Endoscopic Ultrasonography: From the Origins to Routine EUS
- 1.6k Downloads
Endoscopic Ultrasonography (EUS): The Beginning
“The first report of endoscopic ultrasonography [EUS] to my knowledge is that of DiMagno et al.…although images were only obtained in dogs, this work established the feasibility of EUS. …But in 1980 the potential of this hybrid technology was scarcely apparent to anyone probably including these first endosonographers who did not expand on their demonstration of the feasibility of EUS” .
“The resultant high-resolution scans may improve the ultrasonic diagnosis of cardiac, gastrointestinal and renal diseases. A similar adaptation in other endoscopes may be useful in the investigation of genitourinary and respiratory tracts and other areas” .
“…it should be possible to determine whether or not a disease process is mucosal, intramural or extraluminal…point out the wide potential clinical applicability of intracavity endoscopic ultrasonography (e.g. within the gastrointestinal tract, peritoneum, pancreatic ducts, intravascular spaces etc.). It is likely that specialized ultrasonic probes will be developed to detect small lesions anywhere within the human body” .
Brief history of early development of ultrasound (US) in medicine from sound navigation and ranging (SONAR) in diagnosing valvular heart disease, rectal cancer, and urologic disease.
- Time line (Fig. 1) of the development of EUS leading up to the initial performance of EUS in humans, including pre-EUS transgastric US, preclinical proof-of-concept studies in animals using a prototype EUS instrument to assess transesophageal and transgastric imaging of thoracic and abdominal viscera, including cardiac and non-cardiac tissues, vessel enhancement, and three-dimensional imaging.
Design of EUS instruments.
Later studies of diagnosis of foregut duplication cysts, computer-assisted analysis (neural net or artificial intelligence) programs and the combination of EUS and pancreatic function tests for the diagnosis of pancreatic diseases, and some possible future directions of EUS.
Sonar (Ultrasound) and Echocardiography
Spallanzani (1729–1799) discovered (without publishing) that bats used ultrasound-guided navigation.
Jacques and Pierre Curie  discovered the piezoelectric effect in quartz crystals.
Dussik was the first to apply ultrasound to medical diagnosis .
Other advances of intraluminal echography were transrectal and urological endosonography. Here again Wild and Reid led the way when in 1956 they used a mechanically rotating echoprobe to diagnose a recurrent rectal cancer . Other early reports included transrectal ultrasonography of the prostate and seminal vesicles , rectal cancer , and tumor infiltration of the rectal wall . These instruments were US probes without optical capability. The initial attempt to use intragastric US to distinguish between pancreatic cystic and solid lesions compressing the stomach was published by Lutz and Rosch  in 1976. They placed an ultrasound A-mode probe within the stomach by passing the probe through the accessory channel of a therapeutic TGF-Olympus fiber-optic endoscope. The development of EUS occurred shortly thereafter.
Early History of EUS
Prior to our EUS studies, I investigated the accuracy of the diagnostic tests available at that time, including transcutaneous US, to diagnose pancreatic disease . In this study, we reported that transcutaneous US was ~80 % sensitive and specific . The imperfect performance of transcutaneous US led to the formulation of the hypothesis that placing the US probe within the gastrointestinal tract, closer to the pancreas, would improve the accuracy of diagnosing pancreatic diseases.
Likely as a consequence of this study , Philip S. Green of SRI (formerly Stanford Research Institute International) contacted me regarding my interest in an US endoscope. Philip Green and his team at SRI International, including JL Buxton, DA Wilson, and JR Suarez, developed a system incorporating US into endoscopes. Phillip Green is better known for inventing the Green Telepresence System , later called Mona, now called the da Vinci surgical robot in honor of Leonardo da Vinci, who is credited with inventing the robot.
The development, preclinical, and clinical testing of EUS was supported by the NCI from 1978 to 1981 [Contract CB 74l36, Development of Ultrasonic Endoscopic Probes for Cancer Diagnosis]. The rationale of this contract was that current clinical US was hampered by low resolution due to intervening gas and bone. The underlying hypothesis was that EUS could simultaneously visualize the gastrointestinal lumen and accomplish high-resolution scans of adjacent structures with an aim to improve the accuracy of the diagnosis of pancreatic cancer. My responsibility was to perform the initial animal experiments to evaluate the safety and potential applicability of EUS and assess what modifications of the instrument would be necessary to use EUS in humans. The Mayo team that performed the preclinical and clinical studies included co-investigators PT Regan, RR Hattery, B Rajagopalan, JF Greenleaf, JE Clain, and EM James.
Initial Design of the EUS Instrument and Its Experimental Application
EUS Safety, Collaborations, and Early Results
In 1978, after receipt of the animal instrument (Fig. 2a) and after approval by institutional committees, we began the animal studies and presented our preliminary studies in October 1979 in Chicago at the Central Society of Clinical Research and the Midwestern Section of the American Federation of Medical Research . The results of the initial experiments using the experimental instrument in dogs were published in March 1980 in The Lancet . We reported that the instrument was safe in dogs, provided <1-mm resolution real-time images of the heart, great vessels, spleen kidney, porta hepatis, gall bladder, and gastric mucosa, free of bone and air artifacts. Due to the very long 80-mm rigid tip, however, we were unable to enter the duodenum. Because of the limitations of this initial instrument, a second instrument was developed, which we hypothesized would be suitable for human studies.
Vessel and Tissue Enhancement, Three-Dimensional Imaging, Leading to Human Transesophageal Echocardiography
The clinical instrument was a 13-mm-diameter ACMI FX-8 endviewing endoscope that had the same US system as the animal instrument, but had a shorter 35-mm rigid tip (Fig. 2b). The aims of the human EUS study were to determine the safety of the procedure, determine the normal US characteristics of thoracic and abdominal viscera, and the gastrointestinal mucosa viewed from within the gastrointestinal tract, and to assess the ability to visualize pancreatic and extra-pancreatic lesions such as tumors, metastases, and cysts.
Prior to initiating the clinical EUS studies and approval by the institutional review board (IRB), the human instrument was extensively tested by the Mayo Engineering and electrical safety committees to assure the instrument was safe and did not pose a significant shock hazard. During EUS, we were required to continuously monitor cardiac (ECG) and respiratory function (rate and oxygen saturation). After obtaining a signed informed consent, each participant was given intravenous conscious sedation. Finally, the procedure was videotaped (with audio) and limited to 45-min duration (Fig. 3b).
We (gastroenterologists Drs. EP DiMagno, PT Regan, and JE Clain and radiologist EM James) performed 32 EUS examinations in 15 normal persons (4 studied twice), in 10 patients suspected or known to have pancreatic disease (4 with pancreatic carcinoma [1 cystadenocarcinoma, 1 non-functioning islet cell tumor, and 2 ductal adenocarcinomas] and 6 with chronic pancreatitis [1 studied twice]), and in two patients suspected of having pancreatic disease who had a normal pancreas at examination (1 a gastric ulcer and 1 suspected to have a pancreatic abscess). Two gastroenterologists and the radiologist attended each procedure.
Development of Other EUS Instruments
At approximately the same time we were conducting our experimental and human studies, Professor M. Classen and his group were using a prototype instrument consisting of a 5-MHz transducer mounted on the tip of a side-viewing gastroscope (Olympus GR-D3) . The rigid ultrasonic probe was 3 cm long, but the terminal end including the probe was 8 cm long, the same length as the SRI experimental EUS, much longer than the 3-cm rigid end of the SRI clinical EUS. The US image was an 85-degree sector scan generated by a motor-driven 45° acoustic mirror rotating at 4–8 revolutions/s.
I first met Professor Classen in 1980 when we presented our preliminary human data in Hamburg, Germany, at the International Congress of Endoscopy . The Classen group published their findings of 18 patients with known hepatic, biliary, and pancreatic disorders in September 1980 . In nine patients, they identified the aorta and vena cava, landmarks now recognized as necessary for a successful exam, and confirmed known pancreatic malignancies. Apparently, radiologic guidance was necessary to position the instrument.
This Olympus prototype as well as a Pentax-Siemens echoendoscope, which had a rigid metal tip, a Toshiba-Pentax prototype, and the ACMI-SRI echoendoscopes, which we used, were never produced commercially . The latter 2 echoendoscopes employed a linear array transducer.
More Recent Investigations and Possible Future Directions of EUS
We described the diagnosis of foregut duplication cysts by EUS , and more importantly, we reported that artificial neural network (ANN) analysis of EUS images could differentiate between pancreatic malignancy and pancreatitis . We hypothesized that self-teaching ANN (“neural net” or “artificial intelligence”) programs, which were developed to interpret complex waveforms such as electrocardiograms [39, 40] and US images of benign and malignant breast lesions, might aid in their differentiation [41, 42].
Briefly, we scanned and digitized the internal echo texture of EUS images from X-ray film. The rows of pixels were expressed as gray-scale displays, and by computer analysis, patterns could be distinguished that discriminated malignancy from pancreatitis. According to receiver-operated curve analysis, the diagnostic accuracies of differentiating pancreatitis from pancreatic cancer were similar for the computer ANN, the endosonographer performing the examination, an endosonographer who had no clinical or other information, and the CA 19-9 (80, 85, 83 and 78 %, respectively). Since the differentiation of pancreatic cancer from pancreatitis by the ANN program was similar to an experienced endosonographer, ANN could be useful, particularly to endoscopists with limited EUS experience.
Other approaches aimed at increasing the general applicability and reducing observer variability of EUS, besides improving ANN-based analysis, are to add elastography to EUS for evaluation of lymph nodes [43, 44, 45, 46, 47] and pancreatic lesions [43, 47, 48, 49, 50, 51, 52, 53], to combine ANN with EUS elastography , and to use probe confocal endomicroscopy via EUS guidance [54, 55, 56, 57]. Even more futuristic is the idea of “driverless” EUS and image acquisition (as in driverless car). Another approach is to combine EUS with a pancreatic function test to measure pancreatic enzyme activity and/or bicarbonate concentration in duodenal samples after secretin or cholecystokinin-octapeptide stimulation [58, 59]. These tests are highly predictive of the absence of pancreatic disease, but have poor positive predictive value (false positives) . Thus, this test is widely applicable and valuable to exclude pancreatic diseases but has poor positive predictability, which hampers its utility.
In summary, we have presented a personal view of the early history of EUS. From these early studies, EUS has rapidly become an integral part of the gastroenterologist’s toolbox to diagnose and manage gastrointestinal diseases, particularly of the pancreas and biliary system. In this cohort of patients, EUS is a fundamental tool “because of its ability to provide superior visualization of a difficult anatomical region, but also because of its valuable role as a problem-solving tool and ever-improving ability in an interventional capacity” . In addition, the ever-increasing application of EUS has improved the diagnosis of other gastrointestinal and non-gastrointestinal diseases and provided a delivery system for therapies. EUS has an important role in the evaluation of Barrett’s esophagus, cancer of the esophagus, stomach, and rectum, gastric lymphoma, submucosal lesions, fecal incontinence, perianal disease, lymph nodes, mediastinal adenopathy, and heart and vascular structures . Therapeutic EUS is increasingly used to treat pancreatic diseases, including intratumoral chemotherapy drug delivery [63, 64], EUS-guided endoscopic cystogastrostomy [65, 66, 67], EUS-guided celiac plexus blockade [68, 69, 70], or neurolysis [70, 71]. The future certainly will bring further EUS innovations in scope design, probes, and devices to enhance the diagnosis and treatment of our patients.
We are grateful for the thoughtful review and constructive feedback from Phillip S. Green, and Mayo collaborators Drs. Patrick T. Regan, Jonathan E. Clain, E. Meredith James, James F. Greenleaf, Robert D. Hattery, and James B. Seward.
Compliance with ethical standards
Conflict of interest
The authors have no conflict of interest.
- 1.Sivak MVJ. In: Gress F, Savides T, eds. Endoscopic Ultrasonography. 2nd ed. London: Wiley-Blackwell; 2009.Google Scholar
- 5.Curie J, Curie P. Sur l’electricité polaire dans les cristaus hémièdres à faces inclinées. Compt Rend Séances Acad Sci. 1880;91:383–389.Google Scholar
- 6.Chilowsky CM, Langevin MP. Procédés et appareils pour la production de signaux sous-marins dirigés et pour la localisation à distance d’ obstacles sous-marins. French patent no. 502913; 1916.Google Scholar
- 7.Chilowsky CM, Langevin MP. Production of submarine signals and the location of submarine objects. US patent no. 1471547; 1917.Google Scholar
- 8.Dussik KT. Über die Möglichkeit Hochfrequente Mechanische Schwingungen als Diagnostisches Hilfsmitel zu Verwerten. Z Neurol. 1941;174:153.Google Scholar
- 9.Ludwig GD, Struthers FW. Considerations underlying the use of Ultrasound to detect Gallstones and foreign bodies in tissue. Naval Medical Institute Research Reports, Project no. 004 001, Report No. 4 June; 1949.Google Scholar
- 10.Ludwig GD, Struthers AD. Detecting gallstones with ultrasonic echoes. Electronics. 1950;23:172–178.Google Scholar
- 13.Edler I, Hertz CH. The use of ultrasonic reflectoscope for the continuous record of movements of heart walls. Kungl Fysiogr Sallsk Lund Forhandl. 1954;24:1–19.Google Scholar
- 16.Wild JJ, Reid JM. Diagnostic use of ultrasound. Brit Phys Med. 1956;11:248–264.Google Scholar
- 18.Alzin HH, Kolberger E, Schwaiger R, Alloussi S. Valeur de l’echographie endorectale dans la chirurgie du rectum. Ann Radiol. (Paris) 1983;26:334–336.Google Scholar
- 23.DiMagno EP, Buxton JL, Regan PT, Hattery RR, et al. Canine intra-gastroesophageal ultrasonography: the ultrasonic endoscope. Clin Res. 1979;27:682A (Abstract).Google Scholar
- 24.DiMagno EP, Buxton JL, Regan PT, et al. Experimental endoscopic ultrasonic indocyanine green cholecystography and cholangiopancreatography in dogs. Gastoenterology. 1980;78:1157 (Abstract).Google Scholar
- 25.Rajagopalan B, DiMagno EP, Regan PT, et al. Intraesophageal ultrasonic imaging of the heart. In: Proceedings 9th International Symposium on Acoustical Imaging. Houston, Texas. 1979:41 (Abstract).Google Scholar
- 27.Seward JB, Tajik AL, DiMagno EP. Esophageal phased-array sector echocardiography: an anatomic study. In: Hanrath P, Bleifeld W, Souquet J, eds. Cardiovascular Diagnosis by Ultrasound. Transesophageal Computerized Contrast Doppler Echocardiography. Developments in Cardiovascular Medicine, vol. 22. New York: Springer; 1982:270–279.Google Scholar
- 30.DiMagno EP, Buxton JL, Regan PT, et al. The ultrasonic endoscope: preliminary human studies. Gastroenterology. 1980;78:1157 (Abstract).Google Scholar
- 31.DiMagno EP, Buxton JL, Regan PT, et al. Canine and human intra-esophagogastroduodenal ultrasonography: the ultrasonic endoscope. In: Abstracts IV European Congress of Gastrointestinal Endoscopy June 13–14, 1980; Hamburg:51 (Abstract).Google Scholar
- 32.Buxton JL, DiMagno EP, Regan PT, et al. The ultrasonic endoscopic examination of humans: preliminary experience. In American Institute of Ultrasound in Medicine Annual Meeting September 15–19, 1980; New Orleans, LA (Abstract).Google Scholar
- 33.DiMagno EP, Clain JE, James EM. Human endoscopic ultrasonography: preliminary studies in diseases of the upper gastrointestinal tract. Gastroenterology. 1981;80:1136 (Abstract).Google Scholar
- 34.DiMagno EP, Clain JE, James EM. Human endoscopic ultrasonography: preliminary studies in diseases of the upper gastrointestinal tract. In: Midwest Gut Club Meeting March 7, 1981; Cleveland, Ohio (Abstract).Google Scholar
- 35.DiMagno EP, Silverstein F, Giulani D, Ohmori S. An improved ultrasonic endoscope: preliminary canine experiments. Gastrointestinal Endoscopy. 1982;28:129–130 (Abstract).Google Scholar
- 36.Tio TL. Endoscopic ultrasonography of the esophagus. In: Sivak MV, ed. Sivak Gastroenterologic Endoscopy. 2nd ed. London: W.B. Saunders; 2000.Google Scholar
- 55.Nakai Y, Shinoura S, Ahluwalia A, Tarnawski AS, Chang KJ. In vivo visualization of epidermal growth factor receptor and survivin expression in porcine pancreas using endoscopic ultrasound guided fine needle imaging with confocal laser-induced endomicroscopy. J Physiol Pharmacol. 2012;63:577–580.PubMedGoogle Scholar
- 69.Santosh D, Lakhtakia S, Gupta R, et al. Clinical trial: a randomized trial comparing fluoroscopy guided percutaneous technique vs. endoscopic ultrasound guided technique of coeliac plexus block for treatment of pain in chronic pancreatitis. Aliment Pharmacol Ther. 2009;29:979–984.CrossRefPubMedGoogle Scholar
- 72.Hirschowitz BI, Peters CW, Curtiss LE. Preliminary report on a long fiberscope for examination of stomach and duodenum. Med Bull (Ann Arbor). 1957;23:178–180.Google Scholar