In the study by Sickles, initial follow-up was with a unilateral mammogram at 6 months which was then followed by a bilateral mammogram at 6–12 months depending on age appropriate screening protocol, followed by two more annual follow-ups for a total of four follow-up examinations spanning up to 3.5 years. In his study the compliance rate decreased at each of the follow-up exam with compliance dropping to 66 % for the fourth follow-up exam [2]. Others have reported a compliance rate of 94.5–96 % [3–5].

Notwithstanding the extremely low incidence of malignancy among probably benign lesions that underwent surveillance based on these studies, one does encounter a higher incidence of cancers in some practices among those lesions that are categorized as probably benign. This is often due to failure to use the criteria for placing abnormalities in the probably benign category. In one reported series there were 51 malignancies identified in a group of 178 biopsies performed in patients who were assigned a probably benign assessment. On review of these findings, it was noted none of the findings fulfilled strict criteria needed to be placed in a short-interval surveillance group [7]. Calcifications were the most frequent findings in those abnormalities that proved malignant and were seen in 45 % of these cases, followed by noncalcified masses and focal asymmetry in 24 % of cases.

Stavros and others in a prospective study were able to predict benignity of a solid mass with a high degree of accuracy using morphologic criteria based on a retrospective analysis [8] (Fig. 6.5). In their series the negative predictive value for benignity was 99.5 % and was high enough to avoid biopsy. In practice these strict criteria may not be strictly adhered to while assigning solid masses to the probably benign category [9–12]. Also to note that unlike with mammography, the prospective study on probably benign category for ultrasound abnormalities was studied in a group of patients who had both palpable and nonpalpable abnormalities [8].

The validity of the use of probably benign assignment for ultrasound-detected solid masses was studied in a series of 920 solid nodules assigned to the probably benign category based on established criteria. In this series 40.9 % of the solid masses did not fulfill the strict criteria needed for categorizing a mass as being benign. All 14 cancers that were diagnosed in the probably benign group in this study belonged to the group that subsequently was categorized as suspicious [9]. Nevertheless the malignancy rate among these 920 solid nodules was still an acceptable 1.5 % and also to be noted, 8 of the 14 cancers were palpable [9]. For ultrasound-detected masses the criteria for biopsy of a lesion placed in the probably benign finding group are either an increase in diameter beyond 10 % or change in margin characteristics on a follow-up examination [9]. Stable appearance at 2 years leads to a downgrade in final assessment category to a BI-RADS 2 [9]. The cost-effectiveness of follow-up of solid masses with benign features as an alternate to biopsy has been studied and reported to lead a reduction of cost by a third [13]. Recent studies have validated the use of probably benign assessment category for ultrasound-detected solid masses [14, 15]. The malignancy rate for lesions categorized as probably benign is still low and well below the accepted 2 % rate. A series of 4,000 women reported a malignancy rate of 0.8 %. As reported by others on retrospective analysis of the features, 29/32 [90.6 %] were incorrectly classified as benign and the error was mostly due to lack of appreciation of suspicious margin characteristics [14]. The malignancy rate among the probably benign lesions was higher among the lesions that were palpable, 2.4 % [21 of 859] compared to 0.4 % [11 of 3141] among nonpalpable lesions. In a study of 288 probably benign lesions that underwent immediate histological diagnosis, only three cancers were found for a malignancy rate of 1 % which was again within the acceptable range of 2 %. Two of these three cancers was ductal carcinoma in situ and one was an invasive ductal cancer. There were 195 masses in the study group [15]. Sonographic criteria for classifying a solid mass as benign were presence of circumscribed margins, round, oval, or slightly lobulated shape, wider than tall with parallel orientation to the skin surface, internal echotexture that was iso- or slightly hypoechoic to subcutaneous fat and with normal through transmission of sound. As in the study published by Stavros and others, size of the solid mass was considered a criterion for exclusion [8, 15]. Also included in the follow-up category were sonographic findings of hypoechoic masslike abnormalities with low-level internal echoes consistent with complicated and clustered microcysts. It is, however, important to follow these criteria strictly when categorizing as probably benign. Analysis of the margin characteristics is critical and dependency only on presence of an ovoid shape may lead to misdiagnosis (Fig. 6.6a–d).