Introduction
It is easy to speculate where a procedure will be in ten years time but the accuracy
of such a prediction is always uncertain. It is however, an important mental exercise
to consider because training and practice patterns need to evolve in parallel with
the procedure and technology. In order to try to predict where EUS will be in ten
years time, one needs to consider the strengths and weaknesses of EUS as well as contemplate
where gastrointestinal endoscopy in general is moving. Finally, one must consider
how competing technologies will evolve over the next decade.
Assessing the strengths of EUS
Some of the strengths of EUS are listed in Table [1]. In my opinion, the greatest current strength of endoscopic ultrasound is its ability
to image and sample lymph nodes. This can be important from a diagnostic perspective
(evaluation of the patient with enlarged mediastinal lymph nodes) and is also important
in staging malignancies. Currently, response to chemoradiotherapy is a relatively
crude science based on inaccurate imaging. In the future, as minimally invasive therapies
to remove primary tumors evolve (an example is endoscopic mucocsal resection - EMR),
the evaluation of lymph nodes will increase in importance. I believe this application
of EUS will continue for the foreseeable future and with more sophisticated analysis
such as molecular markers for micrometastases evolves, EUS FNA of lymph nodes will
gain even greater prominence in clinical practice [1]
[2]
[3]. I do not see any competing technologies for this type of accurate sampling of lymph
nodes for the foreseeable future.
I believe that endoscopic ultrasound will continue to play a role in tissue sampling
in areas that are difficult to reach by CT scan. These areas include the mediastinum,
the uncinate process of the pancreas and the extrahepatic biliary tree. For mediastinal
processes, though I believe endoscopic ultrasound will continue to play a significant
role, I believe that it is probable that this business will be increasingly taken
over by pulmonary physicians. The reason for this is that transesophageal ultrasound
of the mediastinum is a relatively simple and safe procedure and therefore can be
easily adopted by pulmonologists and thoracic surgeons. There are an insufficient
number of gastroenterologists to efficiently manage these cases and the obstacles
inherent in cross-referring (patients with pulmonary diseases being referred to gastroenterologists)
is too substantial.
A significant growth area for endoscopic ultrasound has been in the evaluation of
pancreatic disease. Data continues to be collected suggesting that EUS can accurately
diagnose early chronic pancreatitis [4]
[5]
[6] and because of the ubiquitous nature of CT scan, we are seeing more and more cystic
lesions of the pancreas [7]. Despite the advent of MRCP (even with secretin stimulation), these images do not
match the resolution of EUS. In addition, because difficult pancreatic cases are managed
in specialist centers which increasingly integrate EUS in evaluation algorithms, the
use of EUS for pancreatic disease will be preserved.
Over the next decade, I think the role of endoscopic ultrasound for therapy will (hopefully)
continue to expand. To date, we have not yet seen the emergence of a ”killer” therapeutic
application.
In terms of injection therapy, I suspect that the utilization of EUS guided celiac
neurolysis may well increase. However, this is a palliative therapy for patients with
terminal disease and therefore its overall impact is relatively small. Hopefully,
more studies will be done on EUS guided celiac blocks, however, this procedure is
only of marginal and transient benefit for patients with chronic pancreatitis. I do
not think that this procedure will be modified to the extent that it can have a major
impact on the pain associated with chronic pancreatitis. There is greater promise
in procedures like total pancreatectomy and islet cell reinfusion as the pain syndrome
associated with chronic pancreatitis is too complex to be solved by celiac injection
therapy. To date, EUS guided injection therapy for cancer has been disappointing [8]. Trials are ongoing using TNFerade but to date, the results for EUS guided treatment
are no better than that produced with CT guided therapy. For any type of EUS intervention
to succeed, it must have a clear and measurable benefit over the radiologic approach
because the number of CT scanners and radiologists overwhelm the number of advanced
endosonographers.
A great deal of enthusiasm and anticipation has surrounded the potential of EUS to
guide the creation of digestive anastomosis. This began with Anand Sahai’s work in
using EUS to create a hepaticogastrostomy [9]
[10]. Further efforts followed in performing gastrojejunostomy (animals only) [11]
[12]
[13]
[14] as well as choledochoduodenostomy [15] and pancreaticogastrostomy [16]. Regrettably, these early efforts have progressed very slowly and none of these
procedures are routinely done.
To my view, progress will continue to be very slow for the development of interventional
EUS. In part, this is due to the relatively small number of endosonographers capable
of embarking on these very sophisticated procedures. The second obstacle is that there
is very little incentive by instrument manufacturers to put substantial resources
into device development; the market is too small. While EUS guided therapy remains
a potential strength of endoscopic ultrasound, the current obstacles have prevented this potential
from being fully realized. The success of EUS over the next decade however, will in
part be dependent upon the development of EUS-guided therapies. For the reasons mentioned,
I suspect that it will be a long and arduous road and success will depend on finding
therapies that are uniquely delivered by EUS and therefore do not compete head to
head with CT, MRI or vascular interventional radiology.
A listed strength for endoscopic ultrasound is the fact that it is performed by physicians.
Thus, a high level of patient care is brought to every procedure that is performed.
This is a significant strength over ”automated” procedures performed by technicians
with images interpreted by physicians who have little in depth understanding of the
patient.
Table 1 Strengths and unique assets of EUS
| 1. |
can image small structures - lymph nodes |
| 2. |
can sample difficult to access structure |
| 3. |
produces unique images of the pancreas |
| 4. |
potential for targeted therapy |
| 5. |
performed by physicians |
Weaknesses of endoscopic ultrasound: (Table [2])
The sophistication and complexity of endoscopic ultrasound is a source of both strength
and weakness. While it is performed by physicians, it takes considerable training
to be expert at all aspects of endosonography particularly EUS-guided FNA and interventions.
While physicians performing the procedure likely enhance patient care, it makes the
procedure quite expensive and the dissemination of the procedure is limited by the
availability of adequately trained physicians. This will continue to be a hindrance
in the development of endosonography over the coming decade. Unlike CT and MRI, the
procedure cannot be automated and fully standardized. There is an infinite number
of unique images that can be generated by endosonography and it requires real-time
interpretation. These factors - physician performed examination and inability to automate,
will continue to erode its ability to compete with MRI and CT in the imaging market.
Over the next decade, endosonography will be overwhelmed by the number of multidetector
CT scanners and every improving MRI scanners. It will become easier and easier to
generate high-quality images and because they are generated in a standardized digital
format, outsourcing of the images for interpretation can be accomplished in a less
costly physician environment (such as in India). These factors will cause endoscopic
ultrasound to be a ”boutique” procedure performed under very specific circumstances
based on a finding by CT or MRI which cannot be resolved by these imaging modalities.
We are even seeing lymph node evaluation being encroached upon by PET scanning and
as this improves, the only remaining niche for EUS will be in tissue sampling.
To warrant the cost of a physician driven technology like EUS, it will have to evolve
and mature its capability for intervention. Unique therapies to endosonography which
demonstrably benefit patients will be an area where EUS can sustain its viability.
The key is finding those niches and then developing safe, reliable and relatively
simple procedures.
A continuing inhibition to successful progression of EUS over the next decade will
be our inability to train a majority of gastroenterology fellows within the context
of a three-year training program. Most training programs relegate EUS training to
a fourth year of fellowship yet these fourth year positions have no formal and systematic
mechanism of funding. Training in EUS involves a patchwork of institutions with the
adequate resources and facilities, nurses and trained faculty who are able to provide
sufficient numbers of EUS cases to ensure competency. While our data is incomplete,
it appears likely that 150 or more carefully supervised examinations are required
to achieve competency in both radial and linear endosonography [17]. Under these circumstances, only a minority of the 300 or so fellows in the United
States will be trained in EUS. With the requirement for hands-on training to achieve
competency, this essentially eliminates gastroenterologists already in practice from
fully adopting this technique. It is very difficult to advance a procedure that few
physicians perform. Those physicians without the availability of endosonography, will
argue vigorously that it is not a necessary procedure in patient care. Healthcare
payers, are not very interested in low volume, highly specialized procedures as they
have little impact on overall patient care and costs. These factors will severely
limit the impact of EUS over the coming decade.
The question then becomes how these obstacles can be overcome. In my opinion, there
are several potential solutions:
-
Begin to focus more exclusively on performing EUS with a linear echoendoscope. This
is the instrument that will serve as a basic platform for therapeutic intervention.
-
Encourage and support the development, implementation and integration of simulators
into EUS training programs to shorten the number of hands-on, supervised examinations
required.
-
Strongly encourage all gastroenterology training programs to offer training to the
level of competency in endosonography within a three year curriculum.
-
Greatly simplify and lower the cost of the technology to encourage its wider adoption.
-
For the United States market, and clearly the most difficult obstacle overcome, is
to secure appropriate and fair reimbursement commensurate with the training required
and time and intensity that the procedure requires from the physician.
Table 2 Weaknesses of EUS
| 1. |
performed by physicians |
| 2. |
not automated |
| 3. |
requires considerable training |
| 4. |
significant competition from CT and MRI (and PET?) |
| 5. |
expensive |
Competition
We continue to see rapid develop within the fields of CT and MRI scanning. There is
no reason to believe that this progress will stop. Market factors will continue to
lower the cost and increase the quality of these imaging procedures. Combined with
competition for outsourcing interpretation of the images, it is logical that these
technologies will dominate the field of GI imaging. In light of this, EUS will need
to significantly increase the number of physicians capable of performing EUS and will
need to refine the application of endosonography to those niche areas in which EUS
has unique capabilities which improve patient outcome. Will we be successful? Only
time will tell but success or failure will depend on the younger generation of endosonographers
as many of us (present author included) will likely be put out to pasture around 2016!
