Abstract P1-07-28: Retrospective analysis of clinicopathologic features predictive of oncotype DX discordance in estrogen receptor positive, node negative breast cancer patients

BACKGROUND: Oncotype DX (ODX) is a validated recurrence score (RS) used to predict the risk of recurrence and benefit of chemotherapy in ER positive, node negative breast cancer patients. Prior to ODX, treatment recommendations regarding adjuvant chemotherapy and mortality approximation have taken into account clinical and pathologic risk factors. A discordance rate of 7-19% between risk allocating pathologic factors and ODX RS has been previously reported with progesterone receptor (PR) negativity noted as a defining clinical feature in numerous cases. The association between other clinicopathologic features and discordance is less certain.

METHODS: ODX data and clinicopathologic features were retrospectively reviewed for 724 breast cancer tumors belonging to 704 patients between 2006 and 2016. ODX discordance was defined as either 1-step discordance or 2-step discordance between ODX risk group (low, intermediate, high) and tumor grade (TG) (well differentiated, moderately differentiated, poorly differentiated). Tumors with 1-step discordance received a discordance score (DS) of 1 while those with 2-step discordance received a DS of 2. The database was subsequently analyzed using Paik's RS cutoffs as well as those outlined in the TAILORx trial. An odds ratio (OR) of >1 was consistent with discordance.

RESULTS: Among 724 tumor samples, ODX ER score (p=0.000), ODX PR score (p=0.000), ODX HER2 score (p=0.000), TG (p=0.000), mitotic count (MC) (p=0.0012), DCIS grade (p=0.0046), DCIS type (comedo necrosis vs. non-comedo necrosis) (p=0.0335) and micropapillary features (p=0.0044) were significantly associated with RS. Median age of cohort was 59 years and median tumor size was 1.2 cm.

Of 724 tumors, 619 from 604 subjects were eligible for assessment of discordance. Median RS was 16. Using Paik's RS cutoffs, 64.3% discordance was observed: 52.5% 1-step discordance (DS 1) and 11.8% 2-step discordance (DS 2). The TAILORx categorization yielded a discordance rate of 44.3%: 40.1% 1-step discordance and 4.2% 2-step discordance.

On univariate analysis and using Paik's RS cutoffs, young age (p= 0.0240), high MC (p=0.0006), large tumor size (>20 mm) (p=0.0209), the presence of DCIS (p=0.0480), high DCIS grade (p= 0.0033), and high ODX PR and ER scores (p= 0.0000) were significant clinicopathologic features predictive of discordance. On multivariate analysis, high MC (p= 0.0000), high ODX PR and ER scores (p=0.0000) remained significant as well as premenopausal status (p=0.026).

Per TAILORx cutoffs, univariate analysis revealed younger age (p= 0.0060), high MC (p= 0.0270), premenopausal status (p= 0.0124), and high ODX PR and ER scores (p= 0.0000) as significant for discordance. On multivariate analysis, high ODX PR and ER scores (p= 0.0000) remained significant.

CONCLUSION: In this retrospective ODX database, premenopausal status, high MC, high ODX PR and ER scores as per Paik's RS cutoffs were significant predictors for ODX discordance while high ODX PR and ER scores were significant predictors per the RS's outlined in the TAILORx trial. RS cutoffs per the TAILORX trial appear to create less discordance between RS and TG than the original cutoffs outlined by Paik and colleagues.

Citation Format: Cascetta KP, Zimmerman BS, Eggert L, Molot MC, Ru M, Nayak A, Bleiweiss I, Tiersten A. Retrospective analysis of clinicopathologic features predictive of oncotype DX discordance in estrogen receptor positive, node negative breast cancer patients [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-07-28.

Abstract P1-07-16: Retrospective analysis of oncotype DX recurrence score (RS) and discordance in patients with node-negative, ER+ breast cancer with recurrence

BACKGROUND: Oncotype RS is a 21-gene assay used to predict the likelihood of distant recurrence and benefit of chemotherapy in patients with node-negative, tamoxifen treated breast cancer. We developed a database to determine tumor recurrence rates and identify cases of discordance between Oncotype RS and tumor grade (TG). Our goal was to recognize patients with discordant tumors who had breast cancer recurrence and to understand the implications for patient management.

METHODS/RESULTS: We analyzed patient and tumor characteristics from 704 breast cancer patients between 2006-2016. Of these patients, there were thirteen recurrences (n=13), or 1.9% recurrence rate at a median follow-up of 4.2 years. When stratified by RS, recurrence rates were 1%, 2.4% and 4.3% in low, intermediate and high-risk groups respectively. Of the 13 patients who recurred, 31% had a low RS (<18), 54% had an intermediate RS (18-30) and 15% had a high RS (>31). The median RS was 23 and median age at time of recurrence was 55 years (62% postmenopausal). Tumor characteristics at time of recurrence were notable for: 77% metastatic, 23% locally recurrent, 85% PR positive, 69% moderately-differentiated (MD) and 31% poorly-differentiated (PD). No well-differentiated (WD) tumors recurred.

We defined Oncotype discordance as either 1-step or 2-step difference between Oncotype risk group (low, intermediate, high) and tumor grade (WD, MD, PD). Prior studies have demonstrated 7-19% “2-step discordance” between TG and RS (i.e. PD tumors with low-risk RS or WD tumors with high-risk RS). Of the 13 recurrences in our database, 46% were at least 1-step discordant, compared with 64.3% in our overall database. Among these recurrences, we compared discordant versus concordant tumors using two-sided T-tests. We found that fewer patients were treated with systemic chemotherapy in the discordant group (p=0.045), which was statistically significant. Among discordant patients, only one received chemotherapy, though all displayed MD or PD tumor grade. Discordant tumors tended to have lower RS (mean 17 vs. 27, p=0.34) and tended to be larger (mean 1.88cm vs. 1.33cm), however this was not statistically significant (p=0.84). Notably, the two largest tumors were both discordant. There were no significant differences in terms of age, Oncotype ER/PR score or mitotic count.

CONCLUSION: Although the sample size of recurrent patients is small, our data may suggest that patients with discordant tumors of low-risk Oncotype RS but higher TG may be receiving inadequate treatment (i.e. no chemotherapy). In addition to RS, other factors such as discordance, TG and tumor size should perhaps be considered when determining treatment plans.

Characteristics of Breast Cancer Recurrences in Oncotype DX DatabaseSubject #Age (years)Oncotype RSTumor GradeDiscordancePath PR%Tumor Size (cm)Chemotherapy14111MDY600.9N25811MDY1001.7N35713MDY301.0N45515MDY902.7N56122MDN301.5Y64522MDN901.8Y75323MDN00.9N83324MDN951.4Y96324PDY802.9U104129MDN801.1Y116530PDY102.1Y124933PDN01.6Y136135PDN51.0YMD=moderately-differentiated, PD=poorly-differentiated, Y=Yes, N=No, U=Unknown

Citation Format: Zimmerman BS, Cascetta KP, Ru M, Eggert L, Molot MC, Nayak A, Bleiweiss I, Tiersten A. Retrospective analysis of oncotype DX recurrence score (RS) and discordance in patients with node-negative, ER+ breast cancer with recurrence [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-07-16.

Correlation of Oncotype DX Recurrence Score with Cyclin D1 and ErbB2

Background: The Oncotype DX assay predicts the risk of recurrence in patients with stage I-II ER+, node negative breast cancer treated with tamoxifen. It is not understood if the Oncotype DX assay predicts the natural aggressiveness of an individual breast cancer or if it identifies a subtype of tamoxifen resistant breast cancer. In clinical practice, a high recurrence score (RS) on Oncotype DX is interpreted as a more aggressive tumor and is used to justify the use of chemotherapy. However, if the RS was actually predictive of tamoxifen resistance, patients may benefit from the use of an aromatase inhibitor, and chemotherapy may be unnecessary. Cyclin D1 and ErbB2 are two biomarkers shown to predict tamoxifen resistance.Cyclin D1 is overexpressed in approximately 35% of breast cancers. The Austrian Breast and Colorectal Cancer Study Group assessed expression of Cyclin D1 in patients taking tamoxifen within the ABCSG trial 05 and ABCSG trial 06. In both trials, Cyclin D1 overexpression correlated with a lower relapse free survival and overall survival compared to patients without Cyclin D1 overexpression.Erb2 is overexpressed in 15-30% of breast cancers. In the Gruppo Universitario Napoletano 1 trial, ER+ patients with early stage node negative breast cancer who overexpressed ErbB2 had no improvement in progression free survival and overall survival with 2 years of adjuvant tamoxifen therapy. Additional retrospective studies have supported initial reports of an association between overexpression of ErbB2 and tamoxifen resistance.Methods: 69 patients who had the Oncotype DX assay performed and had unstained pathology slides available were assessed for ErbB2 and Cyclin D1 expression. ErbB2 overexpression status was also obtained in another 74 patients who had the Oncotype DX assay performed. ErbB2 overexpression was determined from a review of medical records where ErbB2 was defined as being positive if immunohistochemical (IHC) staining intensity was 3+ with >90% of cells expressing ErbB2 or FISH revealed an amplification of >2.0. IHC analysis of Cyclin D1 was performed according to standard protocol and using commercially available antibodies. Scoring of slides for Cyclin D1 staining were performed by blinded pathologists who assessed both extent and intensity of nuclear staining for Cyclin D1.Results: The median Oncotype Dx RS within ErbB2+ patients was significantly higher than ErbB2- patients (36.5 vs. 18 p<0.0001), and approximately 50% of patients within each RS grouping (high, intermediate, and low) overexpressed cyclin D1.Conclusion: ErbB2 overexpression among high RS patients suggests the Oncotype DX assay may predict tamoxifen resistance and other markers for tamoxifen resistance need to be correlated with the RS. Although preliminary analysis of the IHC staining for Cyclin D1 does not correlate with a high RS, high Cyclin D1 expression among patients within the low RS subgroup is concerning since this subgroup may have an increased likelihood of disease relapse when treated with adjuvant tamoxifen alone.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 3035.

Wound Fluid Induces Cancer Cell Growth: A Mechanism for Recurrence?

Background: The process of wound healing after surgical resection of breast cancer may contribute to the development of local recurrence and possibly even metastases. Local recurrence after lumpectomy or mastectomy occurs most frequently near the surgical scar, leading to the hypothesis that the wound itself may promote the growth of residual disease. Some studies even suggest that removal of the primary tumor can accelerate progression of occult metastases. Two studies have shown that wound fluid can stimulate the growth of cancer cells and that differing cancer cell lines may be stimulated by differing wound fluid. A major criticism of these studies is that the activity of the wound fluid was compared to either the patient serum, which is likely less inflammatory than the wound fluid, or to serum free media. Thus, these studies did not define whether the inflammatory component of wound fluid alone is sufficient to stimulate cancer cell growth or whether the ability was specific to breast derived wound fluid. In this pilot study wound fluid was collected from 10 patients who had undergone unilateral mastectomy with abdominal flap reconstruction or bilateral mastectomy where one breast had known cancer and the contralateral did not. In the unilateral mastectomy patients, the fluid from the involved breast was compared to the fluid from the abdominal drain. In the bilateral mastectomy patients the fluid from the breast with cancer was compared with the fluid from the prophylactically removed breast. The effect of wound fluid on the growth of breast cancer cells was analyzed for each patient.Methods: Fluid from each breast and abdominal drain were collected 24 and 48 hours after surgery and their ability to promote growth of 6 different breast cancer cell lines was tested. In addition, the ability of the 48 hour wound fluid to stimulate colony formation of HBL-100 cells (ER-) on matrigel was established. The cytokine profile of the different wound fluid was also analyzed.Results: We found that when the ER – cells were cultured on matrigel, the breast derived fluid led to growth of invasive, branching colonies, while abdominal fluid from the same patient led to small round colonies. Further, we found that wound fluid derived from the breast where a tumor was present led to the formation of more invasive colonies than the wound fluid derived from the normal breast. Thus, we speculate that in addition to the normal cytokines and matrix metalloproteases associated with inflammation, breast wound fluid may contain additional proteins from the tumor bed microenvironment. In support of this model, we found that the protein composition of wound fluid varies from one site to another in the same patient.Conclusion: The biological activity and protein composition of wound fluid from the breast and abdomen is drastically different—such that breast wound fluid promotes growth and invasion of breast cancer cells. Thus, we hypothesize that residual disease may be stimulated to grow during the window of time after surgery when the wound fluid is produced. This transient stimulation may result in local recurrence if the nature of the residual disease is a foci cancer or acceleration of metastasis if the residual disease is disseminated into the surrounding lymph nodes or blood vessels.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 3146.

High prevalence of MMTV-like env gene sequences in gestational breast cancer

Gestational breast cancer (BC) is generally associated with rapid growth and increased mortality. Because the presence of MMTV-like sequences in BC has been associated with laminin receptor expression, a marker of poor prognosis, gestational BCs were analyzed for MMTV env gene-like sequences to explore whether MMTV-like sequences were also associated with its adverse outcome. Whereas 30-38% of sporadic BC have the sequences, in gestational BC the prevalence is 62%. We suggest that hormonal response elements present in the MMTV-like LTR may play a role in promoting cell growth, as they do in the mouse system.

Pleomorphic adenoma of the breast. A case report and distinction from mucinous carcinoma

Pleomorphic adenoma of the breast is a rare, benign tumor accounting for 68 cases in the literature. It is most commonly seen in postmenopausal women and is characterized by an admixture of epithelial and myoepithelial cells embedded in abundant myxomatous stroma. Its clinical and histologic appearance can be challenging and may lead to a misdiagnosis of invasive carcinoma. We report a case of mammary pleomorphic adenoma in an asymptomatic 59-year-old woman and briefly discuss its distinction from mucinous carcinoma through the use of special stains.

The significance of malignancies incidental to microcalcifications in breast spot localization biopsy specimens

BACKGROUND

Incidental breast cancer is occasionally found in spot localization biopsy specimens adjacent to mirocalcifications in benign breast disease. Because this phenomenon could prove problematic for percutaneous sampling of microcalcifications without excisional biopsy, we studied surgical specimens from patients with cancers incidental to microcalcifications and compared them with specimens with microcalcifications within the malignancy.

METHODS

The pathology database at the Mount Sinai Medical Center from January 1993 to July 1998 was reviewed to identify breast cancer patients who underwent spot localization biopsy for microcalcifications. Patients presenting with microcalcifications within malignancy (determinate) were compared with patients with mirocalcifications in benign breast tissue adjacent to malignancy (incidental).

RESULTS

Thirty-two (13%) of the 241 specimens had microcalcifications in benign tissue adjacent to malignancy and 209 (87%) had microcalcifications within the malignancy. Fifty-six percent of the incidental cases and 65% of the controls had ductal carcinoma in situ. Infiltrating lobular carcinoma accounted for 25% of the incidental cancers and 2% of the determinate cancers (P <0.001). Fifty-seven percent of the infiltrating carcinomas incidental to mammographic findings were infiltrating lobular carcinoma compared with 7% of the nonincidental infiltrating carcinomas. None of the incidental invasive carcinomas were poorly differentiated (P = 0.002). There were no significant differences with regard to age, tumor size, stage, differentiation, estrogen and progesterone receptors, type of surgery and final margin status. In none of the patients with incidental malignancies did local or distant recurrences develop.

CONCLUSIONS

Incidental carcinomas were found in 13% of spot localization biopsy specimens obtained for suspicious mammographic microcalcifications and have a favorable prognosis. Infiltrating lobular carcinomas are more commonly found with incidental microcalcifications than with determinate microcalcifications, and incidental invasive carcinomas are less likely to be poorly differentiated. The majority of malignancies, both determinate and incidental to microcalcifications, are due to ductal carcinoma in situ. Incidental malignancies commonly occur adjacent to fibrocystic changes and their other pathologic characteristics are not significantly different from nonincidental carcinomas. Despite the absence of radiographic findings, these patients can be successfully treated with breast conservation.

Lumpectomy margins, reexcision, and local recurrence of breast cancer

BACKGROUND

The diagnosis of breast cancer is often made by excisional biopsy without margin assessment for mammographic findings or palpable masses. Many patients treated with breast conservation undergo reexcision to obtain clear margins although the relationship between clear margins and local recurrence remains controversial.

METHODS

Patients undergoing breast conservation and adjuvant radiation therapy with complete follow-up over 5 years were studied. Factors associated with obtaining clear histopathologic margins and undergoing reexcision to obtain clear margins were studied in relation to the risk of local recurrence.

RESULTS

Clear biopsy margins were associated with diagnosis by fine-needle aspiration cytology (fine-needle aspiration 42%, spot localization 11%, excisional biopsy 10%; P <0.001). Reexcision was significantly related to diagnostic method (spot localization 63%, excisional biopsy 36%, fine-needle aspiration 10%; P <0.001), first margin status (clear 0%, close 11%, positive 46%, unknown 48%; P <0.001), patient age (54 years for reexcised patients and 58 for non-reexcised patients; P <0.001), and tumor size (mean tumor size 1. 4 cm for reexcised patients and 1.7 cm for non-reexcised patients; P = 0.003). Patients undergoing reexcision were significantly more likely to be diagnosed by spot localization, have nonnegative excisional biopsy margins, be younger, and have smaller tumors than patients not undergoing reexcision. Local recurrence was not significantly related to margin status (8% with clear margins, 7% with positive margins, 19% with close margins, and 11% with unknown margins) or reexcision (10% local recurrence rate for patients with negative final margins after reexcision and 12% with positive, close or unknown first margin without reexcision). Estrogen receptor status was the only variable related to local recurrence in Cox proportional hazards model (P = 0.009). Estrogen receptor negative patients with nonnegative margins experienced a 20% rate of local recurrence compared with 10% for estrogen receptor negative patients with negative margins and 7% for estrogen receptor positive patients regardless of margin status (P = 0.021).

CONCLUSIONS

Clear excision margins are facilitated by preoperative diagnosis by fine-needle cytology. For patients with nonnegative margins, reexcision was more commonly performed in young patients with small tumors diagnosed by spot localization biopsy. The relationship of local recurrence to margins and reexcision was not statistically significant. Estrogen receptor negative tumors with nonnegative margins had a significantly higher rate of local recurrence than estrogen receptor negative tumors with clear margins and estrogen receptor positive tumors regardless of margin status.

[Searching for retroviral sequences related to human breast cancer]

The participation of viruses in mammary carcinogenesis has been largely studied in animals. A model similar to the mouse mammary tumor virus (MMTV) was previously proposed. Several lines of research supported the participation of MMTV in human breast cancer, but these evidences were contradicted when further research was performed. One major issue was the presence of human endogenous retroviral sequences that confounded results reporting MMTV-like sequences in human breast cancer. To overcome this problem we selected a 660 bp sequence of the MMTV env gene with low homology to endogenous sequences and search for a sequence to it using the polymerase chain reaction (PCR). The sequence was found in 38% of the human breast cancers and in 2% of the normal breasts studied. The sequence was not present in tumors from other organs. It was 90-98% homologous to MMTV and only 18% to human endogenous retrovirus (HERV) K-10. It was also detected in some of the positive tumors by Southern blot hybridization using one of the cloned 660 bp as a probe. Using reverse transcriptase PCR, it was possible to demonstrate that the 660 bp sequence is expressed in the majority of the tumors. Also, preliminary experiments revealed that sequences related to the LTR and gag genes of MMTV were present in the DNA of breast tumors. The origin of the MMTV-like sequences in tumor DNA could be the result of integrated MMTV-like sequences derived from a human mammary virus or may represent unknown endogenous sequences that can only be detected in breast tumors.

Immunohistologic c-myc protein in benign breast disease and cancer

We have studied the histopathology and differential distribution of the c-myc protein (Myc) in human breast tissues including 17 cases of infiltrating mammary carcinoma, 4 cases of fibroadenoma, 5 cases with fibrocystic changes, and 1 case of reduction mammoplasty (as a control). Using a sensitive immunohistochemical method on frozen tissue sections, both a rabbit polyclonal anti-c-myc antibody and a mouse monoclonal anti-c-myc antibody, H51C116, produced high levels of Myc staining in the nuclei of epithelial cells of infiltrating mammary carcinomas (30-90% of cells stained). In contrast, the nuclei of epithelial cells of fibroadenomas, and breast tissues with fibrocystic changes stained infrequently. We studied benign tissue surrounding the tumors in four cases; three were essentially negative, and one showed nuclear epithelial cell staining throughout the lobules. Sixteen of the tumors were examined in parallel, using formalin-fixed, paraffin-embedded samples. Immunohistological procedures for Myc produced uniform, intense epithelial cell cytoplasmic staining (8 cases); light epithelial cell cytoplasmic staining (5 cases) or were unstained (3 cases). We argue that the differences between frozen and paraffin sections are incompatible with the notion of simple displacement of nuclear Myc to the cytoplasm during fixation. Elevated levels of nuclear Myc in tumor cells and subsets of benign tissue are consistent with a role for Myc in mammary cell proliferation and tumorigenesis.

Proliferative indices and oncoprotein expression in benign and malignant breast biopsies

BACKGROUND

Prognostic factors are used routinely in the management of breast cancer. However, their potential for identifying precursor malignant lesions has not been assessed.

METHODS

We have examined 285 breast biopsy specimens (140 benign, 145 malignant) for DNA ploidy, S-phase fraction, Ki-67 nuclear antigen proliferative indices, and HER-2/neu and epidermal growth factor receptor oncoproteins.

RESULTS

When proliferative indices were compared between the benign and malignant groups, differences were noted for DNA ploidy, S-phase fraction, and cell cycling index (p < 0.0005). When the benign nonproliferative specimens were compared with the atypical/proliferative benign specimens, proliferative indices failed to show any differences. When the specific subset of proliferative/atypical benign breast tissue was compared with the malignant specimens, DNA index, S-phase fraction, and cell cycling index showed significant differences. The mean for epidermal growth factor receptor was greater in the non-proliferative group but not statistically significant (p < 0.1). HER-2/neu oncoprotein failed to show any differences between the benign and malignant groups. Within the atypical benign group, Ki-67 correlated strongly with S-phase fraction and HER-2/neu (p < 0.01).

CONCLUSIONS

We have demonstrated that proliferative indices can differentiate between benign and malignant breast tissues but not among specific subgroups. In addition, epidermal growth factor may differentiate between nonproliferative and proliferative/atypical benign biopsy results. Oncoprotein determination, ploidy, and DNA proliferative indices may be useful in defining malignant and benign breast disease but are not useful in distinguishing between benign and malignant breast disease with an increased likelihood for malignant transformation.

Prognostic variables in invasive breast cancer: contribution of comedo versus noncomedo in situ component

BACKGROUND

Many invasive breast cancers are accompanied by a variety of noninvasive components. Histological distinctions have been made between these components, but to understand their importance, it is essential to examine their molecular biology.

METHODS

Proliferative indices, oncoproteins, and steroid receptor expression were compared for invasive breast cancers containing comedo-type ductal carcinoma in situ (n = 35), noncomedo-type ductal carcinoma in situ (n = 34), and pure invasive cancers (n = 49). Ploidy, S-phase fraction, Ki-67 staining, estrogen receptor (ER), progesterone receptor (PR), and the expression of HER-2/neu and epidermal growth factor receptor (EGFR) were evaluated in these tumors.

RESULTS

The comedo-invasive subgroup differed significantly from the noncomedo-invasive subgroup, demonstrating significantly higher mean ploidy (1.6 vs. 1.3; p = 0.0156), S-phase fraction (7.9% vs. 4.3%; p = 0.0066), Ki-67 staining (20.3% vs. 12.0%; p = 0.0058), and HER-2/neu values (2,247 fm/mg vs. 1,014 fm/mg; p = 0.0412) and lower ER (76 fm/mg vs. 339 fm/mg; p = 0.006) and PR values (99 fm/mg vs. 265 fm/mg; p = 0.0608). A higher percentage of comedo-invasive carcinomas demonstrated aneuploidy 71%; p = 0.0158), elevated levels of S-phase fraction (75%; p = 0.0016) and Ki-67 staining (55%; p = 0.0512), overexpression of HER-2/neu oncogene (47%; p = 0.0011), and were ER negative (35%; p = 0.0148), PR negative (47%; p = 0.0073) when compared to noncomedo-invasive carcinomas. Comedo-invasive and noncomedo-invasive tumors were comparable for nodal status and tumor size, but differences were noted for tumor differentiation and percentage of tumors that were > 1 cm. Comedo-invasive tumors were predominantly poorly differentiated (60 vs. 32%) and were > 1 cm (94 vs. 77%, p < 0.05).

RESULTS

Comedo-invasive cancers were comparable to pure invasive cancers for ploidy, S-phase fraction, Ki-67 staining, and ER, PR, and EGFR expression. However, comedo-invasive carcinomas had greater HER-2/neu overexpression when compared to pure invasive tumors (47 vs. 19%; p = 0.0359).

CONCLUSIONS

These results are consistent with the hypothesis that comedo carcinoma is a more aggressive type of ductal carcinoma in situ and may have independent prognostic value when seen in association with infiltrating ductal carcinoma. In invasive tumors, comedo carcinomas are associated with poor prognostic factors, including higher ploidy, S-phase fractions, Ki-67 staining, negative ER and PR status, poorer differentiation, larger tumors, and presence of HER-2/neu oncogene overexpression.

Interval changes in nonpalpable breast lesions as an indication of malignancy

OBJECTIVE

To determine whether the patterns of interval changes in nonpalpable breast lesions can be used to determine malignancy.

PATIENTS AND METHODS

The authors reviewed the records for 128 nonpalpable breast lesions detected by mammography between March 1990 and May 1992 for which previous imaging studies were available. Changes in size, density and contour of the 59 masses and changes in size, number, configuration and pattern for the 69 cases involving microcalcifications were determined from comparison of the current mammograms with the earlier imaging studies, obtained 6 to 42 months previously. These findings were correlated with the pathological diagnosis based on needle localization and excisional biopsy.

RESULTS

Twelve (55%) of the 22 new masses and 15 (48%) of the 31 masses that changed were malignant, but none of the 6 masses that did not change were malignant (chi 2 test, p < 0.01). All of the 31 masses that changed increased in size (and of these, 15 [48%] were malignant). In addition, 14 (45%) of the masses that changed exhibited increased density (and of these 9 [64%] were malignant), and 11 (35%) became irregular (and all of these were malignant). Eight (47%) of the 17 cases of new calcifications, 18 (43%) of the 42 cases of calcifications that changed and none of the 10 cases of unchanged calcifications were malignant (chi 2 test, p < 0.01). Twelve (44%) of the 27 calcifications that changed in area and 14 (40%) of the 35 that changed in number were malignant. Fifteen (60%) of the 25 calcifications that changed from scattered to grouped and 8 (53%) of the 15 that changed from punctate to linear were malignant.

CONCLUSION

A change in the density or contour of a mass is a fairly definite sign of malignancy. The authors conclude that interval observation is a good method of detecting malignancy in breast masses but is less reliable for detecting malignancy in calcifications.

Human papillomavirus-associated lesions of the vagina and cervix. Treatment with a laser and topical 5-fluorouracil

Twenty women with cervical and vaginal human papillomavirus-associated lesions were treated with CO2 laser ablation followed by eight weekly applications of 5-fluorouracil. Viral subtyping in a majority of patients and histology were obtained before and after treatment. After treatment 88% (15 of 17) had normal vaginal biopsies, and 59% (10 of 17) had normal cervical biopsies. There were no treatment failures with subtype 6/11 infection of the cervix or vagina. All the failures were with viral subtypes 16/18 and 31/35/51. The protocol was effective in treating patients with cervical and vaginal human papillomavirus-associated lesions.