Q. How do you correlate cytologic cases with histologic findings? And how do you correlate cytology with clinical information?
The first issue is addressed in Laboratory Accreditation Program checklist question CYP.07543, which reads: “Is a documented effort made to obtain and review follow-up histologic reports or material available within the laboratory when gynecologic cases with HSIL or malignant cytologic findings are reported?”
The laboratory must compare all gynecologic cytology reports of HSIL or carcinoma with the histology report, if available, and determine any causes of discrepancies. Although cytohistologic correlation is mandated by CLIA ’88, the methods of analysis are not specified. Also, the correlation between cytology and biopsy need not be performed before case sign-out. A systematic approach lab-wide promotes consistency and may improve patient care. The use of an algorithm for cytohistologic correlation, including evaluation of both laboratory and non-laboratory factors, can be useful.1 Reasons for the discrepancy should be pursued. These may include cytologic sampling, histologic sampling, cytologic interpretation, cytologic screening, histologic interpretation, and histologic technical difficulty. Note that while the colposcopic biopsy may be considered the gold standard, both cytology and biopsy are subject to sampling variation, and, in some cases, cytology may better represent the cervical pathology than the biopsy. For example, if a Pap diagnosis of HSIL leads to a colposcopic biopsy and that biopsy is negative, review of the cytology may confirm the original interpretation of HSIL and the cause of discrepancy is attributed to biopsy sampling variation. In that case, the clinician may proceed to definitive treatment such as LEEP cone biopsy. The cytohistologic correlation is critical for appropriate patient management.
In the CAP Q-Probes study by Jones, et al., the majority of cytohistologic discrepancies were due to cytology and biopsy sampling variation rather than screening or interpretive errors. These factors fall outside of the laboratory’s control.
Correlating cytology and biopsy concurrently, if it is possible to do so, may prove most beneficial to physicians and their patients. Biopsies should be adequately sectioned and oriented so the transformation zone can be fully evaluated. When a biopsy appears negative in the face of a cytologic interpretation of HSIL, additional levels may increase the yield of dysplastic epithelium. Conversely, negative cytology cases should also be reviewed, if available, whenever the colposcopic biopsy is positive. It is important that some agreed-upon definition of diagnostic discrepancy be established. A commonly used definition of cytohistologic discrepancy in gynecologic pathology is a two-step difference between the cytologic and histologic interpretations. Examples include a Pap test interpretation of LSIL in which the followup biopsy is cancer, and a Pap test result of HSIL in which the biopsy is negative. It may be useful to consider and review lesser differences in diagnoses primarily for educational purposes. Peer review may be helpful to achieve consensus.
Results of cytohistologic correlations should be documented on a regular basis. A cytology-biopsy correlation log can be used for such purposes.2
Communicating the findings to clinicians is vital. A comment can be included in the report if specimens are correlated concurrently, and when appropriate. In some cases, it is useful to communicate directly with the clinician to discuss individual patient followup. Amended reports may be necessary if screening or interpretive problems are detected that would affect current patient care.
The second issue is addressed in checklist question CYP.07569: “Is an effort made to correlate gynecologic cytopathology findings with available clinical information?”
Methods of clinical correlation should be documented and can include the following: focused rescreening of cases based on clinical history, history of bleeding, or previous abnormality; correlation of the finding of glandular cells with hysterectomy status, age of patient, and date of LMP; and review of previous or current biopsy material.
Documentation of clinical correlation may include policies, problem logs with resolution, or notes in reports. Though cytology and biopsy need not be correlated before sign-out, it can be a very good practice for patient care. Many laboratories follow a procedure of reviewing previous cytology with current biopsies in discrepant cases in order to provide the physician with the most relevant information.
- Jones BA, Novis DA. Follow-up of abnormal gynecologic cytology: A College of American Pathologists Q-Probes study of 16,132 cases from 306 laboratories. Arch Pathol Lab Med. 2000; 124: 665– 671.
- Voytek TM, Mody DR, Davey DD. Quality management in cytopathology. In: Nakhleh RE, Fitzgibbons PL, eds. Quality Management in Anatomic Pathology: Promoting Patient Safety through Systems Improvement and Error Reduction. Northfield, Ill.: College of American Pathologists; 2005: 120–131.
Karen M. Clary, MD
Department of Pathology
Rochester General Hospital