In this study there were no objective responses observed, however three patients (19%) maintained SD or non-CR/ non-PD for at least 6 months and two additional patients (13%) achieved a TTP2 to TTP1 ratio of > 1.3. Despite the absence of any objective responses, the median PFS and OS were 3.5 months and 7.5 months respectively. Taken together, these constitute weak evidence for clinical activity. Palbociclib was well-tolerated and AEs reported in this trial were consistent with other studies (27, 31). This study also reports supplementary translational correlates that can improve our understanding of the totality of effects of CDK4/6 inhibition.
A comparison to other studies that employed molecular selection demonstrates similar outcomes as our trial, with a histologically diverse patient population. The TAPUR study which evaluated palbociclib in a range of advanced cancer types, selected on the basis of CDKN2A alterations reported a disease control rate of 31% and 0% at 16 weeks in the non-small cell lung cancer (NSCLC) and pancreatic/biliary cancer cohorts, respectively (54, 55). The median PFS and OS of our trial (14 weeks and 30 weeks respectively) were also consistent with the TAPUR study for the pancreas (7.2 and 12.4 weeks), biliary (7.3 weeks and 11.1 weeks) and NSCLC (8.1 and 21.6 weeks) cohorts. Additionally, the 12-week PFS rate of 56% in our trial is consistent with the phase 2 liposarcoma trial with 12-week PFS rate of 66% (27). However, given the histological diversity and small representation of each tumor type in our study, comparisons according to histotype specific PFS or OS are limited. In the advanced breast cancer setting, the addition of palbociclib to endocrine therapy has demonstrated significant improvements in PFS (32). The rationale for its addition is founded on a notable emergence of cell cycle pathway alterations associated with estrogen resistance (56). An analysis performed within a subset of patients with PIK3CA mutations, indicating an alternative mechanism of endocrine resistance, demonstrated a comparable improvement in PFS (HR 0.45 and 0.48) to their PIK3CA wildtype counterparts, despite the absence of improved objective response rates (57). Taken together, these data support a predominantly cytostatic, rather than cytotoxic effect of palbociclib and highlight the potential importance of examining co-occurring mutations.
CDK4/6 activity serves as the common pathway between several extracellular signaling pathways and the cell cycle (58, 59). Prior studies have demonstrated that co-occurring molecular alterations can potentiate tumor senescence with CDK4/6 inhibition, where subsequent growth signals do not re-instate cell cycle progression (60–65). Important molecular aberrations include MDM2 downregulation, chromatin-remodeling enzyme ATRX redistribution and repression of HRAS transcription (62, 66). In our trial, all six liposarcomas and an osteosarcoma demonstrated MDM2 amplification, with other common alterations consisting of TP53, KRAS and NF1. Of these genes, particular alteration types in MDM2, KRAS and NF1 are potentially also therapeutically actionable (67–70) and should be investigated in future studies of combinatory treatments to promote senescence. Furthermore, mechanisms beyond canonical cyclin-D–CDK4/6 pathway inhibition, involving immune detection and tumor eradication are increasingly recognized (71, 72). Previous studies have shown that CDK4/6 can enhance effector CD8 T-cell activity. Decreased overall TILS in the tumor microenvironment in two of the cases may reflect a reduction in the proliferation of suppressor regulatory T-cells, as has been observed with CDK4/6 inhibition in preclinical models (73). In the limited number of biospecimens available post-palbociclib treatment, we demonstrated a shift in immune parameters and improved disease control with checkpoint inhibitors as subsequent trial treatment. This provides further rationale for evaluating a combinatory approach – in this case, cell cycle inhibition with palbocliclib and immune checkpoint blockade, in future studies.
This study has several strengths. It recruited rapidly, collected meaningful clinical data, and discovered important hypothesis-generating molecular findings in the translational work that could inform the design of future trials. This study also has limitations. Inherent within a basket study design, the patient population consisted of a diverse range of tumor histotypes, with significant reliance on genomic biomarkers to define a common biological driver or pathway. The broad molecular eligibility may have been insufficient to optimize patient selection, and limited its signal seeking capacity. In the advanced breast cancer setting, no molecular features have emerged beyond estrogen receptor positivity (56), and the true predictive nature of CDK4 amplification in liposarcomas is questionable, since 90% of these cancers display this particular molecular alteration (27, 74). In the absence of any objective tumor responses, quantifying benefits based on long-term outcomes using endpoints such as PFS is challenging, given the absence of appropriate historical controls to benchmark across a range of cancer histotypes. Using patients as their own control requires accurate data on TTP1 at study enrolment, which was not always available on this study. Limiting enrolment to those patients having recently progressed radiologically on their penultimate line of therapy would also facilitate interpretation of disease stabilisation.
The findings of our study have several implications. The translational correlates provide first hand insights into the logical development of future concepts. In this particular instance, the signals of immune activation provide further supportive rationale for tumor priming with CDK4/6 inhibition, followed by immune checkpoint blockade to increase the likelihood of an effective anti-tumor immune response (15, 16, 18). In fact, a MoST trial evaluating palbociclib plus avelumab, an immune checkpoint inhibitor is due to open shortly (Clinical trial registration number: ACTRN12620000568910p). Similarly, co-occurring mutations can also provide insights into worthwhile targeted therapy combinations with potential for synergistic effects (75–77). In the presence of histological diversity, evaluating the effectiveness of a targeted therapy in the presence of a particular genomic target requires sufficient patient numbers and can be achieved more efficiently through international collaboration, particularly in rare and less common cancers.
In conclusion, palbociclib monotherapy has limited clinical activity in this study. Its proposed effects on the tumor microenvironment and immune cells is encouraging and warrants further evaluation of combination therapeutic strategies of CDK4/6 and immune checkpoint inhibition.