Original Research
Women's Imaging
November 21, 2013

Why Are Patients Noncompliant With Follow-Up Recommendations After MRI-Guided Core Needle Biopsy of Suspicious Breast Lesions?

Abstract

OBJECTIVE. The objective of this study was to investigate patient and breast MRI characteristics associated with noncompliance with recommended follow-up after MRI-guided core needle biopsy of suspicious breast lesions.
MATERIALS AND METHODS. A retrospective review was performed of 576 breast lesions biopsied under MRI guidance between 2007 and 2010. Patient follow-up was obtained from the medical record and from contact with referring physicians.
RESULTS. Of 415 women who underwent 576 MRI-guided core needle biopsies for suspicious breast lesions, 123 (29.6%) patients representing 154 of 576 (26.7%) lesions were noncompliant with recommended excision or 6-month MRI follow-up. Referring physicians provided information for 63% (97/154) of lesions in noncompliant patients, of which 49.5% (48/97) were followed by mammography instead of excision or MRI. Noncompliance with MRI follow-up was significantly associated with referral for biopsy by outside hospital physicians (odds ratio [OR], 2.40; p = 0.0001) and with referral for screening MRI (1.46; p = 0.093) and biopsy of a focus or foci lesion (1.63; p = 0.088). Among 178 lesions in patients compliant with follow-up MRI after MRI-guided core needle biopsy, 7.9% (14/178) had abnormal follow-up MRI results, half of which (3.9%, 7/178) were found on repeat biopsy to be high-risk or malignant.
CONCLUSION. Institutions performing MRI-guided core biopsies should be aware that patients referred from outside institutions are more likely to be noncompliant with recommended follow-up. Strategies to improve follow-up should include educating patients on the difference between mammography and MRI follow-up.
Patient compliance with MRI follow-up after MRI-guided breast core needle biopsy is essential to establish stability of concordant benign findings, ensure the lesion has been adequately sampled, and confirm removal of suspicious MRI findings after surgery [13]. Noncompliance with recommended MRI follow-up after biopsy of a benign concordant lesion may result in progression of the lesion or delay in diagnosis [35]. Lesions that decrease in size or disappear on follow-up imaging require no further workup, and lesions that are stable (no change) may indicate missed sampling at biopsy or a benign lesion. But interval growth of lesions requires additional investigation or rebiopsy. Hence, follow-up imaging after biopsy is essential to track the progress of the lesion sampled to decide whether further investigation is warranted.
Although other investigations emphasize the importance of 6-month or 1-year follow-up after MRI-guided core biopsy of breast lesions, patient compliance with MRI follow-up after MRI-guided core needle biopsy has not been specifically studied [2, 3]. In fact, very few articles have studied patient compliance with radiologic or imaging follow-up. Because imaging follow-up is so essential to track the progression of lesions, we think it is important to investigate the reasons that recommendations are not followed.
Previous studies have examined the follow-up status of women who have received stereotactic-guided or ultrasound-guided biopsies of breast lesions and were recommended for follow-up mammography or ultrasound to track changes in the lesion [57]. These studies showed that women with benign histology on biopsy were less likely to comply with follow-up mammography or ultrasound imaging. Likewise, it is recommended that women who have received MRI-guided biopsy should receive MRI follow-up, even when the biopsy shows benign histology [24]. We postulate that women with benign histology after MRI-guided core needle biopsy will also be less likely to comply with MRI follow-up recommendations.
Our study had three main aims. The first was to measure the prevalence of noncompliance with follow-up recommendations after MRI-guided core needle biopsy and investigate the reasons that some patients failed to comply with follow-up recommendations. Our second aim was to identify demographic or lesion characteristics that are associated with noncompliance with recommended MRI follow-up after MRI-guided core needle biopsy. Our third aim was to investigate the follow-up of patients who were compliant with MRI follow-up recommendations with the goal of discovering whether they required additional biopsy or excision because of suspicious findings.

Materials and Methods

We performed a HIPAA-compliant retrospective review of MRI-guided core needle biopsies of breast lesions between 2007 and 2010 with institutional review board (IRB) approval and waiver of informed consent. We reviewed the electronic medical records of 415 patients who underwent 576 MRI-guided core needle biopsies performed at our facility between January 2007 and December 2010 to determine whether the patient was compliant with the radiologist's recommendation for 6-month MRI follow-up or surgical excision of the breast lesion biopsied.
We recorded the reasons the initial diagnostic MRI was performed, which included screening patients at high risk (e.g., BRCA mutation) for breast cancer; diagnostic MRI for a known lesion (e.g., screening mammography finding); staging for existing breast cancer; and assessment of breast lesions before, during, or after chemotherapy treatment.
The diagnostic MRI was performed using a 1.5-T magnet (Signa SP, GE Healthcare) using a protocol and pulse sequence previously described [810]. Diagnostic MRI studies were interpreted by radiologists who are experienced in breast MRI interpretation and MRI-guided procedures and had knowledge of the clinical history of the patient. MRI interpretations of suspicious breast lesions and MRI-guided biopsies have been performed at our institution for more than 12 years by radiologists who are fellowship-trained in breast imaging or MRI body imaging. All patients recommended for MRI-guided core needle biopsy had breast lesions with suspicious morphology and kinetics.
We recorded whether the MRI lesion morphology was a focus or focal area (foci), nonmass enhancement, or a mass and the type of initial rise and delayed phase kinetics as defined by the American College of Radiology (ACR) BI-RADS MRI lexicon [11].
At the time of consent for biopsy, patients are informed that follow-up MRI studies will be recommended even if the core biopsy shows benign findings, similar to our postbiopsy protocols for fine-needle biopsy and core biopsy using other imaging guidance [6, 7]. A variety of protocols exist for MRI follow-up imaging of benign breast lesions. Our facility uses a benign MRI-guided core needle biopsy protocol that recommends follow-up MRI at 6 months, 12 months, and 24 months after biopsy, similar to follow-up standards set for benign core biopsies under ultrasound and stereotactic guidance and on the basis of our experience with MRI. This is compatible with other institutions that perform routine follow-up MRI at 6 months after benign concordant MRI-guided biopsy [2, 3].
For MRI guidance of the core needle biopsy, we used a 0.5-T vertically open horizontal field scanner (Signa-SP, GE Healthcare) using an elevated-platform type quadrature transmit-receive open breast coil using a protocol and dedicated pulse sequence that have been previously described [8, 9]. This protocol and pulse sequence accurately show the lesion to be biopsied at the time of the procedure. The lesion to be biopsied was compared at the time of the biopsy with the diagnostic MRI obtained with a 1.5-T magnet. Almost all of the biopsies were performed with a 9-gauge vacuum-assisted core biopsy probe; only three biopsies were performed using a 14-gauge spring-loaded core biopsy needle.
A freehand biopsy technique was used that was based on previously reported methods for freehand MRI-guided preoperative needle localization and large-core vacuum-assisted core biopsy [10, 1214]. In short, after using T1-weighted fast spin-echo (FSE) imaging to select an appropriate skin entry for the biopsy, the skin was marked, prepared, and draped in a sterile fashion and infiltrated with buffered 1% lidocaine for local anesthesia. Deep anesthesia was achieved with additional buffered 1% lidocaine and bupivacaine HCL 0.75%. A small skin incision was made with an MRI-compatible scalpel. Using T1-weighted FSE imaging, areas of architecture were initially used to guide the vacuum-assisted core biopsy trochar and introducer to the enhancing area of concern seen on the prior MRI. An average of 9–12 core samples were taken for each biopsy site. Contrast enhancement of the target was performed and confirmed with water-selective three-point-Dixon gradient-echo imaging before and after a bolus IV injection of 16 mL of 0.1 mmol/kg gadolinium (gadopentetate dimeglumine) contrast material.
Histopathology and initial diagnostic MRI findings were compared for concordance by the radiologist at the time of the biopsy. If histopathology results and imaging findings were benign and concordant, 6-month MRI follow-up imaging was recommended. If the imaging and histopathology findings were discordant or if histopathology results were malignant, surgical excision was recommended. These recommendations were relayed by telephone or electronic medical record to the referring physician or designate at the time of review, and an addendum was added to the core biopsy report by the radiologist to reflect the biopsy results and follow-up recommendations. There is currently no mechanism in place to ensure that the referring physician and the patient are scheduled for or reminded of follow-up MRI appointments.
Study data were collected and managed using a Research Electronic Data Capture (REDCap) database (open source) hosted at our facility with IRB approval [15]. REDCap is a secure web-based application designed to support data capture for research studies. The database was used to record patient demographic data, initial MRI data, MRI-guided biopsy procedure details, histology from the core needle biopsy, secondary histology information from surgical excision, and follow-up compliance. The demographic data collected for this study were limited by the broad ethnic categories of “Hispanic/Latino” and “Non-Hispanic/Non-Latino” found in the electronic medical record at our institution. Several women in our study had Asian surnames (9.7% of the total cohort, 10.4% of noncompliant patients, and 9.5% of compliant patients), but we cannot discover whether the surname corresponded to the patient's actual ethnicity. We have therefore decided to retain the Hispanic/Non-Hispanic nomenclature present in our electronic medical record. The designation “Pacific Islander” was only applied to one patient in the electronic medical record. Because this occurred only once, we recategorized this patient as Non-Hispanic/Non-Latino for the purposes of statistical analysis.
We defined compliance as the percentage of the study cohort patients who received the recommended excision or 6-month follow-up MRI at our institution during the study period. This definition is consistent with a previous study that examined compliance in patients recommended for follow-up MRI [16].
To assess compliance among patients lost to follow-up in our electronic medical record, we contacted referring physicians by letter to inquire about the status of patients who did not follow up at our institution. We asked the referring physician to tell us why the patient was unable to make her appointment. After 1 month, a second letter was sent to referring physicians who did not respond to our initial letter. If no information was obtained from the referring physician after 2 months and after multiple attempts to contact the referring physician's office by one of the investigators, the patient was considered lost to follow-up. The length of the study from initiation to obtaining all follow-up information from the electronic medical record, letters, and telephone calls to referring physicians was 1 year.
This study specifically examines the follow-up status of patient lesions rather than individual patients because one patient may have 2–4 lesions that were biopsied under MRI guidance on the same day or consecutive days. One lesion may be recommended for excision because the pathology findings are high-risk or malignant, and another lesion may be recommended for MRI follow-up because the pathology and imaging findings are benign. If the patient complies with recommended excision for one lesion but does not return for MRI follow-up of the benign lesion, the patient would be both compliant and noncompliant at the same time. We therefore decided to follow patient lesions in this study because it would tell us whether certain lesion characteristics are associated with noncompliance.
Statistical analysis was completed using Stata SE 12.1 (StataCorp). Unadjusted odds ratios (ORs) were calculated and Pearson chi-square test was used to generate p values and 95% CIs.

Results

Of 576 breast lesions biopsied under MRI guidance in 415 patients, 26.7% (154/576) of biopsies performed in 123 patients did not receive recommended excision or 6-month MRI follow-up at our facility between January 1, 2007, and December 31, 2010, judged before sending letters and making telephone calls to referring physicians requesting additional follow-up information (Fig. 1). Table 1 shows the characteristics of patients who were compliant with follow-up recommendations compared with noncompliant patients. There was no clinically significant difference between the compliant and noncompliant groups with respect to age (mean age, 50.5 and 52.4 years, respectively; p = 0.06). With regard to insurance status, Table 1 also shows that patients who were noncompliant were not significantly more likely to be covered by public insurance (Medicare, Medi-Cal, and Veterans Affairs medical care insurance) than private insurance (p = 0.27). The ethnic-racial composition of a large proportion of patients was not recorded in the electronic medical record. On the basis of the available data, there was no statistically significant difference in noncompliance among patients whose ethnicity-race was recorded as Hispanic/Latino compared with Non-Hispanic/Non-Latino (p = 0.56).
Fig. 1 —Flowchart shows compliant and noncompliant follow-up of 576 biopsies. DCIS - ductal carcinoma in situ.
TABLE 1: Demographic Characteristics of 576 Lesions Stratified by Follow-Up Status at Our Institution After MRI-Guided Core Needle Biopsy
CharacteristicsCompliant (n = 422/576, 73.3%)Noncompliant (n = 154/576, 26.7%)
Mean age (SD) (y)50.5 (10.6)52.4 (10.2)
Ethnicity  
 Non-Hispanic/non-Latino367 (87.0)107 (69.5)
 Hispanic/Latino27 (6.4)6 (3.9)
 Unknown28 (6.6)41 (26.6)
Health insurance  
 Private342 (81.0)131 (85.1)
 Public80 (19.0)23 (14.9)

Note—Except for mean age, values in parentheses are percentages.

Table 2 stratifies lesions in compliant and noncompliant patients by affiliation of referring physician, reason for initial MRI, and recommendation after biopsy. Of the 576 lesions biopsied under MRI guidance, follow-up was recommended by our radiologists as follows: 63.7% (367/576) of benign concordant lesions were recommended for 6-month MRI follow-up, 33.5% (193/576) were recommended for surgical excision or biopsy, 1.4% (8/576) were recommended for other follow-up, 1% (7/576) were recommended for clinical follow-up, and 0.2% (1/576) were recommended for 1-year MRI follow-up. Noncompliant patients were more likely to have lesions recommended for 6-month MRI follow-up compared with those who were recommended to undergo surgical excision after the MRI-guided core needle biopsy (OR = 5.33; 95% CI, 3.13–9.47, p < 0.00001).
TABLE 2: Characteristics of 576 Breast Lesions Stratified by Follow-Up Status at Our Institution After MRI-Guided Core Needle Biopsy
CharacteristicsCompliant (n = 422/576, 73.3%)Noncompliant (n = 154/576, 26.7%)Total
Referring physician for breast lesion   
 Affiliated303 (71.8)71 (46.1)374 (64.9)
  Obstetrician gynecologist0 (0)0 (0)0
  Oncologist50 (11.8)6 (3.9)56 (9.7)
  Primary care physician0 (0)0 (0)0
  Radiation oncologist6 (1.4)6 (3.9)12 (2.1)
  Surgeon247 (58.5)59 (38.3)306 (53.1)
 Nonaffiliated119 (28.2)83 (53.9)202 (35.1)
  Obstetrician gynecologist16 (3.8)2 (1.3)18 (3.1)
  Oncologist56 (13.3)24 (15.6)80 (13.9)
  Primary care physician13 (3.1)11 (7.1)24 (4.2)
  Radiation oncologist5 (1.2)1 (0.6)6 (1.0)
  Surgeon29 (6.9)45 (29.2)74 (12.8)
Reason initial MRI was performed for lesion   
 Screening122 (28.9)63 (40.9)185 (32.1)
 Diagnosis100 (23.7)28 (18.2)128 (22.2)
 Staging178 (42.2)60 (39.0)238 (41.3)
 Chemotherapy10 (2.4)010 (1.7)
 Other12 (2.8)3 (1.9)15 (2.6)
Recommendation after lesion biopsy   
 6-month follow-up MRI232 (55.0)135 (87.7)367 (63.7)
 12-month follow-up MRI1 (0.2)0 (0)1 (0.2)
 Surgical excision or biopsy174 (41.2)19 (12.3)193 (33.5)
 Clinical follow-up7 (1.7)0 (0)7 (1.2)
 Other8 (1.9)0 (0)8 (1.4)

Noncompliant Patients

Of the 154 lesions in noncompliant patients, 12.3% (19/154) were recommended for surgical excision after MRI core needle biopsy and 87.7% (135/154) were recommended for 6-month MRI follow-up.
We contacted by mail the referring physicians of 123 patients who were noncompliant with imaging or excision follow-up recommendations. Fifty-five of 123 (44.7%) patients had a physician response after two inquiry letters and a 2-month hiatus, representing 41.6% (64/154) of the total number of lesions biopsied under MRI guidance in patients noncompliant with follow-up recommendations. After telephone calls to referring physicians asking why the patient did not comply with recommended imaging follow-up or surgical excision, additional information was obtained, resulting in a total of 63% (97/154) of lesions among 123 noncompliant patients with follow-up information after at least two letters and telephone calls to referring physicians. For the remaining 37% (57/154) of lesions for which no follow-up information was obtained, 9.9% (57/576) of total lesions in the entire cohort, the referring physician was contacted multiple times about patient follow-up after MRI-guided core needle biopsy at our facility, but no answer was provided.
Table 3 shows the final disposition of 154 lesions in 123 patients who were noncompliant with recommended follow-up. Patients with four lesions (2.6%, 4/154) recommended for 6-month follow-up MRI underwent MRI elsewhere—but this follow-up was performed outside of the recommended 6-month window at 8, 8, 14, and 26 months. From the letters sent to referring physicians, we discovered that 13.6% (21/154) of lesions were being followed up by mammography, and through an additional chart review, we found another 17.5% (27/154) of lesions being followed up by mammography in our own facility. Thus, a total of 31.2% (48/154) of lesions in patients noncompliant with excision or MRI follow-up recommendations were followed up by mammography, and 34.8% (47/135) of lesions specifically recommended for MRI follow-up (not excision) were followed up by mammography instead. Referring physicians who returned our letters also reported that the follow-up status of 9.1% (14/154) of breast lesions that were biopsied under MRI guidance and recommended for MRI follow-up was unknown.
TABLE 3: Final Disposition of 154 Lesions in Noncompliant Patients Recommended for Excision or Biopsy or 6-Month MRI Follow-Up
Final DispositionFollow-Up RecommendationTotal Lesions (n = 154, 100%)
Excision or Biopsy (n = 19/154, 12.3%)6-Month MRI Follow-Up (n = 135/154, 87.7%)
Doctor contacted but no answer0 (0)57 (42.2)57 (37.0)
Patient underwent MRI elsewhere outside the 6-month window0 (0)4 (3.0)4 (2.6)
Patient lost to follow-up by referring physician0 (0)14 (10.4)14 (9.1)
Alternative follow-up recommended3 (15.8)0 (0)3 (1.9)
Ultrasound chosen instead of MRI or biopsy0 (0)0 (0)0 (0)
Mammography chosen instead of MRI or biopsy1 (5.2)47 (34.8)48 (31.2)
Patient lost insurance0 (0)2 (1.5)2 (1.3)
Patient relocated0 (0)0 (0)0 (0)
Young age indicated clinical follow-up0 (0)0 (0)0 (0)
Patient died0 (0)0 (0)0 (0)
Patient refused follow-up MRI0 (0)2 (1.5)2 (1.3)
Biopsy or excision at outside hospital15 (78.9)0 (0)15 (9.7)
Other disposition0 (0)9 (6.7)9 (5.8)

Note—Data in parentheses are percentages.

Moreover, letters from referring physicians informed us that two patients had no follow-up because of loss of insurance and two patients refused MRI follow-up. No physician letters indicated that patients were noncompliant because of alternative follow-up with ultrasound, death, relocation, or young age that indicated clinical follow-up.
Table 3 also shows that of the 154 lesions in our 123 noncompliant patients, 19 lesions were recommended for surgery after MRI-guided core needle biopsy. Of these 19 lesions, 15 underwent excision elsewhere, three had “alternative” follow-up, and one had follow-up by mammography or ultrasound instead. Many of the outside facilities would not reveal the pathology of the surgical or excisional biopsy or the results of the follow-up imaging tests in these patients recommended for surgical excision.

Compliant Patients

There were 292/415 (70.4%) patients compliant with follow-up recommendations (MRI follow-up, rebiopsy, or surgical excision) representing 73.3% (422/576) of the total number of lesions biopsied under MRI guidance. Of 422 lesions, 55.0% (232/422) were recommended for 6-month MRI follow-up (Table 2).
About three quarters (178/232, 76.7%) of the patient lesions recommended for 6-month MRI follow-up in the compliant cohort underwent 6-month follow-up MRI. For the other 54 of 232 (23.3%) patient lesions recommended for 6-month MRI follow-up in the compliant cohort, 15 of 232 (6.5%) lesions were removed by excisional biopsy and 39 of 232 (16.8%) were removed by mastectomy. These 54 patient lesions in patients recommended for MRI follow-up are included in the compliant cohort because the patient was willing to comply with the recommendation for MRI follow-up but later decided after consultation with her referring physician to undergo excision or mastectomy to remove the breast lesion; hence, the patient no longer required MRI follow-up to assess stability of the lesion. No patients in the compliant cohort who were recommended for follow-up MRI were given alternative follow-up with mammography or ultrasound, unlike the noncompliant cohort.
Of the patients with the 178 lesions who were recommended for follow-up MRI and returned to our institution and underwent the recommended follow-up MRI at 6 months, most (164/178, 92.1%) showed no suspicious findings. The remaining 14 of 178 (7.9%) MRI follow-up studies were recommended either for surgical excision (8/14, 57.1%) or repeat biopsy (6/14, 42.9%) based on suspicious findings. Of those lesions that were recommended for repeat core needle biopsy, if the histopathology revealed high-risk or malignant findings, the radiologist recommended excision, and the patient underwent an excisional biopsy by a surgeon. Thus, all core needle biopsies of suspicious lesions that revealed high-risk or malignant pathology in patients who were compliant with 6-month follow-up MRI were excised.
Half of these 14 lesions showed high-risk or malignant pathology on repeat biopsy or excision, and the remaining half were benign. The seven high-risk or malignant lesions represent 3.9% (7/178) of the total number of lesions that were seen on follow-up imaging in compliant patients. Of the seven high-risk or malignant lesions, one (1/7, 14.3%) was invasive ductal carcinoma and ductal carcinoma in situ, five (5/7, 71.4%) were papillomas, and one (1/7, 14.3%) was a radial scar. Thus, follow-up MRI in compliant patients showed five (5/178, 2.8%) papillomas, one (1/178, 0.56%) radial scar, and one (1/178, 0.56%) invasive malignancy. The lesion that was found to be invasive ductal carcinoma corresponded with the location of the original biopsy site; a metallic biopsy clip was seen just superior to the mass. The patient elected to undergo mastectomy, and a 4.5-cm mass containing invasive ductal carcinoma and ductal carcinoma in situ was described in the pathology report.

Compliant Versus Noncompliant Patients

Sixty-four percent (367/576) of breast lesions were recommended for 6-month MRI follow-up after MRI-guided core needle biopsy (Table 4). Of 367 lesions, 63.2% (232/367) were recommended for 6-month MRI follow-up and underwent either 6-month MRI follow-up or surgical removal of the lesion (compliant), and 36.8% (135/367) of lesions recommended for 6-month MRI follow-up did not undergo the recommended follow-up at our institution (noncompliant). Among those patients compliant with MRI follow-up imaging, 29.7% (69/232 lesions) were referred to our institution by outside physicians. Among noncompliant patients recommended for follow-up imaging, 50.4% (68/135 lesions) were referred to our institution by outside physicians. We hypothesized that patients would be less likely to comply with recommended MRI follow-up if referred by outside physicians rather than physicians affiliated with our institution. We found a strong and statistically significant association between referral from nonaffiliated physicians and lesions in women who were noncompliant with recommended MRI follow-up (OR, 2.40; p = 0.0001).
TABLE 4: Factors Associated With Noncompliance With Recommended 6-Month MRI Follow-Up After MRI-Guided Core Needle Biopsy in 367 Breast Lesions
CharacteristicsCompliant (n = 232/367, 63.2%)Noncompliant (n = 135/367, 36.8%)Odds Ratio (95% CI)p
Referring physician    
 Nonaffiliated (outside referral)69 (29.7)68 (50.4)2.40 (1.5–3.8)0.0001
 Affiliated163 (70.3)67 (49.6)1 (reference)NA
Reason MRI was performed for lesion    
 Screening76 (32.8)56 (41.5)1.46 (0.92–2.31)0.093
 Nonscreening156 (67.2)79 (58.5)1 (reference)NA
 Diagnosis57 (24.6)24 (17.8)NANA
 Staging84 (36.2)52 (38.5)NANA
 Chemotherapy6 (2.6)0 (0)NANA
 Other9 (3.9)3 (2.2)NANA
MRI lesion type    
 Focus or foci or nonmass170 (73.3)107 (79.3)1.39 (0.82–2.41)0.20
 Focus or foci76 (32.8)56 (41.5)1.63 (0.90–2.99)0.088
 Nonmass94 (40.5)51 (37.8)1.20 (0.66–2.20)0.52
 Mass62 (26.7)28 (20.7)1 (reference)NA

Note—Except where indicated otherwise, data in parentheses are percentages. NA indicates not applicable.

Table 4 shows compliance status for the 367 of 576 (63.7%) lesions recommended for 6-month MRI follow-up stratified by reason for initial MRI referral. Of 135 lesions among noncompliant patients recommended for 6-month follow-up imaging, 41.5% (56/135) were discovered by screening MRI. Of 232 lesions among compliant patients recommended for 6-month follow-up imaging, 32.8% (76/232) were discovered by screening MRI. We hypothesized that patients referred for a screening as opposed to nonscreening MRI (i.e., for diagnosis, staging, chemotherapy, or other treatment) would be more likely not to comply with recommended 6-month MRI follow-up. Noncompliance with MRI follow-up was associated with having undergone screening MRI (OR, 1.46; p = 0.093) as opposed to nonscreening MRI, but this association was not statistically significant.
We hypothesized that the MRI lesion type (focus or foci or nonmass compared with mass) may be associated with noncompliance with 6-month MRI follow-up after MRI-guided core needle biopsy (Table 4). Patients with a total of 367 of 576 (63.7%) lesions were recommended for 6-month MRI follow-up after MRI-guided core needle biopsy. Of those 367 lesions, 232 (63.2%) were recommended for MRI follow-up in compliant patients, and 135 (36.8%) were recommended for MRI follow-up in noncompliant patients. Of the 232 (63.2%) lesions in patients compliant with MRI follow-up recommendations, 32.8% (76) were described as focus or foci on the initial MRI, and 41.5% (56/135) of lesions in noncompliant patients were described as focus or foci on the initial MRI. Noncompliance with MRI follow-up was associated with a focus on initial MRI as opposed to a mass (OR, 1.63; p = 0.088), but this association was not statistically significant.

Discussion

To our knowledge, our study is the largest follow-up study of compliance after MRI-guided core needle biopsy. Of the 576 MRI-guided core needle biopsies in 415 patients, we showed that 70.4% (292/415) of patients could be tracked in our facility's electronic medical record to determine compliance with the recommendation of the radiologist to have the lesion excised or followed by MRI surveillance. Patients who were recommended for 6-month MRI follow-up were five times more likely to be noncompliant than those recommended for surgical excision after the MRI-guided core needle biopsy (OR, 5.33; p < 0.00001). Referral by a nonaffiliated physician as opposed to a physician affiliated with our institution significantly more than doubled the odds of noncompliance with recommended MRI follow-up (OR, 2.40; p = 0.0001). If the initial MRI referral was for breast cancer screening (OR, 1.46; p = 0.093) or if the finding on initial MRI was a focus or foci (OR, 1.63; p = 0.088), the association with noncompliance with MRI follow-up was not statistically significant, likely due to lack of power. We also discovered that of 154 lesions in noncompliant patients, 29.2% (45/154) of lesions underwent mammography follow-up instead of MRI, whereas only 2.6% (4/154) of lesions underwent outside MRI. Of 178 lesions in compliant patients who were followed by MRI, 14 had suspicious findings on the 6-month MRI, of which seven were either high-risk or cancer (five papillomas, one radial scar, and one invasive malignancy).
A recent study by Marshall et al. [17] of compliance with follow-up MRI in patients with BI-RADS 3 classification found similar compliance rates as in our study: 72.7% (96/132) of women complied with recommended MRI follow-up or biopsy. They also showed that patients with a BRCA mutation were more likely to comply with follow-up (9/9, 100%) compared with non-BRCA carriers (87/123, 70.7%). Our study, however, is different from the study by Marshall et al. in several respects: We examined follow-up after MRI-guided core needle biopsies, not follow-up MRI in patients with BI-RADS 3 classification; our sample size was larger (576 vs 132); we investigated the reasons for noncompliance with imaging follow-up via contact with the referring physician by telephone and letter; we found different associations with noncompliance; and we investigated the findings in patients compliant with follow-up MRI who required rebiopsy or excision.
Another recent study by Sung et al. [3] investigated patient follow-up after biopsy of MRI-detected lesions. This study emphasized the importance of a 6-month follow-up MRI in detecting new or occult lesions after MRI-or ultrasound-guided biopsies compared with ultrasound or mammography follow-up. Our study, by contrast, highlights issues related to compliance with follow-up imaging or surgery after MRI-guided biopsy, reasons for noncompliance that we managed to elicit from referring physicians, associations with noncompliance, and whether high-risk or malignant findings were discovered in patients who complied with follow-up MRI recommendations.
We found that women recommended for excision were more likely to comply with follow-up than those recommended for 6-month MRI follow-up of concordant benign results (p < 0.00001). Moreover, patients with 79% (15/19) of the lesions recommended for excision who failed to return to our institution followed through with excision at an outside facility. The remaining four of 19 (21%) underwent other follow-up: one by mammography and three by alternative follow-up after consultation with the referring physician. Thus, most high-risk or malignant lesions in the noncompliant cohort underwent appropriate follow-up, which we found reassuring.
We postulate that there are many reasons for noncompliance with 6-month follow-up MRI. Regarding the statistically significant association between outside referral and noncompliance with MRI follow-up, we think that outside referring physicians may not have easy access to facilities that perform MRI examinations. This may account for the large number of patients who underwent mammography follow-up instead of the recommended MRI follow-up. Some outside referring physicians may have thought that follow-up studies after benign biopsy results are unnecessary and therefore did not inform their patients about the necessity of regular MRI follow-up to assess the stability of the lesion sampled. Moreover, nonaffiliated referring physicians may be associated with poor follow-up because their patients live far from our institution and imaging follow-up could more easily be accomplished at an institution closer to the patient's home, even if that follow-up was mammography or clinical examination instead of MRI. Other factors associated with noncompliance with MRI follow-up were undergoing screening MRI as opposed to MRI associated with diagnosis or chemotherapy treatment and having a focus or foci on initial MRI as opposed to a mass. Both of these characteristics may have seemed less serious to a patient or referring physician. Although these findings did not meet strict criteria for statistical significance, the p values were very close to the α level of 0.05. Thus, the failure to reach statistical significance may have been due to our moderate sample size and lack of power rather than the absence of a true effect.
A surprising number of patients who were noncompliant with recommended 6-month MRI follow-up underwent mammography follow-up (48/154, 31.2%). This may be due to several factors, such as a perception that mammography is just as effective as MRI, insurance plans that cover mammography but not MRI, and discomfort with the MRI procedure (two patients refused follow-up MRI in our study). Moreover, some insurance plans require preapproval before obtaining follow-up MRI, and this may explain why some patients did not comply with recommended follow-up. And if the insurance plan does not deem the follow-up MRI study as medically necessary, this, too, may explain noncompliance with the follow-up recommendation.
Although mammography is used to screen for early breast cancer, it is not as sensitive as MRI in detecting occult malignances, nor is it as effective as MRI in finding lesions in dense breasts [1820]. Other studies have shown that MRI is more sensitive than mammography without a significant decrease in specificity [2123]. These studies show that mammography follow-up after MRI-guided core needle biopsy is no substitute for MRI follow-up. Follow-up MRI can immediately be compared with the prior study and is far more likely to detect a change in the lesion that originally prompted the radiologist to recommend biopsy.
Follow-up MRI is important because it can detect interval growth of previously biopsied lesions or detect new lesions not seen on prior MRI. It is therefore important to comply with recommended MRI follow-up at designated time intervals because a small percentage of the follow-up MRI studies may show high-risk or malignant findings after core biopsy, as we found in our compliant patients. A study by Meissnitzer et al. [4] found that follow-up MRI after ultrasound-guided biopsy of suspicious breast lesions found lesions that did not correspond with the sonographic lesions that were originally biopsied in 10 of 80 (12.5%) cases. Subsequent MRI-guided biopsy of nine of 10 lesions found five of nine malignancies. This study supports our finding that timely follow-up MRI after biopsy is important because new or evolving lesions may be detected that require additional biopsy or excision.
Patient compliance with recommended follow-up imaging is a well-recognized problem in the literature [6, 7, 24]. Our results show that patient noncompliance with recommended MRI follow-up after breast biopsy is 29.6% (123/415). Our findings are consistent with the most recent studies of compliance rates of women who were recommended for follow-up mammography. The Digital Mammographic Imaging Screening Trial in 2011 found that 29% of patients with BI-RADS 3 mammography results were noncompliant with recommended follow-up [25]. A 2001 study by Vizcaino et al. [26] found that follow-up noncompliance for benign lesions was 11%. In 1998, Goodman et al. [7] found that only 54% of patients were compliant with imaging surveillance after stereotactic- and ultrasound-guided breast core biopsy, and in 1996, Pal et al. [6] showed that there was 27.8% noncompliance with mammographic surveillance after fine-needle aspiration biopsy. These studies raise an important question with relation to our study: Why is noncompliance with MRI follow-up imaging similar to mammography follow-up noncompliance? It may be that some of the characteristics associated with noncompliance found in our study (e.g., nonaffiliated referring physician) not only pertain to MRI-guided procedures but may also be associated with noncompliance among women recommended for mammography follow-up after stereotactic or ultrasound-guided biopsies.
Originally, patients with 367 of 576 (63.7%) lesions were recommended for 6-month MRI follow-up because they had benign concordant findings on imaging and pathology, but patients with 135 of 367 (36.8%) of these lesions were noncompliant (or did not return to our institution for 6-month MRI) and patients with 232 of 367 (63.2%) lesions underwent follow-up at our institution. Of these 232 lesions, patients with 54 (23.3%) lesions underwent some form of alternative follow-up in consultation with the referring physician, and patients with 178 (77%) lesions actually underwent 6-month follow-up MRI at our institution. We do not know how many lesions from the sizeable proportion of women who underwent some alternative follow-up at our institution (54/367, 14.7%) or who were noncompliant and failed to return to our institution (135/367, 37%) may have harbored high-risk or malignant lesions.
The follow-up MRI studies for patients with the 178 lesions who did return for the 6-month follow-up MRI detected 14 (7.9%) lesions that required additional biopsy or excision because the imaging findings were suspicious for high-risk lesions or malignancy. The reference standard pathology results of seven of 14 (50% positive predictive value) biopsies or excisions were characterized as true-positive findings for high-risk or malignant lesions: one invasive ductal carcinoma, five papillomas, and one radial scar.
On the basis of the data we collected, we unfortunately cannot calculate the true-positive rate (i.e., sensitivity) for 6-month MRI follow-up because we did not continue to follow patients whose 6-month MRI follow-up was negative to see whether they developed high-risk or malignant lesions on their 12-month or 24-month follow-up MRI. One reason we did not follow patients was because women whose 6-month follow-up MRI occurred in 2010 could not be followed into 2011 because we only collected data on biopsies conducted between 2007 and 2010. We cannot comment on sensitivity or the true-positive rate because we do not know the number of false-negative results in our population needed to calculate sensitivity: [sensitivity = true-positive / (true-positive + false-negative)].
In terms of the eventual yield of invasive ductal carcinoma among all women who underwent 6-month MRI follow-up, our finding of 0.56% (1/178) is consistent with a recent study that indicated that 0.6–0.9% of patients who were recommended for follow-up MRI and complied were found to have malignancy [16]. However, recent studies by Liberman et al. [27] and Linda et al. [28] suggest that high-risk lesions, such as papillomas and radial scars seen on follow-up MRI may reveal malignancy on excision and thus should be excised as well. When we compare our finding of seven of 178 (3.9%) high-risk or malignant lesions with recent studies, we think that 6-month follow-up MRI for benign concordant breast lesions is reasonable at this time. Future studies should evaluate the cost-effectiveness of 6-month MRI follow-up studies compared with 12-month follow-up studies in patients with benign concordant lesions after MRI-guided core needle biopsy.
MRI follow-up of suspicious breast lesions after core needle biopsy might be improved by stressing the importance of follow-up among women with the chief characteristics that were found in this study. A tracking system could specifically identify women with the characteristics associated with noncompliance, such as outside referring physician, and ensure that they and the referring physicians are sent reminder notices about their upcoming appointments. The database used to capture the data for the patients in this study is now being used for quality assurance purposes. Patients with the characteristics associated with noncompliance can now be tracked. The next phase of our study is to ensure that patients at risk for noncompliance are sent additional notices regarding the date and reason for their follow-up appointments.
Moreover, it seems that some patients (or referring physicians) may have thought that mammography follow-up after MRI-guided biopsy is just as effective as MRI follow-up. Radiologists might be able to improve follow-up by explaining to the patients and referring physicians why MRI is preferred to mammography in cases of MRI-guided core needle biopsy. At our institution, the radiologist emphasizes to the patient and referring physician that MRI is the best imaging modality to track the progress of the lesion that was originally biopsied under MRI guidance.
The cost in time spent in electronic medical record review, letter communication, and telephone calls to obtain information on noncompliant patients is not trivial. In our study, additional information on follow-up was obtained on 97 lesions only after an exhaustive analysis of all patients undergoing MRI-guided core needle biopsy to search for those who did not return for recommended excision or MRI follow-up. The search for more information required a full-time commitment from one of the investigators for 5 months. One example of the difficulty in obtaining follow-up on these patients is that we contacted three facilities to track the follow-up of a single patient. This patient had gone to two other facilities in our state for imaging or physician follow-up and then went to a third facility in yet another state to obtain the MRI-guided biopsy that was recommended.
Although computerized patient reminder systems may be helpful to inform patients and referring physicians that follow-up is needed, no follow-up was done and no answer to our letters was obtained for patients with 57 lesions despite numerous attempts to contact the referring physician. In our facility, trained personnel inform patients at the time of MRI-guided core needle biopsy of the importance of MRI follow-up. Despite these measures, patients with 9.9% (57/576) of total lesions in the study were lost to follow-up (i.e., we have no information about whether the patient received any type of follow up after MRI-guided core needle biopsy). It is important to acknowledge that after MRI-guided biopsy of breast lesions some patients will not follow-up with MRI studies as recommended and that there are variable types of follow-up in the community. In some cases, we simply do not know why MRI follow-up was not obtained, or why clinical follow-up or other imaging was substituted.
There are limitations to this retrospective study. Our moderate sample size may have undermined our ability to detect statistical significance for several lesion characteristics. Moreover, of 154 lesions in noncompliant patients, we only received information from referring physicians for 97 lesions, a response rate of 63%. We can only speculate as to what happened to the rest of the patient lesions because their follow-up status remains unknown. Additional information about these lesions might have changed the outcome of our data, specifically with regard to the number of patients who opted for follow-up mammography instead of MRI. Another limitation is that we were unable to obtain more specific racial and ethnic information about individuals categorized as Non-Hispanic/Non-Latino in the electronic medical record. We readily acknowledge that a more comprehensive breakdown of the ethnicity of our patients could have yielded additional information regarding the compliance practices of various ethnic groups, especially African Americans, whites, and Asians. Our data can only tell us that there appears to be no difference in the compliance of Hispanic/Latinos and Non-Hispanic/Non-Latinos with respect to follow-up after MRI-guided core needle biopsy of suspicious breast lesions.
With regard to the 63% response rate observed in our study, Visser et al. [29] in 1996, Curtin et al. [30] in 2000, Holbrook et al. [31] in 2007, and Keeter et al. [32] in 2006 compared response rates in the 50% range to studies in the 25% range and found no distinguishable statistical significance between the two. Moreover, Cummings et al. [33] found that the average response rate to a mailed questionnaire was 61% overall and 52% for large studies. Thus, our response rate of 63% is above average for studies of this kind.

Conclusion

The aim of this study was to investigate patient and breast MRI characteristics associated with noncompliance with recommended follow-up after MRI-guided core needle biopsy of suspicious breast lesions and highlight the importance of MRI follow-up in detecting new or progressing lesions. After core needle biopsy, patients recommended for MRI follow-up were significantly more likely not to comply with these recommendations compared with those recommended for surgical excision. We found three associations with noncompliance with 6-month MRI follow-up: a strong and statistically significant association with referral from nonaffiliated physicians at outside institutions, a nonstatistically significant association with referral for screening MRI, and a nonstatistically significant association with focus or foci lesions seen on MRI before biopsy. We also found that follow-up mammography was frequently used in place of follow-up MRI. Lastly, we underscored the importance of MRI follow-up imaging by showing that some lesions visualized on follow-up MRI required subsequent excision or biopsy that revealed high-risk or malignant findings. Therefore, strategies to improve MRI follow-up after MRI-guided core needle biopsy should be focused on explaining the difference between mammography and MRI follow-up to the patient and referring physician and ensuring that a tracking system can identify patients who are more likely to be noncompliant.

Footnote

This work was supported by a grant from the Stanford University School of Medicine Medical Student Scholars Program.

References

1.
Perlet C, Heywang-Kobrunner SH, Heinig A, et al. Magnetic resonance-guided, vacuum-assisted breast biopsy: results from a European multicenter study of 538 lesions. Cancer 2006; 106:982–990
2.
Li J, Dershaw DD, Lee CH, Kaplan J, Morris EA. MRI follow-up after concordant, histologically benign diagnosis of breast lesions sampled by MRI-guided biopsy. AJR 2009; 193:850–855
3.
Sung JS, Lee CH, Morris EA, Comstock CE, Dershaw DD. Patient follow-up after concordant histologically benign imaging-guided biopsy of MRI-detected lesions. AJR 2012; 198:1464–1469
4.
Meissnitzer M, Dershaw DD, Lee CH, Morris EA. Targeted ultrasound of the breast in women with abnormal MRI findings for whom biopsy has been recommended. AJR 2009; 193:1025–1029
5.
Chang SD, Abrahams Z, Gordon PB. Patient compliance with recommended follow-up after fine-needle aspiration biopsy of solid breast masses. Can Assoc Radiol J 2005; 56:36–39
6.
Pal S, Ikeda DM, Birdwell RL. Compliance with recommended follow-up after fine-needle aspiration biopsy of nonpalpable breast lesions: a retrospective study. Radiology 1996; 201:71–74
7.
Goodman KA, Birdwell RL, Ikeda DM. Compliance with recommended follow-up after percutaneous breast core biopsy. AJR 1998; 170:89–92
8.
Leong CS, Daniel BL, Herfkens RJ, et al. Characterization of breast lesion morphology with delayed 3DSSMT: an adjunct to dynamic breast MRI. J Magn Reson Imaging 2000; 11:87–96
9.
Agoston AT, Daniel BL, Herfkens RJ, et al. Intensity-modulated parametric mapping for simultaneous display of rapid dynamic and high-spatial-resolution breast MR imaging data. RadioGraphics 2001; 21:217–226
10.
Daniel BL, Birdwell RL, Butts K, et al. Freehand iMRI-guided large-gauge core needle biopsy: a new minimally invasive technique for diagnosis of enhancing breast lesions. J Magn Reson Imaging 2001; 13:896–902
11.
American College of Radiology. Breast imaging reporting and data system atlas (BI-RADS atlas). Reston, VA: American College of Radiology, 2003
12.
van de Ven SM, Lin MC, Daniel BL, et al. Freehand MRI-guided preoperative needle localization of breast lesions after MRI-guided vacuum-assisted core needle biopsy without marker placement. J Magn Reson Imaging 2010; 32:101–109
13.
Daniel BL, Butts K, Glover GH, Cooper C, Herfkens RJ. Breast cancer: gadolinium-enhanced MR imaging with a 0.5-T open imager and three-point Dixon technique. Radiology 1998; 207:183–190
14.
Brenner RJ, Shellock FG, Rothman BJ, Giuliano A. Technical note: magnetic resonance imaging-guided preoperative breast localization using “freehand technique.” Br J Radiol 1995; 68:1095–1098
15.
Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap): a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009; 42:377–381
16.
Eby PR, DeMartini WB, Peacock S, et al. Cancer yield of probably benign breast MR examinations. J Magn Reson Imaging 2007; 26:950–955
17.
Marshall AL, Domchek SM, Weinstein SP. Follow-up frequency and compliance in women with probably benign findings on breast magnetic resonance imaging. Acad Radiol 2012; 19:406–411
18.
Helbich TH. Contrast-enhanced magnetic resonance imaging of the breast. Eur J Radiol 2000; 34:208–219
19.
Heywang SH, Wolf A, Pruss E, et al. MR imaging of the breast with Gd-DTPA: use and limitations. Radiology 1989; 171:95–103
20.
Berg WA, Gutierrez L, NesAiver MS, et al. Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer. Radiology 2004; 233:830–849
21.
Sardanelli F, Giuseppetti GM, Panizza P, et al. Sensitivity of MRI versus mammography for detecting foci of multifocal, multicentric breast cancer in fatty and dense breasts using the whole-breast pathologic examination as a gold standard. AJR 2004; 183:1149–1157
22.
Warner E, Plewes DB, Hill KA, et al. Surveillance of BRCA1 and BRCA2 mutation carriers with magnetic resonance imaging, ultrasound, mammography, and clinical breast examination. JAMA 2004; 292:1317–1325
23.
Leach MO, Boggis CR, Dixon AK, et al. Screening with magnetic resonance imaging and mammography of a UK population at high familial risk of breast cancer: a prospective multicentre cohort study (MARIBS). Lancet 2005; 365:1769–1778
24.
Helvie MA, Pennes DR, Rebner M, Adler DD. Mammographic follow-up of low-suspicion lesions: compliance rate and diagnostic yield. Radiology 1991; 178:155–158
25.
Baum JK, Hanna LG, Acharyya S, et al. Use of BI-RADS 3-probably benign category in the American College of Radiology Imaging Network Digital Mammographic Imaging Screening Trial. Radiology 2011; 260:61–67
26.
Vizcaino I, Gadea L, Andreo L, et al. Short-term follow-up results in 795 nonpalpable probably benign lesions detected at screening mammography. Radiology 2001; 219:475–483
27.
Liberman L, Tornos C, Huzjan R, Bartella L, Morris EA, Dershaw DD. Is surgical excision warranted after benign, concordant diagnosis of papilloma at percutaneous breast biopsy? AJR 2006; 186:1328–1334
28.
Linda A, Zuiani C, Furlan A, et al. Radial scars without atypia diagnosed at imaging guided needle biopsy: how often is associated malignancy found at subsequent surgical excision, and do mammography and sonography predict which lesions are malignant? AJR 2010; 194:1146–1151
29.
Visser PS, Krosnick JA, Marquette J, Curtin M. Mail surveys for election forecasting? An evaluation of the Colombia Dispatch poll. Public Opin Q 1996; 60:181–227
30.
Curtin R, Presser S, Singer E. The effects of response rate changes on the index of consumer sentiment. Public Opin Q 2000; 64:413–428
31.
Holbrook A, Krosnick J, Pfent A. The causes and consequences of response rates in surveys by the news media and government contractor survey research firms. In: Lepkowski JM, Tucker NC, Brick JM, et al., eds. Advances in telephone survey methodology. New York, NY: Wiley, 2007
32.
Keeter S, Kennedy C, Dimock M, Best J, Craighill P. Gauging the impact of growing nonresponse on estimates from a national RDD telephone survey. Public Opin Q 2006; 70:759–779
33.
Cummings SM, Savitz LA, Konrad TR. Reported response rates to mailed physician questionnaires. Health Serv Res 2001; 35:1347–1355

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: 1391 - 1400
PubMed: 24261382

History

Submitted: November 10, 2012
Accepted: March 30, 2013

Keywords

  1. biopsy
  2. breast imaging
  3. compliance
  4. MRI

Authors

Affiliations

Matthew O. Thompson
Stanford University School of Medicine, 300 Pasteur Dr, Stanford, CA 94305.
Jafi Lipson
Department of Radiology, Breast Imaging Section, Stanford Hospitals and Clinics, Stanford, CA.
Bruce Daniel
Department of Diagnostic Radiology, Stanford Hospitals and Clinics, Stanford, CA.
Chivonne Harrigal
Department of Radiology, Breast Imaging Section, Stanford Hospitals and Clinics, Stanford, CA.
Paul Mullarkey
Department of Radiology, Breast Imaging Section, Stanford Hospitals and Clinics, Stanford, CA.
Sunita Pal
Department of Radiology, Breast Imaging Section, Stanford Hospitals and Clinics, Stanford, CA.
Atalie C. Thompson
Stanford University School of Medicine, 300 Pasteur Dr, Stanford, CA 94305.
Debra Ikeda
Department of Radiology, Breast Imaging Section, Stanford Hospitals and Clinics, Stanford, CA.

Notes

Address correspondence to M. O. Thompson ([email protected]).

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options

PDF

View PDF

PDF Download

Download PDF

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media