Thermoradiotherapy for locally Advanced head and Neck Cancer patients - A phase I dose finding study (TANCA-I)

The prognosis of patients with stage III and IV, so-called locally advanced
(LA), head and neck squamous cell carcinoma (HNSCC) is generally poor, with an
overall 5-year survival of only 50% [1, 2]. In the Netherlands, the total
incidence of head and neck cancer is around 3200 per year (Dutch Cancer
Registration). Stage I and II HNSCCs are treated with a single modality
consisting of surgery or radiotherapy (RT), while stage III and IV (LA-)HNSCCs
are preferably treated with combined modality (surgery, RT, and/or
chemotherapy). This is because of the abovementioned low survival rates which
are for a large part dependent on the ability to achieve local control of the
primary tumor and regional lymph nodes, so the regions with macroscopic tumor
[3, 4]. Thermotherapy has the potential to directly target tthe areas at risk
for local failure, because of the focussed targeting of all macroscopic tumor
sites.

In LA-HNSCC, RT is generally applied using the simultaneous integrated boost
(SIB) technique, delivering 54.25Gy of RT dose to regions with high risk to
contain microscopic tumor cells and 70Gy of dose to the regions with
macroscopic tumor. To enhance the effectivity of the radiotherapy treatment,
especially in the macroscopic tumor regions, Cisplatin chemotherapy or the
epidermal growth factor receptor (EFGR) inhibitor Cetuximab are generally added
to the RT in LA-HNSCC [5, 6]. However, in patients above 70 the addition of
systemic therapy to radiotherapy is debatable due to lack of efficacy and
increased toxicity. The additional effect of Cisplatin on survival is not
observed in patients over 70 years and there is an increase from 6% to 15% in
tube feeding dependency after 6 months by Cisplatin chemotherapy[5-7]. Due to
the lack of an effective and/or low toxicity sensitizer, the cancer specific
survival in elderly patients is worse compared to younger patients with a
Hazard ratio of 1.53 (95% CI 1.04-2.25) [8]. Since RT alone is already
resulting in significant toxicity in the elderly patients, a new sensitizer
with a low toxicity profile is needed for these patients.

Thermotherapy, also called (mild) hyperthermia, is the elevation of tissue
temperature to fever-like levels, between 40 and 44 degrees Celsius.
Thermotherapy is proven to be effective as radio sensitizer and is reimbursed
as part of routine clinical care for cervical carcinoma patients [9, 10].
Preclinical evidence suggests that thermotherapy directly kills cells that are
relatively resistant to radio- or chemotherapy. It inhibits the repair of DNA
damage caused by radio- or chemotherapy and improves oxygenation in the tumour
area, thereby enhancing the effect of RT [11].
In a review of 6 studies using thermotherapy as an adjuvant to RT in HNSCC
patients, complete response rate improved from 39,6% to 62,5%, compared to RT
alone [12]. However, these results, although promising, were obtained using so
called superficial thermotherapy techniques, only allowing heating up to
limited depths [12]. We believe these superficial techniques are less suitable
for HNSCC, because tumors and lymph nodes are mostly located deeper into the
tissues.

Therefore, at the department of radiotherapy at Erasmus MC, the HyperCollar3D
was developed. This is a medical device that -for the first time- allows this
so-called deep thermotherapy in the head and neck region, while avoiding
sensitive structures [13, 14]. The device is equipped with advanced treatment
planning software that enables the physician to mark the tumor area and predict
the energy distribution [15].
We have reported on two retrospective cohorts using the HYPERcollar (27
patients) and the next generation device, the HyperCollar3D (22 patients). We
used the devices to apply thermotherapy in a compassionate use indication;
patients having a recurrent or second primary HNSCC following previous RT in
the HN area [16, 17]. In the latter cohort, using the Hypercollar3D, we
observed an unexpected incidence of acute trismus (limited mouth opening)
maximal grade II in 4 out 22 patients[17]. Notably, following the first 3 cases
of trismus in the first 5 patients, the dose rate to the tumor region was
lowered, after which in only 1/17 patients a trismus was observed, suggesting a
dose-effect relationship. Whether occurrence of trismus or other toxicity was
related to the thermotherapy is not known. Since thermotherapy dosage in the
past was empirical and analysis was performed post-hoc, dose effect relations
could be not determined reliably. In addition, it is also unclear whether the
applied dose in these recurrent HNSCC patients, would be the optimal dose in
previously untreated, primary HNSCC patients.

We are now moving away from a *last resort/compassionate use* treatment in the
re-irradiation setting towards the first line treatment of primary HNSCC
patients and therefore we propose to perform a dose finding phase I trial. The
reasons for a dose finding study are 1.) to obtain more evidence of the safety
of thermoradiotherapy in the HN region and 2.) to find the highest dose that is
tolerable and does not lead to unacceptable acute and/or late side effects. The
latter is highly important given the temperature-effect relation of
thermotherapy, as clearly observed in other tumor sites [10, 18].

Until today, it has not been possible to predict the temperatures in the heated
area due to unknown tumor and normal tissue blood perfusion, which varies
between patients and tumors. Invasive measurement of temperatures in head and
neck tumors is notably difficult, due to inability to reach the primary tumor
in the pharynx or larynx and complication risks like bleeding and infection.
Therefore dosage of thermotherapy in this study will not be based on measured
or modelled temperatures, but instead on the delivered energy, with a physical
unit expressed by Joule/kilogram.

The dose levels in our study were selected based on previous data from our
center and others. We took into consideration the dose to achieve a therapeutic
rise in temperature in the target as well as the dose at which toxicity was
observed in the normal tissues([19, 20] and unpublished data Erasmus MC). As
the primary endpoint of this trial is the MTD consisting of feasibility and
DLT, we decided to base our dose escalation on the maximal dose to the normal
tissues. Since we previously observed that the maximal dose to the normal
tissues is linearly correlated to dose to the target (unpublished data from
Erasmus MC), we also escalate the dose to the target in a linear manner, in
this way.

In thermotherapy a target temperature of at least 40 degrees Celsius is
considered therapeutic. Although the dose escalation is based on energy applied
in this study, we do aim to determine the target temperature whenever possible.
Using ultrasound guidance, it is safe to insert a temperature probe in the neck
close to a pathologically enlarged lymph node (=target). To determine if
therapeutic temperatures are reached during the dose escalation, we would like
to obtain an invasive thermometry measurement in one (out of the seven)
thermotherapy fractions. To this end, a separate informed consent will be asked
to patients to insert a neck catheter and measure target temperature in one
thermotherapy fraction. We have chosen for a separate informed consent for
this, because we do not want to hamper inclusion of the primary endpoint of the
trial. In addition, a few patients willing to obtain invasive temperature
measurements would suffice to reach our goal of determining obtained target
temperatures in the dose escalation process.

The primary endpoint of the phase I dose escalation trial will be the
recommended phase II dose (RP2D) of thermoradiotherapy in primary HNSCC
patients. This RP2D is tolerable and does not lead to increased acute and late
effects of the radiotherapy. This trial will be conduc

Primary Objective:
To determine the recommended phase II dose (RP2D) of thermoradiotherapy in
LA-HNSCC patients.

Secondary Objective(s):

• Evaluate the Local control, Disease Free Survival and Overall Survival of
thermoradiotherapy.

• Objective response rate will be evaluated by radiological response on
CT-scans or MRI at 3 months post treatment, with or without histopathological
confirmation of residual disease. These variables will be categorical, data
will be descriptive in nature. The best response will be reported.

• Evaluate the doctor and patient reported outcome measures on acute and late
toxicity after thermoradiotherapy.

• Measure reached target temperatures (in selected patients)

• Evaluate the safety and performance of the HyperCollar3D.

The safety of addition of thermotherapy to radiotherapy will be evaluated in a
phase I dose finding study. The thermotherapy dose will be increased per group
of patients and the tolerability and toxicity will be closely monitored. The
thermotherapy dose levels are described in section 4. The primary endpoint of
this trial is the recommended (phase II) dose (RP2D) of thermotherapy added to
radiotherapy.

The RP2D is defined as the dose level below the dose level for which any of
these events occur:
1. the incidence of dose limiting toxicity (DLT) Trismus is not higher than
baseline
2. any occurrence of severe DLT
3. the infeasibility to apply the dose level is more than 33%
(Feasibility/tolerability rules are described in section 4).

Rationale of DLT definition
In previous cohorts, we observed no increased late toxicity other than trismus
(limited mouth opening) maximal grade II following thermoradiotherapy in 4/22
patients. Therefore, trismus is a potential toxicity to occur that could be
dose limiting. A recent study by the group of prof. Dijkstra (UMCG) reported on
the incidence of trismus at 6 months post treatment in HNSSC patients treated
with radiotherapy [21]. Trismus was defined as a mouth opening of <=35 mm,
which included potentially all CTCAE gradings, but mostly grade I trismus
(personal communication dr. Dijkstra). The incidence of new trismus in patient
having no trismus before treatment, was 28% until 6 months post-treatment [22].

Therefore, the RP2D is considered the dose for which the incidence of the DLT
trismus (defined by a mouth opening of <=35 mm) until 6 months is at most 28%.
The incidence of trismus as defined above is monitored using the TITE-BOIN
design described below.

Dose-escalation design and DLT trismus
A TITE-BOIN design will be used to guide dose escalations regarding the DLT
*trismus* as defined by a mouth opening of <=35 mm until 6 months. A Uniform
non-informative Prior for the occurrence of DLT was used, (i.e., a priori the
current dose is equally likely to be below, equal to, or above the RP2D.The
target toxicity probability, i.e., the toxicity probability of the RP2D, was
specified to be 28%. A cutoff probability to eliminate an overly toxic dose
for safety of 0.95 was specified. This means that a dose will be eliminated if
the posterior probability that the current dose is above the RP2D is more than
95% (this value is the general recommendation in BOIN designs). There are 5
dose levels defined (see paragraph 4).

The first cohort of three patients will be treated at dose level 2 and after
that cohorts of six patients will be treated.

Operating characteristics
The operating characteristics of the design after the first 6 patients were
explored by simulation using the BOIN software developed by at the MD Anderson
Cancer center, version 1.1.0. Six scenarios were explored, each belonging to a
different true RP2D. The simulation was started at dose level 3, and it was
assumed that 24 more patients would be treated, that accrual would follow a
Poisson process with a rate parameter of 1 patient per month. The operating
characteristics show that the design selects the true RP2D, if any, with high
probability and allocates more patients to the dose levels with the DLT rate
closest to the target of 0.28 (table 3 below).


In addition to trismus as objectified by a mouth opening of <=35 mm until 6
months, we define the following toxicities until 6 months as severe DLT:
• Grade III trismus
• Grade IV mucositis
• Grade IV dermatitis
• Grade III or IV osteo- or soft tissue necrosis
• Grade IV burn wound
• Grade IV tumor hemorrhage
• Grade IV laryngeal edema

These severe DLTs are not expected to occur at all:
Therefore, as soon as a single severe DLT (as defined above) is observed, the
dose level will immediately be closed for inclusion of new patients,
irrespective of any escalation rule from the TiTE-BOIN design. Those rules are
only used for trismus as ordinary DLT (i.e., mouth opening of <=35 mm, but not
trismus grade III (trismus resulting in inability to adequately hydrate or
aliment orally).

In this single center study, the local data manager and local investigator will
ensure to keep the study database (e-CRF) up to date, ensuring all patients
enrolled for a certain dose level (maximal 11/12) are update on DLT at the time
of inclusion of a new patient.

General description of investigation treatment

Treatment preparation:
The preparations for the thermotherapy treatment will be performed in the
radiotherapy department. First, an individualized head cushion and nose bridge
fixation are made to be able to fixate the head but leaving as much space as
possible for cooling of the head. Also, an individualized dental mould will be
made, to fix the oral temperature probe to the mucosa of the cheek. Next,
patients will receive an CT scan in this fixed position. Next, the patient is
positioned in the HyperCollar3D to check and record the optimal positioning in
the device. The latter is necessary for optimal treatment planning. The
mouldings, CT scans and preparation at the device are all performed on the same
day but require one additional hospital visit for the patient.

Before each thermotherapy treatment, a thermometry probe will be placed
intra-orally using the dental mould and thermometers will be placed on the skin
of the neck and the cheeks. Patient will be positioned in the HyperCollar3D
applicator (Figure 1 in the study protocol) and the water bolus will be filled
around the neck and cheeks.

During each thermotherapy treatment, the patient will be instructed to mention
any unpleasant sensation, such as a burning, pressure, or pain. By switching
off the power briefly (30 sec), it will be established if the pain was caused
by the radiated power. If this is the case, the treatment settings (phase and
amplitude) and/or applied power will be adjusted. In case of skin complaints,
water cooling bags can be applied. The treatment includes a warm-up time of
approximately 15 minutes in which the energy is slowly increased to the maximal
dose rate. This is followed by a steady state phase of 60 minutes, in which the
energy deposition is kept at a more constant level.
Thermotherapy is applied weekly following the radiotherapy fraction, within
approximately 2 hours.

If during treatment the anatomy of the patient changes significantly due to
tumor regression or weight loss, the initial plan may not be suitable anymore.
In this case an new planning-CT will be made for replanning of the
thermotherapy. This will NOT require an additional visit to the hospital, as
the new CT will be made before or after a radiotherapy fraction.

Upon additional informed consent for invasive thermometry (see paragraph 9.6),
at one or two of the thermotherapy sessions, a catheter will be placed inside
or close to a pathological neck lymph node by a head and neck surgeon or
intervention radiologist under local anaesthesia. This catheter will be used
for invasive thermometry during one or two thermotherapy fractions and will be
removed directly afterwards. Therefore, the catheter will only be in place
while the patient is in the hospital.

Dose escalation and feasibility rules
The primary target of the thermotherapy treatment is the gross tumor volume of
the primary tumor and any pathological lymph nodes, named the *target* from now
onwards. Thermotherapy dose is expressed as the Specific Absorption Rate (SAR),
quantifying the delivered electromagnetic energy to the target and healthy
tissues. To optimize the SAR to the target and minimize the SAR to healthy
tissues, in-silico treatment planning is performed based on the acquired CT
scan in the treatment position. From this thermotherapy treatment plan, the
maximal SAR to the target and the maximum SAR to healthy tissues (also called
SAR hotspot) are important parameters. From previous clinical thermotherapy
plans it was shown that the relation between SAR to target and SAR hotspot are
strongly associated and is linear. The tolerability and likely toxicity are
more related to SAR to healthy tissues compared to SAR to the target. We
therefore choose the SAR hotspot to be the dose parameter to dose escalate our
thermotherapy treatments on.

The intended dose levels are:
1. SAR to normal tissue in the treatment plan= 40 W/kg
2. SAR to normal tissue in the treatment plan= 75 W/kg
3. SAR to normal tissue in the treatment plan= 100 W/kg
4. SAR to normal tissue in the treatment plan= 125 W/kg
5. SAR to normal tissue in the treatment plan= 150 W/kg

Tolerability of the investigational treatment
In case of pain or discomfort during the thermotherapy treatment, optimization
of the thermotherapy treatment plan will take place in real-time/online,
resulting in replacement of the maximal SAR to another place within the normal
tissues. To maintain the SAR to normal tissue, optimization is only allowed if
the delivered energy to normal tissues as a whole is maintained. If maintenance
of the delivered energy to normal tissues is not possible, a maximum decrease
of 20 percent is allowed, for no more than 10 out of 60 minutes during the
steady state heating.
When more than 20 percent power decrease is needed and/or when 10 minutes with
less power (until 20 percent) are exceeded, the fraction is considered not
feasible to apply.
If 2 or more out of 4 thermotherapy fractions in the HYDRA fractionation scheme
(20 fractions = 4 weeks) or 3 or more out of 7 thermotherapy fractions in the
normofraction scheme (35 fractions = 7 weeks) are not applied due to
infeasibility as defined above, that dose level is considered not feasible for
that individual patient.
Thermotherapy fractions that were not applied due to logistical and/or social
reasons and/or medical reasons not related to the study treatment are excluded
for these rules.
If it is not feasible to apply the intended dose in >33% patients per dose
level group (for example 2 out of 3 patients or 3 out of 6 patients or 5 out of
12 patients), the intended dose level is considered intolerable and elimated
(see paragraph 2).
Of note, patients for which the dose level is considered infeasible to apply
are offered to finish any remaining thermotherapy fractions with the previous
dose level that was feasible and did not show DLT. These patients will not be
evaluated for DLT (paragraph 7.5).

Patients with a locally advanced (LA)-HNSCC are at high risk for locoregional
recurrence following primary radiotherapy and therefore have an indication for
radio-sensitisation, mostly done by using Cisplatin. Due to age and
co-morbidity, Cisplatin cannot be applied in all patients. We select these
sensitizer *unfit* patients for our study because of 1.) the need for treatment
intensification, while 2.) no effective radio sensitizer is available.
Potential known acute side effects of thermotherapy in other disease sites are
increased fatigue, local pain, and a small chance of a self-limiting burn
wound. In salvage/re-irradiation HNSCC patients, we previously applied
thermotherapy in a compassionate use setting. We observed no acute or late
toxicities that could potentially be attributed to the thermotherapy, except
for acute grade I-II trismus in 4/22 patients. Going to first line treatment,
we believe a dose finding study is therefore warranted for two reasons: 1.) to
proof the safety of thermoradiotherapy in the head and neck region and 2.)
finding the highest dose that is tolerable and does not lead to increased acute
and late toxicity compared to radiotherapy alone.
The latter is important, as a clear dose/temperature-effect relationship has
been reported for thermotherapy in other tumor sites. Ensuring clinical safety
and knowing the maximal tolerable dose of thermotherapy are both pivotal
factors, considering a potential future phase II trial -not in this protocol-
to further explore the effectivity of thermoradiotherapy in HNSSC patients.

In summary, we will perform a phase I dose finding study of thermoradiotherapy
in HNSCC patients amenable to first line radiotherapy treatment. Patients
undergoing thermoradiotherapy in this study could have a potential benefit, as
they might experience a higher chance of tumor control compared to radiotherapy
alone.