Negative Predictive Value Of MRI Research Articles

In newly diagnosed breast cancer, is a contralateral prophylactic mastectomy necessary following a negative MRI?

627 Background: With more centers using MRI to aid in initial staging, there is apprehension that the highly sensitive MRI exam may lead to “unnecessary” mastectomies. We evaluated whether a negative contralateral MRI could be used to rule out a synchronous contralateral carcinoma and hence avoid a prophylactic contralateral mastectomy (CPM) at initial breast cancer diagnosis. Methods: Data of women with a new diagnosis of breast cancer (invasive or DCIS) from February 2003-November 2007 at our institution were reviewed. Women who had a negative MRI of the contralateral breast and who subsequently underwent CPM were included. A CPM was one in which there was no histological confirmation of cancer in the contralateral breast before surgery. All mastectomy specimens were analyzed pathologically. True negatives were mastectomy samples which contained no cancer following pathologic examination. False negatives were mastectomy samples which were found to have cancer after pathologic review. Results: 538 women were identified. 51 women met the study criteria. Median patient age was 56 years (range, 33–77). Of the 51 women who underwent a CPM, 2 were found to have pathologic confirmation of cancer. One occult cancer was a 0.3 cm grade 1, invasive lobular carcinoma; the second a 0.5 cm, low grade DCIS. The overall prevalence of synchronous occult contralateral cancer after a negative MRI was 3.9% (95% CI 0.5 - 13.5%). The negative predictive value of MRI was 96.1% (95% CI 86.5 - 99.5%). Conclusions: This study provides pathologic data regarding the utility of a negative contralateral breast MRI in women with newly diagnosed breast cancer. The prevalence of a synchronous occult contralateral cancer after a negative MRI in women with newly diagnosed breast cancer is very low. Physicians and patients can be assured that the risk of occult cancer is low in the setting of a normal contralateral MRI, an important fact when considering CPM after a diagnosis of unilateral breast cancer. No significant financial relationships to disclose.

Labyrinthitis ossificans: How accurate is MRI in predicting cochlear obstruction?

Objective To determine the accuracy of preoperative MRI in predicting cochlear obstruction in pediatric patients with a history of bacterial meningitis. Methods A case series with chart review was performed at a tertiary care multidisciplinary cochlear implant program. Forty-five children with hearing loss that resulted from bacterial meningitis were implanted from 1991 to 2006. Twenty-five children had preoperative MRI with high-resolution axial T2-weighted images to assess for cochlear patency. Results Seventeen of 25 patients (68%) had surgical evidence of cochlear obstruction. Six patients (37.5%) required circummodiolar drill-outs, and one patient (6.25%) underwent placement of a double array cochlear implant. The nine remaining patients (56%) with cochlear obstruction required removal of fibrous tissue or drilling of the inferior basal turn, but did not require manipulation of the ascending basal turn to achieve full electrode insertion. The sensitivity, specificity, and positive and negative predictive value of MRI predicting intraoperative cochlear obstruction with 95 percent confidence intervals was 94.1 percent (71-99), 87.5 percent (47-99), 94.1 percent (71-99) and 87.5 percent (47-99), respectively. Conclusion Preoperative high-resolution T2 MRI may be useful in predicting cochlear obstruction in patients with a prior history of bacterial meningitis.

MRI features of the radial meniscal tears

Objective To access the ability of MRI for detecting the radial meniscal tears.Methods Arthroscopy of the knee was performed by a single orthopedic surgeon on 1085 patients. On the basis of the results of arthroscopy, 107(9.9%) knees had radial tears of menisci. Through interval sampling of the patients with meniscal tears but not radial tears, 100 patients were selected as a controlled group. All MR images were analyzed by two radiologists independently. Four radiological signs were used to detect radial tears: cleft and marching cleft sign, detached sign, ghost sign, and small meniscal sign. All the features of MRI were compared with the results of the arthroscopy. Results (1) The most frequent location of radial tears was the midbody of lateral meniscus (71/107, 66. 4%)followed by the boundary of anterior horn and midbody of the lateral meniscus (25/107, 23.4%). (2) One hundred and two of 107 patients were identified as radial tears on MRI. However, 11 of 100 patients with non-radial tears were misdiagnosed as radial tears by MRI. The sensitivity, specificity, positive predictive value, and negative predictive value of MRI were 95.3% (102/107),89.0% (89/100),90.3% (102/113), and 94.7% (89/94),respectively. (3) The cleft sign (59/102, 57.8%) and detached sign (43/102, 42.2%) were the most frequent and important MRI signs. Conclusion The accuracy of MRI to identify the radial meniseal tears'is very high and MRI is the best imaging modality for patients suspected of radial meniscal tears. Key words: Menisci, Tibia; Lacerations; Magnetic resonance imaging; Arthroscopy

Caractéristiques IRM de l’endométriose profonde : corrélation aux résultats cœlioscopiques

Deep pelvic endometriosis is an invalidating disorder affecting the retrocervical region, rectosigmoid colon and urinary bladder generally requiring surgical management. MRI is the preoperative imaging modality of choice. The purpose of this paper is to describe the MR imaging features of deep pelvic endometriosis with laparoscopic correlation. Thirty-five patients with clinical suspicion of deep pelvic endometriosis underwent pelvic MRI. Results of MRI, including morphological and signal characteristics features of the lesions were compared to laparoscopic fidings. Laparoscopy detected lesions of deep pelvic endometriosis of the uterosacral ligaments (n=10), torus uterinum (n=9), rectosigmoid (n=11), Douglas pouch (n=9), recto-vaginal septum (n=6), bladder (n=4) and posterior vaginal cul-de-sac (n=2). The sensitivity, specificity, positive predictive value and negative predictive value of MRI were assessed for each localization. MRI allows diagnosis of deep pelvic endometriosis of the bladder, rectosigmoid and Douglas pouch and with lower sensitivity for lesions of the uterosacral ligaments, posterior vaginal cul-de-sac and rectovaginal septum.

Real-time ultrasound elastography for monitoring tumor response to neoadjuvant chemotherapy in primary breast cancer

14023 Background: Response to neoadjuvant treatment is vital to predict a patient’s long-term survival. Precise detection of residual tumor cells after neoadjuvant chemotherapy would allow a better cosmetic results avoiding over surgery and reduce second operation due to positive margin status. Moreover, accurate prediction of pathological CR will yield no surgical intervention in certain population. Recently, a new generation of ultrasound platforms with real-time freehand elastography that enables the imaging of elasticity of the lesion by using the extended combined autocorrelation method (CAM) has become available. We are currently applying this technology to our patients with primary breast cancer in an attempt to assess response to neoadjuvant chemotherapy in comparison with MRI, conventional ultrasound and pathological findings. Methods: A total of 38 patients with primary breast cancer who underwent neoadjuvant chemotherapy and following surgical resection From May 2005 to Dec 2006 were included in this study. Board certified radiologists assessed the tumor response by MRI, US and US Elastography prior to surgery. Positive predictive value (PPV), and negative predictive value (NPV) for pathological CR (pCR) was assessed. Tsukuba Elastography score was applied for the assessment of Elastography. Results: 11/38 patients (28.9%) achieved a pCR in breast to neoadjuvant chemotherapy while no patients demonstrated progressive disease. The PPV for pCR of MRI and US was 54.5% and 36.3% respectively. The NPV of MRI and US was both 90.9%. None of the residual tumor mass with score 4 or 5 cases diagnosed by Elastography achieved pCR. When residual tumor image was detected by US, pCR was present in all 4 cases with score 1 or 2 Elatography. If the cut-off line is determined between score 3 and 4, the PPV and NPV for pCR by Elastography was 100% and 66.6% respectively. Conclusions: Elastography is easy to perform and it can provide an inexpensive, non-invasive, real-time tool for assessment of response to neoadjuvant chemotherapy among patients with primary breast cancer. In particular, Elastography might more effectively diagnose pathological CR. More patients are needed to evaluate the sensitivity and specificity of this new technology. No significant financial relationships to disclose.

Deficiencies of MRI in the diagnosis of chronic symptomatic lateral ankle ligament injuries

Background: Stress radiography and more recently magnetic resonance imaging have been used to study the integrity of lateral ankle ligaments in chronic symptomatic instability after injury. Aim: Our aim was to see if magnetic resonance imaging was as good as examination under anaesthesia and stress radiography, for diagnosing injury to the lateral ankle ligaments. Methodology: Fifty eight patients, 47 men and 11 women, who were athletes or military personnel, with symptomatic instability of their ankle were included in the study. This cohort of patients had MRI scans, stress radiography and arthroscopy of their ankle. Integrity of the calcaneo-fibular ligament (CFL) was recorded arthroscopically.The sensitivity, specificity, positive and negative predictive value of MRI and stress views, in assessing integrity of the CFL, were compared against arthroscopic findings. Results: Stress radiography under anaesthesia and MRI had sensitivity of 94% and 47% respectively and specificity of 98% and 83% respectively, for diagnosing injury to the CFL. Stress radiography has a higher accuracy in diagnosing CFL injuries as compared to MRI. Conclusion: The results of this study casts doubt on the efficacy of MRI in the diagnosis of serious ankle ligament injuries.

Effectiveness of dynamic MRI for diagnosing pericicatricial minimal residual breast cancer following excisional biopsy

The purpose of this study was to investigate the effectiveness of dynamic MRI for diagnosing pericicatricial minimal residual breast cancer following excisional biopsy. Twenty-six patients who underwent excisional biopsy of a tumor or calcified lesion of the breast underwent gadolinium-enhanced dynamic MRI by the fat-saturated 2D fast spoiled gradient echo (SPGR) sequence (group 1), 24 patients by the spectral IR enhanced 3D fast gradient echo (Efgre3d) sequence (group 2). Pericicatricial residual cancer was confirmed histologically in 29 of the 50 patients. The overall sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of MRI for residual cancer diagnosis was 66, 81, 72, 83 and 63%. A nodular, thick and discontinuous enhanced rim around the scar is indicative of a residual tumor. However, false-positive findings due to granulation or proliferative fibrocystic change remain limitations.

Prospective Evaluation of Magnetic Resonance Imaging in the Management of Acute Diabetic Foot Infections

Infectious complications of the foot are a major cause of morbidity and mortality in diabetic patients. This prospective study evaluated the ability of MRI to adequately direct the medical and surgical management of 41 diabetic patients with acute foot infections. Forty-seven MRI scans of the foot in question were performed and classified as consistent with osteomyelitis, abscess, cellulitis or diffuse soft tissue infection, or any combination of these. Twenty-seven scans were negative or showed ill-defined soft tissue infection, or superficial cellulitis. Nineteen of these infections were treated nonoperatively and 17 resolved without surgical intervention. MRI was unsuccessful in directing management in one patient in whom an abscess spontaneously drained but was not seen on an MRI scan 4 days earlier. Eight scans revealed focal osteomyelitis and all eight of these patients were successfully managed with one operation. MRI showed a focal abscess in 12 patients, and adequate drainage was achieved without excessive disruption of uninvolved tissue planes in 11 of these patients. The remaining patient required a major amputation from the outset. Based on clinical outcome during the acute hospitalization period, operative findings, and/or pathologic confirmation, the positive predictive value of MRI in defining infectious pathology in the foot was 100% in this series of 20 positive scans. The negative predictive value of MRI was 96%. On the basis of this experience, we conclude that MRI is a diagnostic modality particularly well suited to evaluate acute diabetic foot infections and reliably aids in the management of acute infection to avoid exploration and debridement of uninvolved tissue.(ABSTRACT TRUNCATED AT 250 WORDS)