Diagnosis

Visualization of the mucosal surfaces is essential to diagnose most early lesions in the hollow organs of the gastrointestinal tract. The conventional upper gastrointestinal (UGI) series, small bowel follow-through, and barium enema (or endoscopic evaluation of the upper and lower tracts) are superior to either CT or MRI for detecting mucosal lesions (Table 30D.2). On the other hand, CT, MRI, and sonography are superior to barium studies for evaluating the walls and extramural portions of the GI tract for deep tumor involvement, and they may also help in detecting distant metastases.^1–4

Table 30D.2

Imaging Techniques for Diagnosis.

While both CT and MRI can be used to detect neoplasms involving the solid organs of the abdomen and pelvis, CT plays a dominant role in the pretreatment evaluation of these patients. CT-guided biopsies of masses can be extremely useful (Figure 30D.1). In conjunction with endoscopic retrograde cholangiopancreatography (ERCP) or magnetic resonance cholangiopancreatography (MRCP), CT is being used to help evaluate most patients with suspected or known pancreatic neoplasms.

Figure 30D.1

Cystic pancreatic mass. Contrast enhanced CT image of upper abdomen demonstrates large cystic mass (M) in tail of pancreas, anterior to left kidney and posterior to stomach. Operative specimen was diagnosed as pancreatic cystadenoma.

CT scanners normally acquire body images in the axial plane, and scan times have now been reduced to 1 to 2 seconds per image, eliminating most motion artifacts. Spiral CT with single detector arrays will now allow continuous scanning for periods of up to 20 seconds so that larger tissue volumes can be scanned during a single breath-hold. Multiple-detector rings in more advanced scanners that are now becoming available will compound this imaging advantage.

MRI technology has also been advancing rapidly. The initial enthusiasm for MRI was due not only to the absence of ionizing radiation but, more importantly, to its potential for improved tissue characterization. When MRI was introduced, it was hoped that differences in the signal intensities of abnormal tissues would enable the radiologist to differentiate benign masses from malignant ones. Unfortunately, tissue characterization has not proven to be as reliable as once hoped, and in the abdomen and pelvis, the only tumor that was found to have a relatively pathognomonic appearance was adrenal pheochromocytoma. With most other tumors and inflammatory processes, the MR signal characteristics overlap.

The standard pulse sequences that have been used with MRI have been spin echo.⁵ T₁-weighted spin echo images provide better spatial resolution than T₂-weighted images and, therefore, also yield more detailed anatomic information. T₂-weighted images show differences in contrast more clearly between normal and abnormal tissues. In general, pathologic foci tend to appear as relatively dark (low-signal intensity) areas on T₁-weighted images and as bright (high-signal intensity) areas on T₂-weighted ones. Unfortunately, many different pathologic processes (not just tumors) can produce this dark-bright appearance.^6,7 For instance, edema around a tumor, as well as the tumor itself, may be indistinguishable with certain scanning sequences. A simple cyst may also be dark-bright onT ₁- and T₂-weighted images, as will abscesses and some benign tumors. Therefore, while MRI may be quite sensitive in detecting an abnormality, absolute characterization of the abnormality is usually not possible.

Newer pulse sequences allow some MRI data to be acquired now in 20 seconds or less (e.g., during one breath-hold), eliminating respiratory motion artifacts.⁸ In general, breath-holding MRI techniques produce images with lower spatial resolution than standard spin-echo techniques. Therefore, they are useful as problem-solving tools but have not replaced spin echo or fast spin echo techniques for tumor detection and staging.

Intravenous contrast media are now being used frequently during MRI as well as CT scans. Gd-DTPA and other agents have proven useful as intravenous contrast agents during T₁-weighted pulse sequences. The combination of faster MRI scanning sequences and intravenous contrast media has shown promise and may eventually replace contrast-enhanced CT as the preferred imaging modality for solid organs in the abdomen and pelvis. Organ-specific contrast agents may prove to be especially helpful with MRI scans of the liver, but the overall role of intravenous contrast agents with MRI is still being evaluated.

Staging

Various imaging techniques can be useful in tumor staging (Table 30D.3). The cross-sectional imaging capabilities of both CT and MRI can be helpful in detecting extramural spread of a tumor within the gastrointestinal (GI) tract, with certain limitations.⁹ For instance, either CT or MRI can detect abnormal tissue infiltrating the fat that surrounds the rectum, but neither may be able to differentiate inflammatory strands from neoplastic invasion. Another limitation of both CT and MRI is their inability to detect tumor in normal-sized lymph nodes. Both imaging techniques are more useful for detecting lymph node involvement in selected types of cancer, such as seminomas and lymphomas, which often produce lymph node enlargement. On the other hand, it is not uncommon for certain GI carcinomas to replace a lymph nodeís architecture without enlarging the node. Neither CT nor MRI can ascertain the depth of tumor invasion accurately in the wall of the GI tract. Therefore, the major role of CT and MRI for staging abdominopelvic malignancies is in detecting distant metastases.

Table 30D.3

Imaging Techniques for Diagnosis.

Post-treatment Surveillance

CT is now the standard imaging technique in most institutions for monitoring cancers of the abdomen and pelvis, and it is used routinely for detecting local recurrences or liver metastases.^10,11 While MRI now approaches or equals CT in its ability to depict liver metastases, its sensitivity for detecting extrahepatic metastases in the abdomen is limited. CT examinations of the upper abdomen have the potential for detecting more metastatic cancer sites than MRI studies. The higher cost and restricted availability of MRI have also enabled CT to retain its role as the primary abdominal imaging modality for post-treatment surveillance of cancer in the abdomen and pelvis.

Gastrointestinal Tract: Hollow Organs

While evaluation of the mucosal surfaces with barium studies and endoscopy still plays a primary role in the diagnosis of cancers of the stomach, small bowel, and colon, CT is now used widely as the principal imaging test for staging and monitoring these tumors.^4,10,12 CT is somewhat limited in its ability to show mesenteric masses because of difficulty in differentiating unopacified loops of bowel from tumors, but it can be useful for detecting metastases to solid organs or to (enlarged) retroperitoneal lymph nodes. MRI has only limited utility for evaluating cancer in the upper abdomen because of the lack of a suitable oral contrast medium as well as the prolonged imaging times.

As described in another section, endoscopic ultrasonography (especially transrectal sonography) is an evolving technology which shows promise for evaluating gastrointestinal wall invasion more accurately.^2,3,13,14 However, endoscopic sonography is limited by its inability to depict lesions that are further than 5 cm from the gastrointestinal lumen and by problems of access for the ultrasound probe: tumors that prevent passage of the probe because of lumenal obstruction or severe narrowing cannot be evaluated adequately by endosonography.

The role of nuclear medicine scans for the diagnosis, staging, and follow-up of GI tract cancers has diminished as advances in other imaging modalities have occurred. For instance, CT, MRI, and ultrasonography are more sensitive and reliable now for demonstrating liver metastases than radionuclide scanning. Positron emission tomography (PET) is a very effective tool for the detection of metastases and will play an increasingly important role as PET scanners become more available.

Liver Metastases

The diagnosis, staging, and follow-up of primary and metastatic liver lesions can now be done with the assistance of multiple imaging techniques. CT is the most widely used technique in the United States for ascertaining liver involvement by cancer. In Europe, many institutions use hepatic ultrasonography first, and CT or MRI as second studies for problem cases. Transabdominal ultrasonography is limited in its abily to detect smaller hepatic lesions: its reported sensitivity is only 20% for lesions < 1 cm in diameter, compared with 31% for MRI and 49% for CT.¹⁵ With the development of faster scanning techniques and new contrast media, MRI has the potential to replace CT and ultrasonography as a primary diagnostic tool for liver disease, but CT and MRI of the liver are complementary procedures at present (Figure 30D.2).^6,7 Each study has its limitations. Lesions that are missed by CT are often detected by MRI and vice versa. Therefore, when it is important to detect all metastatic lesions in the liver before attempting a hepatic resection, both CT and MRI may be indicated, and intraoperative ultrasonography should also be considered.

Figure 30D.2

Hapatocellular carcinoma. A. Arterial phase of a contrast-enhanced CT of the liver showing splenic infarct (arrow) and vague area of decreased attentuation in the left lobe of the liver (arrowheads). B. T₂-weighed non-contrast-enhanced MR image showing (more...)

Helical CT, which allows continuous scanning (volumetric imaging) while the patient is moved through the scanner, is now becoming widely available. Helical CT facilitates the rapid acquisition of multiple images, which, in turn, enables repeated scanning through the same body region. For instance, many institutions now routinely scan the liver twice, once during the arterial phase of hepatic contrast enhancement and seconds later during the portal venous phase. Hypervascular tumors, such as hepatomas, are usually more visible during the hepatic arterial phase, while hypovascular lesions in the liver, such as colon cancer metastases, are more easily identified during the portal venous phase.

CT can not only detect but may also help characterize some hepatic masses. Cavernous hemangiomas, the most frequent benign hepatic masses, may appear on CT scans as lesions with well-defined margins. They will occasionally demonstrate globular enhancement during a contrast infusion, which refers to small areas of parenchymal enhancement at the rim of a liver mass, which are similar in density to vessels in the same area. When the CT findings are suggestive, but are not characteristic, of a hemangioma, an MRI study can also be useful. The characteristic MRI appearance of a liver hemangioma is a well-defined, low-signal-intensity mass on T₁-weighted images and a high signal intensity or light bulb effect in the mass on T₂-weighted pulse sequences.^6-81617 Unfortunately, the same MRI appearance can sometimes be encountered with necrotic neoplasms or with liver metastases from hypervascular primary tumors, such as islet cell carcinomas.

A radionuclide scan performed with technetium-99m–labeled red blood cells (Tc-RBC) can also be used to make a specific diagnosis of hemangioma of the liver. With a hemangioma, diminished radionuclide activity is observed in the liver lesion during the early (vascular) phase of the red cell infusion, and increased activity (a “hot spot”) is seen on delayed or blood-pool scan images. This flip-flop appearance of a mass during a red cell infusion is diagnostic of a liver hemangioma, but the Tc-RBC study is limited in its ability to detect hemangiomas that are < 2 cm in diameter.

CT or ultrasound guidance can also be helpful for fine-needle biopsies of liver lesions. While sonographically guided biopsies may require more technical expertise than biopsies that are guided with CT, the ability to image the needle tip in realtime with sonography, while advancing the needle into a lesion, is a definite advantage. A biopsy procedure with CT guidance can be considered to be partially blind, since the needle tip cannot be visualized at the same time as it is being advanced within the liver. Sonographically guided biopsies also tend to be less expensive than CT-directed biopsies.

Biliary Obstruction and Lesions of the Pancreas

CT and sonography can both play important roles in evaluating the jaundiced cancer patient. Sonographic techniques are accurate for diagnosing bile duct dilatation and helping to determine whether “medical” or “surgical” jaundice is present. CT can be used subsequently to confirm duct enlargement. It is also superior to sonography in many cases for determining the actual cause of an obstruction because of its ability to show the extrahepatic biliary tree and to outline the actual obstructing lesion (e.g., a gallstone or pancreatic mass).

CT is the primary imaging modality for evaluating suspected pancreatic disease.¹⁸ Using a bolus of intravenous contrast material and thin collimation, this technique can be used to detect a majority of pancreatic adenocarcinomas and islet cell tumors. A CT-guided biopsy can then be employed to confirm the suspected diagnosis (see Figure 30D.1). Helical CT now makes thin-section images of the entire pancreas possible during one breath-hold, and it has added greatly to our ability to recognize tumor encasement of peripancreatic vessels. The CT appearance of a tumor which is in direct contact with a major vessel and which wraps around more than one-half of the vesselís circumference has been shown to be highly specific (98%) for vascular invasion, with a high positive predictive value (95%).¹⁹ Ultrasonography has also been used to evaluate the pancreas, but it is limited by intervening bowel gas, which makes evaluation of the entire pancreas difficult in most patients. Recent technical advances including the use of paramagnetic contrast agents have improved the ability of MRI to image the pancreas. While one large multi-center study could not confirm the superiority of MRI over CT for staging pancreatic tumors,²⁰ another study found that MRI could detect more small tumors than CT.²¹ Dynamic MRI (rapidly repeated scans during, and immediately after, vascular contrast infusions) may prove to be superior to helical CT for the preoperative evaluation of tumor stage, including vascular invasion, in patients with pancreatic adenocarcinomas.

Kidney

While ultrasound is the preferred imaging modality to detect hydronephrosis because of its low cost, CT (performed first without contrast, and then with intravenous contrast) is the primary study to use when a solid mass in the kidney has been demonstrated or is suspected. CT can also be used to differentiate certain benign kidney tumors from malignant lesions. For instance, a renal mass that contains fat and no calcification can be presumed to be an angiomyolipoma, since renal carcinomas have rarely been reported to contain substantial amounts of fat without calcification on CT. CT is also useful for staging known renal carcinomas (reported accuracy: 91%).²² MRI also has considerable potential as a modality for staging renal tumors, especially with intravenous Gd-DTPA (Figure 30D.3). Two major advantages of MRI for tumor staging, as opposed to CT, are better delineation of tumor thrombus in the renal vein and superior ability to distinguish collateral vessels in the renal hilus from lymph nodes. In two studies that compared CT with MRI for staging renal cell carcinoma, the staging accuracy of CT was 70 to 78%, while that of MRI was 92 to 96%.^23,24

Figure 30D.3

Renal cell carcinoma. A. Postcontract CT showing enhancing mass in the left kidney (arrow). B. Immediate post-gadolinium T1 image from MR study showing same lesion (arrow)

Tumors of the Female Reproductive Tract

Sonography remains the primary imaging modality to screen for masses in the pelvis. It is the ideal method also for distinguishing cystic from solid masses. Transabdominal sonography uses the distended bladder as an acoustic window to evaluate the deep pelvic structures, but transvaginal sonography is being used increasingly now as the initial examination when a pelvic mass is suspected. It provides excellent visualization of the lower pelvic structures (see separate section on ultrasound techniques).

When further evaluation is warranted, pelvic MRI can also be helpful (Figure 30D.4). Because of the relative lack of motion in this body area, MRI is often more effective in the pelvis than in the upper abdomen. The ability of MRI to obtain direct coronal, sagittal, and oblique images may also be valuable when trying to depict uterine and adnexal masses.²⁵ Mitchell and co-workers found that MRI provided additional information or increased diagnostic confidence in 25 out of 35 patients with these masses who had undergone prior ultrasonography or CT.²⁶ MRI can also be used to determine the depth of myometrial invasion by a uterine carcinoma, and it can help detect extramural spread of the disease as well.²⁷

Figure 30D.4

Lymphoma of uterine cervix. A. CT demonstrates ill-defined mass (M) between the bladder (containing layered urine and contrast material) and the rectum (R). Note that mass extends lateral to the right ureter (arrow) and is inseparable from body of uterus. (more...)

Prostate

After physical examination, transrectal sonography is the most helpful technique for detecting nodules in the prostate. While sonography has not proved to be effective in screening asymptomatic men for prostate cancer, sonographically guided biopsies of suspicious prostate nodules can be valuable in cancer diagnosis. The role of sonography in staging prostatic cancer is still uncertain, and CT has not been particularly helpful in assessing the extent of local disease. With its multi-planar imaging capability, MRI is probably the most accurate imaging modality to assist in staging prostate cancer. An endorectal receiving coil can also be used to increase anatomic definition in the area. In a study of 46 patients with prostatic carcinoma who subsequently underwent radical prostatectomy, staging with CT had an accuracy rate of 65%; with MRI, the staging accuracy increased to 83%.²⁸

Conclusion

Conventional imaging techniques, including barium studies and intravenous urography, are still the primary diagnostic tools to detect and initially evaluate neoplasms in the lumens of the GI and genitourinary tracts. Evaluation of tumors in the solid abdominopelvic organs requires the use of a cross-sectional imaging method, such as CT, MRI, or sonography, often followed by a guided biopsy. While CT has replaced conventional radiographic studies for staging and monitoring most abdominopelvic malignancies, MRI may assume a more important role in the future.

References

1.

Butch R J, Stark D D, Wittenberg J. et al. Staging rectal cancer by MR and CT. AJR Am J Roentgenol. 1986;146:1155–1160. [PubMed: 3486559]

2.

Hulsmans F J J H. Staging of rectal CA: US and CT. Radiology. 1989;170:319. [PubMed: 2643135]

3.

Hulsmans F J J H, Tio T L. et al. Assessment of tumor infiltration depth in rectal cancer with transrectal sonography: caution is necessary. Radiology. 1994;190:715. [PubMed: 8115617]

4.

Thompson W M, Halvorsen R A, Foster W L Jr. et al. Preoperative and postoperative CT staging of rectosigmoid carcinoma. AJR Am J Roentgenol. 1986;146:703–710. [PubMed: 3485343]

5.

Turnbull L W, Kean D M. Tumour identification using magnetic resonance imaging. Cancer Surv. 1987;6:343. [PubMed: 3322551]

6.

Li K C, Glazer G M, Quint L E. et al. Distinction of hepatic cavernous hemangioma from hepatic metastases with MR imaging. Radiology. 1988;169:409–415. [PubMed: 3174988]

7.

Wittenberg J, Stark D D, Forman B H. et al. Differentiation of hepatic metastases from hepatic hemangiomas and cysts by using MR imaging. AJR Am J Roentgenol. 1988;151:79–84. [PubMed: 3259825]

8.

Glazer G M. MR imaging of the liver, kidneys and adrenal glands. Radiology. 1988;166:303–312. [PubMed: 3275974]

9.

Thompson W M, Halvorsen R A Jr. Computed tomographic staging of the gastrointestinal malignancies. Part II. The small bowel, colon and rectum. Invest Radiol. 1987;22:96–105. [PubMed: 3549618]

10.

Charnsangavej C. New imaging modalities for follow-up of colorectal carcinoma. Cancer. 1993;71:4236–4240. [PubMed: 8508385]

11.

De Lange E E, Fechner R E, Wanebo H J. Suspected recurrent rectosigmoid carcinoma after abdominoperineal resection: MR imaging and histopathologic findings. Radiology. 1989;170:323–328. [PubMed: 2911656]

12.

Sussman S K, Halvorsen R A Jr, Illescas F F. et al. Gastric adenocarcinoma: CT versus surgical staging. Radiology. 1988;167:335–340. [PubMed: 3357941]

13.

Carroll B A. US of the gastrointestinal tact. Radiology. 1989;172:605–608. [PubMed: 2672087]

14.

Rifkin M D, Ehrlich S M, Marks G. Staging of rectal carcinoma: prospective comparison of endorectal US and CT. Radiology. 1989;170:319–322. [PubMed: 2643135]

15.

Wernecke K, Rummeny E, Bongarte G. et al. Detection of hepatic metastases in patients with carcinoma: comparative sensitivities of sonography, CT and MR imaging. AJR Am J Roentgenol. 1991;157:731–739. [PubMed: 1892027]

16.

Yamashita Y, Hatanaka Y, Yamamoto H. et al. Differential diagnosis of focal liver lesions: role of spin echo and contrast-enhanced dynamic MR imaging. Radiology. 1994;193:59. [PubMed: 8090922]

17.

McFarland E G, Mayo-Smith W W, Saini S. et al. Hepatic hemangiomas and malignant tumors: improved differentiation with heavily T2-weighted conventional spin-echo MR imaging. Radiology. 1994;193:43. [PubMed: 8090920]

18.

Freeny P C, Traverso L W, Ryan J A. Diagnosis and staging of pancreatic adenocarcinoma with dynamic computed tomography. Am J Surg. 1993;165:600–606. [PubMed: 8488945]

19.

Lu D S K, Reber H A, Krasny R M. et al. Local staging of pancreatic cancer: criteria for unresectability of major vessels as revealed by pancreatic-phase, thin-section helical CT. AJR Am J Roentgenol. 1997;168:1439–1443. [PubMed: 9168704]

20.

Megibow A J, Zhou X H, Rotterdam H. et al. Pancreatic adeno-carcinoma: CT versus MR imaging in the evaluation of resectability. Report of the Radiology Diagnostic Oncology Group. Radiology. 1995;195:327–332. [PubMed: 7724748]

21.

Ichikawa T, Haradome H, Hachiya J. et al. Pancreatic ductal adenocarcinoma: preoperative assessment with helical CT versus dynamic MR imaging. Radiology. 1997;202:655–662. [PubMed: 9051012]

22.

Johnson C D, Dunnick N R, Cohan R H, Illescas F F. Renal adenocarcinoma: CT staging of 100 tumors. AJR Am J Roentgenol. 1987;148:59–63. [PubMed: 3491524]

23.

Hricak H, Demas B E, Williams R D. et al. Magnetic resonance imaging in the diagnosis and staging of renal and perirenal neoplasms. Radiology. 1985;154:709. [PubMed: 3969475]

24.

Semelka R C, Shoenat J P, Magro C M. et al. Renal cancer staging: comparison of contrast-enhanced CT and gadolinium-enhanced fat-suppressed spin-echo and gradient-echo MR imaging. J Magn Reson Imaging. 1993;3:597. [PubMed: 8347952]

25.

Lee M J, Munk P L, Poon P Y, Hassell P. Ovarian cancer: computed tomography findings. Can Assoc Radiol J. 1994;45:185–192. [PubMed: 8193964]

26.

Mitchell D G, Mintz M C, Spritzer C E. et al. MR imaging observations at 1.5 T, with US and CT correlation and pulse sequence optimization. Radiology. 1988;169:359.

27.

Hricak H, Stern J L, Fisher M R. et al. Endometrial carcinoma staging by MR imaging. Radiology. 1987;162:297. [PubMed: 3797641]

28.

Hricak H, Dooms G C, Jeffrey R B. et al. Prostatic carcinoma: staging by clinical assessment, CT, and MR imaging. Radiology. 1987;162:331. [PubMed: 3797645]