ORIGINAL ARTICLE 
Annals of Nuclear Medicine Vol. 12, No. 6, 307-312, 1998 
Preoperative evaluation of the chemosensitivity of breast cancer by means of double phase 99mTc-MIBI scintimammography 
Hirofumi FUJII,* Kayoko NAKAMURA,* Atsushi KUBO,* Kohji ENOMOTO,** Tadashi IKEDA,** Tetsuro KUBOTA,** Shinjiro Wilson MATSUZAKI** and Masaki KITAJIMA** 
*Department of Radiology and **Department of Surgery, Keio University School of Medicine 
The chemosensitivity of breast cancer is important for its management, but it is difficult to evaluate preoperatively. Tc-99m hexakis-2-methoxyisobutylisonitrile (MIBI) scintimammography has been reported to indicate the expression of P-glycoprotein, which is one factor concerned with multidrug resistance. We developed a chemosensitivity assay by using surgical specimens to investigate whether 99mTc-MIBI scintimammography findings before the operation are related to chemosensitivity according to our assay. Fifteen patients with primary breast cancer were enrolled into the study. Early and delayed images were obtained at 10 and 120 minutes after intravenous injection of 99mTc-MIBI, respectively. Regions of interest were placed on the tumors and the contralateral healthy breasts in each patient to estimate 99mTc-MIBI uptake in the tumor, and retention indices were then calculated to assess the washout of 99mTc-MIBI. Chemosensitivity assay was performed by incubating surgical specimens with anticancer agents such as doxorubicin, epirubicin, pinorubicin, mitomycin C, cisplatin and 5-fluorouracil. 99mTc-MIBI washout on scintimammography was successfully related to inhibition ratios on chemosensitivity tests when compared with 99mTc-MIBI uptake by the tumor. In particular, high correlation coefficients were obtained between the retention index of 99mTc-MIBI and the inhibition ratios of doxorubicin (r=0.75), epirubicin (r=O.60) and pinorubicin (r=0.62), but poor correlation was found for mitomycin C (r=O.44) and cisplatin (r=0.31). Our results indicate that the retention index of 99mTc-MIBI is closely correlated to chemosensitivity to anthracyclines, suggesting that double-phase scintimammography allows preoperative prediction of chemosensitivity of breast cancer. 
Key words: breast cancer, technetium-99m methoxyisobutylisonitrile, scintimammography, anthracycline 
INTRODUCTION 
MANY COMBINATION chemotherapy regimens including anthracyclines have improved the outcome of patients with advanced breast cancer.1 Some cancers, however, have shown an insufficient response because of resistance to chemotherapy. Because chemotherapy of resistant tumors gives rise to serious side effects without any benefits, it is important to estimate chemosensitivity before starting treatment. 

Received May 19, 1998, revision accepted July 2, 1998. 
For reprint contact: Hirofumi Fujii, M.D., Department of Radiology, Keio University School of Medicine, 35 Shinano-machi, Shinjuku-ku, Tokyo 160-8582, JAPAN. 

We have developed an in vitro chemosensitivity test for surgical specimens2 and shown that a single-cell suspension assay by the 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) method can predict the clinical effect of chemotherapy.3 
On the other hand, a recent study has indicated that accumulation of Tc-99m hexakis-2-methoxyisobutylisonitrile (MIBI) in breast tumor cells varies with the level of expression of P-glycoprotein,4 which is involved in the resistance to anthracycline antibiotics.5 
This is a study to assess whether 99mTc-MIBI scintimammography findings are related to the chemosensitivity of breast cancer predicted by our assay methods. 
MATERIALS AND METHODS 
Subjects (Table 1) 
Fifteen female patients aged 30-66 years, diagnosed with breast cancer by mammography, ultrasonography or biopsy, were examined. Pt. 1, Pt. 6 and Pt. 13 received adjuvant chemotherapy before surgery, but no cancer chemotherapy was added between the scintimammography and the surgery. All pathological diagnoses were confirmed by using surgical specimens. 
99mTc-MIBI scintimammography 
99mTc-MIBI, 600 MBq, was injected into the cubital vein of the side opposite to the breast with the cancer lesion. Scintigraphy was performed with a GCA-7200A/DI (Toshiba, Tokyo) with a low-energy, high-resolution collimator. Lateral views of both breasts and an anterior view of the chest were acquired in the prone position. When the lateral view of the breast was acquired, the breast was suspended from the edge of the table and the contralateral breast was compressed onto the bed and a lead block was put between both breasts to avoid the influence of activity from the opposite side. Early and delayed images were obtained at 10 and 120 minutes after injection, respectively. The acquisition time of the early images was 7 minutes, and of the delayed images, 9 minutes, in consideration of the physical attenuation of Tc-99m. Accumulation of 99mTc-MIBI in the tumor was evaluated by placing the regions of interest (ROIs) on lateral scintimammography images. Breast tumors were imaged more clearly on lateral images than on anterior images because the lateral images were less affected by the activity of surrounding organs such as the heart and liver. A ROI was drawn over a tumor (L: lesion) so that it covered the whole tumor and a ROI with the same pixel size was placed on the contralateral healthy breast (N: normal). The retention index was calculated by the following equation: retention index = (delayed L/N ratio - early L/N ratio)/early L/N ratio. 
In vitro chemosensitivity test 
The sensitivity of surgical specimens to anticancer agents was assayed by the method previously reported.6 Briefly, sterile surgical specimens from primary breast tumors were minced quickly with scissors in cold Hanks' balanced salt solution, added to an enzyme cocktail containing 0.2 mg/ml collagenase, 0.2 mg/ml DNase and O.5 mg/ml pronase, and stirred in a water bath at 37deg.C for 30 min. The solution was then filtered and centrifuged at 1,000 rpm for 5 min, the supernatant was removed, and RPMI1640 containing 10% fetal calf serum (FCS) was added to the pellet. The resulting single-cell suspension was plated in 96-well plates with anticancer agents. The anticancer agents examined were doxorubicin (DOX) 10 ug/ml, epirubicin (FAM) 10 ug/ml, pinorubicin (PINO) 10 ug/ml, mitomycin C (MMC) 10 ug/ml, cisplatin (CDDP) 25 ug/ml, and 5-fluorouracil (5-FU) 50 ug/ml. After incubating the plate at 37deg.C in a humidified atmosphere containing 5% CO2 for 48 hours. MTT (5 mg/ml in PBS) 10 ul/well, and sodium succinate (0.1 M PBS) 10 ul/well were added. After an additional 4 hours of incubation, dimethyl sulfoxide, 150 ul/well, was added to dissolve the formazan salt. The absorbance of each well was read at 530-640 nm. The inhibition rate compared with the absorbance of drug-free control wells was calculated as follows: 
inhibition rate (%) = (1- (absorbance of treated wells/ absorbance of control wells)) x 100. 
Statistical analysis 
Values are shown as the means +-s.d. Statistical comparisons were made by a Mann-Whitney U-test. A p value of < O.05 was considered statistically significant. 
RESULTS 
Scintimammography 
The cases are summarized in Table 1. All of the tumors were invasive carcinomas and consisted of thirteen invasive ductal carcinomas, one mucinous carcinoma and one invasive lobular carcinoma. Seven of the 13 patients with invasive ductal carcinomas showed a scirrhous pattern. The primary tumor in all the patients was greater than 2 cm in diameter, and the primary tumors were classified as T2 or a higher grade according to the TNM clinical classification proposed by UICC in 1987. LN metastases in the axillary and parasternal regions were detected in eleven patients by ultrasonographic studies or palpation, and some of them were smaller than 1 cm in diameter. Pt. 1 had cervical node metastases. 
The primary breast tumors showed high 99mTc-MIBI accumulation on the early images in all 15 cases, and the ROIs could be placed on the tumors without any problem. Some tumors showed high washout of 99mTc, and with faint activity noted in the primary tumors on the delayed images. There were no cases in which 99mTc was completely washed out, and we were therefore able to place the ROIs on the primary tumors in both early and delayed images. 
Clear abnormal 99mTc accumulation was detected in axillary node metastases in 6 of 11 node-positive patients, weak activity in 4, and no visible activity in 1 patient. No accumulation of 99mTc-MIBI in the node metastases was evaluated. 
Typical scintimammograms are shown in Figures 1 and 2. Pt. 1 did not respond to FAM, and her scintimammogram (Fig. 1) showed strong 99mTc accumulation in the tumor on the early image, with faint activity on the delayed image, indicating the rapid washout of 99mTc. Figure 2, on the other hand, shows that 99mTc had accumulated in her tumor on the early image and the retention of 99mTc was observed on the delayed image. 
Quantitative analysis 
The scintimammography indices and results of the chemosensitivity tests are summarized in Table 2. No statistically significant differences between scirrhous and non-scirrhous carcinomas were observed in the early L/N ratio (4.47+-2.16 versus 3.41+-1.19, p = 0.35), delayed L/N ratio (3.55+-1.52 versus 3.03+-1.12, p = O.42) or retention index (-0.16+-0.13 versus -0.10+-0.18, p = O.39). The correlation between retention indices and inhibition ratios of anti-cancer agents are shown in Figure 3. The best positive correlation was shown between retention indices and inhibition ratios of DOX (r = 0.75, p = 0.008) (Fig. 3-a), followed by FAM (r = 0.60, p = 0.02), PINO (r = 0.62, p = 0.02) (Fig. 3-b, c) and 5-FU (r = 0.63, p = 0.01 ), although the effect of 5-FU was inferior to that of the other drugs (Fig. 3-f). A poor correlation between retention indices and inhibition ratios was noted for MMC (r = 0.44, p = 0.10) and CDDP (r = 0.31 , p = 0.11) (Fig. 3-d, e). The correlation coefficients of the early and delayed L/N ratios and the inhibition ratios for DOX were -0.53 and -0.28, respectively, without statistically significant correlation. 
DISCUSSION 
In this study, all of the breast cancers examined had primary tumors greater than 2 cm in diameter, and all of them could be detected by both early and delayed scintimammographies. Some metastatic nodal lesions, however, were small and not visualized well enough to put ROIs on the lesions, and therefore we did not evaluate node metastases in this study. 
We used early L/N- and delayed-ratios and retention indices to evaluate tumor uptake and washout quantitatively, and retention indices were devised to evaluate the activity of lung cancer to enable differentiation between benign and malignant lesions.7 No significant differences were noted between scirrhous and non-scirrhous carcinoma in the early L/N ratios, delayed ratios or retention indices, suggesting that these indices are independent of the pathological findings of breast tumors. A similar observation was described by Del Vecchio et al.8 
The response of breast cancer to drugs should be evaluated by the reduction in the measurable tumor size, but this clinical correlation was not observed in the present study because of small cases treated with preoperative chemotherapy. As a result, in vitro chemosensitivity testing was applied to compare the chemosensitivity of the breast tumors and 99mTc retention. Our chemosensitivity test has mainly been used for tumors of the digestive system, and the results have been shown to predict the clinical response to chemotherapy.9 We have also reported that our chemosensitivity test is effective in predicting the outcome of breast cancer.10 When 6 agents, DOX, FAM, PINO, MMC, CDDP and 5-FU were examined, a good positive correlation was observed between the retention indices and inhibition ratios of all three anthracycline antibiotics, DOX, FAM and PINO, whereas MMC and CDDP did not yield good correlations between retention indices and inhibition ratios. It is important that the preoperative scintimammography predict the chemosensitivity of breast cancer to DOX, FAM, and PINO, which are often used as the main anticancer agents for breast cancer. 
While the predictability of chemosensitivity was lower for MMC and CDDP, these drugs are not usually used for the initial treatment of breast cancer. Although 5-FU is often used for breast cancer and a strongly positive correlation between the retention index and inhibition ratio has been shown, the inhibition ratio of 5-FU was low. Because 5-FU inhibits thymidylate synthetase, tumor cells should be incubated for a long period to evaluate the anticancer effect of 5-FU. But the tumor cells from the surgical specimens did not grow well during the incubation period of the MTT method in the present study, which may explain the low inhibition ratio of 5-FU. 
Although previous studies have reported a positive correlation between P-glycoprotein expression and the Tc-99m MIBI washout ratio, many factors besides P-glycoprotein are involved in multidrug resistance, complicating the mechanism. In this study we suggested that the washout of Tc-99m MIBI from breast cancers might predict overall chemosensitivity. 
The coefficients for the correlations between the inhibition ratios and the early or delayed L/N ratios were lower than the coefficients for the correlations between the inhibition ratios and retention indices in breast cancer. The retention index, i.e., 99mTc-MIBI wash-out, is a better parameter than the early or delayed L/N ratios, i.e., the absolute uptake of 99mTc-MIBI, for predicting the response to chemotherapy. 
In conclusion, a good correlation was observed between the in vitro chemosensitivity of anthracycline anticancer agents and the retention indices on the 99mTc-MIBI scintimammograms. Our results suggest that double-phase 99mTc-MIBI scintimammography can predict the sensitivity of breast cancer to chemotherapy in the preoperative stage. 
ACKNOWLEDGMENTS 
This work was supported by Grants-in-Aid for Scientific Research (Grants-in-Aid for Encouragement of Young Scientists 09770716) from the Ministry of Education, Science, Sports and Culture. 
We wish to thank radiological technologists, Koshi Okabe and Toshikazu Sanmiya for their technical assistance in imaging and quantification. 
We are grateful to International Medical Information Center for their linguistic review. 
REFERENCES 
1. Carter SK. Cancer treatment today and its impact on drug development, with special emphasis on the phase II clinical trial. J Natl Cancer Inst 57: 235-244, 1976. 
2. Kubota T, Saikawa Y, Furukawa T, Kitajima M. Chemosensitivity test is useful to prolong the survival of the postoperative patients with gastric and colon carcinomas. Jpn J Cancer Digestive Organs 3: 287-293, 1993. 
3. Suto A, Kubota T, Shimoyama Y, Ishibiki K, Abe O. MTT assay with reference to the clinical effect of chemotherapy. J Surg Oncol 42: 28-32, 1989. 
4. Cordobes MD, Starzec A, Delmon-Moingeon L, Blanchot C, Kouyoumdjian JC, Prevost G, et al. Technetium-99m-sestamibi uptake by human benign and malignant breast tumor cells: correlation with mdr gene expression. J Nucl Med 37: 286-289, 1996. 
5. Endicott JA, Ling V. The biochemistry of P-glycoprotein-mediated multidrug resistance. Annu Rev Biochem 58: 137-171, 1989. 
6. Saikawa Y, Kubota T, Furukawa T, Suto A, Watanabe M, Kumai K, et al. Single-cell suspension assay with an MTT end point is useful for evaluating the optimal adjuvant chemotherapy for advanced gastric cancer. Jpn J Cancer Res 85: 762-765, 1994. 
7. Tonami N, Shuke N, Yokoyama K, Seki H, Takayama T, Kinuya S, et al. Thallium-201 single photon emission computed tomography in the evaluation of suspected lung cancer. J Nucl Med 30: 997-1004, 1989. 
8. Del Vecchio S, Ciarmiello A, Potena MI, Carriero MV, Mainolfi C, Botti G, et al. In vivo detection of multidrug-resistant (MDR 1) phenotype by technetium-99m sestamibi scan in untreated breast cancer patients. Eur J Nucl Med 24: 150-159, 1997. 
9. Furukawa T, Kubota T, Watanabe M, Kase S, Takahara T, Yamaguchi H, et al. Chemosensitivity testing of clinical gastrointestinal cancers using histoculture and the MTT end-point. Anticancer Research 12: 1377-1382, 1992. 
10. Furukawa T, Kubota T, Tanino H, Oura S, Yuasa S, Morita K, et al. Clinical application of histoculture drug response assay (HDRA)-distribution pattern of inhibition index and predictability of clinical chemotherapy effects-. J Jpn Soc Cancer Ther 32: 400-409, 1997.