Unexpected toxicity of CDK4/6 inhibitor palbociclib and radiotherapy

Abstract Background Cyclin‐dependent kinase (CDK) 4/6 inhibitors have recently been approved for the treatment of hormone receptor–positive and HER2‐negative metastatic breast cancer in association with endocrine therapy in postmenopausal women. Data on the interaction of CDK4/6 inhibition and radiotherapy are scarce, but some studies show unexpected toxicity. Cases We report three cases of unexpected severe or prolonged soft tissue, skin, and gastrointestinal toxicity in patients treated with a combination of radiotherapy and the CDK4/6 inhibitor palbociclib. Conclusion These cases indicate a possible interaction between radiotherapy and palbociclib. Therefore, we recommend using radiotherapy cautiously when combined with CDK4/6 inhibitors.


| BACKGROUND
Cyclin-dependent kinase (CDK) 4/6 inhibitors like abemaciclib, ribociclib, and palbociclib have recently been approved for the treatment of hormone receptor-positive and HER2-negative metastatic breast cancer in combination with endocrine therapy in postmenopausal women. [1][2][3] Many of these women are candidates for radiotherapy during palbociclib use due to (oligo)progression or local complaints. 4 However, data on the interaction between CDK4/6 inhibition and radiotherapy are scarce.
Two case reports describe a G3 colitis after 10 Â 3 Gy while on palbociclib. 13,14 Two retrospective studies show that the combination is generally safe, but the authors also report a G3 ileitis in one patient (10 Â 3 Gy) and G3 skin toxicity in two patients (radiation dose not mentioned). 15,16 Grade 3 esophagitis is described in two case reports (30 Â 2 Gy and 5 Â 4 Gy), combined with G3 dermatitis in one patient. 17,18 Also, a conference abstract reports unexpected pronounced pulmonary fibrosis and radiation pneumonitis. 19 We present three patients with unexpected toxicity after radiotherapy in combination with palbociclib (Table 1), which indicates a possible severe interaction between these treatments.

| Patient 1
A 64-year-old woman with recently diagnosed bone metastases of a previously treated breast cancer was treated with letrozole and palbociclib. Two months after the start of systemic treatment, palliative radiotherapy (20 Gy in 5 fractions, once daily) was prescribed to symptomatic bone metastases in the right pelvis, using two opposing 10 MV fields. An EQD 2,α/β = 10 Gy (equivalent dose in 2-Gy fractions) of 23 Gy was delivered to bowel loops inside the field. Palbociclib was continued during radiotherapy. Days after the last fraction, our patient experienced a severe enterocolitis, for which she was hospitalized for 10 days. On a diagnostic computed tomography (CT), bowel loops inside the radiotherapy field were clearly swollen ( Figure 1). Treatment consisted of analgesics, antibiotics, and nil per os. Two months after radiotherapy, she was still not fully recovered, with occasional diarrhea and use of fortified drinks.

| Patient 2
A 60-year-old woman with metastatic breast cancer was treated with fulvestrant and palbociclib. At the start of her systemic treatment, she received palliative radiotherapy (8 Gy in a single fraction, field shown in Figure 2A) to a painful bone metastasis in the left hip, resulting in a short-term reduction of pain. Hoping to increase effectiveness, she received a second course of radiotherapy 2 months later (16 Gy in two fractions). Radiotherapy was administered using two opposing fields (10 MV), resulting in a total dose just below the skin of EQD 2,α/β = 3 Gy 32 Gy, increasing to 53 Gy around the femur.  imaging) sequences showed increased signal intensity in the area of the original radiotherapy field, most pronounced on the gadolinium-enhanced T1-weighted SPAIR (spectral-attenuated inversion recovery) MRI ( Figure 2C,D). Six months after radiotherapy, she was using an equivalent of more than 250 mg oral morphine per 24 h and 20 mg of prednisone, resulting in some reduction of swelling but no satisfactory pain relief.

| Patient 3
A 58-year-old woman with metastatic breast cancer was treated with fulvestrant and palbociclib. In the past, she was treated with 50 Gy (tangential fields) + 15 Gy electron boost to the left breast. she received radiotherapy (17 Â 3 Gy, once daily, field shown in Figure 3A,B) to the mediastinum and right hilum, using volumetric modulated arc therapy (VMAT). The V50 α/β = 10 Gy (relative volume receiving an EQD 2,α/β = 10 Gy of at least 50 Gy) was 13.3%, correlating with a G ≥ 2 esophagitis risk of around 25%. 20 Palbociclib was stopped 3-4 days before radiotherapy and restarted 8-9 days after the last fraction. During the radiotherapy treatment, she developed progressive dysphagia with an inadequate response to analgesics and sucralfate. After an initial slight improvement at 2 weeks after radiotherapy, symptoms worsened and did not improve with pantoprazole. An esophagoscopy was performed 3 months after the last radiotherapy, showing an ulcer with a pinpoint stenosis ( Figure 3C). The location of the ulcer corresponded with the high-dose region of the last radiotherapy course. 21 A biopsy was taken and showed squamous epithelium with some inflammation and neither dysplasia nor carcinoma. Palbociclib was then stopped, which resulted in a gradual improvement of her symptoms. A second esophagoscopy was performed 1.5 months later, which only showed a small, superficial ulcer and a decrease of the stenosis ( Figure 3D). Palbociclib was restarted 5.5 months after the last radiotherapy, following a third esophagoscopy, which showed visible improvement ( Figure 3E). Palbociclib was continued for five more months until disease progression, without new worsening of her symptoms.

| DISCUSSION
We present three cases of unexpected severe or prolonged soft tissue, skin, and gastrointestinal toxicity that occurred in patients receiving palbociclib and radiotherapy. We did not discover alternative explanations for the observed toxicities. Dose delivery during treatment was verified using electronic portal image dosimetry and was delivered as planned. Side effects were only present inside the radiotherapy field and are extremely uncommon for equivalent doses to bowel and soft tissue/skin in the first two cases. In the third case, dysphagia was not unexpected (especially not after previous radiotherapy), but the recovery pattern was unusual. An initial recovery was followed by worsening after restart of palbociclib, and improvement was seen only after discontinuation of palbociclib. This emphasizes the possible negative role of palbociclib during repopulation of healthy tissues after radiotherapy.
A pre-existing increased intrinsic sensitivity to ionizing radiation is highly unlikely in these patients, since they had radiotherapy for their primary breast cancer without serious unexpected toxicity, and there were no known risk factors for increased radiotherapy toxicity.
Additionally, there was no reason to assume different pharmacodynamics in these patients.
It is worth noting that the patients in this case series received hypofractionated radiotherapy, which may lead to a less beneficial therapeutic ratio with regard to normal tissue complications, but this is particularly the case for late-responding tissues. 22,23 Furthermore, hypofractionated radiotherapy is often applied showing several cases of G3 skin and gastrointestinal toxicity. [13][14][15][16][17][18] However, the nature and limited size of these studies, including our study, make it difficult to reliably predict the absolute risk of increased toxicity.

| Possible (radio)biological mechanism
Palbociclib is a CDK4/6 inhibitor that inhibits cell division by causing a G1/S block. 24 While this is advantageous for the elimination of fastdividing cancer cells, cell division is pivotal to repair/repopulate normal tissues after radiotherapy. Inhibition could therefore lead to increased and prolonged normal tissue damage. Although the main mechanism of palbociclib is cell cycle inhibition, the possibility of radiosensitization by an off-target effect cannot be excluded. A report

AUTHOR CONTRIBUTIONS
All authors had full access to the data in the study and take responsi-

ETHICAL STATEMENT
All patients agreed to anonymized publication of their data. As this is a case series, no institutional approval was required. All authors contributed to the collection of the data, the analysis of the data, and the writing of this manuscript. All authors had full access to the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

DATA AVAILABILITY STATEMENT
Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.