Safety of TKI concurrent with cranial radiotherapy in NSCLC patients; the SATIN platform trial

Non-small cell lung cancer (NSCLC) is one of the major causes of cancer related
mortality worldwide.
Increasingly, new molecular features of NSCLC are being discovered, leading to
an unprecedented growth of targeted agents, such as tyrosine kinase inhibitors
(TKIs). Currently, TKIs are approved for metastasized NSCLC patients with a
driver mutation (EGFR, ALK, ROS1). Approximately 20-35% of these patients are
diagnosed with brain metastasis at initial diagnosis and are often amenable for
initial treatment with a TKI. A relatively high percentage will also develop
new brain metastases or progression of brain metastases during the course of
their disease, often while on TKI treatment.
In patients with brain metastases both whole brain radiotherapy (WBRT) and
stereotactic radiosurgery/stereotactic radiotherapy (SRS/SRT) can be considered
as a local therapy. Neurotoxicity after brain radiotherapy is especially seen
with WBRT, but also SRT. The radio-induced neurocognitive impairment evolves in
a biphasic pattern: a subacute transient decline with a peak at four months,
and a late delayed irreversible impairment several months or years after
completion of WBRT. However, these results were obtained in unselected patients
with often a poor overall prognosis, and it is not clear whether the
deterioration resulted from radiotoxicity or intracranial progression.
Moreover, the brain metastases itself can cause neurological complaints, also
before initiation of radiotherapy. Patients with a driver mutation have a
superior prognosis compared to those without, even in the presence of brain
metastases. It has also been suggested that EGFR-mutated and ALK-translocated
NSCLC cells have a higher radiosensitivity than wildtype NSCLC but it is
unclear what the impact is on neurotoxicity after cranial radiotherapy.
In current guidelines, no advice regarding TKI use during cranial radiotherapy
is given. As the TKI may still be active on extra-cranial sites, clinicians are
faced with the question whether to continue the TKI or not. Especially because
rapid flare of the disease is a known phenomenon after interruption of a TKI.
Preclinical studies suggest that TKIs enhance radiation effects but the effects
on normal tissues are unclear.
In daily practice TKI*s are given concurrent with cranial radiotherapy or they
are discontinued during cranial radiotherapy because of (neuro)toxicity
concerns, depending on the treating physician. When the TKI*s are being
discontinued, they are stopped for approximately one week before, during and
one week after the radiotherapy (i.e. approximately 3 weeks) with the risk of a
systemic disease flare-up.
Advantages of combining TKI with cranial radiotherapy would be a possible
synergistic effect on the brain metastases and the prevention of a systemic
disease flare-up.
As cranial radiotherapy will be indicated for a significant number of these
patients we want to extensively evaluate neurotoxicity in patients with a
driver mutation, treated with concurrent brain radiotherapy and a TKI. Even
though this concurrent treatment is rapidly becoming standard practice,
detailed neurotoxicity data are not available for this patient group.

Primary Objectives:
- To assess whether there is an increase in acute severe toxicity (per TKI
cohort) 2 weeks after completion of cranial radiotherapy measured with CTCAE v
4.0.
- To assess whether there is an increase in neurotoxicity (measured with
neurocognitive testing, per TKI cohort) 4 months after completion of cranial
radiotherapy compared to the baseline measurement.

Secondary Objectives:
- To assess whether there is an increase in neurotoxicity (per TKI cohort) 6
months after completion of cranial radiotherapy in relation to the baseline
measurement.
- To assess whether there is an increase in incidence of severe toxicity (per
TKI cohort) 4 months after completion of cranial radiotherapy, measured with
CTCAE v 4.0, compared to baseline data.
- To assess whether there is an increase in incidence of severe toxicity (per
TKI cohort) 6 months after completion of cranial radiotherapy, measured with
CTCAE v. 4.0, compared to baseline data.

Exploratory:
- To assess what the intracranial PFS is per TKI cohort at 4 and at 6 months
after completion of cranial radiotherapy.
- To assess what the cerebrospinal fluid penetration potential is, measured by
the unbound cerebrospinal fluid-to-plasma ratio (Kp,uu), before and 2 weeks
after cranial radiotherapy and to correlate this ratio with toxicity (CTCAE v
4.0) 2 weeks after cranial radiotherapy, evaluated separately for each TKI
cohort
- To assess the plasma concentration of the TKI before and 2 weeks after
cranial radiotherapy and to correlate this concentration with Kp,uu and
toxicity (CTCAE v 4.0) 2 weeks after cranial radiotherapy (if there is also
consent for liquor sampling), evaluated separately for each TKI cohort

Phase IV observational trial
- Duration: 2 year per TKI
There will be different cohorts, every TKI will be assessed separately.
(erlotinib, gefitinib, afatinib, osimertinib, crizotinib, ceritinib, alectinib,
lorlatinib and brigatinib). For every TKI there is a WBRT (N=10) and a SRT
(N=10) cohort. As osimertinib also became registered for the first line
treatment of EGFR mutated NSCLC (previously only for EGFR T790M) and prognosis
including brain progression rates, is different for these subgroups, two
separate osimertinib groups will be made (osimertinib first line, WBRT [N=10]
and SRT [N=10], and osimertinib beyond first line for T790M+ patients, WBRT
[N=10] and SRT [N=10]. Patients are already treated with a TKI, the TKI is not
part of the study. Patients are already scheduled for WBRT/SRT, this decision
is not influenced by this trial. When a new TKI becomes available a new cohort
will open and an amendment will be presented.
o Duration per patient will be approximately 6 months.
o Normally about 20 patients with a driver mutation are seen in 1 year at 1
centre. Of these patients about 5 will develop brain metastasis and are
eligible for the study. There will be 6 centres that participate.
- Setting: the 6 centres (MUMC, VUmc, NKI/ AVL, UMCG, Erasmus MC, Radboud MC)
that are NVALT acknowledged as specialised driver mutation centers.

The MRI and questionnaires that are used in this study are non-invasive except
for a venapuncture to administer gadolinium-contrast for the MRI (and half of
the MRIs are standard care for these patients). The risks of a MRI-scan are
negligible because it is a magnetic field and does not involve ionizing. The
venapuncture can cause a hematoma. The MRI will be performed twice (once
already as standard care), preferably the same day as regular visits. Time per
MRI is approximately 30 minutes. The neurocognitive testing will take about 60
minutes. This will also be done at the same day as regular visits.
Obtaining CSF by lumbar puncture is optional and is an invasive investigation.
It will take about 10 minutes and will be done by an experienced neurologist.
As a possible complication of the puncture a temporary headache can occur.