False negative mammography: causes and consequences

Radiology review of the UKCCCR Breast Screening Frequency Trial: potential improvements in sensitivity and lead time of radiological signs

AIM

To review all previous mammograms of breast cancer cases diagnosed during a randomized trial comparing 3 yearly to annual mammographic screening with a view to identifying and quantifying cases that might potentially have been diagnosed earlier.

METHODS

Mammograms of 602 breast cancer cases (399 screen-detected and 203 interval cases) were reviewed in chronological order and suspicious radiological features noted for each mammogram, up to and including the diagnostic mammogram.

RESULTS

Of the 602 cases, 79 (13%) had features at diagnosis that were visible on previous mammograms, suggesting a sensitivity of interpretation of 87%. A similar proportion of screen-detected (14%) and interval cancers (11%) had signs at diagnosis that were visible on previous mammograms. The potential for improvement was particularly noted for asymmetric density (sensitivity = 77%, average time visible before diagnosis 14 months) and parenchymal deformity/stellate lesion (sensitivity = 81%, average time visible before diagnosis 12 months).

CONCLUSION

The highest sensitivity was observed for comedo-type microcalcifications (sensitivity = 97%, average time visible before diagnosis 5 months). By improvements in sensitivity to asymmetric density and parenchymal deformity/stellate lesion, 4% of tumours could have their time of diagnosis advanced substantially.

The relevance of reported symptoms in a breast screening programme

AIM

Although the National Health Service (NHS) Breast Screening Programme is aimed at asymptomatic women, inevitably patients attending screening report symptoms. The study aim was to assess the usefulness of recall based on clinical symptoms.

MATERIALS AND METHODS

Information on breast symptoms is recorded at screening and radiologists can make recall decisions based on mammography and symptom history. We identified 1394 women with significant symptoms, between 1991 and 1996. The majority (54%) complained of a lump, 21% had breast distortion, 18% breast pain alone and 6% reported nipple discharge.

RESULTS

Of the 1394 women, 262 were recalled because of mammographic suspicion and of these, 45% had breast cancer. The other 1132 women had symptoms and benign mammograms and 44% of these were recalled. Seven breast cancers were diagnosed; all had complained of a breast lump. In two the cancer would have been seen on two-view mammography. Of 638 not recalled, five women went on to develop an interval cancer.

CONCLUSION

The results indicate that collecting details on symptoms is useful given the high rate of breast cancer in those with mammographic abnormality. When mammography is benign, however, the low rate of cancer detection means recall should be selective based on only the most relevant symptoms.

In-hospital delay in the diagnosis of breast cancer

BACKGROUND

There is evidence that delay in the diagnosis of breast cancer may prejudice survival. The aim of this study was to determine the incidence, time trends and causes of delay in a dedicated breast clinic.

METHODS

The interval between first breast clinic visit and a definitive diagnosis was recorded in all patients with invasive breast cancer between 1988 and 1997. In all patients with a delay of 3 months or more, the case notes were reviewed for evidence of a triple assessment (clinical examination, imaging and needle biopsy). The principal cause of delay was identified.

RESULTS

Of 1004 patients with invasive breast cancer, there was a delay in diagnosis of 3 months or more in 42 patients between 1988 and 1997, an incidence of 4.2 per cent. The median delay was 6 months and the median age at diagnosis was 53 (range 27-89) years. Triple assessment was undertaken in 30 patients; ten did not have a needle biopsy performed and three patients had no mammography. The principal cause of delay was: false-negative or inadequate fine-needle aspiration cytology (FNAC) in 19 patients, failure of follow-up in eight, clinical signs did not impress in five, FNAC not carried out in four, false-negative mammogram in three, failure of needle localization in two and one patient did not accept clinical advice. The annual incidence of delay in diagnosis did not change significantly over the 10-year interval.

CONCLUSION

Triple assessment is not sufficiently sensitive to detect every breast cancer and a small incidence of diagnostic delay is therefore inevitable with current techniques.

Women with normal mammography describing symptoms at screening

A retrospective study from May 1989 to June 1995 identified 4327 women with symptoms and normal mammography, from a total population of 139,852. The women with symptoms were divided into two groups: those who had been recalled for assessment, because of potentially significant symptoms, and those who had not. Out of 448 women recalled for assessment with significant symptoms, 8 malignancies were detected and 8 benign biopsies were performed. Review of the remaining 3879 women with symptoms showed that 4 had developed interval cancers in the same breast as the symptoms described at the previous screening attendance. No cancer was detected at subsequent screening examinations. This review shows that breast cancer is so uncommon in women attending for screening with symptoms that a policy of only recalling women with significant symptoms is justifiable.

Does open-access mammography and ultrasound delay the diagnosis of breast cancer?

There is a consensus that open-access breast imaging for general practitioners is inappropriate since the process omits an expert clinical examination and fine needle aspiration cytology. However, it was decided to test this hypothesis by comparing the outcome of breast referrals in a district with both an open-access imaging and a one-stop clinic. The time from referral to definitive diagnosis in all women with breast cancer was compared over a 12 month period in 1996. Of 1049 women referred for open-access imaging 20 (2%) were found to have breast cancer compared with 91 (9%) of 995 women referred to the one-stop breast clinic. There was a longer interval before the diagnosis of breast cancer was made in cases referred for open-access imaging compared with cases referred to the one-stop breast clinic (mean 63 vs 35 days). However, if patients with advanced disease are excluded the mean interval was 63 vs 44 days. The mean tumour size of breast cancers in cases referred to open-access imaging was smaller (1.5 vs 2.3 cm), there were fewer grade 3 tumours (10% vs 39%), and there was a lower rate of axillary lymph node metastases (20% vs 32%) compared with cases referred to the one-stop breast clinic. The longer interval for the open-access patients was largely due to administrative delay and the 3 patients with a delay of more than 3 months had all had a triple assessment which was false negative. This study does not support the hypothesis that open-access mammography is unsafe and should be withdrawn.