Introduction
The working group for abdominal and gastrointestinal diagnosis (www.ag-gastro.drg.de) is a group of the German Radiological Society (DRG) that focuses on one of the main core areas of radiology. The working group is a representative of German abdominal radiology and a comprehensive platform for interested researchers and clinicians with a focus on radiological abdominal diagnosis and therapy.
In addition to active work on guidelines that is impressively supported by all members of the working group and the board in recent years and currently, the working group would also like to promote and support advanced training in abdominal radiology in Germany.
For this reason we tried to make the expertise of our active members regarding research and clinical practice in abdominal radiology available to the members of the German Radiological Society. Therefore, a current overview of the literature with a focus on the developments and trends of gastrointestinal and abdominal radiology in Germany in recent years was created in the following publication. To improve clarity, we tried to structure the topic of gastrointestinal and abdominal diagnosis with respect to content and topography. Therefore, we will focus on hollow gastrointestinal organs in the first part of this overview. The parenchymal abdominal organs will be discussed separately in another publication. Our intention was to represent and discuss the issues and radiological problems that have been increasingly researched in recent years in the individual organ and hollow organ regions and review the relevant international literature. The important German publications subjectively selected by the authors are then presented and discussed. The paper is structured anatomically from cranial to caudal. In particular, the diagnosis of the esophagus and stomach based on current guidelines is discussed. This is followed by an overview of small intestine diagnosis with subclassification of the issues and studies according to the individual working groups. In colorectal diagnosis, in addition to CT and MRI colonoscopy, primarily MRI-based evaluation of rectal cancer is discussed. As in the case of the small intestine, this topic is also analyzed here in the context of the updated new German guidelines. In addition to the hollow gastrointestinal organs, mesenteric bleeding and ischemia are discussed. The overview concludes with additional varia regarding gastrointestinal radiology with a focus on graft versus host disease and the current assessment of conventional radiography in the context of modern cross-sectional imaging of the abdomen.
The individual chapters regarding the particular organ and intestinal regions were written by the experts and the board of the working group for abdominal and gastrointestinal diagnosis and were edited, supplemented, and compiled in consensus under the supervision of the chairmen of the working group.
Small intestine and chronic inflammatory bowel diseases
The main focus of radiological imaging of the small intestine is the diagnosis of rare tumors of the small intestine as well as the diagnosis of inflammatory changes in chronic inflammatory bowel diseases. The majority of scientific publications on the topic of small intestine diagnosis regarding imaging modalities are based on the recommendations of the updated German S3 guidelines for Crohn's disease [45]
[46]. In principle, high-resolution ultrasound of the small intestine and MRI examination of the small intestine such as enterography or enteroclysis are postulated as the primary comprehensive small intestine visualization methods. Therefore, many publications and German working groups address the selection of the optimal radiological modality for detecting inflammation in chronic inflammatory bowel diseases. Other important issues in the international literature in recent years include the possible differentiation of inflammatory and fibrous changes in affected intestinal segments.
In the Regensburg working group of Schreyer, magnetic resonance enterography (MRE) with and without biphasic contrast enema was compared with conventional ileocolonoscopy in patients with Crohn's disease [47]. Therefore, improved detection of inflammatory changes in the terminal ileum is postulated in examinations performed with a rectal water enema. In the same working group, diagnosis with high-resolution ultrasound and MRE in patients with chronic inflammatory bowel disease was retrospectively compared [48]. It was shown that ultrasound can miss clinically relevant changes since some anatomical regions are difficult to visualize. However, in a study using contrast-enhanced ultrasound, a significant correlation with histopathological changes in patients with Crohn's disease regarding the quantitative evaluation of intestinal wall vascularization was found [49]. There was a strong correlation of contrast-enhanced ultrasound regarding clinical inflammation markers [50]. In patients with known chronic inflammatory bowel disease and acute abdominal pain, the same working group additionally showed that regular emergency CT for intestinal wall evaluation is sufficient and no additional MRE studies are necessary since an advanced finding is often present in these patients [51]. In a further study of the Regensburg group, it was able to be shown in patients undergoing bowel resection that MRE is an excellent imaging modality to correctly evaluate the Montreal classification of disease behavior [52]. The extent to which MRE can be used as an independent objective imaging method in the diagnosis and treatment of Crohn's disease was evaluated in an additional study [53]. MRE showed high interobserver agreement in the diagnosis of inflammatory activity while bowel distension and lymphadenopathy had only moderate interobserver agreement, which however had only a minimal effect on the diagnostic significance of the method in the clinical routine. The Münster working group of Lenze et al. compared MRE with 18F-FDG PET/CT and ultrasound with endoscopy and histopathology in the detection of active and fibromatous changes of Crohn's disease [54]. The detection rate of strictures was not significantly different but neither PET/CT nor MRE nor ultrasound could differentiate inflammatory from fibrotic strictures. Moreover, the combination of MRE and ultrasound as well as PET/CT and ultrasound had a 100 % detection rate of strictures requiring surgical or endoscopic dilatation therapy. Holtmann et al. also analyzed 18F-FDG PET in the diagnosis of Crohn's disease [55] and found that FDG-PET detected inflammation of the mucosa in Crohn's disease with high sensitivity and specificity and proved the value of the evaluation of inflammatory activity with Gallium-DOTATOC-PET/CT for the diagnosis of neuroendocrine tumors in the small intestine [56].
Additional groups addressed the application of new abdominal MR sequences: Kinner et al. were able to show in the Essen working group that diffusion-weighted imaging (DWI) in bowel diagnosis with MRI ([Fig. 2]) increases diagnostic reliability [57]. The same working group published that DWI showed inflammatory lesions of the bowel with higher reliability in children and adolescents with chronic inflammatory bowel disease than contrast-enhanced T1-weighted sequences alone [58]. In addition, there is a study by Neubauer et al. [59] that also concluded that contrast administration when using DWI is no longer necessary.
Fig. 2 Patient with active inflammatory changes of the small intestine in Crohn's disease: The a axial fat-saturated 3 D gradient echo sequence after intravenous contrast application shows three-layer intestinal wall thickening with increased contrast enhancement in terms of acute inflammation. The same information is provided in the b axial native diffusion-weighted sequence/DWI (b = 800).
Multiple German working groups studied intestinal motility and its correlation in chronic inflammatory bowel disease. In 2013 Bickelhaupt et al. showed the first options for the automatic measurement of small bowel contractions [60] and demonstrated that automatic measurement is better than manual [61]. Moreover, the same working group was able to show that small bowel motility correlates with the histopathology of the terminal ileum [62] as well as with inflammatory markers [63] and can differentiate active from chronic inflammatory processes [64]. In addition, the group analyzed prelesional and intralesional motility and found that the extent of the inflammation did not correlate with restricted movement [65]. Animal studies of this group also showed the value of the imaging of intestinal movement [66]. Hahnemann et al. evaluated an automatic color-coded algorithm for evaluating intestinal motility [67] and were able to show an increased detection rate of inflammatory lesions using this technique [68].
Additional studies addressing MRE at 3 Tesla in the diagnosis of (neo-)ileitis terminalis [69] have been published in the last three years. The authors again highlight the value of MRE at 3 Tesla for the evaluation of foci of inflammation in the ileum in chronic inflammatory bowel diseases. An animal study with gadofluorine M showed a higher correlation between intestinal wall enhancement and histopathological grading than Gd-DTPA-enhanced imaging [70].
Colon and rectum
Virtual colonoscopy
For virtual colonoscopy the current S3 guidelines [71] provide consensus-based recommendations in the case of incomplete colonoscopy due to a stenosed tumor (level of evidence 4, recommendation level 0, coordination in plenum: strong consensus) and as a result of other causes (e. g. adhesions, level of evidence 4, recommendation level B). Virtual colonoscopy has almost completely replaced double contrast examination of the colon in the clinical routine. The most important indications include incomplete or not implementable colonoscopy and patients under anticoagulation therapy. The European Society of Gastrointestinal Endoscopy (ESGE) and the European Society of Gastrointestinal and Abdominal Radiology (ESGAR) recently created joint guidelines for determining the indication for virtual colonoscopy [72]. The guidelines view virtual colonoscopy as an acceptable and equally sensitive alternative method in the case of suspicion of colorectal cancer and not implementable conventional colonoscopy (strong recommendation, high quality evidence). The role of virtual colonoscopy in colorectal preventive care remains controversial. Virtual colonoscopy is not taken into consideration for colorectal preventive care in the current S3 guidelines [82]. Virtual colonoscopy is also not recommended as the primary method in colorectal preventative care in Europe [72]. However, there is agreement that virtual colonoscopy can be recommended to the patient on an individual basis after the provision of sufficient clarification and information (weak recommendation, moderate quality evidence). Based on the increasing amount of scientifically sound data, multiple professional associations including the American Cancer Society and the American College of Radiology have argued for the use of virtual colonoscopy as part of preventive care. However, in March 2009 Medicare Service took an adverse stance to the financing of virtual colonoscopy in colorectal preventive care. Medicare's arguments were mainly based on three points of criticism: A) Risks of radiation exposure, B) Transferability of current data to persons older than 65 years and C) The unclear data regarding the relevance of extracolonic findings, i. e., possibly clinically relevant findings outside the colon. These points have been systematically and scientifically addressed: Regarding A): With respect to radiation exposure, an effective dose < 3 mSv can be assumed under screening regulations. According to Berrington de Gonzáles [73] from the U. S. National Cancer Institute, the risk/benefit ratio is 1:24 to 1:35. Accordingly, the cancer risk in a patient who is over 30 years old and receives a virtual colonoscopy every 5 years is significantly less than 1 % and is thus lower than the lifetime risk of getting colon cancer without preventive care of approximately 6 %. Regarding B): Data regarding the transferability to older patients were published as part of subgroup analyses of the ACRIN 6664Trial and in follow-up studies [74]. In people over 65 years, the rate of advanced adenomas of 7.6 % is slightly higher than in younger people (6.1 %). In contrast, the size, histology, and distribution of colorectal polyps was found to be identical. Regarding C): From 2008 to today, 10 studies have addressed the frequency and relevance of extracolonic findings. As a result, E3 findings (further workup needed) and E4 findings (clinically severe consequences) are seen in 10 – 16 % of patients and require additional examinations in 6 – 10 % of cases [75]. In addition to quality standards regarding implementation and interpretation, the ESGAR (European Society of Gastrointestinal and Abdominal Radiology) and the ACR (American College of Radiology) defined a minimum number of CT colonoscopy examinations in training programs [76]. The finding standards relate both to intestinal abnormalities and to pathological findings and their significance outside the bowel. The U. S. Preventive Services Task Force as the key regulatory authority in the USA together with the CMS (Centers for Medicare and Medicaid Services) are currently reevaluating the available data regarding colon cancer prevention and in particular regarding virtual colonoscopy. Final recommendations in this regard are expected in 2015.
Rectal cancer
Current scientific topics highlight the value of imaging in the local staging of rectal cancer. The identification and validation of prognostically relevant image parameters that in the future will allow better and more precise evaluation of the course and prognosis and, if necessary, treatment and aftercare stratification are in the foreground here. A further topic involves the role of imaging after neoadjuvant radiochemotherapy. In addition to better evaluation of the local tumor finding and its differentiation from reactive treatment-induced changes, a search for surrogate parameters for prognosis, course, and treatment modifications in the framework of multiparametric MRI imaging is also performed here. The role of imaging in the case of relapse will not be further discussed here.
Local staging of rectal cancer
Depending on the position of the cancer, there are different minimum requirements for the preoperative diagnosis of spreading according to the current S3 guidelines from 2013 [71]. Taking into account some limitations in the differentiation of T1 substages (T1 sm1 – 3), T1 tumors of the rectum continue to be the domain of endoscopic ultrasound (recommendation level B, level of evidence 2b – de novo). The staging of rectal cancer stage T2 and higher is more differentiated and also more complex [71]. The method of choice here is MRI, with 1.5 Tesla units with body coils representing the standard. To date, MRI at 3 T or the use of endorectal coils could not yield a significant diagnostic advantage [77]
[78]. The minimum diagnostic requirements are sagittal and paraaxial, thin-slice (voxel size at least 3 × 1 × 1 mm3) T2-weighted sequences and diffusion-weighted and contrast-enhanced T1-weighted sequences [78]
[79]. In the case of low-lying rectal cancer, coronal T2 and contrast-enhanced T1-weighted sequences must be additionally performed for the evaluation of a possible infiltration of the sphincter complex/extralevator and intersphincter fascia [79]. 3 T MRI with the option to acquire 3 D datasets of the pelvis will make it possible in the future to cover the entire T-staging with one sequence [78].
Extramural tumor extension largely determines the probability of distant metastases and is thus a decisive independent prognostic factor. In the current TNM classification for rectal cancer, a differentiated view of the T3 stage is currently becoming established. Based on the mesorectal infiltration depth, 4 subgroups are defined (T3a: < 1mm; T3b: 1 – 5mm; T3c: > 5 – 15mm; T3d: > 15 mm) [80]. According to current data, MRI has an accuracy equivalent to that of histology [81]. Regardless of the nodal status, T3a and T3b tumors have a similar 5-year survival rate to that of T2 tumors with 85 %. This number drops to 54 % for T3c tumors. Exact classification via MRI is thus to be considered a valid prognostic factor [78]
[81].
The same is true for the obligatory determination of the tumor distance from the mesorectal fascia. In the case of a positive circumferential resection margin (CRM+) – per definition the distance of the tumor from the mesorectal fascia is less than 1 mm – the local relapse rate is 3.5 times and the mortality rate is 2 times higher than in patients with a negative circumferential resection margin (CRM-) [82]. According to the currently available data including T2w and T1w post-contrast-enhanced images, MRI is capable of determining this distance with a sensitivity of 92 % and a negative predictive value of 94 % [83].
The determination of the so-called extramural vascular invasion (EMVI) via MRI was identified as a further independent prognostic factor. Such a situation can be assumed in approx. 25 % of patients in stage UICC II and approximately 45 % of patients in stage UICC III [84]. EMVI-positive patients with tumor infiltration into the perirectal vessels have a 4 times higher risk of distant metastases [85]. Sohn et al. [38] were able to recently show that the risk of distant metastases increases particularly in the case of infiltration of larger vessels (≥ 3 mm). For MRI they found a sensitivity, specificity, and accuracy of 28 %, 94 %, and 80 %, respectively (n = 447 patients, 79 with distant metastases) with respect to the evaluation of extramural vascular infiltration. MRI is currently the only preoperative imaging modality capable of visualizing extramural vascular invasion [79].
N-staging continues to be a major challenge. The main problem is frequent micrometastases in normal-sized lymph nodes. In 1991, Dworak et al. showed that affected perirectal lymph nodes have an average diameter of only 3.3 mm [86]. Mönig et al. found tumor cell nests in lymph nodes with a diameter of < 5 mm in up to 53 % of cases [87]. Due to the low tumor mass of micrometastases and the associated low tracer uptake, PET-CT has only a low sensitivity of 29 % [88]. Only insufficient differentiation of benign or metastatic lymph nodes can currently be achieved with diffusion-weighted sequences in MRI [89]. According to Birkhäuser et al., USPIO (ultrasmall superparamagnetic particles of iron oxide) contrast agents can detect lymph node infiltration with a specificity of 94 – 97 % at least in prostate cancer or bladder cancer [90]. However, it remains to be seen whether the FDA and BfArM will approve USPIO contrast agents. Liu et al. [91] recently reviewed the value of dual-energy CT in the differentiation of benign and malignant lymph nodes in 55 rectal cancer patients. Compared to only the determination of the short-axis diameter of lymph nodes (SAD), the additional determination of the normalized iodine concentration (NIC) in this collective increased the accuracy of metastasis determination to 83 %. However, the importance of dual-energy imaging for the differentiation of benign and malignant lymph nodes is currently not sufficiently clarified.
Even when using a routine MRI protocol, it is possible under consideration of certain criteria (form, margin definition, and lymph node signaling) to detect infiltrated lymph nodes with a relatively high accuracy of up to 85 % [92]. In certain constellations, nodal status is less decisive for prognosis than previously assumed. Nodal-negative and N1 nodal-positive tumors have the same local relapse rates after neoadjuvant therapy and total mesorectal excision [93]. The same is true for T3a and T3b tumors with similar survival rates as T2 tumors regardless of nodal status [81]. Therefore, it remains to be seen in which cases exact preoperative N-staging is absolutely necessary for the planning of the proper therapy.
Evaluation of disease course under neoadjuvant therapy
Neoadjuvant radiochemotherapy of locally advanced rectal cancer has been a proven therapy for years. The downstaging rate is 60 %. Pathohistologically verified complete remission can even be achieved in 15 – 27 % of cases. Neoadjuvant radiochemotherapy regimes have resulted in a significant improvement of the relapse rate and the survival rate. Moreover, the rate of sphincter/continence-preserving operations has risen.
A current challenge from a radiological standpoint is the MRI follow-up of the local finding under neoadjuvant therapy. Reactive changes of peritumoral and tumoral tissue with corresponding fibrotic and desmoplastic reactions during therapy complicate differentiation from vital tumor tissue. Accordingly, the accuracy of correct T-staging in this situation decreases to only 43 – 54 % or correct N-staging to 64 – 68 % [78]. Overestimation of the T-stage can be expected in 38 % of cases and a positive circumferential resection margin is incorrectly assumed in 56 % of cases [78]. In cases of complete pathohistological remission after radiochemotherapy, the necessity for subsequent surgical treatment is currently being questioned and imaging biomarkers that are suitable for reliably identifying such a situation are being sought.
Scientific developments currently focus particularly on multiparametric imaging via MRI. The significance of diffusion imaging in terms of the evaluation of cell integrity and dynamic contrast-enhanced examinations in terms of the evaluation of microcirculation or vascular permeability is in the foreground. While tumor tissue has a high signal in diffusion-weighted sequences ([Fig. 3]) with a low signal in the ADC (apparent diffusion coefficient), the reverse is true in reactively changed tissue. Cancers that respond to therapy show an ADC signal increase in intermediate staging. Non-responders do not show any changes or only a minimal increase in the ADC signal. Birlik et al. [94] found significantly lower ADC values prior to the start of treatment in the group of non-responders compared to late responders and recently defined corresponding prognostic ADC threshold values of b = 600 and b = 1000. By comparing DWI before and after neoadjuvant therapy, treatment response can currently be derived with an accuracy of approx. 80 – 88 % [95]. Texture analyses and their significance as biomarkers of tumor response at 3 Tesla and T2-weighted plain images are still in an early stage of development [96]. Current studies also compare the evaluation of tumor response via F18 FDG PET-CT and F-18 FLT PET-CT. While a significant SUVmax decrease after radiochemotherapy can be observed in both methods, significantly more histopathologically verified responders can be identified in metabolically negative F-18 FLT [97].
Fig. 3 Low-lying and difficult to delimit rectal cancer. In T2w a and T1w (fat saturation post-Gd, center), the tumor is unusually difficult to delimit. The fusion of T1w post-Gd with axial DWI images c shows improved tumor detection and ability to evaluate tumor extension. The example (T1w post-Gd) of another tumor d shows its infiltration in to the extramural veins (EMVI).
Mesenteric bleeding and ischemia
Both mesenteric bleeding and mesenteric ischemia are potentially life-threatening diseases. Therefore, quick diagnosis is extremely important. Due to its high availability and speed, modern cross-sectional imaging, in particular multidetector CT (MDCT), is predestined for the diagnosis of abdominal vessels. In contrast, angiography has been largely replaced by MDCT with respect to the diagnosis of mesenteric vascular diseases. However, it continues to play an important role in the therapeutic care of vascular diseases of the mesenterium in part due to the further development of the catheter technique.
Mesenteric bleeding
Intraabdominal bleeding is potentially life-threatening and can result in an acute abdomen but in general is a rare cause. Modern cross-sectional imaging has become increasingly important, particularly in the diagnosis of gastrointestinal bleeding [98].
While upper gastrointestinal (GI) bleeding is usually caused by ulcers, varices, and erosions, the most common causes for lower gastrointestinal (GI) bleeding are diverticular bleeding, hemorrhoids, and angiodysplasia. Ell et al. [99] also developed the term mid-gastrointestinal bleeding describing the region distal to the major duodenal papilla to proximal to the ileocecal valve. While endoscopic methods in particular for lower GI bleeding continue to be very important for diagnosis and treatment, mid-GI bleeding is often the domain of cross-sectional imaging. MDCT with CTA has become significantly more important in recent years for the diagnosis of gastrointestinal bleeding. It was able to be shown on the basis of an animal model that MDCT can detect active bleeding starting at a bleeding rate of 0.5 ml/min [100]. This value is comparable with that of angiography. The sensitivity for detecting GI bleeding is approx. 85.2 % and the specificity is 92.1 % [101]. In addition, the location of bleeding can be identified with high diagnostic accuracy [102]. Active contrast extravasation can often be directly identified.
Intraperitoneal bleeding is caused, for example, by rupture of a visceral aneurysm as part of a rupture of parenchymatous abdominal organs or in the framework of bleeding due to arrosion caused by inflammatory processes or tumor infiltration. In pancreatitis, bleeding due to arrosion occurs in approx. 1.3 % of cases [103]. Moreover, pseudoaneurysms representing a new potential source of bleeding can form.
MDCT has largely replaced digital subtraction angiography (DSA) as the diagnostic method of choice in the search for intraperitoneal bleeding. Angiography is currently used more as a therapeutic tool than as a diagnostic instrument. Hyare et al. [104] were able to show that MDCT with CTA has a high sensitivity (94 %) and specificity (90 %) for the detection of arterial bleeding in pancreatitis. A study of the Ulm working group [105] detected the source of bleeding in 9 of 10 patients with intraperitoneal bleeding via MDCT.
Mesenteric ischemia
In 1926, A. J. Cokkins wrote the following about mesenteric ischemia: "The diagnosis is impossible, the prognosis hopeless and the treatment useless..." [106]. This statement was unfortunately often true at that time. Thankfully, radiology has been able to help to change at least the first point of the statement.
Arterial occlusion
Acute mesenteric ischemia is a rare but often life-threatening disease with a mortality rate between 60 % and 80 %. The mortality rate can increase to 95 % if the cause is thrombotic occlusion of the superior mesenteric artery [107]. It is therefore essential to detect arterial mesenteric ischemia before a transmural bowel infarction develops. The diagnosis of patients with clinical suspicion of arterial mesenteric ischemia includes lab tests (serum lactate and hyperkalemia) and imaging. In particular conventional angiography of the mesenteric arteries was the method of choice in the past for detecting an occlusion. However, the diagnostic algorithm has increasingly changed in recent years as a result of the development of CT. In particular, biphasic CT angiography (CTA) began to become increasingly established with the first published studies in 2001 [108]
[109]. The advantage of MDCT is that not only the arteries and veins are visualized via CT angiography but also the perfusion of the abdominal and visceral organs can be shown. A pictorial essay by Wildermuth et al. [110] illustrates this very clearly. In an animal experiment with pigs, it was able to be shown that CTA can effectively differentiate between open and embolically occluded branches of the superior mesenteric artery with a positive predictive value of 92 % [111]. In a further study in which CT units with 16 rows or more were used, excellent sensitivity and specificity values for the diagnosis of arterial mesenteric ischemia were able to be achieved [112]. This study additionally examined indirect signs of the presence of mesenteric ischemia. The most sensitive CT signs of mesenteric ischemia are the detection of intestinal pneumatosis or gas in the mesenteric vein or the portal vein. Both of these signs were 100 % sensitive.
In contrast to mesenteric angiography using the catheter technique, MDCT has a series of further advantages: It is widely available, fast, noninvasive, and yields reproducible results. As such, it is ideally suited as the primary imaging modality for identifying patients who need immediate surgical therapy. Due to the high negative predictive value of multidetector CT, a watch-and-wait strategy can be used in patients with a negative CT. Such a diagnostic algorithm helps to save time and thus makes it possible to lower the mortality rate of patients with mesenteric arterial occlusions.
Special form: non-occlusive mesenteric ischemia (NOMI)
Already in 1995, Lock [113] defined NOMI as intestinal ischemia not caused by arteriosclerosis, arterial or venous thrombosis, arterial embolism, or vasculitis of the mesenteric vessels. This form of intestinal ischemia is therefore not based on a mechanical occlusion but rather on vasospastic events caused by various factors. Hypotension, intraabdominal pressure elevation, low cardiac output, and treatment with vasopressors or catecholamines represent the most common factors, i. e., concomitant circumstances that change hemodynamics and thus cause low blood flow to the mesenteric vascular system [114]
[115]. Primarily multimorbid patients are mainly affected by NOMI with a prevalence of 2 – 10 %, with a mortality rate of approx. 58 % being defined here [116].
MDCT has also become important for the diagnosis of NOMI. According to recent data, its sensitivity of 96 % is sufficient but specificities of only between 33 % and 60 % are achieved [117]. Angiography is the treatment method of choice, particularly for the treatment of non-perforated NOMI. Potent vasodilators can be locally applied over multiple days so that the non-occlusive ischemia can be treated [118].
Venous occlusion
With 5 – 15 % of cases, venous occlusion ranks third after arterial occlusion and non-occlusive disease as a cause of mesenteric ischemia. Portal hypertension and portal vein thrombosis, abdominal trauma, previous abdominal surgical interventions, inflammation in the abdominal cavity, and hypercoagulability are considered to be predisposing factors [119]. Approximately 20 % of mesenteric vein thromboses remain etiologically unclarified and are referred to as idiopathic. The disease can take different forms. In addition to a chronic form that is usually clinically inapparent, the subacute form in which patients complain of abdominal pain for weeks and months is differentiated from acute venous mesenteric ischemia. The acute course is clinically similar to the arterial form of mesenteric ischemia. The symptoms coincide with the clinical picture of an acute abdomen. The main symptom is abdominal pain in more than 90 % of patients [120]. Bloody diarrhea, hematochezia, and hematemesis are signs of a mesenteric infarction that has already occurred.
Analogously to arterial mesenteric ischemia, both direct detection of a venous mesenteric occlusion on CT as well as indirect signs of stenosis are seen. These relatively nonspecific signs can be dilated intestinal loops, ascites, and mesenteric fatty tissue imbibition. Intestinal wall changes are also ultimately nonspecific. Intestinal wall thickening with a lack or at least reduced contrast enhancement and a target sign is often seen [121]
[122]
[123]. The lack of contrast enhancement of the intestinal wall with detection of intramural gas formation or intestinal gas in the mesenteric outflow vessel is an indication of transmural ischemia. Specific signs associated with transmural ischemia include an indistinct intestinal wall and larger quantities of ascites [124]. Depending on the location of the occlusion of the venous outflow in the periphery, morphologically normal and pathological intestinal loops can occur next to one another. In addition to indirect signs, direct thrombus detection can be achieved via contrast-enhanced MDCT. CT is a highly specific examination modality with values of 100 % [112]
[125] and also a negative predictive value of 100 %. However, the sensitivity is significantly lower and depending on the study is only up to 21 % [112].
Angiography does not play a role in the diagnosis of mesenteric or portal vein thrombosis and also the radiological intervention options are very limited even if small case series have promising results [126].
Varia in the gastrointestinal tract
Acute intestinal graft versus host disease (GvHD) is a serious complication of allogenic stem cell transplantation and occurs in 30 – 75 % of cases. A Mannheim working group described the typical clinical picture on MRI based on the data of 9 patients with confirmed GvHD [127]. Extensive wall thickening of the intestine is typical for GvHD with the terminal ileum always being affected in the examined group. In addition, the submucosa was not enhanced in all cases while strong contrast enhancement was seen in the mucosa. The Regensburg working group of Schreyer was able to show using CEUS (contrast-enhanced ultrasound) in 2 studies involving patients with GvHD that a transfer of microbubbles to the intestinal lumen in patients with acute GvHD probably due to the damaged mucosal barrier can be detected [128]
[129]. With a positive predictive value (PPV) of 100 %, this method could be a good specific instrument for early and specific diagnosis of GvHD and for differential diagnosis from viral enteritis, but further evaluation in a larger cohort is not yet available.
Conventional radiology in gastrointestinal and abdominal radiology was a further topic that was critically addressed and discussed in recent literature in Germany. Therefore, the Regensburg working group was able to show that there was no diagnostic added value of scans acquired in a lying position in the retrospective analysis of 2148 patients who underwent a conventional survey scan of the abdomen in a standing position or in a left-lateral position with an additional scan in a lying position [130]. A pathological finding (obstruction, free air, etc.) was seen in 10.8 % (n = 232) of patients on these conventional scans. Even if in total 3.5 % (n = 75) more anatomical structures or foreign material was able to be detected on images acquired in a lying position, this information was not diagnostically relevant in a single case so that routine acquisition of images in a lying position as a supplement to images in a standing position or in a left-lateral position is no longer to be viewed as indicated in the majority of cases. A further publication from the same working group retrospectively examined the value of conventional fluoroscopy of the gastrointestinal passage in 300 patients of a university hospital [131]. In this study the authors found that a further diagnostic imaging method such as endoscopy, CT, or ultrasound was initiated in over 87 % of patients who underwent gastrointestinal passage prior to a change in diagnosis or treatment. The already known diagnosis such as intestinal stenosis or partial obstruction was confirmed in a majority of patients with the patients being exposed to an average radiation dose of 7 mSv which is approximately in the range of modern CT systems. Based on the analysis, the authors concluded that the indication for gastrointestinal passage should be made extremely conservatively and critically given the availability of modern cross-sectional imaging methods such as CT, MRI, and ultrasound since indicative therapy-relevant decisions are made based on conventional methods with relatively high radiation exposure in only rare cases.
Several publications analyzing the value of conventional abdominal radiography compared to low-dose CT in body packers provide a similar critical analysis of conventional techniques [132]
[133]. In principle, the studies determine that the detection of illegal packets of drugs in the bowel is more successful with low-dose CT than with conventional abdominal radiography. The fact that low-dose CT uses a lower biological radiation dose than conventional radiography with the same diagnostic value is interesting. Average values of 2 mSv for conventional radiography compared to 1.2 mSv for low-dose CT in women are specified here as an example. This data can be largely extrapolated from "healthy" body packers to patients with pathological changes.
Summary and further developments
In recent years, there have been notable scientific and clinical advances in abdominal radiology in Germany that have significantly influenced global radiological literature. A definite paradigm shift in the diagnosis of the esophagus and stomach was able to develop based on this new literature so that cross-sectional imaging methods such as MRI and primarily MSCT with their high resolution are put at least on par with endosonographic methods in some cases. This literature and thus evidence-based data were able to be included in the new S3 guidelines for esophageal and gastric cancer. With respect to the diagnosis of the small bowel, a similarly strong influence of the current scientific literature on the updated care guidelines was noted. The scientific focus in recent years was on the diagnosis and more precise evaluation of chronic inflammatory bowel diseases. The scientific focus on radiation-free methods such as MRI and ultrasound was the main driving force for clinical care and diagnosis in Germany and Europe so that MRI in addition to ultrasound was specified in the German guidelines on Crohn's disease in 2008 as the basic examination for this disease of typically young patients [45]. Fortunately, these findings have been integrated in European and global recommendations and guidelines in recent years. New techniques such as movement imaging in MRI as well as diffusion-weighted imaging of the bowel for the detection of inflammatory changes are interesting examples of the optimization of existing protocols that will hopefully soon become established in clinical practice given sufficient evidence from studies.
In the diagnosis of the colon, virtual colonoscopy has replaced conventional intestinal examinations. However, there is still a limited indication spectrum (stenosed tumors, contraindications to colonoscopy) according to the guidelines. A reevaluation of the literature regarding prevention and screening by the CMS (Centers for Medicare and Medicaid Services) is expected in the USA over the course of 2015. In the diagnosis of rectal cancer, primarily high-resolution MRI plays a decisive role in treatment stratification and staging. Diffusion techniques may allow potential tools for the evaluation of tumor response for treatment.
In the diagnosis of mesenteric ischemia and bleeding, multidetector CT (MDCT) is increasingly becoming the established method of choice. Modern cross-sectional imaging methods such as MRI of the abdomen and contrast-enhanced ultrasound (CEUS) can be a specific diagnostic agent also in the evaluation of GvHD and its differential diagnosis. The influence of traditional conventional abdominal radiography is further reduced under consideration of the high diagnostic performance of modern cross-sectional imaging methods such as CT, MRI, and CEUS with low or no radiation exposure.
