PAX8 Reactivity in Lower Urothelial Tract Tissues and Neoplasms: A Comparison of Two Monoclonal PAX8 Antibodies

While PAX8 positivity is a well-recognized diagnostic pitfall in upper urinary tract urothelial carcinomas, evidence suggests that both normal urothelium and urothelial carcinomas of the urinary bladder may also exhibit PAX8 staining, potentially leading to diagnostic confusion. Anecdotally, we have encountered PAX8 reactivity in lower urinary tract specimens, including in both non-neoplastic and neoplastic bladder urothelium.

Prior studies documenting this phenomenon are limited and sometimes yield conflicting results. One study determined PAX8 expression in the majority of urothelial carcinoma specimens. ¹ Other reports indicated positive PAX8 expression (by polyclonal antibody) in approximately 7%-17% of invasive urothelial carcinomas of the bladder,^2–4 with subsequent smaller studies suggesting increased PAX8 reactivity in the nested subtype of urothelial carcinoma.^5,6 More recently, a large scale study demonstrated that PAX8 expression (by MSVA708R and clone MRQ-50) occurs in a small subset of urothelial neoplasms, with a predilection for noninvasive and low-grade tumors. ⁷ These findings raise the possibility that PAX8 expression reflects persistence of developmental lineage markers rather than true aberrant expression. To further evaluate PAX8 reactivity in lower urinary tract tissues and neoplasms, we compared two commonly used monoclonal antibody clones, SP348 and BC12, both of which exhibit significantly less cross-reactivity to other PAX proteins than the polyclonal form or the MRQ-50 clone. ⁸ This comparison aimed to better characterize the frequency and diagnostic implications of PAX8 staining in bladder urothelium.

Immunohistochemical staining with two PAX8 monoclonal antibody clones, SP348 and BC12, was performed on two tissue microarrays (TMA) composed of 692 bladder, 13 upper urinary tract, 14 prostate, 6 soft tissue, 2 lung, 2 ENT, and 14 placenta cores arranged in duplicate. Tissues that were lost during processing, lacked tumor, or contained non-urothelial tumors were excluded in the final analysis. Duplicate cores were considered a single specimen. A total of 341 bladder specimens (331 from tissue microarray cores and 10 from whole slides) were scorable with clone SP348, and 331 tissue microarray cores were scorable with clone BC12. Staining results were evaluated semi-quantitatively for both intensity and extent, using a 0-3 scale, and subsequently classified as negative or positive for PAX8 expression.

Both the SP348 and BC12 PAX8 antibody clones demonstrated comparable performance in neoplastic urothelium, with positive staining observed more often in noninvasive urothelial carcinoma (32%) compared to invasive urothelial carcinoma (9%), both of comparable staining intensity and extent (Figures 1 and 2). This distribution is in keeping with the suggestion that PAX8 expression in urothelial carcinoma appears to be particularly enriched in noninvasive, low-grade tumors. ⁷ The SP348 PAX8 clone more often stained normal urothelium compared to the BC12 clone, although the number of specimens tested was limited.

OPEN IN VIEWER

Figure 1.

Representative H&E, PAX8 clone SP348, and PAX8 clone BC12 images showing moderate (A) and strong (B) staining intensity in morphologically noninvasive bladder tumors.

OPEN IN VIEWER

Figure 2.

The table summarizes the staining results of PAX8 clones SP348 and BC12 in normal, noninvasive, and invasive bladder urothelium.

Overall, both PAX8 antibody clones SP348 and BC12 showed concordant positive staining in approximately 13% of all bladder specimens, consistent with prior estimates of PAX8 positivity in bladder urothelial carcinoma reported in a smaller number of specimens. ² These findings reinforce the importance of antibody selection and immunohistochemical interpretation in urothelial lesions.

Although PAX8 positivity has been more extensively reported in the upper urinary tract,^4,9,10 its expression in the lower urinary tract has been less consistently documented. Our findings further support that PAX8 consistently stains a subset of both normal and neoplastic lower urinary tract epithelium. Interestingly, this data also suggests that PAX8 is expressed more frequently in noninvasive bladder carcinomas compared to their invasive counterparts (more than a 3-fold increase), a finding that warrants further investigation. Particularly in small biopsies, awareness of PAX8 positivity in some bladder tissue is essential to avoid misclassification as renal, Mullerian, or (less anatomically likely) thyroid malignancy. As a clinical tool, PAX8 alone cannot be reliably used to distinguish urothelial carcinoma from renal carcinoma or nephrogenic adenoma in urinary tract lesions, or from Mullerian type carcinomas in pelvic masses; rather, a combination of PAX8 and other antibodies such as CDH16, GATA3, p63, or SOX17 may be helpful in better differentiating the origins of the neoplasm. CDH16 when used in combination with PAX8 may have higher specificity for renal, thyroid, and mullerian origin tumors. ⁷ SOX17 has demonstrated higher specificity for Mullerian type carcinomas without staining renal or theyroid tumors. ¹¹