng/ml for letrozole).

Neither the pharmacogenetic analysis, with no defective polymorphisms in the CYP3A4, CYP2C9, ABCB1 (T allele found in heterozygous form in the three analyzed loci), and ABCG2 genes, nor the DDIs analysis explained the patient overexposure to the drug. The patient also presents a decreased function SLCO1B1 genotype-predicted phenotype (SLCO1B1*1/*5).

A dose reduction could have been considered for that patient with the recommendation to monitor ribociclib plasma levels through TDM analysis to increase the chance of a safer treatment and to ameliorate treatment compliance.

Key takeaway: The pharmacological evaluation highlighted the presence of a risk factor for the development of neutropenia consisting of the overexposure to 600 mg/day ribociclib and ruled out the presence of pharmacogenetic variants as a cause. An early TDM-guided dose reduction could have been better tolerated by the patient and exposure to reduced dosages could have been monitored, but the patient was reactively referred to our service, after several episodes of toxicity.

Case V was diagnosed with luminal MBC with nodal and bone metastases, therefore a first-line therapy with palbociclib (125 mg/day, 3 weeks on/1 week off) and letrozole (2.5 mg/day) was started. After 3 months from initiation, pharmacological counselling was required to monitor therapy because of the underlying neutropenia. A blood sample was taken from the patient for the assessment of the concentration of palbociclib 23 h after the last drug intake (C_min) at steady state. The reported average C_min in patients treated at the standard dose of palbociclib 125 mg/day is 61 ng/ml and 85.6 ng/ml for letrozole. The concentration of palbociclib found in Case V was 36.2 ng/ml, while letrozole C_min was 31.4 ng/ml therefore, both concentrations were lower than the target C_min.

The pharmacogenetic analysis revealed no defective polymorphisms in the CYP3A4, CYP2C9, SLCO1B1, ABCB1 (T allele found in heterozygous form in the three analyzed loci), and ABCG2 genes except for a heterozygous CYP3A5*3/*1 genotype. The resulting CYP3A5 intermediate metabolizer status could potentially be responsible for an accelerated metabolic inactivation of palbociclib which could, in turn result in reduced plasma concentration. The analysis of coadministered drugs revealed no potential DDIs. It was also verified that the drug was taken with food, excluding this additional source of variability for palbociclib capsules (Ruiz-Garcia et al., 2017).

After 5 months, grade 3 protracted neutropenia required a dose reduction of palbociclib from 125 mg/day to 100 mg/day despite the relatively low drug plasma level. Other factors not related to drug exposure could have been the cause of the neutropenia. It should be noted that the patient presented a baseline ANC of 2.39 × 10³/mm³, possibly concurring with toxicity development (Iwata et al., 2021). The patient developed again neutropenia grade 3. No more blood samples were collected for C_min quantification of palbociclib and letrozole. It was hypothesized that hypersensitivity of the patient to the toxic effect of the drug, unrelated to the pharmacokinetic profile of palbociclib, may have been the cause of the neutropenia. A switch to abemaciclib have been considered due to the lower incidence of treatment-associated hematologic toxicity, particularly neutropenia (Sledge et al., 2020).

Key takeaway: The pharmacological evaluation ruled out overexposure as a reason for the observed neutropenia and highlighted the presence of a baseline ANC median value below 3.60 (×10³/mm³) as a risk factor for it. The counselling also identified an underexposure to 125 mg/day palbociclib and the CYP3A5*1 allele as a potential risk factor for this. The dose reduction put the patient at risk for suboptimal exposure to the drug. A different strategy to manage recurrent hematologic toxicity, as switching to a compound less associated with bone marrow suppression, like abemaciclib, could have been considered.