Intraoperative Consultation and Optimal Processing

Types of Neurosurgical Specimens

The physical size of tissue specimens submitted for frozen section evaluation varies greatly ( Box 3.1 ). Specimen size is directly related to the goal of the surgical procedure. Endoscopic and stereotactic biopsies are typically performed for diagnostic purposes only, whereas at the other end of the spectrum, hemispherectomies and lobectomies are often performed for surgical cure under the best of circumstances. Between these two extremes lies a large range of specimen sizes encompassed by open biopsies, partial resections, and gross total resections. Despite this variation in specimen size, some general principles are more or less universally applicable ( Box 3.2 ). The two most helpful rules are (1) perform a cytologic preparation as a complement to the frozen tissue section, and (2) never freeze all of the lesional tissue; that is, always save some tissue for formalin-fixed paraffin-embedded (FFPE) permanent sections. A rare exception to the latter rule occurs when the surgeon had intended to obtain additional tissue for optimal preservation, but was unable due to intraoperative complications that arose.

Sometimes, definitive tumor is identified on the frozen section but unusual or unexpected morphologic features are apparent. In such cases it is prudent, if sufficient tissue is available, to place a small representative fragment in glutaraldehyde for ultrastructural examination, if needed. If glutaraldehyde is not available, retaining a small amount of tissue in formalin is the next best option. Retrieving tissue from paraffin blocks followed by postfixation and processing for electron microscopy is possible, but often yields suboptimal ultrastructural detail. In the current era of advanced molecular diagnostics, freezing a small portion of tumor tissue might be considered by the pathologist or requested by treating physicians, as frozen tissue provides better preservation of nucleic acids than does FFPE tissue. Tissue to be frozen must be selected carefully as the reliability of downstream molecular tests depends on it. It is essential to preserve tissue representative of the lesion for histologic examination, but to the extent that gross examination permits, one must also include tissue representative of the lesion in the samples to be frozen. Frozen tissue has the limitation that the tumor content is unknown, since it has not been evaluated microscopically. Therefore, results of downstream molecular testing should be interpreted with caution, as little or no tumor content in the frozen tissue might lead to false negative results.

The pathologist is sometimes faced with a situation in which a minute stereotactic biopsy tissue fragment has been entirely frozen and reveals the presence of neoplasm or other lesional tissue, but immunostaining will likely be required for further classification or grading. If the surgeon is reluctant to take additional tissue cores due to the risk of intracranial hemorrhage (a rare scenario in the setting of thorough preoperative planning and open communication with the pathologist), it is advisable to have unstained tissue sections cut from the frozen section block before removing it from the cryostat, since there is no guarantee that lesional tissue will still be present in the paraffin block after processing and refacing. Similarly, if small biopsy cores are provided by the surgeon for permanent sections, it is prudent to request unstained sections initially (i.e., cut at the same time as the section for hematoxylin-eosin [H & E] staining) in order to avoid loss of lesional tissue secondary to block refacing.

An additional safeguard in the setting of scant lesional tissue is the preparation of several unstained cytologic touch preparations from the fresh tissue before submitting for processing.

Intraoperative Cytologic Preparations as a Complement to Frozen Tissue Sections

At some institutions, historical practices have dictated that only frozen sections are performed for intraoperative consultation; at others, intraoperative diagnoses are made primarily by cytologic preparation, with frozen sections being only rarely employed. However, a large number of surgical neuropathologists worldwide routinely use a combination of cytologic preparations and frozen sections to render intraoperative diagnoses, and these two procedures are viewed as complementary, with the cytologic preparation providing exquisite nuclear and cytoplasmic detail free of freeze artifact and distortion, and the frozen section providing architectural details, including the relationship between the disease process and the host tissue (e.g., solid versus infiltrative tumor).

Several types of technical procedures are available for generating the cytologic preparation, and the choice depends on a number of factors, including the disease type and the consistency of the tissue submitted for intraoperative evaluation ( Box 3.3 ). For example, touch (imprint) preparations are generally optimal for pituitary adenoma. The concept is to take advantage of an integral aspect of adenoma pathobiology—the clonal expansion of loosely cohesive adenoma cells together with the attendant effacement and loss of the fibrovascular septa that compartmentalize normal adenohypophyseal tissue into acini. Adenoma cells tend to shed profusely on touch preparations compared with normal adenohypophysis. Touch preparations also work well on other hypercellular, loosely cohesive or dishesive tumors, such as lymphoma and melanoma. However, for more cohesive neoplasms and disease processes, the touch preparation is frequently hypocellular and often consists primarily of uninformative red blood cells. For this reason, many pathologists prefer the smear (squash, crush) preparation for soft tissue specimens, including most primary and metastatic central nervous system (CNS) tumors.

Cytologic preparations are of particular value for specific problematic tissue specimens ( Table 3.1 ). The pathologist sometimes receives a minute biopsy specimen that is too small to divide into separate portions for cytologic smear and frozen tissue section. In such cases, the tissue fragment can be first dragged across a slide to yield a cytologic preparation (cytologic drag preparation) before freezing. Another problematic specimen is the extensively cauterized tissue fragment. Although some of these may be beyond salvage, an attempt at diagnosis can be made by bisecting the tissue and performing a cytologic drag preparation before freezing the tissue. Fibrous or desmoplastic tissue samples often do not smear well. In such situations, a scrape preparation may be optimal. With this procedure, the tissue is grasped with forceps, and a scalpel blade is repeatedly drawn across the surface. The collected debris is then spread across a glass slide. Another challenging situation is presented by a Petri dish full of grossly necrotic tissue fragments. Arbitrarily choosing one or two fragments for frozen section often yields only nonspecific necrosis. In this situation, it is advantageous to sample as much tissue as possible for any viable tumor cells or cell clusters, yet preparing 10 or 20 frozen sections is impractical. One solution is to perform a cytologic drag preparation on multiple tissue fragments. With this procedure, one fragment after another is quickly dragged across the same, single slide; 5, 10, 15, or more fragments can be sampled in a matter of only seconds. If any viable tumor cells are present, this procedure is an efficient and cost-effective way to maximize the chances of their detection. Occasionally the pathologist is confronted with a request for “frozen section” on a specimen consisting of bony fragments, without an identifiable soft tissue component. An attempt to obtain a diagnosis can be made by performing a cytologic drag preparation of the bony fragments; this may yield a diagnostic preparation for some pathologic processes, such as metastatic carcinoma.

Fixation and Staining Options for Intraoperative Cytologic Preparations

Depending on the background, experience, and personal preference of the pathologist, cytologic preparations may either be air-dried or immersed immediately in 95% ethanol to avoid drying artifact. For gliomas in particular, the latter method provides superior preservation since nuclear cytology is particularly critical and air drying often produces marked artifacts of size, shape, and chromatin density. Similarly, stain preference varies among pathologists, although many prefer routine H & E. At some institutions, the preference is to perform an initial rapid one-step Diff-Quick stain on the first cytology slide, followed by routine H & E staining of a second preparation.