Hepatic Imaging With Iron Oxide Magnetic Resonance Imaging

June 1, 2000

The management of hepatic tumors presents a challenging problem. The natural history of primary and metastatic liver lesions portends a poor prognosis. However, surgical resection and newer ablative techniques have had a

ABSTRACT: The management of hepatic tumors presents a challenging problem. The natural history of primary and metastatic liver lesions portends a poor prognosis. However, surgical resection and newer ablative techniques have had a great impact on cure rates. Unfortunately, the majority of newly diagnosed patients have surgically unresectable disease. Advances in hepatic imaging have improved the preoperative evaluation of malignant lesions and greatly assisted in selecting patients for surgical resection or other interventions. Currently, a number of modalities are available for the evaluation of hepatic tumors. This article provides an overview of some of the modalities currently in use, examines the role of iron oxide magnetic resonance imaging (MRI), and relates experience with its use at Baylor University Medical Center. [ONCOLOGY 14(Suppl 3): 29-36, 2000]

Primary liver cancers, in particular, hepatocellular carcinoma (HCC), represent perhaps the most common malignancy in the world and account for almost 1.25 million deaths annually. The worldwide geographic variation in incidence is well known; in the United States, the annual incidence remains relatively low, with roughly 2,500 cases reported per year. However, a diagnosis of HCC carries a poor prognosis. Okuda et al reported the results of a large multi-institutional review of 850 patients with HCC that demonstrated an overall median survival of 4.1 months. Left untreated, patients had a rather poor prognosis, with a median survival of 8.3 months for stage I disease, 2.0 months for stage II, and 0.7 months for stage III.[1]

Patients who received treatment fared somewhat better. Those with stage I and stage II disease treated surgically had a median survival of 21.9 months, while those treated with medical management had a median survival of 5.0 months. Other researchers have reported 5-year survival rates ranging from 26% to 40% and 5-year cancer-free rates of 55%.[2-4]

The surgical resection of hepatic tumors has historically been a daunting task, however. Prior to 1970, surgical resection was associated with mortality rates of 35% to 45%. In contrast, more recent series have demonstrated considerably lower mortality, routinely less than 10%.[5]. Yet, despite advances in reducing operative mortality, hepatic resection still carries significant morbidity. Complications include wound infection, bile leaks, bleeding, subphrenic abscess, liver failure, renal failure, pleural effusions, and pneumothorax; rates of complications vary from 11% to 74%.

In the United States, primary liver cancer represents a minority of hepatic tumors—only 2.5% of all new cancers. By far, the majority of hepatic tumors seen are metastatic lesions, most often of colorectal origin.

Again, with colorectal metastases to the liver, surgical resection has proven to be an effective means of treatment. Patients with unresected tumors seldom survive beyond 5 years, with a median survival of 3 to 24 months.[6] In contrast, reports of surgical resection have demonstrated a 20% to 50% overall 5-year survival rate.[7,8]

Furthermore, resection of recurrent hepatic metastases has proven beneficial as well. Nearly 80% of patients develop a recurrence of disease after hepatic resection, and yet, in 35% to 40%, recurrence is limited to the liver[9] and re-resection may provide long-term survival.[10,11]

Hepatic Imaging

Accurate preoperative imaging studies are paramount in determining appropriate treatment and predicting outcome. The most common imaging methods include ultrasound, computed tomography (CT), contrast-enhanced CT (CECT), and/or CT arterial portography (CTAP).

Recent advances in diagnostic imaging have greatly improved the detection and characterization of hepatic lesions. Advances in CT include the development of helical scanners that allow for rapid sequence imaging and dynamic intravenous (IV) contrast enhancement. Unfortunately, the prognostic abilities of contrast-enhanced CT are still limited, as evidenced by the fluctuations in management resulting from the findings of intraoperative ultrasound (IOUS).[12] This inability to accurately depict lesions preoperatively can have an enormous impact on patient care, with respect to therapeutic options and outcome, the patient’s psychological state, and the cost of care delivered.

Advances in magnetic resonance imaging (MRI) offer great promise for improving preoperative imaging capabilities. New imaging techniques, including fast spin-echo and gradient-echo techniques, permit rapid breath-hold image acquisition, thus eliminating significant motion-induced artifact.[13-16] In addition, a number of IV contrast agents have been developed that enhance the capabilities of MRI, namely, iron oxide agents and specific hepatobiliary agents.

Three iron oxide contrast agents for liver imaging have been developed thus far: AMI-25 (Feridex I.V.), SHU-555A (Resovist Injection), and AMI-227 (Combidex). Feridex I.V. was the first liver specific contrast agent developed. It uses iron oxide particles as negative contrast agents to enhance hepatic imaging. Resovist Injection is a contrast agent similar to Feridex I.V., but it can be administered by bolus injection rather than infusion. Combidex differs from the other two iron oxide agents because it consists of smaller iron-oxide particles. The primary indication for Combidex is lymph node imaging, but it can also be used for liver imaging.

Hepatic Imaging With Ferumoxides

Feridex I.V. is the only ferumoxides contrast agent currently available commercially in the U.S. It has been studied extensively, and its pharmacologic and radiographic properties are well known. Feridex I.V. is currently administered as a dilute IV infusion over a 30-minute period. Overall, side effects from Feridex I.V. infusion occur in 10% to 15% of patients and are well tolerated, although hypotension may still occur in 1% to 2% of patients.

The most common side effects are lower back pain (4%), flushing (2%), various combined gastrointestinal complaints (5.6%), and an assortment of other sporadic discomforts.[17] Back pain typically resolves spontaneously and permits continued contrast infusion.

The theory behind image enhancement with iron oxide contrast agents rests on the magnetic properties of iron oxide and its affinity for the reticuloendothelial system. The contrast agent is composed of crystalline iron oxide particles coated with a surface molecule, typically a polysaccharide that helps stabilize the particles in aqueous solution. Paramagnetic ions, such as Fe2+ and Fe3+, produce domains of spontaneous magnetization when packed closely in a crystalline structure. The number of domains depends on the particle size. Larger particles produce multiple domains of magnetization, while smaller particles produce single domains.

Groups of single-domain particles are particularly susceptible to external magnetic fields, resulting in super-paramagnetic properties. When such a field is applied to iron oxide particles, a large heterogeneous magnetic field results that can be used to enhance MR images. More specifically, the particles cause increased spin dephasing upon magnetic resonance (MR) excitation and relaxation, resulting in a significant reduction in normal liver signal, especially on T2-weighted images.[18]

Furthermore, iron oxide particles are particularly suited for hepatic imaging because they are cleared from the blood by phagocytosis. This results in uptake of the particles by the liver, spleen, bone marrow, and lymph nodes. Kupffer cells take up iron oxide particles, which results in a signal loss of T-2 weighted images.

Hepatic tumors, particularly metastatic lesions, cannot take up iron oxide particles because they either do not contain Kupffer cells or their activity is reduced. This difference in uptake of contrast medium results in an improved depiction of metastatic lesions on MR images (Figure 1).

Comparisons With Other Modalities

Reports of hepatic imaging with ferumoxides have demonstrated its improved ability to detect focal lesions by increasing the ratio of lesion-to-background signal intensity. Iron oxide-enhanced MRI has proven to be more sensitive than unenhanced MRI in detecting focal hepatic lesions.[19] In a large multicenter trial, iron oxide enhanced MRI identified 27% more lesions.[17]

Theoretically, iron oxide MRI will improve the detection and characterization of intrahepatic lesions, resulting in more accurate staging, more reliable treatment plans and options, and better selection of surgical candidates. Thus, unnecessary or deleterious surgery could be avoided, realistic patient expectations maintained, and medical resources used more efficiently.

The Baylor Experience

Although a number of studies have compared iron oxide MRI with other imaging modalities, few have attempted to correlate the findings of iron oxide MRI with surgical pathologic specimens, examine its impact on clinical management, or perform a cost-analysis. Furthermore, iron oxide MRI has received little attention in the surgical literature, and its exposure to some areas of the medical community has been limited.

In order to address some of these issues and further evaluate the clinical utility of iron oxide MRI, a retrospective review of the Baylor University Medical Center experience was performed. The goal of this study was to review the current use of and indications for iron oxide MRI for a single institution, compare its findings with CECT, determine its impact on patient management, and report its effect on the cost of care provided.

Methods and Materials

• Patients—The imaging methods applied to all patients undergoing abdominal MRI during a 20-month period (November 1996 to July 1998) were reviewed, All patients who underwent iron oxide MRI for the evaluation of hepatic lesions were selected for further study, and a retrospective chart review was performed.

Data collected included patient demographics, imaging indications, pre- and post-imaging diagnoses, other imaging modalities employed, management course, and cost of care delivered. Patient data were reviewed according to the indications for imaging: known extrahepatic cancer, primary hepatic cancer, or an indeterminate hepatic lesion. The following end points were chosen for changes in management: establishing a benign diagnosis, excluding metastatic hepatic disease, establishing resectability or unresectability, and/or guiding further operative ablative therapy.

• Diagnostic Imaging—All iron oxide MRI studies were performed at Baylor University Medical Center, Dallas, Texas. Prior to MRI examination, contrast-enhanced CT was performed using biphasic intravenous contrast (Omnipaque, 100-mL volume, 5 mL/s, with approximately 20- and 70-ms phase delay imaging), a helical scanner (GE Systems High Speed Advantage) with 7- to 10-mm thickness image acquisition, and a breath-hold technique.

In general, pre- and post-ferumoxides-enhanced MR images (Feridex I.V., 0.05 mL/Kg, 30-minute infusion) were acquired, except in cases with known metastatic disease for which only post-infusion images were obtained. Images were taken at 1.5 T using breath-hold coronal and axial T2-weighted imaging sequences (9-mm-thick slice, 0-mm gap; flip angle, 25° to 40°; repetition time [TR], approximately 150 ms; echo time [TE], 10 ms). In addition, axial spin echo (10-mm-thick slice, 2.5-mm gap; TR, 2,000 ms; TE, 30/80 ms), and optional breath-hold coronal and axial single-shot fast echo (SSFE) imaging techniques were applied.

Imaging studies were interpreted by board-certified radiologists specializing in CT and body MR imaging. Complementary studies were reviewed in conjunction with one another when available. Lesions were characterized according to number, size, and location.

• Data Analysis—The data of all patients selected for study were reviewed. Indications for imaging and changes in management were recorded for all cases. Patients who had concomitant iron oxide MRI and contrast-enhanced CT scans were selected for a comparison of the abilities of these two modalities to detect and characterize focal lesions. Only patients with solid hepatic lesions were included in the analysis.

Contrast-enhanced CT and iron oxide MRI findings were compared according to the number and size of lesions detected. Furthermore, the number of lesions detected by each modality was compared to a reference standard that included the findings of IOUS and pathologic examination, in order to determine the relative sensitivity and positive predictive value of iron oxide MRI and CT.

Changes in management were tabulated according to the indications for imaging, results, and impact of cost. Cost determinations were derived by adding the overall cost of hospitalization as billed to the patient (including the technical fee of imaging) and the professional fee as determined by the appropriate CPT Medicare fee. Statistical analysis was performed using the Primer of Statistics;[20] standard t-tests with a confidence level of 0.95 were used.

Results

• Patients and Indications—During a 20-month period, over 1,000 patients completed contrast-enhanced MRI of the abdomen at a single institution. A hepatic tumor was suspected in 57 of those who had undergone iron oxide MRI. This included 23 men and 34 women, with a mean age of 58 years (range, 31 to 87 years). Indications for imaging included evaluation of metastases from a known extrahepatic cancer (N = 43), the presence of an indeterminate hepatic mass (N = 9), or the evaluation of a primary hepatic cancer (N = 5).

Of those undergoing evaluation for suspected metastatic disease, the majority had a colorectal source (adenocarcinoma of the colon [N = 37]) while a minority had various other sources (breast [N = 3], esophageal [N = 1], prostate [N = 1], undetermined [N = 1]). Of the nine patients evaluated for an indeterminate hepatic mass, five were diagnosed with focal nodular hyperplasia, one with fatty infiltration, one with cholangiocarcinoma, one with hepatocellular carcinoma, and one with a hemangioma. Of the five patients with a known hepatic cancer, four were diagnosed with hepatocellular carcinoma and one with cholangiocarcinoma. Of the five patients with hepatocellular carcinoma, none had known cirrhosis, hepatic fibrosis, or clinical evidence thereof.

Overall, iron oxide MRI identified 157 lesions (mean, 2.8 lesions per patient; range, 0 to 14 lesions). In addition, 50 patients also had a concomitant CECT scan performed; of these, 35 were found to have solid focal hepatic lesions. A comparison was made of the findings of contrast-enhanced CT and iron-oxide MRI for these patients.

• Comparison of Iron Oxide MRI and CT Findings—As summarized in Table 1, among the 35 patients who underwent both CECT and iron oxide MRI, iron oxide MRI detected significantly more lesions (mean, 3.9 lesions per patient; range, 1 to 14 lesions) than did contrast-enhanced CT (mean, 2.2 lesions per patient; range, 0-7 lesions; P = .016). More lesions were detected by iron oxide MRI than contrast-enhanced CT in 16 patients (mean, 6.5 lesions per patient by iron oxide MRI vs 2.8 lesions per patient by CECT; P = .001), while no additional lesions were detected by iron oxide MRI in 18 patients. In one patient, iron oxide MRI failed to detect a lesion previously seen on CECT; however, pathologic examination confirmed that this was a CECT false-positive.

The average size of the lesions detected by the various modalities was compared as well (Table 2). Overall, the average size of the lesions detected by iron oxide MRI was significantly smaller than those detected by CECT (2.5 vs 3.4 cm, P = .018).

• Correlation of Iron Oxide-MRI and CECT Findings With IOUS and Pathology—In order to determine the accuracy of CECT and iron oxide MRI, these modalities were compared to the findings of IOUS and pathologic examination. Abdominal exploration for possible resection or ablation was performed in 22 patients. Of these, 19 were evaluated by IOUS, and 11 underwent hepatic resection (right trisegmentectomy [N = 1], right lobectomy [N = 5], segmental resection [N = 4], left extended [N = 1]). Two additional patients had a resection (right trisegmentectomy) without IOUS. One patient was not eligible for IOUS or resection due to gross extrahepatic disease, confirmed by frozen biopsy specimen.

In all 11 cases for which pathologic specimens were available, the number of lesions identified by IOUS correlated 100% with the pathologic specimens; ie, no additional lesions were found on pathology that were not identified by IOUS. Thus, 22 patients comprised the operative group that had intraoperative data and/or pathologic specimens for analysis.

The preoperative findings of CECT and iron oxide MRI for the operative group were juxtaposed against their corresponding intraoperative and pathologic data in order to evaluate their sensitivity for detection (Table 3). Both the IOUS and pathologic findings were combined as a reference standard for comparison.

First, the findings of iron oxide MRI were examined (Table 3). No additional lesions were detected in 16 patients (total number of lesions, 35); in 4 other patients, IOUS detected more lesions than iron oxide MRI, but difference was not statistically significant (number of lesions, 33 vs 22; P = 0.76). One lesion identified by iron oxide MRI was not seen by IOUS, but was confirmed as an iron oxide MRI false-positive in comparison to pathologic specimen. Overall, iron oxide MRI demonstrated a sensitivity of 86% for detecting solid hepatic lesions (significantly less than that of IOUS [100%; P = .003]), and had a corresponding positive predictive value of 93% (Table 4).

Contrast-enhanced CT findings were then compared to the findings of IOUS and pathology (Table 3). No additional lesions were detected by IOUS and pathology in 11 patients (total number of lesions, 15); however, significantly more lesions were identified by IOUS and pathology than by MRI in 15 patients (number of lesions, 53 vs 21; P = .02). In general, the number of lesions detected by CECT was significantly less than the number detected by IOUS (41 vs 71; P = .003). The same patient who had a false-positive finding by iron oxide MRI had two false-positives by CECT (CECT = 5, iron-oxide MRI = 4, IOUS/pathology = 3). Overall, as summarized in Table 4, the sensitivity of CECT (58%) was significantly lower than that of iron oxide MRI (86%; P < .001) and IOUS (100%; P < .001).

• Changes in Management—Clinical management was altered in 67% of patients who underwent iron oxide MRI for the evaluation of a hepatic lesion, based on the findings of iron oxide MRI (N = 38/57). These changes included: establishing a benign diagnosis and avoiding biopsy (N = 10), excluding malignant disease (N = 8), determining unresectability (N = 8), establishing resectability (N = 8), and guiding ablative therapy (N= 4; Table 5). According to the indications for imaging, the magnitude of changes in management varied from 57% to 100% per group (Table 5).

• Cost Considerations—The economic impact of these changes was then determined. The average total cost for each diagnostic modality was calculated by adding the total hospital charge and CPT Medicare professional fee. For diagnostic imaging and image-guided biopsy, the costs were: $1,625 for CECT, $1,740 for MRI, and $3,746 for image-guided biopsy. The average cost of operative intervention was determined from data on those who underwent such procedures and included: $17,515 for exploratory laparotomy, $19,326 for resection, $22,343 for cryosurgical ablation, and $28,525 for resection/ablation.

Overall, iron oxide MRI negated the need for liver biopsy in 18 patients and laparotomy in another 8; this represents a gross savings of $207,548 (Table 6). Factoring in the cost of iron oxide MRI for all patients studied ($99,180), this still represents an overall savings of $1,901 per patient studied.

Discussion

The evaluation of hepatic tumors is a challenging task for the surgeon. Ideally, preoperative imaging should have a high sensitivity for small lesions, a low false-positive rate, and be able to precisely localize a lesion in relationship to the main anatomic structures of the liver, namely, the main bile ducts, portal vessels, and hepatic veins. Although numerous imaging modalities have been used in an attempt to achieve these goals, no single technique fulfills all of these requirements. Consequently, the methods and goals of hepatic imaging have shifted over time.

Contrast-enhanced CT scanning has clearly replaced transabdominal ultrasound as the optimal preoperative imaging technique. Its sensitivity (68%) surpasses that of ultrasound (53%), but it is still limited even when combined with ultrasound.[20] Other series comparing the two modalities have supported these findings (sensitivity of CECT vs ultrasound, 67% vs33%) in comparison to IOUS as a reference standard.[12]

Dynamic helical CT scanning has improved the sensitivity of lesion detection (81%), but remains a poor method for dectecting lesions smaller than 1 cm (56%).[21] Enhancement of CT with arterial portography (CTAP) has demonstrated more sensitivity for detecting lesions (90%) in comparison to conventional CT (84%),[22] and it can detect lesions smaller than 1 cm.[23-26] However, CTAP is associated with a high-false positive rate (low specificity), flow artifacts (especially in the setting of cirrhosis), complexity of timing, potential morbidity of invasive angiography, and increased cost, compared to conventional CT.

Because of these limitations, most surgeons rely less on preoperative imaging and depend more on exploration with IOUS. Using this approach, lesions detected by IOUS resulted in a change in the planned operative management in 30% to 40% of cases.[12] Unfortunately, this may be associated with increased patient morbidity and increased overall health-care costs.

Impact of Recent Advances in MRI

Recent advances in MRI have brought this modality to the forefront as a promising noninvasive method for imaging hepatic tumors. The physics of MRI is well understood, and imaging studies have been shown to accurately distinguish between benign and malignant lesions.[27,28] Improved techniques and pulse sequences have upgraded the detection capabilities of MRI[13-15] to the extent that the sensitivity of MRI (90% tp 95%) now rivals that of CTAP and contrast-enhanced helical CT.[16,29] Furthermore, MRI has been shown to have a significantly greater impact on clinical management than CTAP and is considerably less expensive.[30]

The development of new IV contrast agents has also dramatically improved the capabilities of MRI. Several hepatic-specific MRI agents, including ferumoxides, have been described and reviewed.[18] Numerous early reports have characterized iron oxide particles and their application in MRI of human subjects.[19,31-38] A pilot study of imaging hepatic lesions with iron oxide particles demonstrated increased lesion detection and decreased size threshold.[19].

A recent prospective, multicenter trial also demonstrated improved lesion detection with iron oxide MRI in comparison to CECT.[17] In this multicenter trial, iron oxide MRI detected additional lesions in 40% of patients, mostly in the 0.5- to 1-cm range, and had a significant impact on clinical management (59%). In addition, adverse reactions to infusion of iron oxide particles were carefully monitored and well tolerated.

Some reports have suggested that the sensitivity of CTAP combined with biphasic contrast-enhanced helical CT (96%) is superior to iron oxide MRI (76%).[39] However, other reports have shown that the sensitivity of iron oxide MRI with certain pulse sequence techniques (spin-echo and gradient-echo) (95%) is superior to non–contrast-enhanced MRI (90%) and CTAP (92%).[40]

The relative merits of these various modalities have been well described and reviewed elsewhere.[41,42] With such changes in the strategies of hepatic imaging, CTAP currently plays a much smaller role as surgeons turn toward CECT and IOUS, and now have iron oxide MRI as a promising new noninvasive adjunct to preoperative planning.

Data from Baylor Support Benefit of MRI

The Baylor data demonstrate several of the important and encouraging capabilities of iron oxide MRI. First and foremost, iron oxide MRI can detect more hepatic lesions than CECT. Overall, 76% more lesions were detected by iron oxide MRI (P = .016), and the average size of lesions detected was significantly smaller than those detected by CECT.

Moreover, in a comparison of IOUS and pathologic findings, iron oxide MRI demonstrated a higher sensitivity than CT (86% vs 58%; P < .001), approaching the sensitivity of IOUS. The ability of iron oxide MRI to detect more and smaller lesions is underscored by its effect on clinical management and by the substantial savings in cost it produces.

Overall, iron oxide MRI altered the management of 67% of patients. Of the patients imaged, 46% were able to avoid further invasive diagnostic procedures, including 17% (N = 10) who were diagnosed with benign lesions. Most commonly, these lesions proved to be follicular nodular hyperplasia with functioning reticuloendothelial cells that appear as a negative contrast, characteristic of follicular nodular hyperplasia. Patients so diagnosed are now followed clinically, reimaged in 3 to 6 months, and routinely avoid invasive biopsy.

The ability to avoid biopsy of benign lesions, and to avoid laparotomy in patients with malignant lesions that are unresectable, provides a significant reduction in morbidity and health care cost. In this study, this benefit represented a savings of $1,901 per patient for all patients undergoing iron oxide MRI.

Thus, iron oxide MRI is a cost-effective, noninvasive method of preoperative evaluation. Furthermore, in comparison to CTAP, iron oxide MRI is less expensive, representing an additional savings of $1,759 per patient (if all patients would have otherwise undergone CTAP). This analysis does not take into account the cost and potential complications of further invasive diagnostic evaluation (biopsy or laparotomy) or invasive preoperative evaluation with CTAP. Thus, iron-oxide MRI likely has an additional clinical and economic impact on patient management that has yet to be measured.

In previous studies, IOUS changed the management in 30% to 40% of patients who were evaluated with preoperative contrast-enhanced CT.[12] This can have a further significant impact on patient outcome by leaving the clinician unprepared when additional unsuspected lesions are found that require other therapeutic modalities. There may be an impact on patient and family psychodynamics as well. In this study, iron-oxide MRI altered the preoperative clinical management of 50% of surgical patients in comparison to preoperative evaluation with contrast-enhanced CT. This impact on clinical management is further illustrated by the low rate of changes in clinical management (5%) based on IOUS when iron-oxide MRI was employed as a preoperative evaluation in this series.

Limitations Associated With Iron-Oxide MRI

Iron oxide MRI is a promising and useful tool in the preoperative evaluation of hepatic neoplasms. Although its imaging capabilities rival those of CECT and CTAP, and offers other distinct advantages over these modalities, iron oxide MRI does have limitations.

Currently, MR imaging with iron oxide contrast agents require prolonged infusion. Newer liver-specific contrast agents that can be administered as a bolus injection are under development (eg, Resovist Injection and Combidex). Magnetic resonance imaging is also limited in imaging the extrahepatic abdomen—an important aspect in the evaluation of metastatic disease for which CECT is better suited.

The sensitivity of MRI can be easily determined. However, caution must be used in interpreting the specificity and accuracy of various imaging modalities. Examination of pathologic specimens is the ideal standard by which to measure the rate of true- and false-positive studies. However, previous studies, as well as this one, are limited by the number of pathologic specimens that can be acquired.

Intraoperative ultrasound is the reference standard to determine the rate of true- and false-positive findings. This study also emphasizes the correlation of IOUS with pathologic findings, as no lesions were found on pathologic specimens that were not identified by IOUS.

Among those patients who underwent exploration with a comparison of CECT, iron oxide MRI, IOUS, and pathologic findings, the measured sensitivity of CECT in detecting lesions 1 cm or less may initially appear rather low (58%), compared to other studies. However, care must be taken when comparing this figure with other studies. In this series, 59 additional lesions were detected by MRI that were not detected by CECT. Their median size was 1 cm. Thus, not surprisingly, there were a considerable number of smaller lesions that were not detected by CECT, which lowered its sensitivity significantly.

Furthermore, the time delay between CECT and iron oxide MRI may result in disease progression, with more lesions reaching the detection size threshold by the time MRI takes place. Patients who have already undergone CT, and are referred for surgical evaluation, do not undergo unnecessary repeat CT, but rather, are then evaluated by iron oxide MRI. In general, patients are considered to have concomitant CECT and MRI if the scans are performed within 30 days of one another. In our study, the total average time interval between studies was 23 days.

Conclusions

Iron oxide MRI is a promising method of preoperative evaluation for resection and/or further treatment of hepatic neoplasms. It clearly has a significant impact on clinical management and can provide substantial cost-savings. It should be used as an integrated method of preoperative evaluation along with ultrasound and CT, but may replace the need for CTAP.

The application of iron-oxide MRI improves preoperative treatment planning and avoids unnecessary laparotomy in unresectable cases. It also helps in the selection of alternative treatment modalities. Consequently, the morbidity and mortality associated with hepatic resection can potentially be lowered.

References:

1. Okuda K, Ohtsuki T, Obata H, et al: Natural history of hepatocellular carcinoma and prognosis in relation to treatment. Study of 850 patients. Cancer 56(4):918-928, 1985.

2. Nagasue N, Kohno H, Chang YC, et al: Liver resection for hepatocellular carcinoma. Results of 229 consecutive patients during 11 years. Ann Surg 217(4):375-384, 1993.

3. Tsuzuki T, Sugioka A, Ueda M, et al: Hepatic resection for hepatocellular carcinoma. Surgery 107(5):511-520, 1990.

4. Arii S, Tanaka J, Yamazoe Y, et al: Predictive factors for intrahepatic recurrence of hepatocellular carcinoma after partial hepatectomy. Cancer 69(4):913-919, 1992.

5. Matsuda Y, Ito T, Oguchi Y, et al: Rationale of surgical management for recurrent hepatocellular carcinoma. Ann Surg 217(1):28-34, 1993.

6. Bengmark S, Hafstrom L: The natural history of primary and secondary malignant tumors of the liver. I. The prognosis for patients with hepatic metastases from colonic and rectal carcinoma by laparotomy. Cancer 23(1):198-202, 1969.

7. Ekberg H, Tranberg KG, Andersson R, et al: Determinants of survival in liver resection for colorectal secondaries. Br J Surg 73(9):727-731, 1986.

8. Hughes KS, Simon R, Songhorabodi S, et al: Resection of the liver for colorectal carcinoma metastases: A multi-institutional study of patterns of recurrence. Surgery 100(2):278-284, 1986.

9. Hughes KS: Resection of the liver for colorectal carcinoma metastases: A multi-institutional study of indications for resection. Registry of Hepatic Metastases. Surgery 103(3):278-288, 1988.

10. Stone MD, Cady B, Jenkins RL, et al: Surgical therapy for recurrent liver metastases from colorectal cancer. Arch Surg 125(6):718-721, 1990.

11. Griffith KD, Sugarbaker PH, Chang AE: Surgery Branch NCIBM. Repeat hepatic resections for colorectal metastases. Surgery 107(1):101-104, 1990.

12. Staren ED, Gambla M, Deziel DJ, et al: Intraoperative ultrasound in the management of liver neoplasms. Am Surg 63(7):591-596, 1997.

13. Moss AA, Goldberg HI, Stark DB, et al: Hepatic tumors: Magnetic resonance and CT appearance. Radiology 150(1):141-147, 1984.

14. Heiken JP, Lee JK, Glazer HS, et al: Hepatic metastases studied with MRI and CT. Radiology 156(2):423-427, 1985.

15. Stark DD, Wittenberg J, Butch RJ, et al: Hepatic metastases: Randomized, controlled comparison of detection with MR imaging and CT. Radiology 165(2):399-406, 1987.

16. Semelka RC, Shoenut JP, Kroeker MA, et al: Focal liver disease: Comparison of dynamic contrast-enhanced CT and T2-weighted fat-suppressed, FLASH, and dynamic gadolinium-enhanced MR imaging at 1.5 T. Radiology 184(3):687-694, 1992.

17. Ros PR, Freeny PC, Harms SE, et al: Hepatic MR imaging with ferumoxides: A multicenter clinical trial of the safety and efficacy in the detection of focal hepatic lesions. Radiology 196(2):481-488, 1995.

18. Hahn PF, Saini S: Liver-specific MR imaging contrast agents. Radiol Clin North Am 36(2):287-297, 1998.

19. Stark DD, Weissleder R, Elizondo G, et al: Superparamagnetic iron oxide: Clinical application as a contrast agent for MR imaging of the liver. Radiology 168(2):297-301, 1988.

20. Primer of Statistics, McGraw-Hill, 1992.

21. Wernecke K, Rummeny E, Bongartz G, et al: Detection of hepatic masses in patients with carcinoma: Comparative sensitivities of sonography, CT, and MR imaging. AJR Am J Roentgen 157(4):731-739, 1991.

22. Kuszyk BS, Bluemke DA, Urban BA, et al: Portal-phase contrast-enhanced helical CT for the detection of malignant hepatic tumors: Sensitivity based on comparison with intraoperative and pathologic findings. AJR Am J Roentgen 166(1):91-95, 1996.

23. Soyer P, Bluemke DA, Hruban RH, et al: Hepatic metastases from colorectal cancer: Detection and false-positive findings with helical CT during arterial portography. Radiology 193(1):71-74, 1994.

24. Heiken JP, Weyman PJ, Lee JK, et al: Detection of focal hepatic masses: Prospective evaluation with CT, delayed CT, CT during arterial portography, and MR imaging. Radiology 171(1):47-51, 1989.

25. Matsui O, Kadoya M, Suzuki M, et al: Work in progress: Dynamic sequential computed tomography during arterial portography in the detection of hepatic neoplasms. Radiology 146(3):721-727, 1983.

26. Nelson RC, Chezmar JL, Sugarbaker PH, et al: Hepatic tumors: Comparison of CT during arterial portography, delayed CT, and MR imaging for preoperative evaluation. Radiology 172(1):27-34, 1989.

27. Lupetin AR, Cammisa BA, Beckman I, et al: Spiral CT during arterial portography. Radiographics 16(4):723-743, 1996.

28. Winston CB, Schwartz LH: Advances in magnetic resonance imaging: Applications in body imaging. Cancer Investig 16(6):413-420, 1998.

29. Mitchell DG: Focal manifestations of diffuse liver disease at MR imaging. Radiology 185(1):1-11, 1992.

30. Reinig JW, Dwyer AJ, Miller DL, et al: Liver metastasis detection: Comparative sensitivities of MR imaging and CT scanning. Radiology 162(1 Pt 1):43-47, 1987.

31. Semelka RC, Schlund JF, Molina PL, et al: Malignant liver lesions: Comparison of spiral CT arterial portography and MR imaging for diagnostic accuracy, cost, and effect on patient management. J Magn Reson Imag 6(1):39-43, 1996.

32. Saini S, Stark DD, Hahn PF, et al: Ferrite particles: A superparamagnetic MR contrast agent for enhanced detection of liver carcinoma. Radiology 162(1 Pt 1):217-222, 1987.

33. Saini S, Stark DD, Hahn PF, et al: Ferrite particles: A super paramagnetic MR contrast agent for the reticuloendothelial system. Radiology 162(1 Pt 1):211-216, 1987.

34. Gore JC, Majumdar S: Superparamagnetic iron oxide particles. Radiology 169(3):657-658, 1988.

35. Weissleder R, Stark DD, Engelstad BL, et al: Super paramagnetic iron oxide: Pharmacokinetics and toxicity. AJR Am J Roentgen 152(1):167-173, 1989.

36. Fretz CJ, Stark DD, Metz CE, et al: Detection of hepatic metastases: Comparison of contrast-enhanced CT, unenhanced MR imaging, and iron oxide-enhanced MR imaging. AJR Am J Roentgen 155(4):763-770, 1990.

37. Marchal G, Van Hecke P, Demaerel P, et al: Detection of liver metastases with super paramagnetic iron oxide in 15 patients: Results of MR imaging at 1.5 T. AJR Am J Roentgen 152(4):771-775, 1989.

38. Fretz CJ, Elizondo G, Weissleder R, et al: Super paramagnetic iron oxide-enhanced MR imaging: Pulse sequence optimization for detection of liver cancer. Radiology 172(2):393-397, 1989.

39. Ferrucci JT, Stark DD: Iron oxide-enhanced MR imaging of the liver and spleen: Review of the first 5 years. AJR Am J Roentgen 155(5):943-950, 1990.

40. Strotzer M, Gmeinwieser J, Schmidt J, et al: Diagnosis of liver metastases from colorectal adenocarcinoma. Comparison of spiral-CTAP combined with intravenous contrast-enhanced spiral-CT and SPIO-enhanced MR combined with plain MR imaging. Acta Radiologica 38(6):986-992, 1997.

41. Seneterre E, Taourel P, Bouvier Y, et al: Detection of hepatic metastases: Ferumoxides-enhanced MR imaging vs unenhanced MR imaging and CT during arterial portography. Radiology 200(3):785-792, 1996.

42. Taylor HM, Ros PR: Hepatic imaging. An overview. Radiol Clin North Am 36(2):237-245, 1998.

43. Paley MR, Ros PR: Hepatic metastases. Radiol Clin North Am 36(2):349-363, 1998.