Randomized controlled trial of the effect of larvae and larval secretions on wound healing in diabetic ulcers

Larval therapy is widely used today for the treatment of acute and chronic
wounds in patients. Annually, more than 15.000 patients receive larval therapy
in Europe. The exact mechanisms of action of larvae in wound treatment are only
recently becoming better understood. The US Food and Drug Administration
registered maggot debridement therapy (510(k) #33391 as a wound treatment
method in 2004. The Inspectie voor de Gezondheidszorg (IGZ) in the Netherlands
approved the larva Lucilia sericata as a medicine in 2014 (September 1st).
Three randomized clinical trials (RCTs) have shown the debridement potential of
larvae, while other beneficial effects of larvae on wounds, including
anti-infection, immunomodulation, angiogenesis, and tissue remodeling and
regeneration, have been widely reported clinically and are supported by
numerous in vitro studies. One (small) RCT showed a lower incidence of wound
infections in vivo during larval therapy, compared to hydrogel application. The
RCT from Opletalova et al., that focused on debridement, showed a faster wound
bed preparation during the first week with larval therapy compared to surgical
debridement, but there was no difference in percentage of slough after 15 days.
The other two clinical studies compared larval therapy with hydrogel
application and both significantly reduced the time to debridement with larval
therapy. The VenUS II trial from Dumville et al. demonstrated that wounds were
debrided within 14 days using larvae versus 72 days using hydrogel application.
This trial was the only clinical study that investigated the time to healing
and bacterial load during larval therapy, and could not show improved wound
healing rate or bacterial reduction in the wound during larval application.
While debridement, according to the current knowledge of the wound healing
process, is essential to progress from the inflammatory phase to the
proliferative phase, an increased wound healing rate was expected. Maybe, the
discrepant result can be explained by the unclear inclusion criteria of the
trial; e.g. different sizes of wound areas and percentages of slough were
included, which finally resulted in a median larger ulcer area in the larval
group comparing to the hydrogel group, and a quantity of debris that varied
from 26 up to 100%. Patients with immune-related diseases and malignancies were
not excluded from the trial, however these underlying diseases could
significantly interfere with the process of wound healing. To conclude, further
research, especially a (randomized) clinical trial, is needed.

In this study, the following questions will be answered:

1. Does larval therapy accelerate wound healing?
2. Do single larval secretions accelerate wound healing?
3. Is the bacterial load in the wound reduced during larval therapy or is the
microbiome changed?
4. Are there systemic immunological and/or anti-inflammatory effects of local
larval application or larval secretion application?
5. Does moisture balance affect the time to wound healing?
6. Is there a difference in reappearance of debris between the three groups?
7. Do patients have adverse effects of the therapy, including pain, crawling
sensations, bleeding, allergic reactions etc.?
8. What is the cost-effectiveness comparing the three therapies?

We hypothesize that larvae and/or their excretions/secretions (ES) accelerate
wound healing (this study does not focus on debridement as primary outcome). In
several trials no difference in (debriding) effect was shown between free
larvae and larvae, captured in small, permeable bags, so-called Biobags, that
allow free exchange of larval ES. Therefore we choose to test only the bagged
larvae.

We would like to compare three groups: larvae in Biobags (1), larval secretions
gel (under GMP controlled conditions produced) (2) and a control group (3).

This protocol is written, following the guidelines for design and conduct of
clinical trials in wound care, as described by Eskes et al. in the journal of
Wound Repair and Regeneration, 2012.

For references, see above in the paragraph 'Background of the study'.

1. Trial setting:
The trial takes place in two general hospitals in the Hague, the Netherlands,
that have recently merged: The Medical Center Haaglanden (MCH) and the Bronovo
Hospital. In the future, probably the Leiden University Medical Center in
Leiden, the Rijnland hospital in Leiderdorp and the Isala Klinieken in Zwolle
will participate.

2. Patients: see paragraphs below.

3. Interventions: see paragraphs below.

4. Outcomes: see paragraphs below.

5. Sample size and statistics
It was determined that 76 participants per group are needed to detect a
reduction in median healing time from 16 weeks to 10 weeks (i.e. a constant
hazard ratio of 1,6 between two groups) with 80% power and a two sided
significance level of 5%. In literature, the median healing varies between 12
and 34 weeks (13 weeks, Warriner et al, 2011; 21-34 weeks, Pickwell et al,
2013; 12-16 weeks, Gul et al, 2006; 14 weeks, Wilarusmee et al, 2014). We think
that in our trial a reduction from 16 weeks to 10 weeks will be a realistic and
clinical relevant difference to measure.

The calculation above assumed an exponential distribution of time to healing
and an assessment of equality of survival curves using the exponential maximum
likelihood test, assuming an accrual period of 156 weeks (3 years) and a
maximum follow-up time of 168 weeks (so follow up ranging from 12 up to 168
weeks for the last resp. first patient included), and no dropouts.

The estimated percentage of patients with completely healed wound after 1 year
is 72-90%; Jeffcoate et al, 2006, Treece et al, 2004. The assumption above
corresponds to an expected percentage of 89% healed wounds after one year in
one group and 97% healed wounds in another group with a positive effect of the
treatment.

The time to healing between the three groups is investigated using survival
analysis (kaplan meier curves and log rank test). Cox analysis is performed to
correct for wound size, duration of ulcer and ulcer type (stratifying factors
during randomization). For the patients with wounds that won*t heal and have an
amputation, the time to wound healing will be censored at the moment of
amputation. The estimated percentage of patients in these groups requiring
amputation is less than 9% (Orneholm et al. 2015; in this publication 9% minor
and major amputations had to be performed in a group of 770 diabetic patients,
however 25% had major peripheral vascular disease which will be excluded in our
trial).
The patients whose therapy has been changed after 4 weeks will be analyzed in
the group in which they are allocated, in conformity with the
intention-to-treat principle. The follow up of these patients will continue
until the wound has been closed.

A secondary analysis will be performed per protocol. Details of the per
protocol analysis will be specified in the statistical analysis plan. In
another secondary analysis, the groups will be compared corrected for the
baseline HbA1c values.

Baseline characteristics of patients and wounds in the three different groups
will be reported using descriptive statistics. These characteristics include:
mean age, sex, mean size of ulcer, median duration of ulcer, toe-brachial
index, smoking, BMI, antibiotic therapy, bacterial colonization. The outcomes
of wound cultures will also be described.

The repeated measurements of inflammatory markers (e.g. cytokines, CRP,
complement factors etc.) will be analyzed with linear mixed models including a
time x treatment interaction effect, using transformations where appropriate.

6. Randomization
We will randomize all patients in a stratified way to prevent large differences
in wound characterics between the groups, such as wound surface, duration of
ulcer and ulcer type (Wagner I-III, Texas AI-III and BI-III). Each patient will
be assigned to a group using randomization software, especially developed for
clinical trials (e.g. on the website: https://www.tenalea.com/nkiavl/ALEA). The
caregiver will randomize the patient and start the therapy. Which therapy will
be started is registered in the patient record.

7. Blinding
The outcome assessor is blinded and will only assess the results of the
photographs and the measurements without knowing which therapy was performed.
Caregivers cannot be blinded. Patients are not blinded. Although coping style
can possibly affect wound healing, psychological factors such as anxiety or
depression do not influence the time to complete healing of chronic wounds
(Vedhara et al. 2010, Diabetologia).

8. Intention-to-treat
All randomized patients are analyzed in the group to which they are allocated.

9. Funding
This study is financially supported by Research Foundation Bronovo, The Hague,
The Netherlands. Laboratory facilities are provided by the Leiden University
Medical Center, Leiden, The Netherlands.

Sterile larvae (Lucilia sericata) and larval secretions gel are a generous
gift from Biomonde, GmBH, Barsbuttel, Germany.

10. Follow-up
All patients will be followed until the wound has been closed, with a maximal
follow-up of 168 weeks. When there is no effect of the treatment after 4 weeks,
the therapy may be changed by the caregiver.

11. Ethics
The ethic committee has to evaluate this protocol.

12. Final remarks
For future trials, we would like to investigate dose-reponse relations of
larvae and larval secretions, and different ways of application and their
effect on wound healing. Of course, we need to analyze the results of this
trial before any other study can be initiated.

3. Interventions:

All patients that are included will have surgical debridement of the wound at
the start of the study, preferably under local anaesthesia (Lidocaine 1%).
Surgical debridement is performed again during policlinical visits if slough
reappears after initial debridement. After initial debridement, patients are
randomized and the fully debrided wound will receive one of the three therapies
described below.

- Application of larvae in biobags. Larvae will be changed twice a week at the
wound care unit (outpatient department).
o 5-8 larvae/cm2 wound surface, which is the adviced dose in literature and by
the manufacturer*s guide. This is the dose which was also used in the other
RCT*s and it was effective for debridement of the wounds. If a patient is not
present at his or her appointment, there is no risk that the larvae will
puppate and become flies in the Biobags. Larvae need a dry environment and
enough oxygen to puppate and develop, which they do not have in the bags. They
won*t survive.

- Application of larval secretions in (hypromellose) gel once a day. Patients
receive a tube with larval secretions gel 0.05% (this is not the definitive
dosage, first a pilot study will be performed to test the optimal dosage, see
also C1 protocol, July 2015, version 2), which can be stored at room
temperature for a month. Previous and recent studies have shown that larval
secretions are temperature tolerant. Once a day, a thin layer of the gel is
applied on the wound by the patient or the wound nurse at home. The wound is
covered by a wound film dressing. Patients come back at the policlinical
outpatient department every one or two weeks for wound inspection. If the wound
of a person has almost healed, the controls will be twice a week to prevent
bias in the primary outcome (time to wound healing).
o Larval secretion gel, comparable with 5-8 larvae/cm2 wound surface/24 hours),
produced under GMP controlled conditions by Biomonde, Cardiff, UK. More
information about the dosage can be found in the IB and IMPD of this RCT.

- Control: standard care. Dry wounds are treated by hydrogel and a hydrocolloid
dressing. On more exudative wounds alginate is applied. Wounds will be
inspected at the wound care unit every one or two weeks. If the wound of a
person has almost healed, the controls will be twice a week to prevent bias in
the primary outcome (time to wound healing).

If a patient has multiple wounds, all wounds receive the same therapy. The
largest wound of this patient will be included in the study for analysis.
Surgical debridement is performed if there is reappearance of slough (after
initial debridement) and will be scored in each group. Other basic wound care,
that will be the same for these 3 groups, includes compression therapy if there
is peripheral edema (arterial blood supply is not compromised in these groups
with a minimal toe-brachial index of 0.7). All wounds will be finally covered
with an absorbing bandage and a hydrophilic bandage for fixation.

All patients are treated until the wound has closed, with a maximal follow-up
of 168 weeks. When there is no effect of the treatment after 4 weeks, the
therapy may be changed by the caregiver. In acute settings, e.g. if a patient
develops a systemic infection or an acute imflammation, of course the therapy
can/must be changed if necessary.

There are no severe side-effects known for any of these therapies. Patients who
receive larvae on their wounds complain sometimes of pain, but pain killers,
such as paracetamol, are enough to reduce the complaints. Allergies for larvae
are not known, but could happen. Allergies for hydrogel or alginates are
possible as well. In the case of an allergy, the therapy will be discontinued
immediately.

The persons in the study will have four extra venous punctures: one before
start of the therapy, one after the first and third week of the initial
treatment and the last puncture one week after ending the therapy. At the same
time of the puncture a wet gauze is applied on the wound for five minutes to
obtain wound fluid for laboratory research. Several times photographs of the
wounds will be taken. Once, the person will be asked to fill in a
questionnaire. Finally, for the group which receives live larvae, it is
necessary to come to the policlinical department for a change of the biobag
with larvae twice a week.