Univariate and Multivariate Analysis of Histological Features Associated with the Use of Immunohistochemistry for the Diagnosis of Prostatic Carcinoma in Needle Biopsy Material
Omar Hameed, Leonard I Bloom, Chura Salih, Peter A Humphrey. Vanderbilt University, Nashville, TN; Washington University, St. Louis, TN
Background: Immunohistochemistry (IHC) is increasingly being used in the evaluation of prostate needle biopsy material. Although it is intuitive to consider that IHC is more frequently needed for smaller and/or lower grade carcinomas, it is unclear whether the presence of certain histological features, if any, are particularly associated with the use of IHC, especially since, to the best of our knowledge, there are no studies that specifically look at this issue.
Design: 38 consecutive PCs diagnosed with the aid of IHC were reviewed to determine the Gleason grade/score, muliple detailed glandular, architectural and cytological features, as well as adjacent findings. These same features were also evaluated in 81 consecutive "extent-matched" PCs diagnosed without utilizing IHC.
Results: The frequency of Gleason score 6 and 7 PCs was very similar in the two groups (84% and 16% vs 81% and 14%) with higher grade PCs limited to those without IHC (5%) (P>.05). There were also no significant differences between the two groups in the predominant glandular pattern seen (separate/loosely aggregated; crowded; poorly formed), luminal content (none; mucin; secretions; crystalloids), the frequency of perineural invasion, and the presence/absence of non-amphophilic cytoplasm, nuclear enlargement, hyperchromasia and prominent or multiple nucleoli. PCs for which IHC was used for the diagnosis were significantly (P<.0001) more likely to display less malignant glands (mean, 30 vs 56), a lower "gland density" (#malignant glands/linear extent) (mean, 27 vs 43); larger-sized glands (>100µ) (45% vs 4%) and an inflammatory or desmoplastic stroma (21% vs 0%). Dificulty in evaluation of the basal layer was also more likely in these PCs (53% vs 25%; P=.004), as was the finding of inconspicuous nucleoli (42% vs 20%; P=.019) and adjacent HGPIN (21% vs 5%; P=.008). Factors significantly associated with the use of IHC on multivariate analysis are displayed in the table.
|FACTOR||OR (95% CI)||P VALUE|
|Glandular Density||0.92 (0.87-0.96)||.0016|
|Larger Glands||20.7 (4.7-129)||.0002|
|Difficulty with Basal Layer||9.2 (2.7-38.2)||.0008|
|Inconspicious Nucleoli||4.8 (1.4-18.8)||.0164|