Punam Bajaj

Memorial Sloan-Kettering Cancer Center, New York City, NY, United States

Are you Punam Bajaj?

Claim your profile

Publications (4)6.34 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE To analyze a possible correlation between morphology prior to treatment, RECIST response, volumetric reduction, shrinkage pattern and disease free survival (DFS) in locally advanced breast cancers (LABC). METHOD AND MATERIALS This is a retrospective analysis of 52 women (mean age 44 years, range: 31-69 years) with LABC who had breast MR before and after neoadjuvant chemotherapy. Dynamic breast MR was acquired using 1.5 or 3 T scanners with dedicated breast coils. 3D-tumor volume measurements were done with automated segmentation of MR images using Sentinelle Aegis 2.0.1. Morphology of tumor at diagnosis (solitary, grouped in same quadrant , separated in different quadrants , total breast replaced by tumor), response to treatment according to RECIST, volumetric response (complete, partial volume reduction: 65-99%, stable: volume reduction < 64%) as well as shrinkage pattern (no residual enhancement, concentric shrinkage, concentric with surrounding lesions, multinodular lesions, diffuse enhancement of the breast) were evaluated and correlated to DFS defined in months after the end of chemotherapy. DFS was evaluated on mammography and PET-CT or total- body CT. Receptor status was also correlated to DFS. Pathological response was defined according to pathology report. Mantel-Cox test was used for the statistical analysis. RESULTS Agreement in the evaluation of treatment response between diameter at MR and pathology was observed in 38/52 (73%), overestimated in 4/52 (8%) and underestimated in 10/52(19%). Tumor morphology at diagnosis did not significantly correlate to DFS (p=0.33). RECIST versus volumetric response in women with complete, partial or stable disease did not significantly affect the DFS (p=0.6, 0.24 and 0.31 respectively). DFS was independent of shrinkage pattern of tumors (p=0.69). No statistical difference in DFS was found between complete and residual pathologic response (p= 0.76). The percentage of volume reduction was higher in triple negative cancers compared to others (p=0.003), but DFS was significantly lower (27 months) than the other group (66 months)( p=0.04). CONCLUSION DFS in LABC is not related to pre and post treatment breast MR morphology. CLINICAL RELEVANCE/APPLICATION DFS in locally advanced breast cancer is better predicted by tumor receptor status than tumor morphology at MR, both prior to and after treatment.
    Radiological Society of North America 2013 Scientific Assembly and Annual Meeting; 12/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE To determine if suspicious breast MRI lesions proven to represent invasive ductal carcinoma with an ultrasound correlate are of different histological grade compared with ultrasound occult lesions. METHOD AND MATERIALS Institutional review board approved retrospective study of 310 MRI examinations performed between 2008 and 2011 yielded 350 suspicious lesions for which biopsy was recommended. Subsequent high resolution targeted ultrasound was performed and histopathological grade of carcinomas was recorded as I (low), II (intermediate) or III (high). Statistical analysis was performed applying the Fisher's exact test ,Kruskal-Wallis test and exact Wilcoxon rank sum test. RESULTS Targeted ultrasound demonstrated a correlate in 181/350 (52%) suspicious MRI lesions yielding 63/181 (35%) malignant lesions. The remaining 169 (48%) lesions which were sonographically occult, yielded 25/169 (15%) malignant lesions. Sonographic correlates were seen for 72% (63/88) of malignant lesions. Of these, 87% (55/63) were invasive carcinomas and 13% (8/63) were ductal carcinomas in situ. Histological grade was available for 46 invasive ductal carcinomas with ultrasound correlate (3(6.5%),13(28.3%) and 30(65.2%) were histological grade I,II and III, respectively)and 8 without correlate (4(50%),3(37.5%) and 1(12.5%) were histological grade I,II and III, respectively). There was no statistically significant difference in the size of tumors with or without an ultrasound correlate (p=0.163). In the group with an ultrasound correlate, no significant difference was observed in tumor size between the recorded histological grades (p=0.052). A grade III tumor was more likely to be present in the group with an ultrasound correlate (p <0.001). CONCLUSION When a suspicious breast MRI lesion has an ultrasound correlate, it is more likely to represent invasive carcinoma of higher histological grade. CLINICAL RELEVANCE/APPLICATION The presence of an ultrasound correlate for a suspicious breast MRI lesion may indicate a more aggressive cancer.
    Radiological Society of North America 2013 Scientific Assembly and Annual Meeting; 12/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE To compare the value of the Apparent Diffusion Coefficient (ADC) of normal breast tissue in high risk women who underwent mantle radiation before ages 30 years and a screening control group, matched for breast tissue density. METHOD AND MATERIALS This was a retrospective analysis of breast MRI examinations performed between 2008 and 2013 of 21 women who were treated with mantle radiation. Based on breast tissue density evaluated on pre contrast T1W fat suppressed sequence, cases were divided into two groups: fatty breasts (n=10) and dense breasts(n=11). A total of 41 breasts were evaluated. One breast with a benign mass lesion in the retroareolar region was excluded. Breast MRI exams in 21 controls, matched for breast tissue density, were compared. Diffusion Weighted Imaging(DWI) with b=0, 1000 was performed for all cases and controls and the ADC maps were evaluated on advantage workstation (GE). ADC value of normal breast tissue were calculated by placement of the region of interest (ROI) in bilateral retroareolar regions. Normal breast tissue was defined as absence of any finding on dynamic MRI (BIRADS-1). Statistical analysis was performed using Mann- Whitney unpaired t test. RESULTS Quantitative DWI study showed that median ADC value of irradiated breasts was lower ( 1.32 x 10-3mm2/sec) compared to the non-irradiated control group ( 1.63 x 10-3mm2/sec) ( p value= 0.0004). When matched for breast tissue density, irradiated fatty breasts had lower median ADC value (1.23 x 10-3mm2/sec) compared to control group (1.54 x 10-3mm2/sec) (p=0.0018). The irradiated dense breasts also showed a similar, statistically significant, lower median ADC value (1.59 x 10-3mm2/sec) compared to the controls with dense breasts (1.77 x 10-3mm2/sec) (p value=0.0435). CONCLUSION Mantle radiated breasts have lower ADC values compared to non-irradiated breasts. This difference in ADC value is independent of the breast tissue density. CLINICAL RELEVANCE/APPLICATION Women who have received mantle radiation have lower ADC values, probably due to post-radiation fibrosis which affects the tissue diffusion.
    Radiological Society of North America 2013 Scientific Assembly and Annual Meeting; 12/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: To assess the utility of screening magnetic resonance (MR) imaging in the detection of otherwise occult breast cancers in women with a history of lobular carcinoma in situ (LCIS). This HIPAA-compliant study received institutional review board approval. The need for informed consent was waived. Retrospective review of the database yielded 670 screening breast MR studies obtained between January 2003 and September 2008 in 220 women with a history of LCIS. MR and mammographic findings were reviewed. Number of cancers diagnosed, method of detection, and tumor characteristics were examined. The cumulative incidence of developing breast cancer as detected with MR imaging and mammography was calculated. Breast density was examined as a prognostic factor in the cumulative incidence analysis. Biopsy was recommended in 63 lesions seen in 58 (9%) of 670 screening MR studies. Eight additional lesions were identified at short-term follow-up MR imaging for a total of 71 lesions in 59 patients. Twelve cancers (20%) were identified in 60 lesions sampled. Biopsy was recommended in 26 additional lesions identified at mammography; biopsy was performed in 25 of these lesions and revealed malignancy in five (20%). Overall, 17 cancers were detected in 14 patients during the study period. Of these, 12 were detected with MR imaging alone, and five were detected with mammography alone. Of the 12 cancers detected at MR imaging, there were nine invasive cancers and three cases of ductal carcinoma in situ (DCIS). Of the five cancers detected at mammography, two were invasive and three were DCIS. MR imaging is a useful adjunct modality with which to screen women with a history of LCIS at high-risk of developing breast cancer, resulting in a 4.5% incremental cancer detection rate. Sensitivity in the detection of breast cancers with a combination of MR imaging and mammography was higher than sensitivity of either modality alone.
    Radiology 09/2011; 261(2):414-20. · 6.34 Impact Factor