FOOD
AND DRUG ADMINISTRATION
+
+ + + +
OBSTETRICS
AND GYNECOLOGY DEVICES PANEL
+
+ + + +
SEVENTY-FIRST
MEETING
+
+ + + +
Monday,
March 27, 2006
+
+ + + +
The meeting came to order at 10:00 a.m. in the Gaithersburg
Hilton,
PRESENT:
Kenneth Noller, MD Panel Chair
Paula Hillard, M.D. Voting Member
Hugh Miller, M.D. Voting Member
Jonathan Weeks, M.D. Voting Member
Howard Sharp, M.D. Voting Member
Diana Romero, Ph.D. Consumer Representative
Elisabeth George Industry Representative
Scott Emerson, M.D.,
Ph.D. Consultant
Nasser Chegini, Ph.D. Consultant
Keith Isaacson, M.D. Consultant
Nancy Sharts-Hopko,
R.N. Ph.D. Consultant
Russell Snyder, M.D. Consultant
Michael T. Bailey,
Ph.D. Executive Secretary
Nancy C. Brogdon Division Director
| Kenneth Noller, MD | Panel Chair |
| Paula Hillard, M.D. | Voting Member |
| Hugh Miller, M.D. | Voting Member |
| Jonathan Weeks, M.D. | Voting Member |
| Voting Member | |
| Howard Sharp, M.D. | Voting Member |
| Diana Romero, Ph.D. | Consumer Representative |
| Elisabeth George | Industry Representative |
| Scott Emerson, M.D., Ph.D. | Consultant |
| Nasser Chegini, Ph.D. | Consultant |
| Keith Isaacson, M.D. | Consultant |
| Nancy Sharts-Hopko, R.N. Ph.D. | Consultant |
| Russell Snyder, M.D. | Consultant |
| Michael T. Bailey, Ph.D. | Executive Secretary |
| Nancy C. Brogdon | Division Director |
Call to Order
The Chairman called the meeting to order at 10:05 a.m. and recognized the presence of a quorum. The Panel members present introduced themselves by name, expertise, position, and affiliation.
The Executive Secretary, Dr. Bailey, gave the tentative dates for the remaining 2006 meetings: June 5th and 6th, August 28th and 29th, and November 13th and 14th. He then read the deputization statement, signed March 14, 2006, by Daniel Schultz, for the temporary voting members: Russell Snyder, Nancy Sharts-Hopko, Keith Issacson, and Nassar Chegini. A second statement, signed by Jason Brodsky on March 14, appointed Scott Emerson as a temporary voting member.
He read the conflict of interest statement into the record and noted that no COI waivers had been issued for this meeting. Elisabeth George, the non-voting industry representative, is employed by Phillips Medical Systems. He asked all members to recuse themselves from discussions involving products or firms they have a financial interest in and requested that all participants state their financial interests.
Introductory Remarks
Colin Pollard, Chief of the Obstetrics
and Gynecology Devices Branch, noted that Dr. Noller will be leaving the Panel
to become the President of the
He then
turned to the PMA from Innovata for its Adept Adhesion Reduction Solution. The Center has approved three PMAs for
adhesion barrier products with a gynecological indication: Interceed in 1988,
Seprafilm in 1996, and Intergel in 2001.
None of these were approved for laparoscopic use and only Interceed and
Seprafilm remain on the market. Gynecare
removed its product from the market due to adverse event reports.
In January of 2000, the Panel discussed several key study
design issues for adhesion barrier products.
These discussions led to FDA's issuance of a guidance document on the
type of studies FDA expects to see for adhesion barrier products. One key point in the discussion was that,
although each patient gets pelvic surgery for a specific clinical reason, it
should be sufficient for a pre-market primary outcome measure to look at and
properly evaluate the presence of adhesions at second laparoscopy, since this
could be a clinically meaningful outcome.
The Panel recognized the value of downstream clinical outcome measures
such as pain, infertility, and small bowel obstructions, but it did not believe
that those must be the pre-market outcome measures.
In a closed session in May of 2001, the Panel considered the
study design for Innovata’s product.
This discussion led to the protocol employed for the pivotal study. The Panel liked the size of the study, the
blind randomization, and the entry criteria, but it commented that counting
adhesions without considering severity would be insufficient. The panel also did not like the shift table
analysis and commented there must be a significant reduction in adhesions.
The members emphasized the importance of independent video
scorers, recommended looking at the AFS score, and recommended data to collect,
including age and the kind of infertility.
The market claim should be specific to what the study showed.
Working with the Panel and the FDA, the sponsors set a
stringent definition of patient success at the individual level. FDA set a mark for the minimum difference
between the study and control arms and the proportion of patients achieving
success. That was Primary Endpoint
One. Primary Endpoint Two reflects a
change at the subject level in the overall number of adhesion sites. Primary Endpoint Three is a change in the
number of patients with dense adhesions.
In this study, many patients met the definition for individual
success in Endpoint One but did not pass the five percent confidence interval
and were not counted. The study
succeeded for Endpoint Two. In Endpoint
Three, there was no difference between the experimental and control arms. For the study to succeed, it had to succeed
on three separate hypotheses. The Panel
must decide, after taking into account how the study fared biostatistically on
the three hypotheses, whether the clinical findings from this study are of
sufficient merit. The decision must be
based on valid scientific evidence, demonstrated safety, and demonstrated
effectiveness.
Open Public Hearing
The Chairman opened the floor for public comment. Seeing none, he moved to the sponsor presentations.
Sponsor Presentations
Lorna Clisby introduced her presentation team. The sponsor is seeking approval for the Adept Adhesion
Reduction Solution as an adjunct to adhesiolysis in gynecological laparoscopic
surgery.
Dr. Colin Brown, Medical Director of
Innovata, described Adept as a non-viscous, clear fluid made of a four percent
icodextrin glucose polymer solution buffered in an electrolyte solution. It is isotonic to blood. The icodextrin glucose polymer is linked at
the 1,4 position so that it is metabolized by amylase to maltose, then to
glucose, unlike a dextran glucose polymer, which is linked at 1,6. Because there is no amylase in the peritoneal
cavity and due to the size of the molecule, icodextrin remains in the
peritoneal cavity for a prolonged duration and is not metabolized until it is
absorbed into the vascular system by way of the lymphatics. As a result, the solution remains in the peritoneal
cavity for a prolonged duration.
Icodextrin solution was originally
developed as Extraneal, a 7.5 percent solution used for continuous peritoneal
dialysis. It is used in a higher
concentration and frequency than Adept in about the same volume. Extraneal is approved in the
In a study of patients undergoing
intra-peritoneal chemotherapy, the subjects were infused with either two liters
of saline solution or of 4 percent icodextrin solution. After four days, only half the solution was
absorbed, compared to saline solution, which was three-quarters absorbed in one
day. This is important because the first
three days are the critical adhesion formation time.
Adept was approved in
Adhesions can cause pain, infertility, and small bowel
obstructions. Other consequences include
readmission of patients and re-operations.
It is with this in mind that the sponsor began the Adept Pivotal
Study.
Dr. Elizabeth Peers,
director of Clinical Development at Innovata, reported on the trial. The Adept program has been running since
1997. It is the largest study in
adhesion reduction and the only one that has been double-blind. The study was set up to determine the safety
and efficacy of Adept in the reduction of adhesions after gynecological
laparoscopic surgery that included adhesiolysis. The device is used as an intra-operative
irrigate of at least 100 ml every 30 minutes and as a one liter instillate left
in after closure.
The control device was Lactated Ringer’s Solution (LRS), which
looks the same as Adept. The study was
held in 16 centers in the
The primary eligibility criteria considered in the first visit
were that the patient be undergoing laparoscopic peritoneal surgery for a
gynecological procedure which included adhesiolysis and that the patient agreed
to a second-look laparoscopy four to eight weeks later. The most important four exclusion criteria
were that the patients must have at least three adhesions lysed at the time,
must not have an anatomical site removed, and all the sites must be visible;
this was determined in the operating room.
Of 777 women who consented to take part in the trial, 449 were
eligible. Adept was used in 227 patients
and LRS in 222, and this set of patients was studied for the first
endpoint. The second endpoint was
studied in 402 patients, 203 with Adept and 199 with LRS, because 29 patients,
14 LRS and 15 Adept, withdrew and 18 were excluded. The primary reasons for the surgery were well
balanced between the two arms of the study.
When the safety assessment was completed, the number of adverse
events and types of adverse events were very similar between the LRS and Adept
arms for the top ten AEs: post-procedural pain, headache, nausea, leakage at
port-site, dysmenorrheal, constipation, pelvic pain, arthralgia, flatulence,
and vomiting. In many cases where one
device caused an AE more often, the results were not statistically significant. Even in cases such as dysuria or pyrexia,
which were more common in Adept patients, the investigator found that most of
the events were unrelated to the device, so the result was still statistically
insignificant. Vaginal bleeding is more
common among Adept patients to a degree approaching statistical significance,
but was also found to be mild and unrelated.
Vaginal, vulvar, and labial swelling was statistically greater in Adept
users, but it occurred immediately after surgery and went away with no
intervention within a few days. There
were no deaths in the study. The Adept
group had 8 serious adverse events. The
LRS group had 11. Only bladder and bowel
perforation and bleeding from a nicked vessel were not close on both sides, and
those events are not related to the device.
Abdominal pain, pelvic pain, and urinary retention are considered
adverse events related to Adept, though they occur nearly as frequently with
LRS.
The laboratory values at visits 1, 3, and 4 found that there
was no mean difference between the two groups, most patients remained within
reference ranges, and there was no difference in blood or urine glucose
levels.
The safety data shows that the adverse events and serious
adverse events were largely related to the procedure or underlying condition,
not the device. Adept was well
tolerated, though labial swelling was observed in around six percent of the
patients. That may need to be included
in the product labeling. The study is
consistent with the experience in
Dr. Peers then moved to the primary efficacy results. In the adhesion assessments, each of the 23
anatomical sites were assessed and videoed by a trained investigator at the 1st
and 2nd surgery for the incidence, extent, and severity of
adhesion. The consistency of the scoring
was audited by a blinded video reviewer, and he reported good consistency both
among reviewers and for individual reviewers throughout the study.
Efficacy was judged by three primary endpoints. First, the study should show a difference
between Adept and LRS patients’ successes, success being defined as a decrease
in adhesions of at least three sites or 30 percent of sites lysed, whichever is
greater. Second, Adept-treated patients
should not have more sites with adhesions at the 2nd surgery than at
the 1st. The third primary
endpoint is the difference between Adept and LRS in the percentage of patients
having fewer sites with dense adhesions at the second surgery than at the first.
For the first primary endpoint, the confidence interval showed
a 95.2 percent likelihood that the difference between Adept and LRS would be
between 0.7 percent and 18.9 percent.
Indeed, the LRS group showed 35 percent success and Adept showed 45 percent,
which is a statistically significant difference and falls near the middle of
the confidence interval. For the second
primary endpoint, there was a 95.2 percent confidence that the result would lie
between -2.83 and -1.62 for the difference between the first and second
surgeries. The data showed that the
endpoint was met. Adept showed a 23
percent reduction in the mean number of sites with adhesions at the second
surgery. That’s a mean reduction of 2.4
sites. For the third primary endpoint,
50 percent of Adept patients had fewer dense adhesions at the second surgery,
but so did 49 percent of the LRS group.
That is no real difference. Dense
adhesions are especially challenging, and both devices showed a meaningful
result. In sum, the first primary
endpoint did not meet the lower bounds of the confidence interval but had a
statistically significant result. In the
second, there was a statistically significant result that met the confidence
interval requirement. In the third,
there was not a statistically significant result, but half the patients had
fewer dense adhesions.
Dr. Gere diZerga commented that the surprising part of the
study was not how well Adept worked but how well LRS worked. He summarized the available literature on
placing liquid in the pelvis to separate organs and prevent adhesions. The literature contains first and second-look
laparoscopies and are measured by an AFS score, which is a measure of adnexal
adhesions. The literature shows a
significant reduction in adhesions as LRS instillate is increased. Though when both solutions are used at 1
liter, Adept users have fewer adhesions.
The device has two components.
The first is frequent irrigation during surgery to remove the
progenitors of adhesion formation: fibrin, fibrinogen, and blood clots. Adept and LRS had equal benefit in that
situation. In the post-surgery
instillate function, Adept shows fewer adhesions due to its longer
interperitoneal residence. Adhesion
formation begins in the first 36 hours after surgery. A liter of LRS is absorbed in 20 to 30 hours,
so it prevents adhesions through a portion of that time. Adept protects through the entire time.
The secondary endpoints were measured at the second-look
laparoscopy. An odds ratio display shows
Adept is more likely to help a patient for all of the secondary endpoints but
“free of reformed adhesions.” Over 50
percent of Adept patients were free of de novo adhesions, compared to just over
40 percent of LRS patients.
The AFS score, developed by the American Fertility Society,
measures adnexal adhesions to the tube and ovary and the extent of the
adhesion, describing it as dense or filmy and giving a score for the adhesion’s
severity, 0-32. Patients with moderate
to severe scores, 11-32, are unlikely to conceive. Adept patients had a mean reduction in AFS
score of 35 percent while LRS patients had a mean reduction of 15 percent. On the second look, 26 more Adept patients
were classified with minimal or mild AFS scores than on the first look, over
twice as many as in the control. This is
important because a classification of minimal or mild makes a woman much more
likely to conceive.
Of patients who presented to the study with infertility, Over
50 percent of Adept patients showed a reduction in AFS score, compared to 30
percent in LRS patients, a difference of 23 percent. Adept infertility patients showed a 34
percent mean reduction in AFS score, compared to 12 percent in LRS patients. Just over 35 Adept infertility patients had
moderate or severe AFS scores on the first look. On the second look, it was just over 20, a
reduction of 16. LRS patients, starting
with 35, showed a reduction of 5.
On the first primary Endpoint, fifty percent of Adept patients
were successful, compared to 30 percent of LRS patients.
Over two thirds of the patients in the study had
endometriosis. Looking at the number of
sites with treated endometriosis, the rates of success were high, and the
advantage of using Adept becomes more pronounced as the number of sites
increases.
Among patients undergoing adhesiolysis, Adept does not show an
advantage over LRS when the adhesion burden is low. However, Adept shows significantly greater
effectiveness than LRS as the adhesion burden increases.
Although it was difficult to meet some of the primary
endpoints, the endpoints were unique to this trial, and LRS did better than the
sponsor had expected. However, compared
to LRS, Adept has a greater success rate, showing greater adhesion reduction,
reduction in visceral sites with adhesions, reduction in AFS score for
infertility patients. Adept use results
in more patients free of de novo adhesions and more patients with a reduction
in AFS score.
Adept offers many benefits: a significant reduction in
adhesions compared to baseline, significant reduction in dense adhesions
compared to baseline, a reduction in sites with dense adhesions in 50 percent
of patients, and an efficacy that is maintained with an increasing burden of
adhesions and adhesiolysis for all patients.
Adept shows the ability to preserve and improve fertility
potential. In endometriosis patients, it
significantly reduces adhesions compared to baseline and maintains efficacy
with increasing disease. It reduced
pelvic pain in 80 percent of patients with pelvic pain. It has an extensive safety record, so it has
a high benefit/risk ratio.
Michael Kuchinski, the lead reviewer on this application, started the FDA’s presentation and introduced the review team and the other presenters. The IDE Pilot trials began in 1999 under the Classics and Rapids protocols. The IDE for the pivotal trial was submitted and approved April 2001.
The sponsor is seeking an indication that will allow it to be used as an adjunct to good surgical technique for adhesion reductions and for use during gynecological laparoscopy, as an irrigate during surgery and as a post-surgical instillate.
Adept is essentially identical to Extraneal, but with a lower
concentration of icodextrin, 4% versus 7.5%, respectively. Extraneal is used in
a different population, is considered a drug, and was reviewed by
CDER. Adept is a device because of its
principle mode of action: temporarily separating peritoneal tissue surfaces. Adept acts as a colloid, and draws fluid from
the surrounding tissue, retaining the fluid reservoir for up to 96 hours.
The sponsor submitted validation and verification testing on
the device sterility and proposed a shelf life
of, two years, that the FDA found
acceptable. Material safety testing was conducted to include biocompatibility
pursuant to ISO 10993 Biological evaluation of
medical devices. The Animal
testing generally focused on key testing for device safety including: Testing for delay of/or prevention of
healing, Infectivity testing, Reproductive toxicity testing,
Carcinogenesis/metastasis effects (justification) and Pharmacokinetics studies. Our manufacturing review looks at
compliance with design controls and this aspect of the PMA has been determined
acceptable and has been closed.
Baxter Healthcare voluntarily recalled selected lots of Extraneal
due to cloudy dialysate in peritoneal dialysis patients in
Dr. Julia Carey-Corrado presented on the clinical review
issues. She pointed to Extraneal’s
safety record and added that skin rashes are the most common side effect. In the case of the recall, the patients were
asymptomatic. She discussed the two pilot studies. The first was the CLASSIC study, a
prospective, randomized open label, multi-center controlled study. The procedure was laparoscopic gynecological
surgery, and a liter of fluid was used, Adept in 34 patients, LRS in 28. LRS was used as the control device because it
has an off-label use for adhesion prevention and is visually identical to
Adept. However, LRS is not FDA-approved
for adhesion prevention, has not demonstrated effectiveness, and is rapidly
resorbed. Adept patients had an observed
reduction in adhesion number, extent and severity, and there were two adverse
events of labial or vulvar edema. This
study was too small to be conclusive.
The sponsor also conducted a RAPIDS study, which was similar to
the CLASSIC except that there were 25 Adept patients to 12 LRS and there was a
larger volume of test solutions used, since the solution was used as an
irrigant and not removed before the instillate was added. In Adept patients, there was one case of
dyspnea associated with abdominal distension and one case of vulvar edema. There were also complaints of bloating,
distension, and oozing from the incision.
There was an observed reduction in the number, extent and severity in
the Adept subjects, but there weren't any huge differences in the two groups.
There was a closed Panel discussion in 2001, at which the Panel
weighed in on the Primary Endpoints for the study. The Panel said that it was acceptable to look
at fertility post-market, depending on how the pivotal trial worked out. Any labeling claims would have to be tied to
the pivotal trial data.
She reviewed the primary and secondary endpoints for the
pivotal trial and stated that the patients were well balanced for baseline
adhesion assessment, primary diagnosis, and demographics. .
Dr. Xuefeng Li gave a statistical review of PMA for Adept. The study was randomized, double-blind, and
had a randomization ratio of one to one, 227 patients each for Adept and
LRS. The overall study would be
successful in terms of effectiveness if all co-primary hypotheses were met.
There were three primary endpoints with corresponding statistical
hypotheses. The first primary endpoint,
success rate, was defined as a decrease at the second-look laparoscopy of at
least three sites if 10 or fewer sites with adhesions present at the first
look, or decreased by at least 30 percent if more than 10 sites with adhesions
were realized at the first look. The
study hypothesis was that the success rate of the Adept group would be larger
than that of the control group by at least five percent. In statistical terms, this endpoint is deemed
successful if the lower limit of the confidence interval for the difference in
success rates between the Adept and control groups is greater than five
percent.
The second primary endpoint was the number of sites with
adhesions. The corresponding study
hypothesis was that the Adept patients have fewer sites with adhesions at the
second look compared to the first look.
This endpoint is deemed successful if the confidence interval on the
difference between the number of adhesions at the second look from the number
at the first look in the Adept group had an upper limit of less than zero.
The third primary endpoint was the percentage of patients
with fewer sites with dense adhesions at the second look. The study hypothesis for this endpoint is
that the percentage of patients with fewer sites with dense adhesions in the
Adept group is greater than in the control group. The confidence interval for the difference
between Adept and control groups would have to have a lower limit greater than
zero.
The sample size calculation for this superiority trial was
based on the first primary endpoint. The
expected success rates were 40 percent for the Adept group and 25 percent for
the control group. An acceptable
clinical difference of five percent was specified in the protocol. The resulting sample size was 410, and
assuming a loss or fallout rate of 10 percent, the total approved study sample
size was 450. For the pivotal trial, 777
patients were screened. Of the 449
patients who passed the screening test, 227 were randomized to the Adept and
222 were randomized to the control.
Twenty-nine patients withdrew after treatment and 18 patients were
excluded due to protocol deviations, so the protocol population consisted of
203 Adept and 199 control patients. The
two groups had very similar demographic and prognostic characteristics.
For the first primary endpoint, the Adept group had a success
rate of 45.4 percent and the control group had a success rate of 35.6
percent. The difference between groups
is 9.8 percent, and the confidence interval ranges from .7 percent to 18.9
percent. The confidence interval is
above zero, which means that the Adept group had a statistically higher success
rate than the control group. However,
the lower limit of this interval is not greater than five percent; therefore,
the first primary endpoint did not meet the success criterion.
For the second primary endpoint, the Adept group had an average
decrease of 2.22 sites with adhesions.
The confidence interval is below zero, so the second primary endpoint
matched the success criterion. The
control group also had a statistically significant reduction in the number of
sites with adhesions, but the Adept group had a slightly larger reduction than
the control group. Although 158 Adept
patients had fewer sites with adhesions on the second look, 69 had the same or
more. This is compared to 78 in the
control.
For the third primary endpoint, there was about a 50 percent
reduction in the number of patients with dense adhesions for both groups. The difference between the two groups is 1.12
percent, and the P value is .73, which is not statistically significant; hence,
the third primary endpoint did not meet the success criterion.
As a principle of statistics, generally, if the primary
endpoints fail, the secondary endpoints should not be used to show
effectiveness. If it is necessary to
evaluate secondary endpoints, there should be a pre-specified plan to adjust
for the multiple endpoints before any statistical conclusions can be
reached.
Although the study failed, on the success criteria for the
primary endpoints, the results are consistent with the analysis of the primary
endpoints. The Adept group showed an
improvement over the control group in most of the secondary endpoints, though
some P values were less than .05. After
adjustment for multiplicity, only the endpoint “percentage of patients with
reduction in AFS” might be significant.
Ultimately, the study met one of the three primary
endpoints. For the first, the difference
in success rates was not shown to be greater than five percent. In the second, there was a significant
decrease in the number of sites with adhesions over baseline in the Adept
group. In the third, there was no
significant difference in the percentage of patients with fewer sites with
dense adhesions. No firm statistical
conclusion can be drawn from analysis of the secondary endpoints.
Dr. Carey-Corrado gave a clinical review. She presented adverse event data. Within seven days of the first laparoscopy,
the most common adverse events were headache, abdominal pain, disurea, vaginal
bleeding, vulvar edema, vomiting, diarrhea, and fever. LRS and Adept have similar rates, except with
vulvar edema, which is 5.7 percent for Adept and .5 percent for LRS.
There were two readmissions for Adept-related serious adverse
events, labial swelling and pelvic pain.
Two LRS patients were also readmitted for LRS-related serious adverse
events, abdominal pain and decreased urinary output.
She read the proposed indication and the first three discussion
questions to the Panel so they could consider them. The primary endpoints and hypotheses were
challenging, and LRS performed better than anticipated, but there have been no
serious Adept-related safety issues.
Dr. Baoguang Wang reported on the experiences outside the
Data collection was done with a five-page physician data
collection form that collected patient demographics, medical history, surgical
procedures, use of Adept, and surgeon's inspection with handling Adept and
their clinical observations as well as complications and adverse events during
and after surgery. Adverse events data
were also collected post-discharge, but the post-discharge adverse events were
collected on the basis of spontaneous patient self-reporting.
There were 755 adverse events reported, a 16 percent adverse
event rate. The lowest AE rate was found
in the gynecologic laparoscopic patients, where the AE rate was 5.5 percent,
and the top five events were abdominal pain, pyrexia, leakage, abdominal
distension, and urinary retention. The
highest AE rate was found in general surgery laparotomy patients, where there
were ten deaths, six peritonitis cases, and an overall adverse event rate of
28.4 percent. This may serve as a
warning against off-label use in the
Although the registry data show that the AE ratio is relatively
low in gynecologic laparoscopic patients, providing some assurance of safety,
the data was collected on a voluntary basis, and that should be taken into
account, as well as the fact that post-discharge adverse events are self-reported. The registry also does not have race or
ethnicity data.
Post-approval studies are mentioned in panel questions 4 and
6. The objective of a post-approval
study is to evaluate device performance and problems in a broader patient
population and over a longer period of time after a determination of reasonable
safety and effectiveness. They are not
intended to evaluate unresolved issues of safety and effectiveness from the
pre-market phase. Post-approval studies
can gather post-market information on community performance, study the
effectiveness of training programs, find rare adverse events, and monitor
performance in subgroups not represented in the trial.
Questions and Answers
Dr. Cedars asked the sponsors whether there was any difference due the the surgeons’ judging the primary endpoint and whether there was any difference in viscosity intra-abdominally or in the interaction between the substance and blood. Dr. Anthony Luciano, one of the principle investigators, said that in over 50 patients he could not tell the difference.
Dr. Miller asked the FDA about the five
percent lower boundary for the confidence interval and why that boundary was
chosen. He asked the sponsor for any
speculation about why one cohort had more adhesion formation. He asked both about interpreting the safety
of Extraneal relative to Adept. Dr.
Carey-Corrado addressed the first question.
Five percent was chosen because lower bars have been set in the past and
FDA wanted to change that practice. The
number was chosen because it seemed achievable.
Dr. Steven Piantadosi addressed the question for the sponsor. He commented that having a 95 percent
confidence interval along a five percent tolerance is equivalent to having a
99.9 percent confidence interval above zero.
The operational consequence of a five percent rule is to restrict the
Type 1 error for the
primary comparison to 0.1 percent rather than to the usual 2.5 percent that
would be expected from a two-sided five percent rule. The five percent boundary has skewed the
results. Dr. diZerega answered the
question about the bad cohort. There is
nothing linking the population who responds poorly except that they are women
who tend to form adhesions, and this has been a problem since the first laparoscopic
study in 1979. There are predisposing
factors relating to alternations in plasma and activator activity that
predisposes a small population to form more adhesions than the general
population.
Ms. Clisby addressed Dr. Hillard’s request to hear more about
ARIEL. The registry ran between
September 2000 and December 2003 and was a voluntary program. All the data was collected while the patients
were in the hospital, and there were specific data collection forms for
gynecology and general surgery. The
forms just collect data on events and do not check whether or not the events
are Adept-related. It collected
demographics, surgery performed, presenting conditions, and symptoms.
Dr. Peers addressed the Panel’s question about the video audit
procedure and the effect the process had on the outcomes of the study. The study was double-blind. Half of the patients had no reviews of their
videos, and their results were the same as the half who did. Audits did not benefit Adept. Dr. Cedars asked further about the blinding
process. Dr. Peers agreed that there was
no blinding as to which video was first look and which second, since the
procedure would have made that clear.
Dr. Emerson asked about the second primary endpoint, the decrease in
adhesions, and whether the process protected from bias on that endpoint. Dr. Peers said that if people expected fewer
adhesions on the second look, that would affect both arms the same. Dr. Emerson pointed out that the secondary
endpoint was a single-arm comparison.
Dr. Davies said that those who have done adhesions studies before would
be anticipating more adhesions on the second look.
Dr. Peers addressed the Panel question of how well matched the
groups were for analysis of the secondary endpoints. The patients in the larger group and in the
subgroups were approximately the same.
Dr. diZegra addressed the Panel’s question about overlapping
diagnoses. The distribution of the
patients was well balanced. Generally,
the more complex the adhesion-related diseases were, the better Adept did,
compared to LRS.
Dr. Peers addressed the Panel’s question about labial
edema. All of the labial edema started
within two days of surgery, and the majority resolved within three days. They were all resolved and were all
mild. One started 30 days later and was
severe, but it was probably not a related event. Dr. Luciano said that labial edema is related
to the volume of fluid left inside.
Under certain conditions, fluid can sometimes travel into the
labia. The same mechanism can cause
vaginal swelling. The sponsor intends to
include this in the labeling. Dr. Carey-Corrado
asked about patient 637 at Site 13, who was readmitted for inability to void
and labial swelling. Dr. Peers replied
that the inability to void was severe but that everything else was
moderate. The investigator reported the
event as unrelated. Dr. Miller asked
about the frequency of adverse events with Extraneal. Dr. Brown commented that peritoneal dialysis
is more common in men, but men occasionally get scrotal swelling on one
side.
Dr. Martin addressed the question of the irrigant and whether
or not it was aspirated from the peritoneal cavity. The part of the irrigant placed in the
cul-de-sac was aspirated immediately because it was used for cleaning as well
as irrigation. The part that went over
the pelvic brim into the abdomen could not be reached, but the patient was
reversed to bring the solution back so it could be removed before the liter was
instilled. In an experiment with a
two-liter instillation of 4 percent Icodextrin, there is no increase in fluid
over time. However, at 7.5 percent,
there is a dialysis effect. That’s why 4
percent is used, so that there can be a gradual volume decrease in the
peritoneal cavity. At 2 percent, the
decrease is not gradual enough.
Dr. Peers addressed the top 10 most common adverse events
between surgeries. The reporting of
adverse events in both the Adept and the LRS groups were statistically tested
and were not found to be statistically significant.
Dr. Scrimgenour addressed the stratification issue. The randomization was not stratified by any
patient characteristics, but it was stratified by center. The supplies were
sent out in blocks of six.
Dr. diZegera addressed the Panel’s question about data on
inclusion of the irrigant. Previous
studies had shown that frequent irrigation with a balanced solution that was
isomatic and could absorb hydrogen ions would probably reduce adhesions. A study in rabbits using instillate only
showed a benefit with Adept. Both Adept
and LRS showed additional benefits with both the irrigation and instillate
step. Dr. Snyder pointed out that the
data does not show the value of using Adept for both as opposed to using LRS as
an irrigant and Adept as the instillate.
Dr. diZegera said that could be looked at going forward.
Dr. Martin addressed the Panel’s question about labeling and
training necessary for investigators. In
Dr. Luciano addressed the Panel’s question on uterine
manipulations. Due to the number of
adhesions in most patients, uterine manipulators were used nearly universally
in operative laparoscopy.
Dr. Li addressed Dr. Weeks’ question to the FDA regarding the
center effect and how much of it was due to the small sample size. There was a significant center effect for all
of the co-primary endpoints. The FDA
adjusted for the center effect according to the protocol. However, in the sponsor’s analysis of the
difference in proportion and the confidence interval for the difference in
proportion, the results were almost the same, were there adjustment or not.
Dr. diZegera addressed Dr. Chegini’s question about patients
with more than one diagnosis. Patients
with adhesions but no endometriosis had a treatment effect of fifteen
percent. Patients with endometriosis and
adhesions showed a higher treatment effect.
The study did not address infertility.
Icodextrin does not contain glucose, so it does not leave glucose in the
peritoneal cavity.
Dr. diZegeral addressed the Panel’s
question on how the scoring sheet was derived.
The scoring sheet represents everything the sponsor thought would be
useful for understanding an adhesion reduction device with a long
interperitoneal dwell time and clinical benefit. The sheet captured 23 anatomical sites to see
if the reduction was general or site-specific.
The reduction was general. The
sheet then looked at the types, severity, and extent of adhesions as well as
the likelihood of reformation. Dense
adhesions were, by definition, vascular.
Last, the study looked at endometriosis, which two thirds of the
patients had. The idea was to collect as
much information as was possible in a reasonable amount of time. Dr. Isaacson asked about the statistical
difference between the AFS and the modified AFS. Dr. diZegera explained that the AFS was
developed in 1988, and the modified AFS was developed a few years ago. The idea was to use the severity and extent
of adhesion to the ovary and tube. Those
two parameters were then extrapolated to other sites. There was no statistical difference between
the AFS and modified AFS. This relates
to the large number of adhesions removed.
Dr. Romero asked if the questionnaire is filled out interoperatively by
dictation. It is done in that manner,
with a research assistant present to read each of the sites to the investigator
and to write down the investigator’s responses.
Ms. Clinsby addressed the Panel’s question relating to the
labeling and incidence that was quoted, the 28 percent of adverse events in the
ARIEL registry. That rate was in the
general surgery laparotomy and included more surgeries that were not
clean. The proposed labeling for Adept
states that the safety and effectiveness has not been evaluated in clinical
studies in the presence of rank infections in the abdominal pelvic cavity and
that the safety of Adept has not been established after unintentional
enterotomy or bowel perforation.
Dr. Romero asked the FDA why the pregnancy outcomes data was
given, since the sponsor did not touch on the issue. Dr. Carey-Corrado responded that the
pregnancy outcomes were discussed at the 2001 closed Panel meeting as an
obviously meaningful clinical endpoint worth pursuing post market, but the
Panel had agreed that it was not a practical primary endpoint for the pivotal
clinical trial. The FDA was following up
on the Panel’s previous comments.
Dr. Miller asked Dr. Carey-Corrado to revisit the issue of the
5 percent threshold. Mr. Pollard
responded that the Panel would have to decide whether or not the number had
been set too high during the discussion.
Dr. Emerson asked whether the number was due to the surrogacy of the
endpoint and whether it is a pivotal trial.
Mr. Pollard said that there were many factors, including those. The Panel will have to decide whether or not
the results are clinically significant.
Dr. Hillard asked about the use of LRS as a control and that one of the
conclusions seems to be that LRS works pretty well. Mr. Pollard said that was a matter of
discussion.
Dr. Snyder asked about patients with faster absorption rates
and the effect on outcomes. Dr. Brown
said that there was no ethical way to measure the residual volume in a patient
in this study. However, he noted that
there is a lot of variation. Dr. Snyder
said that daily transvaginal ultrasound could have been done to determine at
what point less than 30 ccs of fluid were left.
Dr. Brown agreed that that could have been done and was not. That had been tried at the
Panel Discussion
The Chairman called the
Panel’s attention to the discussion questions.
The first question was, “Although
the statistical hypothesis for only one of the co-primary endpoints was met,
please discuss each of the primary endpoints considering the objective, the
statistical test, and the clinical significance.”
Dr. Issacson stressed the difference between statistical and clinical significance. A 9.8 percent difference between the placebo and product is statistically significant, but it is not clinically significant. Dr. Sharts-Hopko disagreed, saying that the control device worked better than expected, but that should not work against Adept, since LRS was not really a placebo.
Dr. Emerson expressed concern about the need to maintain the
integrity of the experimental process.
He said that the P value does not match the confidence interval,
probably due to strong effects in the center.
As for the 5 percent threshold, he felt that pivotal studies should have
stronger evidence than normal studies.
The question is not so much meeting an arbitrary endpoint but meeting a
meaningful endpoint that has a clinical effect.
Adept shows fewer adhesions, but there is no way of knowing what that
means.
Dr. Snyder said that the control did not turn out to be a
placebo and there is evidence that Adept decreases the number of
adhesions. However, Dr. Cedars said that
without a control or placebo, there is nothing to make certain that the
experiment is valid and to give some sort of parameter to judging
improvement. LRS is already on the
market for irrigation purposes, so Adept should show some degree of superiority. No matter how well LRS did, it is not a study
without a control arm.
Dr Sharts-Hopko asked if the instillation of a 1,000 ccs, of
anything was standard practice. The
Chairman said it is not and that the problem seemed to be that the placebo
worked better than expected in this procedure.
Dr. Emerson replied that the problem was that there was not enough
distance between the efficacy of the placebo and of the experimental arm. Had LRS done better, Adept would have had to
have done better.
Dr. Emerson asked if anyone knew how good a measure AFS score
is. The sponsor used it as a surrogate
to predict fertility, and he wanted to know the validity of that. Dr. Issacson said that there is a correlation
between adhesions and infertility, but it’s not absolute and not linked to a
number that indicates clinical significance.
Dr. Emerson commented that relying on a correlation may be treating
symptoms and not diseases. The Chairman
said that AFS scores have been used for a long time but were not intended as an
evaluation, just a way of counting adhesions.
Dr. Chengi commented on the subjectivity of the measurement, since it
relies on observation and description.
The Chairman reminded the Panel that infertility was not the only
issue. Pain and other symptoms were also
important.
Dr. Issacson said that the definition of success was very
confusing and unclear. Dr. Snyder said
that the product should not be expected to reform established dense
adhesions. Dr. Cedars said that that was
exactly what the product was supposed to do.
The Chairman said that the product is supposed to improve all
adhesions. Dr. Snyder said that part 3
of the first question focuses on fewer dense adhesions between the first
surgery and the second one; however, AFS scoring looks at total adhesions. Dr. Cedars said that the differences in AFS
scores achieved were not significant.
Dr. Sharp said that the data is clinically relevant, since the
product prevents some of the damage done by surgery in making adhesions. If Adept did not go to market, He might use
LRS to prevent adhesions. Dr. Miller added that the product may not meet all of
the benchmarks of success, but there has been a trend toward helping women and
reducing adhesions.
Dr. Weeks said that the data shows a reduction in the total
number of adhesions, but mostly in the group with mild adhesions. Those are the adhesions that don’t make a
clinical difference in terms of pain or endometriosis. While the product moves things in the right
direction, it has not been shown to improve quality of life.
Dr. Hillard said that she could not separate safety from
efficacy. Dr. Isaacson said that the
study was excellent and difficult. The
safety profile lowers the threshold for needing significant clinical
benefit. The study was designed to show
a statistical benefit, not a clinical one.
Dr. Sharp agreed that, in part three of the first question,
dense adhesions are a problem, but it looks like the reduction was 50
percent. While there might not be much
difference compared to placebo, it did seem to benefit the outcome. Dr. Emerson pointed out that if adhesions are
reduced half the time, the other half of the time they are either the same or
worse. He felt that a 50/50 shot doesn’t
mean much.
Dr. Sharts-Hopko pointed out that there was a reduction in
pain. Dr. diZergera said that eighty
percent of patients with pelvic pain who received Adept had a reduction in
pain. Still, the same effect was found
with LRS.
The Chairman turned the discussion to co-primary 2, the number
of sites with adhesions. Dr. Snyder said
that regardless of the known clinical significance, he would prefer fewer
adhesions to more. Dr. Weeks said that
would depend on risk, but he was still uncertain that the reduction in filmy
adhesions had a clinical effect. Dr.
Miller commented that point 2 was the least controversial because the sponsor
hit their target. Dr. Emerson said that
point 2 may be clear, but it may not be relevant in an unblended assessment of
the change. The Chairman summarized that the Panel seemed to believe the results and
does not question the methods, but it is struggling with the issue of clinical
significance.
The second question was:
“Please discuss the statistical and the
clinical significance of the above secondary outcomes. In particular, please focus on the data for
subjects with primary focus on infertility.” It was accompanied by a table. There were a number of secondary
endpoints.
Dr. Emerson wondered whether there was a reason why the
treatment would work better in patients who had adhesions causing
infertility. Dr. Isaacson said that
there’s probably no reason and that he does not see a great difference in the
modified AFS scores. Dr. Cedars said
that people with pain were more likely to have more dense adhesions, while
infertility patients might have less dense adhesions. He said that it is necessary to control for
the multiplicity in the number of outcomes.
Dr. Snyder commented that all of the outcomes were highly correlated,
and that could reduce the necessary adjustment.
He added that that AFS score is used for more than infertility patients,
since it is a quantifiable measure.
Dr. Hillard commented that the infertility group might be
different in that the adhesions are ovarian as opposed to the abdominal wall,
and the mechanism of infertility is related to PID as opposed to
endometriosis. Dr. Chegini agreed that
that seems related to the percentage of patients with reductions. The separation caused by the volume of
solution can prevent adhesions. The
separation also would reduce pain.
Additionally, there are nerve endings in some adhesions, so removing
them would reduce pain.
The Chairman summarized
that the Panel is not thrilled with the secondary endpoints. Some of the endpoints were interesting, but
the primary endpoints were more important.
Question Three dealt
with safety: “Please discuss the safety
data from the pivotal trial and identify any adverse events, including vulvar
edema, you believe may be related to Adept.
Also please discuss whether you believe that the risk posed by Adept is
outweighed by the clinical benefit as discussed under questions 1 and 2 above.”
Dr. Romero asked about the dramatic difference in the rate of
vulvar edema in
The Chairman summarized
that the Panel consensus is that Adept is a safe product.
Question Four said: “Please discuss whether the safety data from
the Ariel registry supports the safe use of Adept as an adhesion prevention
solution.”
The Chairman said that a
positive answer to number four logically flows from a positive answer to number
three. Dr. Emerson said that the data is consistent
with safe use, but it does not necessarily support it, since he was unsure of
the accuracy of a voluntary registry.
Ms. George commented that there is an MDR reporting process in
Question Five was on
labeling and training: “Does the Panel
have any comments on the labeling provided by the sponsor?”
Dr. Isaacson commented on Section One, draft labeling, suggesting that "a significantly greater percentage of patients, 45 percent versus 35 percent" be amended to say, “statistically significant greater.” He also noted that the definition of success under A should be clarified. Dr. Emerson commented that “success” is not a good word to use on the labeling, since it is an editorial comment. It would be better to say that it met the threshold for reduction in lysis. Dr. Sharp pointed to language on page 3 of volume one: "It is to be used for reduction of post-surgical adhesions in patients undergoing gynecologic laparoscopic surgery which may include adhesiolysis." He wanted that language clarified, since the use described is broader than the use in the trial. Dr. Miller commented that the 9 percent reduction in de novo adhesions might make people want to expand the therapies in which the device is used. The Chairman asked the sponsor for comment. Ms. Clisby said that the sponsor was seeking a broad indication for any gynecological surgery laparascopically performed in which it was felt that adhesions might form. Dr. Sharp was concerned about the infection rates in procedures involving a clean contaminated wound. The Chairman agreed that the ARIEL data on general surgery showed contamination to be a problem.
Dr. Snyder said that the device should be indicated for the
population in which it was tested, though off-label use may occur. Dr. Isaacson said that clinical significance
was already a struggle, so he wanted to see it limited to clean cases. Ms. George directed the Panel to the
precautions on page 4 and said that might be the place to address clean
contamination, under “the safety has not been established.”
Dr. Chegini commented that the testing had been on adhesion
reduction, not prevention, so there was no clear basis for using the device in
patients who were not known to form adhesions.
Dr. Weeks said that the patients in the study already had adhesions, and
the device should be limited to use in similar patients. Dr. Cedars disagreed, saying that the
inclusion criteria were too stringent for clinical use. Besides, it was chosen for statistical power
rather than clinical importance. Dr.
Miller suggested changing the labeling to indicate use in laparoscopic surgery
requiring adhesiolysis. Dr. Emerson said
that in the absence of clinical contraindications there was no reason to not
use it for the prevention of adhesions.
Dr. Isaacson said that the problem is that no patients were studied in
many of the surgeries in which one would expect adhesions to form, so it is
impossible to know whether it is safe or not.
Dr. Snyder suggested specifically saying in the data that there is no
data to establish safety in a clean contaminated case.
Dr. Emerson said he still didn’t know whether or not treating
adhesions matters, but preventing adhesions, if possible, is better than
treating them. Dr. Romero suggested that
the sponsor collect post-approval data before seeking to expand the
indication. Dr. Emerson said that
cherrypicking of which events are significant and which are not, as with
vaginal bleeding, is inappropriate. Ms.
George said that much of the data was inappropriate for labeling because it
would be outdated soon. Mr. Pollard said
that such labeling is common in implantable devices.
Dr. Chegini commented that, in table 4, “control” might be
misleading, since a control could be nothing at all and is usually not
something as effective as LRS.
Dr. Cedars wondered why vulvar edema was not in the list of
things that occurred in more than 5 percent of patients below table one. Dr. Snyder suggested that vulvar edema and
all adverse events that occurred in 5 percent or more of the patients be listed
under precautions. Dr. Isaacson noted that on page 9 only the
secondary endpoints that favored Adept were shown, not the ones that showed no
difference or worse.
Dr. Snyder pointed out that there was no evidence that Adept
was a superior irrigant. Dr. Isaacson
agreed but added that where Adept is used as an instillate, it should be used
as an irrigant, since that was the procedure in the study.
Dr. Chegini wanted to see the pain reduction data on page ten
handled in a more balanced way.
Dr. Cedars pointed out that the directions for use addressed
removing packs and sponges, which is appropriate to laparotomy, not
laparoscopy.
Dr. Weeks said that in the less than 1 percent group on page 6:
urinary retention, vulvar edema, vulvar vaginal edema but severe enough to
occasionally require cause urinary retention, should be included.
The Chairman stated that the FDA wanted more input on the
problem about use in patients undergoing gynecologic laparoscopic surgery that
may include adhesiolysis.
Dr. Sharp said it would include tubal sterilization,
exploratory laparoscopy, diagnostic laparoscopy, tubal ligations, and
Oophrectomy. Dr. Sharp said it could
extend to removing an adnexa. When large
portions of the peritoneum are exposed, you are likely to get adhesions.
The
Chairman asked about situations in which the vagina or bladder is open but
adhesions form. Dr. Sharp advocated
limiting it to clean cases. Dr. Weeks
said that the labeling should reflect what has been investigated and that
expansion would occur naturally as clinical experience develops.
Question Six asked “Does the Panel have input regarding any
issues that should be addressed in a post-approval study?”
Dr. Weeks asked about looking at future pregnancies in patients
undergoing laparoscopy for infertility.
Open Public Hearing
and Final Comments
The Chairman called for public comment. There being none, he opened the floor for final comments from the FDA and the sponsor. The FDA had no further comments.
For the sponsor, Dr. diZegera said that the Panel was considering issues the sponsor had spent years working on. The endpoint the sponsor is looking at is adhesions, and the evaluation of those endpoints was done in a blinded fashion. LRS had a statistically significant effect in reducing adhesions, and that made it an unusually active control. However, since LRS is not approved, the question is not whether Adept is better than LRS but whether Adept is better than nothing.
Adept showed many benefits. Fertility potential was preserved, demonstrated by lowered AFS scores. The sponsor’s position is that all adhesions are important, and clinical consequences of adhesions become more apparent as the number of adhesions increase, so it is important that Adept showed benefit throughout all the groups of adhesions. Efficacy was maintained with increase in extent of endometriosis. Pelvic pain was reduced.
The fundamental issue with dense adhesions is subsequent
surgery from the standpoint of causing additional problems on an ongoing basis,
and there's a 50 percent reduction there, which this is the first time that
degree of reduction has been shown.
The safety record is established, and the device is easy to
use. The device fills an unmet medical
need and shows an unusually high benefit to risk ratio.
Panel Deliberations
and Vote
Dr. Bailey read the
Panel recommendations options for Pre-Market Approval Applications. The Chairman called for a motion, and Dr. Sharp moved for approval with
conditions. Dr. Isaacson seconded the
motion.
The first condition, moved by Dr. Sharp, was that the device be
indicated for clean cases and that the language of the labeling reflect that
thus on page 4: "Adept Adhesion Reduction Solution is intended for use as
an adjunct to good surgical technique for the reduction of post surgical
adhesions in patients undergoing gynecologic laparoscopic surgery which
excludes breach of the gastrointestinal tract or vaginal mucosa." Dr. Sharts-Hopko seconded the motion. After discussion of the limitations of the
language, Dr. Sharp retracted the
motion.
Dr. Cedars moved to take out the statement "surgery which
may include," so that it reads "Adept Adhesion Reduction Solution is
intended for use as an adjunct to good surgical technique for the reduction of
post surgical adhesions in patients undergoing gynecologic laparoscopic
adhesiolysis." Dr. Weeks seconded the motion and it carried
unanimously.
Dr. Sharp moved to state under “precautions” that the use of Adept
has not been studied in patients wherein the vaginal epithelium is
breached. Dr. Cedars seconded the motion and it carried unanimously.
Dr. Snyder moved to include in the safety and effectiveness
statement that Adept has not been studied for primary prevention. Dr.
Isaacson seconded the motion and it carried unanimously.
Dr. Snyder moved to amend the precaution section to include the
statement regarding labial edema or swelling, as well as urinary retention. Dr.
Cedars seconded the motion and it carried unanimously.
Dr. Emerson moved to take out the language on the success rate, “decreased by at least three,” and "there was a success rate," and replace it with "first primary efficacy endpoint was defined as the proportion of patients for whom the number of sites with adhesions decreased by at least the larger of three sites, or 30 percent of the number of sites," at the bottom of page 7 of the labeling. Dr. Snyder seconded the motion, and it carried unanimously. However, after further discussion, Dr. Snyder suggested an amendment to Dr. Emerson’s amendment removing the material inside the parentheses on page 8 to clarify the meaning of success and eliminating A underneath. Dr. Emerson accepted the change and added that the language at the top of page 8, “significantly greater percentage of patients” should read “the Adept met this first primary endpoint. There was also discussion about changing the accompanying charts and graphs. The Chairman added the amendment to change the word “control” to LRS throughout the document, which Emerson accepted. When Emerson read back the motion, it read: “On page 7, it would be the first primary efficacy endpoint was defined as ‘the proportion of patients for whom,’ so I deleted the word ‘success rate which was.’ The top of page 8, ‘a significantly greater percentage of patients, 45.4 percent in the Adept group met the first primary endpoint compared to 35 percent.’ Notice at the top of figure 1, we just have to say ‘pivotal study first primary efficacy endpoint (percentage of patients).’ Similarly, the axis can be ‘percent of patients meeting first primary efficacy endpoint.’ Table 3, the title is okay. And then down, ‘success’ is replaced with ‘first primary efficacy endpoint.’ And you could say ‘difference in percent of patients meeting threshold.’ And under A say, ‘the first primary efficacy endpoint was met if the number of sites with adhesions decreased.’" Dr. Miller seconded the changes and the motion carried unanimously.
Dr.
Isaacson moved to, at the bottom of page 9, secondary efficacy, include the
secondary endpoints in which there was no difference between Adept and
LRS. Dr. Weeks seconded and the motion carried unanimously.
Dr. Cedars moved to remove, on page 9, the language in italics:
“removed all packs and sponges.” The
motion was seconded and carried unanimously.
Dr. Sharts-Hopko moved that on page 10 under secondary efficacy
to indicate pain reduction two months post-procedure in the LRS group after the
language “Eighty-three percent of Adept patients.” Dr. Emerson seconded. Dr. Isaacson pointed out that the change may
be covered by her motion. Dr. Weeks said
that the issue was that the statement concerning the 83 percent rate in the
text favored Adept and should be eliminated.
The motion was restated thus: to provide a table of the secondary
efficacy endpoints along with P values comparing the control to treatment arms
with an explicit denotation that it's not adjusted for multiple
comparisons. Dr. Romero added that table
13 would be used to replace the narrative.
The motion carried unanimously.
Dr. Emerson moved to remove the phrase regarding the
attribution of the vaginal bleeding events on page 4 of 10. The specific phrase is the parenthetical:
“The vaginal bleeding effects were not considered to be related to Adept or
control, and none was considered severe.”
Dr. Snyder seconded and the
motion carried nine to one, Dr. Isaacson dissenting.
Dr. Emerson moved that there be a post-market surveillance database for infections following accidental bowel perforation. Dr. Sharp seconded the motion. Dr. Isaacson suggested that the database include intentional bowel anastomosis. Dr. Sharp suggested that the motion be made broad to address infection. After some discussion, including the question of whether Adept was contraindicated in cases of bowel perforation, the motion was withdrawn.
Dr. Snyder moved to require a post-market survey study on
fertility rates following use. Dr. Weeks
seconded. Dr. Romero suggested that a
pregnancy study collect data on births.
Dr. Isaacson suggested following to the point of intrauterine
pregnancy. There was also the concern of
ectopic pregnancy. After discussion of
what to track and the burden to to sponsor as well as the usefulness of the
data, Dr. Snyder withdrew the motion.
Dr. Hillard moved to have the product studied in a group for
the primary prevention of adhesions. Dr.
Emerson seconded the motion. The FDA
commented that that is not the group included in the indication for use. Dr. Emerson added that is would be difficult
to define a group that you’re sure does not have adhesions. It would be a study for a new indication. The Chairman said that the Panel could
suggest the study but that it might not be possible for the FDA to make the
sponsor seek an indication. The FDA said
that a different study for a different indication did not belong in this
discussion, especially since the indication being sought was not yet approved. The Chairman noted that once the indication
sought was approved, the meeting would be over, so there was no way to address
the issue. Dr. Hillard withdrew the motion.
There being no further conditions, the Chairman called for a vote that on the motion that Innovata’s Pre-Market Approval Application number P050011 for the Adept Adhesion Reduction Solution be conditionally approved with the above nine conditions that passed. The motion carried unanimously.
The
Chairman then asked each Panel member to state the reason for his or her
vote. Dr. Cedars stated that the safety
data is reassuring, the data supports efficacy, and that there is no
currently-available system for preventing adhesion formation during
laparoscopy. Dr. Sharp said that the
device is safe and that two of the three co-primary endpoints were
satisfactory. Dr. Hillard cited the
safety, statistical significance, and probable clinical significance. Dr. Chengini expressed reservations for the
efficacy compared to LRS but acknowledged that the device does help some
patients. Dr. Weeks was convinced about
safety, but less about efficacy and voted to support the device because the
indication was limited to adhesiolysis.
Dr. Sharts-Hopko said that the safety data was compelling and that he
anticipates the competition that will emerge.
Dr. Snyder said that it will be years before medicine understands
adhesions, but the trial showed efficacy in decreasing them and he would want
this product to be available. Dr.
Emerson said that clinical effectiveness was unclear due to the activity of the
control, but the reduction was sufficient for approval. Dr. Isaacson said that there’s no way to
predict clinical significance at this point, but he felt compelled to vote for
it because of Point Two. Dr. Miller
echoed everyone’s comments, adding that it is difficult to look at the fluid as
a device. Dr. Romero spoke from the
consumer perspective, citing the reassurance of the safety profile. He hoped that the manufacturer and clinicians
would be responsible in what information they give the patients
regarding realistic expectations. Ms. George concurred with what the members
had said and was glad that the conditions did not require a post-market study,
since the sponsor already has an incentive to do the studies and try to expand
the indications for use.
Adjourn
The day’s agenda completed, the Chairman
adjourned the meeting at 6:08 p.m.
I certify that I attended this meeting of the
Obstetrics and Gynecology Devices Panel
on March 27, 2006, and that these minutes
accurately reflect what transpired.
Michael Bailey, Ph.D.
Executive Secretary
____________________________________
I approve the minutes of this meeting as recorded in this summary.
_____________________________
Kenneth Noller, M.D.
Panel Chair
Summary prepared by
Eric Hendrixson
Neal R. Gross & Co., Inc.
(202) 234-4433