Purpose: Functional imaging with tomographic near-infrared diffuse optical spectroscopy (DOS) can measure tissue concentration of deoxyhemoglobin (Hb), oxyhemoglobin (HbO2), percent water (%water), and scattering power (SP). In this study, we evaluated tumor DOS parameters and described their relationship to clinical and pathologic outcome in patients undergoing neoadjuvant therapy for locally advanced breast cancer.

Experimental Design: Ten patients were enrolled and intended to undergo five scans each. Scans were taken up to 3 days before treatment and at 1, 4, and 8 weeks after neoadjuvant treatment before surgery. Changes in volume of interest weighted tissue Hb, HbO2, %water, and SP corresponding to the tumor were compared with clinical and pathologic response.

Results: All patients' tumor volumes of interest were significantly different compared with background tissue for all parameters. Five patients had a good pathologic response. Four patients were considered nonresponders. One patient initially did not respond to chemotherapy but, after a change in chemotherapy, had a good response. In the five patients with a good response, the mean drop in Hb, HbO2, %water, and SP from baseline to the 4-week scan was 67.6% (SD = 20.8), 58.9% (SD = 20.3), 51.2% (SD = 28.3), and 52.6% (SD = 26.4), respectively. In contrast, the four nonresponders had a mean drop of 17.7% (SD = 9.8), 18.0% (SD = 20.8), 15.4% (SD = 11.7), and 12.6% (SD = 10.2) for Hb, HbO2, %water, and SP, respectively.

Conclusions: Responders and nonresponders were significantly different for all functional parameters at the 4-week scan, except for %water, which approached significance. Thus, DOS could be used as an early detector of tumor response. Clin Cancer Res; 16(9); 2605–14. ©2010 AACR.

See Commentary by Tromberg and Cerussi, p. 2486

Translational Relevance

This study is a prospective evaluation of the use of functional tomographic near-infrared diffuse optical spectroscopy in monitoring tumor response in patients undergoing neoadjuvant treatment for locally advanced breast cancer. We show that as early as 4 weeks, diffuse optical spectroscopy parameters of oxygenated hemoglobin, deoxygenated hemoglobin, percent water, and scattering power are predictors of final whole-mount mastectomy histopathology. Early identification of treatment response in patients undergoing neoadjuvant therapy can potentially improve clinical outcomes if ineffective treatment is changed to a more efficacious treatment.

Breast cancer is the second leading cause of cancer-related mortality in women in North America. In 2007, there were an estimated 262,810 in situ and invasive cases of breast cancer in the United States and Canada (1, 2). There locally advanced breast cancer (LABC) represents 5% to 20% of all newly diagnosed breast cancers (3). This diagnosis typically includes tumors >5 cm or involving the skin or chest wall, or inflammatory breast cancer. It also includes patients with supraclavicular, infraclavicular, internal mammary, or fixed or matted axillary lymph nodes. All stage III and a subset of stage IIB (T3N0) tumors are usually considered locally advanced (4).

Women with LABC, owing to the extensive nature of the disease, typically have poor outcomes in terms of local and systemic control. Long-term survival is ∼50% (5). The emerging standard treatment for LABC involves multimodality treatment with neoadjuvant chemotherapy. Although there is no evidence to suggest that there is a survival benefit of neoadjuvant chemotherapy when compared with adjuvant treatment, some studies have shown an overall survival benefit for those patients who achieve a pathologic complete response after neoadjuvant chemotherapy compared with patients without a complete response (611). An advantage of neoadjuvant chemotherapy is that it decreases tumor bulk and may allow breast-conserving therapy. Finally, assessment of the effect of chemotherapy on gross disease may indicate the potential response of distant micrometastases—something that is not possible in the adjuvant setting (12).

Traditional methods of monitoring tumor response, including clinical palpation, X-ray mammography, ultrasonography, and magnetic resonance imaging (MRI), have been typically used as anatomic measures of disease. Conventional computed tomography uses ionizing radiation, produces artifact due to cardiac and breathing motion, and lacks fine detail in spatial resolution (13). It has not been used yet in routine clinical practice for the detection or monitoring of breast tumor responses. X-ray mammography is commonly used for disease detection but is not as useful in the presence of fibrotic tissue (14) or high mammographic density (15). In addition, some studies show its poor correlation with pathology (10, 16). Mammography has not found use in disease treatment monitoring. Standard anatomic ultrasound is not very sensitive for disease detection. Offering high resolution and high sensitivity, anatomic MRI is the standard in most cancer centers for monitoring LABC. However, it remains relatively costly and no standard dynamic contrast-enhanced criteria of response have yet emerged. Yeh et al. (17) found that palpation, X-ray mammography, ultrasound, and MRI showed agreement with pathologic response in 19%, 26%, 35%, and 71% of cases, respectively. On its own, clinical palpation is used routinely in clinic but remains a relatively subjective and inaccurate method of assessing tumor response and has been found to correlate poorly with pathology (14).

The need for a noninvasive and inexpensive imaging modality to both diagnose disease and monitor treatment response has given recent new enthusiasm to optical imaging. Optical imaging in breast cancer can be traced as far back as 1929 when light was first used with breast tissue to observe absorption patterns (18). Due to low resolution in the breast and poor differentiation between tumor and benign tissue, this method did not gain clinical interest. However, in recent decades with advances in computing, in addition to source and detection technology, this imaging modality has been revisited (19).

Near-infrared (IR) diffuse optical spectroscopy (DOS) is a noninvasive three-dimensional tomographic technique that quantitatively measures near-IR absorption and scattering spectra across tissue. There are three types of DOS systems: “continuous wave,” “time domain,” and “frequency domain.” Continuous wave systems emit light at constant amplitude and measure its attenuation, whereas time domain systems emit short picosecond-range pulses of light. Measured temporal distributions of the photons are used to infer the optical properties of breast tissue. Finally, frequency domain systems emit light continuously but the amplitude is modulated. Amplitude decay and phase shift are used to infer the optical properties of the breast tissue (20). Through calculations, optically derived functional images of the major absorption chromophores of oxyhemoglobin (HbO2), deoxyhemoglobin (Hb), and percent water (%water) can be obtained. Although functional imaging with DOS in comparison with conventional imaging modalities provides less spatial resolution, it can potentially provide valuable functional information shortly after the initiation of cancer treatment.

Angiogenesis, tumor cell proliferation, and hypoxia are known factors involved in tumor growth that alter HbO2, Hb, %water, and lipid measurements (21). In a review of DOS breast imaging, ∼2,000 women were evaluated from multiple studies, with ∼85% of breast lesions detectable. Breast lesions were found to contain approximately twice the concentration of Hb compared with background tissue. Breast tissue optical parameters [Hb, HbO2, %water, %lipid, and scattering power (SP)] varied with age, body mass index, premenopausal or postmenopausal status, and fluctuations in the menstrual cycle. Despite this, it remains difficult to distinguish benign (such as fibroadenomas) from malignant tissue (20).

Given its limitations in spatial resolution and difficulty in using DOS as a screening tool, interest has shifted to the use of DOS in monitoring treatment response. LABC is an obvious candidate for study because of the use of neoadjuvant treatment and the pathologic data that are available after surgery. Multiple studies, primarily case reports (2226), have been published that show the utility of DOS in predicting early tumor response to neoadjuvant treatment. In the study here with 10 patients, we report to our knowledge the first study that relates functional DOS parameters (Hb, HbO2, %water, and SP) at multiple times during neoadjuvant therapy to clinical and final whole-mount pathologic outcome.

Our ultimate goal is to identify early optical parameters of response that predict for a good clinical response. In cases where there is a predicted poor response, chemotherapy regimens could be changed earlier (after weeks rather than months) based on functional imaging. We show here a concordance between early parameters from functional imaging within 4 weeks and clinical/histopathologic response. We show that optical parameters could be used to discriminate between patients who responded to chemotherapy and those who did not.

Protocol

This study was conducted in accordance with research ethics approval. Ten patients with LABC were recruited to take part in this study. Informed consent was obtained and five scans were planned for each patient. The first scan took place before the start of neoadjuvant therapy, and a subsequent three scans were to take place at 1, 4, and 8 weeks after the start of treatment. The final scan was scheduled after completion of neoadjuvant therapy before surgery. The timing of scans was intentionally weighted toward the beginning of neoadjuvant therapy to investigate early optical markers of response.

DOS images were obtained with the patient lying prone with the pendant breast suspended and stabilized between polymethylmethacrylate plates in an aquarium. Optical compensation medium with optical properties similar to breast tissue was added to the aquarium. DOS images were acquired with the SoftScan platform (Advanced Research Technologies) and took ∼1 hour per study.

All patients were biopsied to confirm a cancer diagnosis before treatment. Baseline imaging of these patients included a clinical MRI of the breast and metastatic work-up as necessary as part of the institutional standard of care for such patients. Patients were followed clinically in a routine fashion by an oncologist blinded to the DOS data. An MRI was obtained before mastectomy. Clinical examination was done at each follow-up, and an assessment of tumor size was recorded. Assessment of clinical response, duration, and changes in therapy was made blinded to DOS information. Pathology was examined after mastectomy on full-mount pathology slides (27), and data on tumor size, grade, histologic subtype, and tumor response were recorded. Pathologists were blinded to DOS data, and vice versa.

All studies were examined in the same manner by the same pathologist (J.Z.), with expertise in reading whole-mount mastectomy slides, who interpreted all LABC pathology. The slides were examined for residual tumor size, content of cellularity, and whether residual cells seemed to exhibit a chemotherapy effect or seem viable. Tumors were categorized as having either a complete pathologic response, good pathologic response, or minimal pathologic response. Complete response was defined as not having any residual invasive or in situ tumor in the pathologic specimen. Good pathologic response meant an over 50% diminishment in size (compared with pretreatment size) with a significant decrease in cellularity (≤10% cells seem invasive and viable). Minimal pathologic response referred to a small change in tumor size (enlargement or diminishment to at best 50%).

Instrument

The acquisition platform used in the study was a time-resolved, optical imaging device used to measure photon migration through the breast in the near-IR range (28). The laser emission assembly was composed of four individual pulsed semiconductor diode lasers (LDH-P, PicoQuant) operating at 690, 730, 780, and 830 nm with a pulse duration (full width at half-maximum) of <150 picoseconds, an average output of 0.5 mW when driven at 20 MHz (PDL 808, PicoQuant), and an oscillator module to synchronize the drivers. The pulses were time multiplexed (12.5 ns) through a single fiber on the emission side of the breast. Photons were collected through five optical fibers positioned on a mobile detector head in an “M” configuration on the collection side of the breast and detected by a photomultiplier (H7422P-50, Hamamatsu). Patients were positioned prone on a tabletop with a pendant breast scanned in an aquarium in a raster pattern. The optical compensation medium was an oil-in-water emulsion that mimicked average optical properties of the human breast (coefficient of absorption = 0.04 cm−1 and coefficient of scatter = 10 cm−1). The acquired data were reconstructed, and tomographic functional images were created from the optical parameters.

Data analysis

Transverse tomographic images of the breast in a craniocaudal direction were reconstructed from scattering data. Image reconstruction was carried out to provide sagittal and coronal views. The DOS system provided three-dimensional images of the breast with measurements of HbO2, Hb, %water, and SP with a 3 mm by 3 mm voxel resolution.

The tumor was identified in the pretreatment DOS images for each of the functional parameters and verified using information from each patient's clinical exam, mammogram, ultrasound, and MRI with regard to tumor size and location. For HbO2, Hb, %water, and SP, the tumor visually corresponded quite well with the areas of highest measured values. A volume of interest (VOI) for HbO2, Hb, %water, and SP on all of the DOS scans was created by obtaining a threshold value that was a fraction of the maximum measured value. This threshold value was adjusted such that the VOI corresponded in size and location as close as possible to the actual tumor as known through other imaging and clinical exam (see Fig. 1). The threshold value was kept consistent across all of the studies, at each time point, for the same optical parameter for each patient. Mean measured values in the VOI of each of the functional parameters were obtained.

Fig. 1.

Breast tumor on functional DOS and MRI. Left, transverse tomographic slice with measurements of HbO2. Red, area of high HbO2; blue, area of low HbO2. The white line represents the boundaries of the VOI. The threshold value defining the boundary of the VOI was 14.5 μmol/L. Right, MR image of the same patient with a 5.5 × 6.5 × 7.5 cm invasive ductal carcinoma. Scale bar, 1 cm.

Fig. 1.

Breast tumor on functional DOS and MRI. Left, transverse tomographic slice with measurements of HbO2. Red, area of high HbO2; blue, area of low HbO2. The white line represents the boundaries of the VOI. The threshold value defining the boundary of the VOI was 14.5 μmol/L. Right, MR image of the same patient with a 5.5 × 6.5 × 7.5 cm invasive ductal carcinoma. Scale bar, 1 cm.

Close modal

Patient characteristics

Ten patients were recruited to this study. Seven patients had all five scans as planned. Two patients underwent four scans each and one patient had three scans. Pathology results were available for all but one patient. The one patient without pathology developed bone metastases and did not undergo a mastectomy. All but one patient had pretreatment and presurgery MRI. Patient, tumor, and treatment characteristics are presented in Table 1.

Table 1.

Patient characteristics

CharacteristicValue
Mean age 50 years (range, 38-64) 
Menstrual status before treatment 
    Premenopausal 6 patients 
    Postmenopausal 4 patients 
Neoadjuvant treatment 
    Chemoradiotherapy 3 patients 
    Epirubicin and docetaxel 2 patients 
    AC + T 2 patients 
    FEC + D 1 patient 
    Sunitinib/trastuzumab→docetaxel/trastuzumab/pamidronate 1 patient 
    Docetaxel/carboplatinum/trastuzumab 1 patient 
Mean maximum tumor size 7.7 ± 2.4 cm (range, 5-11.2) 
Tumor histology 
    Lobular carcinoma 2 patients 
    Ductal carcinoma 8 patients 
Hormone receptor (estrogen or progesterone receptor) 
    Positive 7 patients 
    Negative 3 patients 
Mean body mass index 25.2 ± 7.0 
Her-2-neu  
    Positive 4 patients 
    Negative 6 patients 
Triple negative (estrogen, progesterone, and Her-2-neu) 2 patients 
Grade 
    1 0 patients 
    2 10 patients 
    3 0 patients 
CharacteristicValue
Mean age 50 years (range, 38-64) 
Menstrual status before treatment 
    Premenopausal 6 patients 
    Postmenopausal 4 patients 
Neoadjuvant treatment 
    Chemoradiotherapy 3 patients 
    Epirubicin and docetaxel 2 patients 
    AC + T 2 patients 
    FEC + D 1 patient 
    Sunitinib/trastuzumab→docetaxel/trastuzumab/pamidronate 1 patient 
    Docetaxel/carboplatinum/trastuzumab 1 patient 
Mean maximum tumor size 7.7 ± 2.4 cm (range, 5-11.2) 
Tumor histology 
    Lobular carcinoma 2 patients 
    Ductal carcinoma 8 patients 
Hormone receptor (estrogen or progesterone receptor) 
    Positive 7 patients 
    Negative 3 patients 
Mean body mass index 25.2 ± 7.0 
Her-2-neu  
    Positive 4 patients 
    Negative 6 patients 
Triple negative (estrogen, progesterone, and Her-2-neu) 2 patients 
Grade 
    1 0 patients 
    2 10 patients 
    3 0 patients 

Abbreviations: AC, Adriamycin/cyclophosphamide; T, trastuzumab; FEC, 5-fluourouracil/epirubicin/cyclophosphamide; D, docetaxel.

Patients received a variety of neoadjuvant treatment regimens. Three patients received treatment under a study at our institution receiving weekly docetaxel (25 mg/m2) concurrent with radiotherapy including the breast, axilla, and supraclavicular area of 50 Gy in 25 fractions over 5 weeks. One patient had docetaxel (75 mg/m2) and carboplatinum (AUC, 6), with trastuzumab (6 mg/kg). Another patient had 5-fluourouracil (500 mg/m2), epirubicin (100 mg/m2), and cyclophosphamide (500 mg/m2) followed by docetaxel (100 mg/m2; 5-fluourouracil/epirubicin/cyclophosphamide + docetaxel). Two patients received Adriamycin (60 mg/m2) and cyclophosphamide (600 mg/m2) followed by paclitaxel (175 mg/m2; Adriamycin/cyclophosphamide + trastuzumab). Another two patients received epirubicin (90 mg/m2) and docetaxel (75 mg/m2). The final patient received sunitinib (37.5 mg) and trastuzumab (4 mg/kg, day 1, and then 2 mg/kg weekly) initially, but this was changed after three cycles to docetaxel (75 mg/m2), trastuzumab (4 mg/kg, day 1, and then 2 mg/kg weekly), and pamidronate (90 mg). Doses were adjusted and treatment was delayed under the medical oncologist's discretion depending on toxicity and response.

Image measurement

Reconstructed images of optical absorption and scattering at each of the four wavelengths were produced as above, and images of the functional parameters, Hb, HbO2, %water, and SP, were generated. VOIs were placed in the area of the tumor and over adjacent normal tissue to extract quantitative measurements on all of the DOS studies. Table 2 displays the differences between tumor and the background tissue for each of the functional parameters. All parameters showed a significant difference compared with the background normal breast tissue.

Table 2.

Differences between tumor and normal tissue

ParameterMean tumorMean normalP
Hb (μmol/L) 10.23 ± 2.47 6.41 ± 1.30 <0.0001 
HbO2 (μmol/L) 23.82 ± 5.85 14.01 ± 5.15 <0.0001 
%water 0.52 ± 0.17 0.15 ± 0.10 <0.0001 
SP 1.57 ± 0.45 0.55 ± 0.28 <0.0001 
ParameterMean tumorMean normalP
Hb (μmol/L) 10.23 ± 2.47 6.41 ± 1.30 <0.0001 
HbO2 (μmol/L) 23.82 ± 5.85 14.01 ± 5.15 <0.0001 
%water 0.52 ± 0.17 0.15 ± 0.10 <0.0001 
SP 1.57 ± 0.45 0.55 ± 0.28 <0.0001 

Clinical and pathologic response to neoadjuvant therapy

No patient in this study had a complete response. Table 3 displays patients' initial tumor size, the treatment regimen, and pathologic results. Patient 1 had a weak partial response with the tumor initially measuring 5.0 × 4.5 × 4.5 cm [anterior/posterior (AP) × medial/lateral (ML) × superior/inferior (SI)] on MR before treatment with final surgical pathology revealing a 4.5 × 4.0 × 1.8 cm lesion after chemoradiotherapy. Despite having almost no invasive disease left behind, there was significant in situ tumor remaining. Patients 3, 5, and 7 were considered minimal pathologic responders; the tumors on initial MR measured 6.1 × 7.8 × 3.3 cm, 11.0 × 6.3 × 5.5 cm, and 7.5 × 6.5 × 5.5 cm, respectively. Final surgical pathology revealed minimal response with the tumors remaining relatively unchanged in size. Patients 4, 6, 8, 9, and 10 had good pathologic responses. Although the volume occupied by the remaining lesion on final pathology was in some cases large, the appearance of decreased tumor cellularity and only scattered islands of cells indicated a good response (as defined in Materials and Methods).

Table 3.

Individual patient results

Patient no.Neoadjuvant treatmentMRI dimensions of tumor pretreatment (AP × ML × SI)MRI dimensions of tumor presurgery (AP × ML × SI)Pathologic dimensions of tumor (AP × ML × SI)NotesResponse
Chemoradiotherapy 5 × 4.5 × 4.5 3.0 × 3.4 × 5.5 1.8 × 4 × 4.5 Some response but significant in situ tumor present Weak pathologic response 
Sutent/Herceptin→docetaxel/trastuzumab/pamidronate 11.2 × 7.3 × 8.3 3.3 × 1.4 × 2.1 N/A Clinically good response with second chemotherapy regimen Initial poor clinical response then good clinical response with change in chemotherapy 
Epirubicin and docetaxel 6.1 × 7.8 × 3.3 3.7 × 6.2 × 2.8 5.2 × 7 × 1.8 Minimal response to chemotherapy Minimal pathologic response 
Chemoradiotherapy 10 × 6* No palpable disease 1 × 1 × 0.7 Very small volume of disease remaining Good pathologic response 
Epirubicin and docetaxel 11 × 6.3 × 5.5 10.6 × 7.7 × 7.1 4.5 × 4 × 9 Minimal response to chemotherapy Minimal pathologic response 
AC + D 7.8 × 4.6 × 5.5 0.5 × 0.5 × 0.5 2 × 7 × 3 Despite nests of tumor occupying a large volume, very minimal tumor cellularity overall Good pathologic response 
AC + D 7.5 × 6.5 × 5.5 6.5 × 3.5 × 2.7 8.9 × 7.7 × 3 Tumor cellularity remains high Minimal pathologic response 
Chemoradiotherapy 5.1 × 5.5 × 4.5 4.4 × 3.2 × 3.9 2.5 × 4.8 × 3.6 Numerous small foci of invasive disease; only 10% of tumor is invasive Good pathologic response 
Docetaxel/carboplatinum/trastuzumab 4.8 × 3.1 × 5.5 2.8 × 0.8 × 3.9 2 × 6 × 3 Very small nests of cells, very marked response Good pathologic response 
10 FEC + D 10.2 × 7.2 × 6.8 4.3 × 2.3 × 1.9 1.7 × 6 × 4.5 Invasive tumor is present as single cells only, rare groups of cells Good pathologic response 
Patient no.Neoadjuvant treatmentMRI dimensions of tumor pretreatment (AP × ML × SI)MRI dimensions of tumor presurgery (AP × ML × SI)Pathologic dimensions of tumor (AP × ML × SI)NotesResponse
Chemoradiotherapy 5 × 4.5 × 4.5 3.0 × 3.4 × 5.5 1.8 × 4 × 4.5 Some response but significant in situ tumor present Weak pathologic response 
Sutent/Herceptin→docetaxel/trastuzumab/pamidronate 11.2 × 7.3 × 8.3 3.3 × 1.4 × 2.1 N/A Clinically good response with second chemotherapy regimen Initial poor clinical response then good clinical response with change in chemotherapy 
Epirubicin and docetaxel 6.1 × 7.8 × 3.3 3.7 × 6.2 × 2.8 5.2 × 7 × 1.8 Minimal response to chemotherapy Minimal pathologic response 
Chemoradiotherapy 10 × 6* No palpable disease 1 × 1 × 0.7 Very small volume of disease remaining Good pathologic response 
Epirubicin and docetaxel 11 × 6.3 × 5.5 10.6 × 7.7 × 7.1 4.5 × 4 × 9 Minimal response to chemotherapy Minimal pathologic response 
AC + D 7.8 × 4.6 × 5.5 0.5 × 0.5 × 0.5 2 × 7 × 3 Despite nests of tumor occupying a large volume, very minimal tumor cellularity overall Good pathologic response 
AC + D 7.5 × 6.5 × 5.5 6.5 × 3.5 × 2.7 8.9 × 7.7 × 3 Tumor cellularity remains high Minimal pathologic response 
Chemoradiotherapy 5.1 × 5.5 × 4.5 4.4 × 3.2 × 3.9 2.5 × 4.8 × 3.6 Numerous small foci of invasive disease; only 10% of tumor is invasive Good pathologic response 
Docetaxel/carboplatinum/trastuzumab 4.8 × 3.1 × 5.5 2.8 × 0.8 × 3.9 2 × 6 × 3 Very small nests of cells, very marked response Good pathologic response 
10 FEC + D 10.2 × 7.2 × 6.8 4.3 × 2.3 × 1.9 1.7 × 6 × 4.5 Invasive tumor is present as single cells only, rare groups of cells Good pathologic response 

NOTE: Results with anatomic MRI imaging indicated a mean volume change of 72 ± 70% (mean ± SD) from pretreatment to presurgery in pathologic nonresponders and a mean volume change of 16 ± 19% (mean ± SD) in pathologic responders. The difference between the two groups based on anatomic MRI was not significant (P = 0.18).

*No MRI was done; measurements are based on clinical and mammographic assessment.

Patient 2 was a unique case. This 38-year-old female presented with metastatic disease in the bone. Initial MR revealed a right-sided invasive ductal carcinoma measuring 11.2 × 7.3 × 8.3 cm. The patient was started on sunitinib/trastuzumab chemotherapy and, after three cycles, was noted clinically to be getting worse with a right-sided axillary node increasing in size. When symptoms were observed, a whole-body bone scan was done and metastatic bone progression was noted. The patient's chemotherapy was then changed to docetaxel/pamidronate/trastuzumab. Symptomatically and clinically, the patient began to improve almost immediately. On clinical exam, the large breast mass had subsequently decreased in dimensions to 3 cm × 3 cm (ML × SI).

DOS response to neoadjuvant therapy

For each patient, the mean measured value for the functional parameter in the VOI was multiplied by its volume. Figure 2 shows the changes in parameters resulting from neoadjuvant treatment as a percentage of the initial value, in both a typical nonresponder (patient 5) and a typical responder (patient 6). Alternatively, this was carried out on the basis of the volume occupied by each functional parameter/chromophore alone with nearly identical results (Supplementary Fig. S1).

Fig. 2.

Responders versus nonresponders for each of the optical parameters as measured in breast tumors. The graphs illustrate the %change measured within tumor volume plotted over time for a typical nonresponder (patient 5) and responder (patient 6). The corresponding pretreatment and presurgery DOS images of the Hb are also shown along with one whole-mount slide of the pathologic specimen for each of the responder and nonresponder. Note that in the nonresponder, there is minimal change in the Hb from pretreatment to presurgery. The area of residual disease in the nonresponder and responder is identified by the small black arrow. In the responder, there was only a few nests of cells in three pathology slices surrounded by fibrosis. The remaining slices were clear of disease. Scale bars, 1 cm.

Fig. 2.

Responders versus nonresponders for each of the optical parameters as measured in breast tumors. The graphs illustrate the %change measured within tumor volume plotted over time for a typical nonresponder (patient 5) and responder (patient 6). The corresponding pretreatment and presurgery DOS images of the Hb are also shown along with one whole-mount slide of the pathologic specimen for each of the responder and nonresponder. Note that in the nonresponder, there is minimal change in the Hb from pretreatment to presurgery. The area of residual disease in the nonresponder and responder is identified by the small black arrow. In the responder, there was only a few nests of cells in three pathology slices surrounded by fibrosis. The remaining slices were clear of disease. Scale bars, 1 cm.

Close modal

A representative whole-mount pathology slide illustrating the pathologic response is also presented. Patients 1, 4, 6, 8, and 9 showed significant decreases in HbO2, Hb, %water, and SP compared with values before treatment. Clinically and pathologically, these patients were determined to be good partial responders, except for patient 1. Patients 3, 5, and 7 were clinical and pathologic nonresponders, and they exhibited only a minimal response to the functional parameters HbO2, Hb, %water, and SP. Patient 2 exhibited an initial increase in all the parameters. However, after the change in chemotherapy in this patient before the 8-week scan, there seemed to be a significant decrease in the optically derived parameters.

The DOS parameters corresponded well with the clinical and pathologic outcomes of these patients. Figure 3 displays graphs of each of the parameters for patients who were clinical and pathologic responders (patients 4, 6, 8, 9, and 10) versus the nonresponders (patients 3, 5, and 7). The graphs for the responders and nonresponders seem to separate at ∼4 weeks after initiation of neoadjuvant treatment. Table 4 presents the values as a percentage of the pretreatment scan for each of the optical parameters at 4 weeks of time. The responders and nonresponders were significantly different (using a two-sided t test) at 4 weeks of time for HbO2 (P = 0.049), Hb (P = 0.012), and SP (P = 0.050) and approached significance for %water (P = 0.081).

Fig. 3.

Changes in functional DOS parameters with treatment as measured in breast tumors. The four graphs illustrate each of the functional parameters plotted as a %change from pretreatment of responders versus nonresponders at 4 weeks of time within breast tumors. Bars, SE. At 4 weeks, HbO2, Hb, %water, and SP were all lower for responders compared with nonresponders. All differences were statistically significant, except for %water, which approached significance. The measured values within adjacent normal tissue were relatively invariant for the functional parameters. For instance, the mean variability for the %water parameter was 2.21% (SE) for all patients and for the SP parameter was 2.56%. The remaining parameters were comparable.

Fig. 3.

Changes in functional DOS parameters with treatment as measured in breast tumors. The four graphs illustrate each of the functional parameters plotted as a %change from pretreatment of responders versus nonresponders at 4 weeks of time within breast tumors. Bars, SE. At 4 weeks, HbO2, Hb, %water, and SP were all lower for responders compared with nonresponders. All differences were statistically significant, except for %water, which approached significance. The measured values within adjacent normal tissue were relatively invariant for the functional parameters. For instance, the mean variability for the %water parameter was 2.21% (SE) for all patients and for the SP parameter was 2.56%. The remaining parameters were comparable.

Close modal
Table 4.

Results of statistical analysis

Optical parameterResponders at 4 weeksNonresponders at 4 weeksP
HbO2 33.6% 82.0% 0.03 
Hb 29.6% 82.3% 0.0044 
%water 40.4% 84.6% 0.60 
SP 39.3% 87.4% 0.036 
Optical parameterResponders at 4 weeksNonresponders at 4 weeksP
HbO2 33.6% 82.0% 0.03 
Hb 29.6% 82.3% 0.0044 
%water 40.4% 84.6% 0.60 
SP 39.3% 87.4% 0.036 

NOTE: The table shows the percentage at 4 weeks of the pretreatment value for all of the functional parameters in responders and nonresponders. At 4 weeks, HbO2, Hb, %water, and SP were all lower for responders compared with nonresponders. All differences were statistically significant, except for %water, which approached significance.

A repeated-measures analysis was done using a general linear mixed-model statistical analysis. The mean percent change for Hb in responders was −41.9% (SE = 2.4) versus nonresponders of −12.0% (SE = 3.1). This difference was significant (P < 0.001). The mean percent change for HbO2 in responders was −51.6% (SE = 5.8) versus nonresponders of −10.0% (SE = 7.5). This difference was also significant (P = 0.005). The mean percent change for HbO2 in responders was −43.8% (SE = 6.8) versus nonresponders of −12.0% (SE = 8.7) and significant (P = 0.028). The mean percent change for SP in responders was −41.6% (SE = 7.4) versus nonresponders of −7.0% (SE = 9.5). This difference was in addition significant (P = 0.028).

Standard treatment for LABC usually involves the use of neoadjuvant chemotherapy. Patients with a complete pathologic response have significantly higher disease-free and overall survival rates compared with nonresponders (611). The need for a straightforward, inexpensive, and noninvasive method of early treatment monitoring is essential to ensure that patients receive appropriate treatment. Traditionally, tumor response has been assessed clinically or radiographically through changes in tumor size, which may take several weeks to months to manifest, despite a positive response. The corollary is that patients may continue on treatment that is ineffective because response assessment with clinical or radiographic means is dependent on late-occurring anatomic changes. DOS is gaining new enthusiasm in breast imaging, and studies, primarily case reports (2226), have shown its feasibility for monitoring neoadjuvant treatment in LABC. The largest trial (11 patients), with patients receiving neoadjuvant chemotherapy with Adriamycin and cyclophosphamide, with or without taxane, showed statistically significant changes in Hb (33 ± 7%), HbO2 (27 ± 15%), and %water (11 ± 15%) in pathology-confirmed responders compared with parameters measured before treatment (22). Patients in that study had a scan before treatment (mean of 1.8 days before chemotherapy) and 1 week after the start of the first cycle of chemotherapy. In that study, six patients were considered responders to neoadjuvant therapy and the rest were nonresponders.

In our study of 10 patients with LABC, we were able to use each of the DOS optical parameters to identify an area that corresponded to the tumor detected with MRI, clinical exam, and other imaging. Measurements of all the optical parameters differed between tumor and background breast tissue. We have shown using DOS that as early as 4 weeks after therapy initiation, responders and nonresponders, as finally evaluated by whole-mount pathology, can be differentiated. One patient (patient 1) had significant changes in the DOS parameters despite a weak pathologic response. In that patient, there was a significant diminishment in her invasive disease with almost none left, which was consistent with what the optical parameters indicated. However, there was a significant amount of residual in situ carcinoma left in place, which caused the pathologist to rate the response as weak. We suspect that the optical parameters changed coincident with the changes in the invasive component but were potentially insensitive to the presence of ductal carcinoma in situ, as this would only affect the linings of the breast ducts and possibly not have an effect on HbO2 and Hb.

The patients in this study intentionally received different types of chemotherapy. Results were obtained after normalizing for age, body mass index, premenopausal or postmenopausal status, and fluctuations in the menstrual cycle (20). The results were statistically significant at 4 weeks after the commencement of therapy, and we believe that this makes the method generally applicable, as many types of therapy induced the same response in optical parameters, which corresponded to tumor response. The results were generally consistent and independent of how tumor cell death was induced. The patient numbers used (n = 10) were appropriate given the magnitude of response in the optical measures that were obtained. Pathologic complete response can exceed 40% in hormone receptor–negative patients and is lower in hormone receptor–positive patients (2932). Our patient population exhibited more aggressive disease with no complete response.

Patient chemotherapy in LABC is already to a degree risk adaptive (3335) and customized based on individual clinical circumstances in women with LABC. All 10 patients received a taxane drug. The majority of patients (8 of 10) did receive a combined anthracycline and taxane regimen, presently considered the clinical standard for such women with cancer, or combined radiation and a taxane. We believe that DOS changes in vivo are surrogates of cell death occurring from such exposures and how death is induced, so long as it is induced, is less important to the optical changes than having death induced. We hope that using a “real-life” patient population in such a study better potentially ascertains the use of the imaging technology under investigation.

Other functional imaging modalities have been used to assess treatment response. Positron emission tomography, using radiopharmaceuticals such as fluorodeoxyglucose (FDG PET), has been shown to detect and stage breast cancer with high sensitivity and specificity (3639). In a study by Kim et al. (40), 50 patients with LABC receiving neoadjuvant treatment were assessed for the correlation between peak standardized uptake values of FDG PET with clinical and pathologic outcome. There was a statistically significant difference in the reduction of peak standardized uptake values between pathologic complete responders, partial responders, and nonresponders. However, there was no detected difference between the different groups using clinical assessment. Investigations of FDG PET as a tool for treatment monitoring in multiple cancer sites are ongoing, but the cost to produce the radionuclides can limit its widespread use. Dynamic contrast-enhanced MRI can be used to measure tissue microvasculature and is sensitive to blood volume and vascular permeability. It can monitor the effectiveness of a variety of treatments, including chemotherapy, hormonal manipulation, radiotherapy, and novel therapeutic approaches such as the use of antiangiogenic drugs (41). However, no standard quantification approach has yet emerged. Prospective studies comparing the various quantification approaches are needed (42). Finally, contrast-enhanced ultrasound using microbubbles can measure and assess intravascular volume and, therefore, tumor response. However, approval of microbubble contrast agents for this indication is still pending. Other imaging modalities are being developed but remain mainly experimental at present (43).

DOS is an encouraging functional imaging modality because of its relatively low cost. There are a number of related challenges remaining. Identifying the boundaries of the tumor VOI on DOS tomographic images is not yet standardized and could show considerable variation from one method to the next. In this work, we have defined the VOI on the initial scan and found the minimum value for each functional parameter to be used as a cutoff value to define the tumor VOI for each subsequent scan. This provided for a standardized approach to assess response for each patient. Alternatively, combining the superior spatial resolution of other imaging, such as MRI or ultrasound, with DOS imaging may be useful.

For all the functional parameters tested, the pathologic responder and nonresponder groups could be differentiated at 4 weeks of time; this continued for the 8-week and presurgery scans where the groups remained separated. At 1 week, the responders and nonresponders were not statistically significantly different. With these large tumors, there may be no significant response at 1 week for these patients or greater patient numbers may be required. However, ascertainment of response or lack thereof at 4 weeks into several months of therapy could spare women months of ineffective therapy and permit more effective salvage regimens to be undertaken in a timely manner after one to two cycles of chemotherapy. Nevertheless, results, with the majority of patients receiving a present-day clinical standard of combined anthracycline and taxane regimen, indicated that DOS was able to measure statistically significant changes in women who had pathologic responses compared with those who did not.

No potential conflicts of interest were disclosed.

We thank ART Canada for providing optical coupling media and technical support and also for providing an in-kind contribution toward the purchase of SoftScan device and unrestricted research funds to cover partial study costs, the Odette family who provided a donation to Sunnybrook Health Sciences Centre that financed the equipment used in this study, the Department of Radiation Oncology and Dr. Shun Wong and Sheila Robson at Sunnybrook Health Sciences Centre for helpful support, and Bradbury D. Bear for stimulating scientific discussion and mentoring.

Grant Support: G.J. Czarnota is supported by a Cancer Care Ontario Research Chair in Experimental Therapeutics and Imaging. Support was obtained from the Natural Sciences and Engineering Council of Canada to G.J. Czarnota and the Ontario Institute of Cancer Research to M.J. Yaffe for this research with support from Sunnybrook Health Sciences Centre.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
American Cancer Society. Breast cancer facts and figures 2007-2008. Atlanta: American Cancer Society, Inc.
2
Canadian Cancer Society/National Cancer Institute of Canada. Canadian cancer statistics 2007. Toronto: Canadian Cancer Society/National Cancer Institute of Canada; 2007.
3
Mankoff
DA
,
Dunnwald
LK
,
Gralow
JR
,
Ellis
GK
,
Drucker
MJ
,
Livingston
RB
. 
Monitoring the response of patients with locally advanced breast carcinoma to neoadjuvant chemotherapy using [technetium 99m]-sestamibi scintimammography
.
Cancer
1999
;
85
:
2410
23
.
4
Singletary
SE
,
Allred
C
,
Ashley
P
, et al
. 
Revision of the American Joint Committee on Cancer staging system for breast cancer
.
J Clin Oncol
2002
;
20
:
3628
36
.
5
Therasse
P
,
Mauriac
L
,
Welnicka-Jaskiewicz
M
, et al
. 
Final results of a randomized phase III trial comparing cyclophosphamide, epirubicin, and fluorouracil with a dose-intensified epirubicin and cyclophosphamide + filgrastim as neoadjuvant treatment in locally advanced breast cancer: an EORTC-NCIC-SAKK multicenter study
.
J Clin Oncol
2003
;
21
:
843
50
.
6
Kuerer
H
,
Newman
L
,
Smith
T
, et al
. 
Clinical course of breast cancer patients with complete pathologic primary tumor and axillary lymph node response to doxorubicin-based neoadjuvant chemotherapy
.
J Clin Oncol
1999
;
17
:
460
9
.
7
Chatuverdi
S
,
McLaren
C
,
Schofield
AC
, et al
. 
Patterns of local and distant disease relapse in patients with breast cancer treated with primary chemotherapy: do patients with a complete pathological response differ from those with residual tumor in the breast?
Breast Cancer Res Treat
2005
;
93
:
151
8
.
8
Guarneri
V
,
Broglio
K
,
Kau
SW
, et al
. 
Prognostic value of pathologic complete response after primary chemotherapy in relation to hormone receptor status and other factors
.
J Clin Oncol
2006
;
24
:
1037
44
.
9
Sataloff
DM
,
Mason
BA
,
Prestipino
AJ
,
Seinige
UL
,
Lieber
CP
,
Baloch
Z
. 
Pathologic response to induction chemotherapy in locally advanced carcinoma of the breast: a determinant of outcome
.
J Am Coll Surg
1995
;
180
:
297
306
.
10
Feldman
LD
,
Hortobagyi
GN
,
Buzdar
AU
,
Ames
FC
,
Blumenschein
GR
. 
Pathological assessment of response to induction chemotherapy in breast cancer
.
Cancer Res
1986
;
46
:
2578
81
.
11
Machiavelli
MR
,
Romero
AO
,
Perez
JE
, et al
. 
Prognostic significance of pathological response of primary tumor and metastatic axillary lymph nodes after neoadjuvant chemotherapy for locally advanced breast carcinoma
.
Cancer J Sci Am
1998
;
4
:
125
31
.
12
Carlson
RW
,
Favret
AM
. 
Multidisciplinary management of locally advanced breast cancer
.
Breast J
1999
;
5
:
303
7
.
13
Nelson
TR
,
Cervino
LI
,
Boone
JM
,
Lindfors
KK
. 
Classification of breast computed tomography data
.
Med Phys
2008
;
35
:
1078
86
.
14
Segel
MC
,
Paulus
DD
,
Hortobagyi
GN
. 
Advanced primary breast cancer: assessment at mammography of response to induction chemotherapy
.
Radiology
1988
;
169
:
49
54
.
15
Moskovic
EC
,
Mansi
JL
,
King
DM
,
Murch
CR
,
Smith
IE
. 
Mammography in the assessment of response to medical treatment of large primary breast cancer
.
Clin Radiol
1993
;
47
:
339
44
.
16
Vinnicombe
SJ
,
MacVicar
AD
,
Guy
RL
, et al
. 
Primary breast cancer: mammographic changes after neoadjuvant chemotherapy, with pathologic correlation
.
Radiology
1996
;
198
:
333
40
.
17
Yeh
E
,
Slanetz
P
,
Kopans
DB
, et al
. 
Prospective comparison of mammography, sonography, and MRI in patients undergoing neoadjuvant chemotherapy for palpable breast cancer
.
Am J Roentgenol
2005
;
184
:
868
77
.
18
Cutler
M
. 
Transillumination as aid in the diagnosis of breast lesions
.
Surg Gynecol Obstet
1929
;
48
:
721
9
.
19
Ntziachristos
V
,
Chance
B
. 
Probing physiology and molecular function using optical imaging: applications to breast cancer
.
Breast Cancer Res
2001
;
3
:
41
6
.
20
Leff
DR
,
Warren
OJ
,
Enfield
LC
, et al
. 
Diffuse optical imaging of the healthy and diseased breast: a systemic review
.
Breast Cancer Res Treat
2008
;
108
:
9
22
.
21
Tromberg
BJ
,
Shah
N
,
Lanning
R
, et al
. 
Non-invasive in vivo characterization of breast tumors using photon migration spectroscopy
.
Neoplasia
2000
;
2
:
26
40
.
22
Cerussi
A
,
Hsiang
D
,
Shah
N
, et al
. 
Predicting response to breast cancer neoadjuvant chemotherapy using diffuse optical spectroscopy
.
Proc Natl Acad Sci U S A
2007
;
104
:
4014
9
.
23
Zhu
Q
,
Kurtzma
SH
,
Hegde
P
, et al
. 
Utilizing optical tomography with ultrasound localization to image heterogeneous hemoglobin distribution in large breast cancers
.
Neoplasia
2005
;
7
:
263
70
.
24
Jakubowski
DB
,
Cerussi
AE
,
Bevilacqua
F
, et al
. 
Monitoring neoadjuvant chemotherapy in breast cancer using quantitative diffuse optical spectroscopy: a case study
.
J Biomed Opt
2004
;
9
:
230
8
.
25
Choe
R
,
Corlu
A
,
Lee
K
, et al
. 
Diffuse optical tomography of breast cancer during neoadjuvant chemotherapy: a case study with comparison to MRI
.
Med Phys
2005
;
32
:
1128
39
.
26
Shah
N
,
Gibbs
J
,
Wolverton
D
,
Cerussi
A
,
Hylton
N
,
Tromberg
BJ
. 
Combined diffuse optical spectroscopy and contrast-enhanced magnetic resonance imaging for monitoring breast cancer neoadjuvant chemotherapy: a case study
.
J Biomed Opt
2005
;
10
:
051503
.
27
Clarke
GM
,
Eidt
S
,
Sun
L
,
Mawdsley
GE
,
Zubovits
JT
,
Yaffe
MJ
. 
Whole-specimen histopathology: a method to produce whole-mount breast serial section for 3D digital histopathology imaging
.
Histopathology
2007
;
50
:
232
42
.
28
Intes
X
,
Djeziri
S
,
Ichalalene
Z
, et al
. 
Time-domain optical mammography SoftScan: initial results on detection and characterization of breast tumors
.
Proc SPIE
2004
;
5578
:
188
97
.
29
Kaufmann
M
,
von Minckwitz
G
,
Bear
HD
, et al
. 
Recommendations from an international expert panel on the use of neoadjuvant (primary) systemic treatment of operable breast cancer: new perspectives 2006
.
Ann Oncol
2007
;
18
:
1927
34
.
30
Ring
AE
,
Smith
IE
,
Ashley
S
,
Fullford
LG
,
Lakhani
SR
. 
Oestrogen receptor status, pathological complete response and prognosis in patients receiving neoadjuvant chemotherapy for early breast cancer
.
Br J Cancer
2004
;
91
:
2012
7
.
31
Colleoni
M
,
Viale
G
,
Zarieh
D
, et al
. 
Chemotherapy is more effective in patients with breast cancer not expressing steroid hormone receptors: a study of preoperative treatment
.
Clin Cancer Res
2004
;
10
:
6622
8
.
32
Hennessy
BT
,
Hortobagyi
GN
,
Rouzier
R
, et al
. 
Outcome after pathologic complete eradication of cytologically proven breast cancer axillary node metastases following primary chemotherapy
.
J Clin Oncol
2005
;
23
:
9304
11
.
33
Wenzel
C
,
Bartsch
R
,
Hussian
D
, et al
. 
Invasive ductal carcinoma (IDC) and invasive lobular carcinoma (ILC) of breast differ in response following neoadjuvant therapy with epidoxorubicin and docetaxel + G-CSF
.
Breast Cancer Res Treat
2007
;
104
:
109
14
.
34
Tubiana-Hulin
M
,
Stevens
D
,
Lasry
S
, et al
. 
Response to neoadjuvant chemotherapy in lobular and ductal breast carcinomas: a retrospective study on 860 patients from one institution
.
Ann Oncol
2006
;
17
:
1228
33
.
35
Cristofanilli
M
,
Gonzalez-Angulo
A
,
Sneige
N
, et al
. 
Invasive lobular carcinoma classic type: response to primary chemotherapy and survival outcomes
.
J Clin Oncol
2005
;
23
:
41
8
.
36
Adler
LP
,
Crowe
JP
,
Al-Kaisi
NK
,
Sunshine
JL
. 
Evaluation of breast masses and axillary lymph nodes with [F-18] 2-deoxy-2-fluoro-D-glucose PET
.
Radiology
1993
;
187
:
743
50
.
37
Nieweg
OE
,
Kim
EE
,
Wong
WH
, et al
. 
Positron emission tomography with fluorine-18-deoxyglucose in the detection and staging of breast cancer
.
Cancer
1993
;
71
:
3920
5
.
38
Wahl
RL
,
Zasadny
K
,
Helvie
M
,
Hutchins
GD
,
Weber
B
,
Cody
R
. 
Metabolic monitoring of breast cancer chemohormonotherapy using positron emission tomography: initial evaluation
.
J Clin Oncol
1993
;
11
:
2101
11
.
39
Jansson
T
,
Westlin
JE
,
Ahlstrom
H
, et al
. 
Positron emission tomography studies in patients with locally advanced and/or metastatic breast cancer: a method for early therapy evaluation?
J Clin Oncol
1995
;
13
:
1470
7
.
40
Kim
SJ
,
Kim
SK
,
Lee
ES
,
Ro
J
,
Kang
SH
. 
Predictive value of [18F]FDG PET for pathological response of breast cancer to neo-adjuvant chemotherapy
.
Ann Oncol
2004
;
15
:
1352
7
.
41
Padhani
AR
,
Husband
JE
. 
Dynamic contrast-enhanced MRI studies in oncology with an emphasis on quantification, validation and human studies
.
Clin Radiol
2001
;
56
:
607
20
.
42
Hylton
N
. 
Dynamic contrast-enhanced magnetic resonance imaging as an imaging biomarker
.
J Clin Oncol
2006
;
24
:
3293
8
.
43
Brindle
K
. 
New approaches for imaging tumor responses to treatment
.
Nat Rev Cancer
2008
;
8
:
94
107
.

Supplementary data