Diagnostic Utility and Significance of Performing Multistep Level Sections in Breast and Gynecologic Biopsy Specimens

Al Amin, MD; Sayeeda Yasmeen, MD, MPH; Devi Jeyachandran, MD; Thaer Khoury, MD; Mohamed Mokhtar Desouki, MD, PhD


Am J Clin Pathol. 2021;156(4):620-624. 

In This Article


The senior authors are specialized breast (T.K., M.M.D., and S.Y.) or gyneologic (M.M.D. and D.J.) pathologists. Therefore, they were interested in exploring the diagnostic utility and significance of performing multistep level sections in breast and gynecology biopsy specimens in a single report.

Performing multistep level sections is a well-known practice in the laboratories to increase the diagnostic yield and to minimize the diagnostic mistakes largely caused by sampling inadequacy.[7] Theoretically, cutting multiple levels increases the chance to explore lesions that become evident deeper in the tissue block.[8–11] On the other hand, multistep level sections may waste the valuable sample (eg, small tubular adenoma in colonoscopic biopsy specimen).[12] With personalized medicine, a clinician may not pay attention to the detailed morphologic classification of a lung tumor on a small lung biopsy specimen and prefer PD-L1 or EGFR molecular testing rather than wasting the tissue to run many IHC markers as multiple panels.[13]

This subject has been studied on genitourinary biopsy specimens,[6] sentinel lymph nodes from breast cancer,[14] endometrial cancer,[15] melanoma,[16] gynecology biopsy specimens,[3,9,10] and other organs.[11]In the current study, we reported the significance of obtaining multistep level sections on atypical and malignant breast and gynecology biopsy specimens. The number of tissue sections mounted on the three H&E-stained slide levels in a single case varies with a range of 3 to 12. A diagnosis could have been missed in 3.3% of cases with no diagnostic material present on the first level. Eighteen cases were caught on the second level while the other 8 cases had diagnostic material on the third level. It should be noted that these results are valid for our specific protocol—namely, slides 1, 5, and 9 from the breast; 1, 3, and 5 from the gynecology biopsy specimens stained with H&E, and the intervening unstained sections (saved for further studies, if needed), and 50 μm discarded in between except for very small biopsy specimens. Fadare and Rodriguez[3] reported that if sectioning were limited to only one level, 17.5% of the dysplastic lesions of the cervix would have been missed, including 5.6% of high-grade dysplasia.

Multistep level sections resulted in the diagnostic material to show up in subsequent multistep level sections or became larger in 15.16% when we stained a second level and an extra 3.31% when we stained a third level. On the other hand, the diagnostic material had been depleted in 15 (1.9%) of 785 cases, of which 1 case had invasive breast carcinoma, 2 cases had DCIS, and the remainder of the 12 cases had ALH, FEA, cervical dysplasia, and atypical endometrial hyperplasia. The small fraction of cases among our cohort in which the intervening unstained sections were used (3.4%) for IHC was primarily due to our ordering system, which is not properly set up to request IHC on the upfront prepared unstained sections. Instead, when needed, the pathologist has to contact the histology laboratory in every case to request performing the IHC on the intervening unstained levels.

The optimal number of multistep level sections and establishing a policy for the laboratory may be a challenging task that depends on multiple factors, including, but not limited to, financial budget, personnel staffing, insurance reimbursement, and the characteristics of the lesions in the population serviced by a laboratory. At our cancer institution, most of the biopsy specimens are from high-risk cancer patients who have a personal or family history of cancer or highly suspicious lesions detected clinically or by radiologic investigations.

The limitations of this study are (1) the retrospective design; (2) not including biopsy specimens from other organs; (3) the role of subjective assessments in diagnosing atypical lesions, with one pathologist recognizing the lesion at level 1 and another pathologist recognizing it only at level 3; and (4) the subjectivity in assessing subtle changes in size of tissue between consecutive levels. In addition, we did not address the cost of adding more slides and tissue levels, which has been the subject of multiple studies.[17–19]

In conclusion, we are reporting the significance of obtaining multistep level sections on breast and gynecology biopsy specimens at our institution. Staining two level sections with H&E significantly affects the diagnosis, while preparing a third level does not significantly improve diagnosis. Therefore, preparing tissue sections on two levels is the recommended procedure. A universal protocol should be considered to standardize the handling of biopsy specimens among laboratories.