Chromosome abnormalities in benign hyperproliferative disorders of epithelial and stromal breast tissue

Cytogenetic analysis of short-term cultures from 15 cases of benign proliferative breast disease (PBD), 10 diffuse PBD and 5 papillomas, and 15 fibroadenomas of the breast revealed clonal chromosome abnormalities in 7 diffuse PBD lesions, 4 papillomas and 5 fibroadenomas. The remaining 14 cases had a normal female chromosome complement. Cytogenetically unrelated abnormal clones were seen in 4 fibroadenomas and 2 PBDs. A single abnormal clone was found in 9 PBDs and 1 fibroadenoma. Three clonal abnormalities were seen as recurrent changes in 6 cases, namely interstitial deletions of 3p with 3p 12-14 as the minimally common deleted segment (in 1 papilloma, 1 diffuse PBD with atypia and 1 mixed-pattern lesion with both papilloma and atypical diffuse PBD features), r(9)(p24q34) (in 1 diffuse PBD and 1 fibroadenoma), and del(1)(q12)(again in 1 diffuse PBD and 1 fibroadenoma). Intriguingly, 6 of the 16 abnormal cases had chromosome changes that have been seen repeatedly as primary abnormalities in breast carcinomas: der(16)t(1;16)(q10;p10), del(3)(p12p14), and del(1)(q12). We conclude that some of the chromosome anomalies frequently found in breast carcinomas are also present in PBD and fibroadenomas. These aberrations may be accepted as early, neoplasia-relevant mutations. However, they do not seem to be sufficient by themselves to unleash a malignant process.

Karyotypic changes in phyllodes tumors of the breast

Cytogenetic analysis of short-term cultures of five phyllodes tumors of the breast-classified as benign (one tumor), borderline malignant (two tumors removed from the same breast in 1991 and 1993), and malignant (two tumors)--revealed clonal changes with simple structural abnormalities in the benign tumor, the borderline malignant tumors, and one malignant tumor in which benign areas and areas of borderline malignancy were also present. In contrast, the malignant tumor without admixed borderline malignant or benign areas had a complex karyotype. The karyotype of the benign phyllodes tumor was 46,XX,del(12)(p11p12)/46,XX,t(8;18)(p11;p11)/46,XX. The first borderline malignant phyllodes tumor had t(3;20)(p21;q13) as the sole abnormality. When the tumor recurred, this was no longer the only clone detected and the tumor karyotype was now 46,XX,t(3;20)(p21;q13)/46,XX,t(9;10)(p22;q22)/46,XX,t(1;8) (p34;q24)/46,XX,del(11)(q22-23)/46,XX. The malignant/borderline malignant/benign tumor had t(1;6)(p34;p22) as the sole clonal abnormality. Finally, the karyotype of the malignant phyllodes tumor which contained no benign or borderline malignant areas was 42,XX,der(1)t(1;4)(q21;q21),der(3)t(3;17)(q29;q21), -4,i(8)(q10), -10, -13,i(13)(q10),der(14)t(1;14)(q21;p11),der(14)t(4;14) (p12;p11), -17/80-90,idemx2, +del(1)(q12), +i(1)(p10), +dic(5;5)(p14;p14), +i(6)(p10), +del(7)(p11), +dup(7)(q11q36), +i(15)(q10),inc/46,XX. The findings indicate some cytogenetic similarities between benign/borderline malignant phyllodes tumors and fibroadenomas of the breast, presumably reflecting similar pathogenetic mechanisms in the two types of mixed-lineage tumors.

Cytogenetic analysis of multifocal breast carcinomas: detection of karyotypically unrelated clones as well as clonal similarities between tumour foci

Cytogenetic analysis was performed on short-term cell cultures of two foci (A and B) from each of three multifocal breast carcinomas. In case I, four clones (three related and one unrelated) were detected in sample A. In sample B, two of the three related clones and the unrelated clone seen in A were found, as was also a third subclone showing a pattern of clonal evolution slightly different from that detected in A. In cases II and III, multiple cytogenetically unrelated clones were found in A and B, with only one clone being shared by both foci in each case. Our finding of cytogenetic similarities between macroscopically distinct tumour lesions indicates that the multifocality reflects intramammary tumour spread rather than the synchronous emergence of pathogenetically independent carcinomas within the same breast. On the other hand, the detection of karyotypic heterogeneity in the form of cytogenetically unrelated clones in all foci suggests that human breast carcinoma may be polyclonal. This polyclonality may be part of the explanation for the cellular heterogeneity commonly seen at the phenotypic level in breast cancer.

Chromosome abnormalities in adenolipomas of the breast: karyotypic evidence that the mesenchymal component constitutes the neoplastic parenchyma

Cytogenetic analysis of adenolipomas of the breast, a tumor type that has not been chromosomally characterized before, revealed the karyotypes 47,XX, +del(1)(p22) in one tumor and 46,XX, t(12;16)(q15;q24) in the other. Breast adenolipomas thus seem to be karyotypically identical to sporadic lipomas in other locations: rearrangements of 12q13-15 are the most common cytogenetic aberrations in lipomas, and also breaks in and around 1p22 have been reported in such tumors. The similarity with lipoma could be documented further in case 2, in which epithelial and mesenchymal cells were cultured separately; the t(12;16) was present in the latter but not in the former. This is evidence that the connective tissue is the neoplastic parenchyma in adenolipomas of the breast, whereas the glandular elements show concomitant but nonneoplastic proliferation.

[The duct ectasia syndrome. A prospective clinical study of patients with breast diseases]

Mammary duct ectasia is a benign condition of the mammary gland histopathologically characterized by inflammatory changes in the collecting ducts. A prospective clinical and histopathological study was done of 316 consecutive patients with surgical mammary diseases. The syndrome of mammary duct ectasia (DES) was found in 42 patients (13%). The mean age was 44 years and 81% of the patients were pre- or menopausal. Forty-eight percent of the patients had a history of previous benign breast lesions and abnormality associated with lactation was also frequently represented in the history. Major symptoms were tumour formation, localised in the subareolar region in more than half of the cases, subareolar abscess and nipple discharge. Minor symptoms were mastalgia, mammary fistula, nipple retraction, general symptoms and enlarged axillary lymph nodes. Treatment for an abscess in DES seems to fail more often if only incision and drainage are performed. Excision of the entire focus and corresponding duct may prove to be better.

[The duct ectasia syndrome--an overlooked disease entity]

Mammary duct ectasia is a benign disease of the mammary gland, characterized by a frequently long history of tumour formation, nipple discharge, nipple retraction and mastalgia. Non-puerperal mammary abscess, which may be the presenting symptom, is also part of the syndrome. Diagnosis can often be made on the basis of the history and the clinical findings of nipple discharge, nipple retraction, tenderness on palpation, fistula formation and subareolar tumour/abscess formation. Mammography may guide diagnosis. Breast cancer is the most important differential diagnosis. If the clinical picture resembles cancer, it is necessary to perform diagnostic biopsy. Causal therapy of mammary duct ectasia is not available. Until now excision of the central mammary tissue and larger ducts has been used as treatment for the clinical manifestations of abscess, fistula and nipple discharge, apparently with good results.

Needle biopsy of breast cancer. Appearance of tumour cells along the needle track

Appearance of tumour cells along the needle track in patients with breast cancer diagnosed by SURECUT needle biopsy is found in two out of 47 consecutive cases. It is not known whether these cells are early implantation metastases or are harmless. To avoid any unnecessary risk to the patients, we recommend that needle biopsy is so performed, that it is possible to remove the track during the definite surgical procedure, and that penetration into the muscles of the thoracic wall during the biopsy procedure is avoided.

Lobular carcinoma in situ of the female breast. Short-term results of a prospective nationwide study. The Danish Breast Cancer Cooperative Group

In a Danish nationwide prospective study of in situ carcinomas and atypical lesions of the breast, 88 women, comprising 69 patients with lobular carcinoma in situ (LCIS) and 19 patients with combined lobular and ductal carcinoma in situ (LCIS + DCIS), were accrued from 1982 through 1987. All cases were treated with excision only. Within a median follow-up time of 61 months, a recurrence rate of 17% (15 cases) was found, excluding nine cases of refinding of LCIS. No contralateral recurrences occurred. The recurrences were in eight cases invasive carcinomas (IC), in six cases LCIS + DCIS, and in one case DCIS alone. The recurrence rates among cases of LCIS and of LCIS + DCIS were not significantly different. The histopathological review included an estimate of the number of lobules with LCIS and nuclear size, both of which were significantly related to recurrence. The risk of developing IC was calculated to be increased by a factor 11 as compared with the reference population.

Ductal carcinoma in situ of the female breast. Short-term results of a prospective nationwide study. The Danish Breast Cancer Cooperative Group

In a Danish nationwide prospective study of in situ carcinoma of the breast, 112 women with ductal carcinoma in situ, treated with excision only, were registered from 1982 to 1987. Within a median follow-up of 53 months, a crude recurrence rate of 22% (25 cases) was found, of which five cases recurred as invasive carcinomas and 20 cases as in situ carcinomas. The histopathologic review included a single-parameter analysis of histological growth pattern, size of lesion, nuclear size, presence of comedonecrosis, and subhistologic type. A strong interrelationship was found for histological growth pattern, nuclear size, and comedonecrosis. These parameters were also significantly related to recurrence. Cases that had clinical symptoms had a high recurrence rate as compared with cases that were discovered by mammography only or incidentally.

Primary non-Hodgkin's lymphoma of the breast (PLB): a clinicopathological study of seven cases

Seven cases of PLB in females were reviewed. Six cases involved the breast alone (stage 1E), whereas one case also involved the ipsilateral lymph nodes (stage 2E). None had B symptoms. The age range was 37-70 years (mean 52 years). The clinical course was indistinguishable from that of breast carcinomas, though the tumors were relatively large and the duration of symptoms was relatively short. Histologically, all cases were non-Hodgkin's B-cell lymphomas, comprising five cases of diffuse centroblastic type, one case of diffuse centroblastic-centrocytoid type and one case of follicular centroblastic-centrocytic type, according to the updated Kiel classification, corresponding to six cases of the diffuse large cell type and one case of the follicular mixed, small cleaved and large cell type, according to the modified International Working Formulation (IWF). None of the cases was a lymphoma of mucosa-associated lymphoid tissue (MALT). The mammary glandular tissue outside the tumor showed ductectasia in four cases. Changes resembling fibrous disease of the breast were seen in four cases, too. The inflammatory infiltration in all these cases was mainly composed of T cells. At the time of diagnosis, all patients were receiving cytoreductive treatment. The mean follow-up time was 34 months (range 5-79 months). At present, six of the patients are alive and well.

[Axillary dissection in primary surgical treatment of breast cancer: risk of false-negative axillary status]

The present study evaluates the extent of axillary dissection as part of the primary surgical treatment of operable breast cancer. Data are from the period January 1979 to August 1990 and were collected prospectively as part of the Danish Breast Cancer Cooperative Group protocols for low-risk mammary carcinoma. The series consists of 6774 breast cancer patients aged 69 years or younger. The number of axillary lymph nodes removed was related to the frequency of ipsilateral axillary recurrence, recurrence-free survival, and overall survival after a median of five years follow-up, respectively. The recurrence-free survival and overall survival rate were directly related to the number of axillary lymph nodes removed. The difference in outcome is believed to be caused by false-negative classification of axillary-positive high-risk patients in groups of patients where only a few axillary lymph nodes were removed.

Foetal antigen 2 (FA2) in relation to wound healing and fibroblast proliferation

During wound healing of human skin the deposition of foetal antigen 2 (FA2) and basement membrane (BM) components collagen type IV and collagen VII was followed. FA2 appeared on day 8 in the cytoplasm of proliferating fibroblasts and around newly formed blood vessels. As granulation tissue was formed, FA2 was seen diffusely in the loose matrix and in proliferating fibroblasts. Re-establishment of FA2 as a broad diffuse band along the BM at the dermo-epidermal junction was seen on day 22. In contrast, type IV and VII collagen were found along the BM and in the basal cells of the newly formed epithelium and a continuous linear BM distribution of these two components were re-established by days 12 and 14, respectively. The molecular weight (Mr) analysis of FA2, isolated from human skin fibroblast culture supernatants using SDS-PAGE, revealed a Mr of 27 kDa, corresponding to that of FA2 isolated from amniotic fluid. The presence of FA2 in proliferating fibroblasts and diffusely in the newly formed matrix of granulation tissue, as well as its late appearance at the BM after this had become established, suggests that FA2 takes part in connective tissue metabolism and perhaps tissue morphogenesis.

Foetal antigen 2 (FA2) in the stromal reaction induced by breast carcinoma

An indirect immunoperoxidase technique was used to examine the distribution of foetal antigen 2 (FA2), a recently described basement membrane (BM)-associated antigen, in invasive breast carcinoma (n = 34), fibroadenoma (n = 5) and normal breast tissue (n = 5), and to compare its distribution with that of laminin and collagen type IV. In normal breast tissue, FA2 was detected in the intralobular stroma as a broad band around acini and ducts, but was not present in the interlobular stroma. In areas of carcinoma in situ, FA2 was present diffusely around and in close contact with the glandular elements, the staining being more intense than that found around normal glandular structures. Two distinct patterns of FA2 distribution were found in adenocarcinomas of the breast. In the fibroblast reaction type, fibroblast staining dominated, whilst in the stromal reaction type, intense and extensive staining of the surrounding stroma dominated. Significant correlation was found between the degree of fibroblast activity and the degree of anaplasia (p = 0.005). FA2 extracted from breast carcinoma tissue was shown to be immunologically identical to FA2 fractions extracted from second trimester amniotic fluid (AF). The Mr of FA2 isolated from AF was estimated to be 26 kD, whereas the Mr of FA2 extracted from breast carcinoma tissue was slightly higher. The apparent Mr under reducing conditions were higher and three bands ranging from 26 to 29 kD were seen. FA2 was found to be immunologically distinct from collagen types I, III and IV, laminin, fibronectin and fibrinogen. The increased production and widespread distribution of FA2 in breast carcinomas suggest that FA2 is involved in the stromal changes which occur in response to tumour growth and/or invasion.

[Prognostic factors in pigmented villonodular synovitis]

By means of the diagnostic register, all of the patients in the County of Funn with the diagnosis pigmented villonodular synovitis (PVS) during the years 1973 to 1987 were localized. Out of 22 patients, 19 were followed-up clinically and radiologically. Fifteen patients had PVS in a knee, 63% were women. The diagnosis was established late. Repeated punctures of the knee with dark aspirate without relevant trauma should raise the suspicion of PVS. The clinical course of the disease could not be related to the radiographic findings or the histology but depended rather on the localization. In some patients, PVS ran a more aggressive course and arthroses were observed electively in patients treated early. The prognosis in cases diagnosed early was not better but, as treatment yielded considerable relief of symptoms, early treatment was considered to be of value. This consists of local resection of the nodular PVS elements and as radical synovectomy as possible in the diffuse cases. Treatment of recurrences should depend on the symptoms and the degree of possible arthrosis. One patient who had had three recurrences received irradiation with good results. Restricted joint mobility after synovectomy is treated effectively with brisement.

[Adjuvant chemotherapy in premenopausal and menopausal high-risk patients with breast cancer. 4. Results of the DBCG (Danish Breast Cancer Cooperative Group) 77B study]

From October 1977 to November 1982 premenopausal and menopausal high-risk breast cancer patients were included in a randomized trial (DBCG 77B). The primary surgical treatment was total mastectomy with axillary dissection. In the trial, a total of 1,034 patients were enrolled and received postoperative radiotherapy (RT) and were further randomized to 1) no systemic treatment, 2) cyclophosphamide, or 3) cyclophosphamide + methotrexate + 5-fluorouracil. The chemotherapy was given for one year. With a median observation time of ten years, the survival was 45, 60 and 62%, respectively. Retrospectively, the survival benefit was observed to be most pronounced in the age group less than 40 years, in patients with tumour size less than or equal to 5 cm or with less than or equal to 3 positive lymph nodes. In high-risk premenopausal and menopausal patients adjuvant chemotherapy combined with RT thus resulted in a more than 25% relative reduction in mortality at ten years of observation compared with RT alone.

[Rationalization profits of radiographic diagnosis with systematic follow-up DBCG (Danish Breast Cancer Cooperative Group) programs. 1]

In the follow-up programme of the DBCG-77 control and treatment protocols for primary operable breast cancer, bone scintigraphy and chest X-ray were performed 6 months after surgery and thereafter yearly until the diagnosis of any recurrence or another cancer. In the present study, the value of the two examinations was examined to the 6th year control. As a result of a low incidence of primary recurrence to the bone (0.6-3.9%) and to the chest (0.0-3.0%) the value was found to be low. About two thirds of the patients with primary chest recurrence addressed themselves with or had at the time of control pulmonary symptoms. About one half of those patients, in whom (during a 12-month period after the scheduled bone scintigraphy) bone metastases were diagnosed by another method, had a normal scheduled bone scintigraphy. False positive changes were not infrequent, especially at the bone scintigraphies. The diagnosis of asymptomatic chest recurrence did not reduce the mortality among stage II patients, among whom the greatest value of the control examination was found. It is concluded, that systematic use of bone scintigraphy and chest X-ray is not justified in a follow-up programme for stage I and II breast cancer patients.

[Adjuvant endocrine treatment of postmenopausal patients with breast cancer with high risk of recurrence. 5. Results from the DBCG (Danish Breast Cancer Cooperative Group) 77C randomized trial]

The role of antiestrogen treatment of postmenopausal breast cancer patients with high risk of recurrent disease is evaluated in a nationwide, prospective trial conducted by the Danish Breast Cancer Cooperative Group (DBCG). After total mastectomy and postoperative radiotherapy (RT), 840 patients were randomized to treatment with tamoxifen (RT + TAM) for one year, and 824 were randomized to no further therapy (RT). The recurrenceree survival (RFS) after ten years of lifeable analysis is 31% in the RT + TAM treated group, and 28% in the RT group (p = 0.01). Survival is 38% and 34% in the two treatment groups, respectively (p = 0.04). The data were further analyzed with respect to prognostic factors such as age, number of positive nodes, tumour size, and degree of anaplasia. Survival is prolonged in nearly all subgroups of patients treated with RT + TAM. However, the prolongation is only significant in patients with four or more positive nodes, with tumours of less than 5 centimeters or with tumours of anaplasia grade II. Estrogen (ER) and progesterone receptor (PgR) concentrations were measured in tumours from 309 and 219 patients, respectively. Only patients with ER and PgR values above 100 fmol/mg cytosol protein seemed to have a prolongation of survival. In conclusion, a modest survival benefit is achieved with one year of adjuvant tamoxifen treatment. Nevertheless, this is the first example of a systemic treatment approach being able to change the fatal course of breast cancer in postmenopausal patients. By means of endocrine therapy, and in the context of a new randomized trial, the DBCG will try to improve the survival in these patients even further.

Quality assurance for patients with head injuries admitted to a regional trauma unit

The efficacy of trauma systems in reducing preventable deaths has been established but the methods of auditing care are still evolving. Various "audit filters" to identify which patients' charts should be reviewed have been proposed. An analysis of all patients admitted to the Regional Trauma Unit (RTU) over a 19-month period was conducted. Of 729 patients, 135 were identified as having suffered a traumatic intracranial hemorrhage (TICH). On review, neither delay in transfer from the emergency room to the operating room nor increasing time from the incident to the operating room correlated with increasing mortality. In contrast to delay, the Glasgow Coma Scale (GCS) score on admission correlated well with outcome. The charts of patients with anomalous outcomes based on admission GCS score were reviewed, and two possibly preventable deaths were identified. There were 48 patients with TICH who had no operations but there were no deaths attributable to a missed operation. There were 76 patients for whom the GCS score at the referring hospital and the GCS score on admission to the RTU were available. Seven of 19 patients who worsened on transfer declined because of significant pulmonary injuries. Anomalous outcomes based on admission GCS score and declining GCS scores are recommended as quality assurance filters.

Immunohistochemical studies on the localization of fetal antigen 2 (FA2), laminin, and collagen type 4 in basal cell carcinoma

The localization of fetal antigen 2 (FA2), a recently described basement membrane (BM) associated antigen, was studied by immunohistochemical techniques in 21 basal cell carcinomas (BCC). In both superficial and infiltrating BCC, FA2 was located in a broad diffuse band around the tumor elements and in close contact with the BM. Compared to normal skin, a more extensive distribution of FA2 was seen in BCC. In the infiltrating BCC, FA2 staining was also present in the interstitial stroma between the tumor islands. FA2 was absent in areas with inflammatory cell infiltrates and elastoid degeneration. Epithelial and tumor cells were FA2 negative. The distribution of FA2 was clearly different from that of laminin and collagen type 4. Collagen type 4 and laminin were present as a continuous linear band corresponding to the BM surrounding the tumors. The close contact to the BM and the increased content of FA2 in the reactive stroma around BCC suggest that FA2 is involved in the matrix and/or BM changes taking place during tumor growth and invasion.

Synthesis and antibacterial activity of new C-10 quinolonyl-cephem esters

A series of cephalosporins derived from cephalothin containing an ester-linked quinolonyl substituent at the C-10 position (C-10 quinolonyl-cephem esters) has been prepared and evaluated for in vitro antibacterial activity. The C-10 quinolonyl-cephem esters exhibited a broadened spectrum of activity when compared with cephalothin and the corresponding quinolones, including activity against beta-lactamase-producing bacteria.

Fetal antigen 2 in primary and secondary brain tumors

Immunohistochemical deposition and distribution of fetal antigen 2 (FA2) was examined in normal brain tissue and in primary and metastatic tumors of the brain. In normal brain tissue FA2 was exclusively found linearly around the vessels, along pia and in arachnoidea. A similar localization was seen in primary brain tumors except in gliosarcoma where FA2 was distributed diffusely in the sarcoma region and was absent in the glioma region. In metastatic carcinoma with tumor stroma a diffuse staining reaction was seen in the stroma and with a basement membrane (BM) like staining at the tumor cell/stroma interface. Intracytoplasmic FA2 staining of the tumor cells was seen in areas without tumor stroma. In metastatic melanoma a BM like FA2 staining was seen around and between individual tumor cells. The staining patterns seen in the metastatic tumors were in accordance with that of the corresponding primary tumors.

[Preventive antibiotic treatment in cases of complicated injuries of the extremities. A questionnaire study]

A questionnaire about prophylactic antibiotic treatment of open bone, tendon or joint lesions in the extremities was sent to the 83 casualty departments in Denmark. The questionnaire was answered by 56 casualty departments which treated 72% of the orthopaedic surgical injuries in Denmark. In cases of lesions of the hands and feet, 23% of the patients were treated in casualty departments where prophylactic antibiotic therapy was rarely employed while 24% were treated in casualty departments where prophylactic antibiotic therapy was always employed. In cases of lesions proximal to the wrist or ankle joints, the figures were 26% and 38%, respectively. Penicillin was nearly always employed and treatment was commenced before or during operation. The majority of departments continued treatment for four days or longer. Considerable scatter was reported in the dosage of penicillin employed. The literature about prophylactic antibiotic treatment in cases of complicated lesions of the extremities is insufficient and controlled investigations are necessary to elucidate a series of conditions.

The relative influence of alcohol and seatbelt usage on severity of injury from motor vehicle crashes

Seatbelt usage has been consistently documented to decrease mortality and injury severity from motor vehicle crashes (MVC); however, conflicting results are available comparing mortality and injury severity, and blood alcohol positivity. Prospective testing on all MVC admissions showed that 51.5% of the non-belted, and 22% of the shoulder-belted drivers had a positive blood alcohol content (p less than 0.001). A comparison of belted and non-belted MVC drivers revealed a significantly higher mean length of stay (LOS) (p less than 0.05) and Injury Severity Score (ISS) (p less than 0.01) for the non-belted drivers. A comparison of groups positive and negative for blood alcohol revealed no significant differences in LOS or ISS, suggesting that these parameters are related to seatbelt use and not alcohol consumption.

Significance of incisional biopsy in breast carcinoma: results from a clinical trial with intended excisional biopsy

The recurrence-free survival rates (RFS) after one-stage mastectomy and partial axillary dissection in 1242 low risk breast cancer patients with invasive ductal carcinoma with or without residual cancer tissue (RCT) in the wall of the biopsy cavity were compared. RFS was significantly lower in patients with RCT (RCT-positive) whether premenopausal (n = 416) or postmenopausal (n = 826). By applying the Cox multivariate analysis on RCT and various known prognostic criteria, the incidence rates for RCT-positive patients relative to RCT-negative patients were estimated. The relative risk by RCT-positivity was in the order of 1.45, indicating that RCT is an independent risk factor contributing an increased risk of recurrence of about 45%.

Breast cancer: risk of axillary recurrence in node-negative patients following partial dissection of the axilla

The estimated probability of false negative nodal staging was analysed in breast cancer patients who had lower axillary dissection in conjunction with mastectomy and who were initially characterized as node negative. The series comprised 3128 consecutive female patients with invasive mammary carcinoma treated by surgery alone according to a nationwide DBCG protocol, 77-1a (Danish Breast Cancer Co-operative Group). The median number of removed lymph nodes was four, range 0-30. Median followup period was 6.5 years, quartiles 5.5 and 8.0 years. Ipsilateral axillary recurrences appeared in 178 patients (5.7%) during the followup period. The estimated 5-year probability for developing axillary relapse, calculated on the basis of lifetable analyses, was 19% in patients with no lymph nodes removed, 10% with 1-2 removed and negative nodes, 5% with 3-4, 3% with 5-10, and 3% with more than 10 removed and negative lymph nodes. No detectable increase in estimated probability could be traced in patients with five or more nodes removed during followup beyond 5 years, while in patients with less than five nodes removed the probability of axillary recurrence ascended continuously. The study also revealed a significantly decreased survival in groups of patients with less than five nodes removed and negative compared with groups of patients with five or more nodes removed. It is concluded that lower axillary dissection is associated with a certain risk of overlooking lymph nodes metastases, whereby the patients may be deprived of optimal adjuvant therapy. Misjudgment of the qualitative axillary nodal staging is modest, provided that at least five lower lymph nodes are removed.