Lillian Young

PURPOSE: The presence of disseminated tumor cells (DTC) in the bone marrow of breast cancer patients is an acknowledged independent prognostic factor. The biological metastatic potential of these cells has not yet been shown. The presence of putative breast cancer stem cells is shown both in primary tumors and distant metastases. These cells with a CD44(+)CD24(-/low) phenotype represent a minor population in primary breast cancer and are associated with self-renewal and tumorigenic potential. Recognizing the potential effect of prevalence of putative stem cells among DTC, we evaluated the bone marrow DTC. EXPERIMENTAL DESIGN: We employed the double/triple-staining immunohistochemistry protocol and modified the established bone marrow cytokeratin (CK) staining protocol by adding steps for additional antigens, CD44 and/or CD24. We evaluated 50 bone marrow specimens, previously categorized as CK(+) from early breast cancer patients. CK(+) cells were examined for CD44 and CD24 expression by light microscopy, fluorescence microscopy, and spectral imaging. RESULTS: We detected the putative stem cell-like phenotype in all CK(+) specimens. The mean prevalence of putative stem/progenitor cells was 72% and median prevalence was 65% (range, 33-100%) among the overall DTC per patient, compared with primary tumors where this phenotype is reported in <10% of cells. CONCLUSIONS: This is the first evidence of the existence of the putative stem-like phenotype within the DTC in bone marrow in early breast cancer patients. All patients had a putative stem cell phenotype among the DTC and most individual DTC showed such phenotype. Future molecular characterization of these cells is warranted.

Antigen retrieval (AR) incorporating high-temperature microwave (MW) heating of tissue sections before immunostaining is a revolutionary technique that can unmask the antigens in formalin-fixed tissue sections, thus making them available for immunohistochemical staining. Although high temperature is believed to be the primary mechanism in retrieval of antigens, a variety of chemical solutions have been tested to define an optimal AR solution. We tested the hypothesis that pH of the AR solution may influence the quality of immunostaining by using seven different AR buffer solutions at a series of different pH values ranging from 1 to 10. We evaluated the staining of monoclonal antibodies to cytoplasmic antigens (AE1, HMB45, NSE), nuclear antigens (MIB-1, PCNA, ER), and cell surface antigens (MT1, L26, EMA) on routinely formalin-fixed, paraffin-embedded sections under different pH conditions with MW heating for 10 min. The intensity of immunostaining was graded in a blinded fashion. The pH value of the AR buffer solution was carefully measured before, immediately after, and 15 min after the AR procedure. The influence of pH on AR immunohistochemical staining can be summarized into three patterns. Some antigens (L26, PCNA, AE1, EMA, and NSE) showed excellent retrieval throughout the pH range. Other antigens (MIB1 and ER) showed strong intensity of immunohistochemical staining at very low pH and at neutral to high pH, but a dramatic decrease in the intensity of the AR immunostaining at moderately acidic pH (pH 3-6). Still others (MT1 and HMB45) showed increasing intensity of the AR immunostaining with increasing pH, but only weak immunostaining at low pH. Among the seven buffer solutions at any given pH value, the intensity of AR immunostaining was very similar. However, Tris-HCl buffer tended to produce better results at higher pH, compared with other buffers. Although high-temperature heating is believed to be the most important factor for the AR technique, the pH value of the AR solution is an important co-factor for some antigens. Optimization of the AR system should therefore include optimization of the pH of the AR solution. Our results indicate that AR immunostaining of Tris-HCl or sodium acetate buffer at pH 8-9 may be suitable for most antigens, although certain nuclear antigens show optimal staining at low pH.

We developed a staining protocol for demonstration of androgen receptor (AR) in formalin-fixed, paraffin-embedded tissue sections. The method is based on the antigen retrieval microwave (MW) heating technique. Results are compared with different types of enzyme digestion pre-treatments. The strongest immunostaining signal and clearest background were obtained by MW heating of dewaxed paraffin sections in 5% urea or citrate buffer solution (pH 6); pure distilled water gave less consistent results. Enzymatic digestion with pepsin (0.05% in 2 N HCl) for 30 min at room temperature, or trypsin followed by pronase, or pronase digestion alone, also produced enhanced staining of AR in some cases, but there was more nonspecific background, and specific reactivity was less intense. The antigen retrieval MW method can be used to demonstrate AR epitope in prostate tissue after fixation in formalin for as long as 7 days. AR immunolocalization was also compared in frozen and paraffin sections processed from the same specimen of prostate carcinoma tissue and was found to be qualitatively and quantitatively similar. This study also provided new information concerning the basic principles of the antigen retrieval MW method that may be helpful in further development of this technique.

PURPOSE: To determine the combined effects of p53, p21, and pRb alterations in predicting the progression of bladder transitional cell carcinoma. PATIENTS AND METHODS: p53, p21, and pRb expression was examined immunohistochemically on archival radical cystectomy samples from 164 patients with invasive or high-grade recurrent superficial transitional cell carcinoma (TCC; lymph node-negative, 117 patients; lymph node-positive, 47 patients). Median follow-up was 8.6 years. Based on percentage of nuclear reactivity, p53 was considered as wild-type (0% to 10%) or altered (>10%); p21 was scored as wild-type (>10%) or altered (<10%); and pRb status was considered wild-type (1% to 50%) or altered (0% or >50%). RESULTS: As individual determinants, the p53, p21, and pRb status were independent predictors of time to recurrence (P<.001, P<.001, and P<.001, respectively), and overall survival (P<.001, P=.002, and P=.001, respectively). By examining these determinants in combination, patients were categorized as group I (no alteration in any determinant, 47 patients), group II (any one determinant altered, 51 patients), group III (any two determinants altered, 42 patients), and group IV (all three determinants altered, 24 patients). The 5-year recurrence rates in these groups were 23%, 32%, 57%, and 93%, respectively (log-rank P<.001), and the 5-year survival rates were 70%, 58%, 33%, and 8%, respectively (log-rank P<.001). After stratifying by stage, the number of altered proteins remained significantly associated with time to recurrence and overall survival. CONCLUSION: This study suggests that alterations in p53, p21, and pRb act in cooperative or synergistic ways to promote bladder cancer progression. Examining these determinants in combination provides additional information above the use of a single determinant alone.