Annals of Oncology Advance Access originally published online on June 13, 2008
Annals of Oncology 2008 19(9):1573-1577; doi:10.1093/annonc/mdn185
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
breast cancer |
Preoperative [18F] FDG–PET after chemotherapy in locally advanced breast cancer: prognostic value as compared with histopathology
1 Department of Nuclear Medicine and PET Research, VU University Medical Center, Amsterdam
2 Department of Medical Oncology, Amstelland Hospital, Amstelveen
3 Department of Epidemiology and Biostatistics, Amsterdam
4 Department of Surgical Oncology, Amsterdam
5 Departmentof Medical Oncology, VU University Medical Center, Amsterdam, The Netherlands
* Correspondence to: Dr O. S. Hoekstra, Department of Nuclear Medicine and PET Research, VU University Medical Center, PO Box 7057, Amsterdam 1007 MB, The Netherlands. Tel: +31-20-4442886; Fax: +31-20-4444329; E-mail: os.hoekstra{at}vumc.nl
 |
Abstract |
|---|
 Top Abstract introductionmaterials and methodsresultsdiscussionReferences |
|---|
Background: Established prognosis-based criteria determine the need for further treatment after primary surgery for breast cancer. Such criteria are lacking after neo-adjuvant chemotherapy. We determine the prognostic value of preoperative [18F]-2-fluoro-2-deoxy-D-glucose–positron emission tomography (18FDG–PET) after chemotherapy in locally advanced breast cancer (LABC), both as independent indicator and as add-on to postoperative histopathology.
Patients and methods: Preoperative PET was carried out in 40 LABC patients. Two expert readers assessed residual 18FDG uptake in the primary tumor. At histopathological examination of the surgical specimen, chemotherapy response was graded using the Honkoop criteria. Cox proportional hazards analysis was used to determine prognostic relevance of PET and histopathology.
Results: Median follow-up was 60 months (range 15–94), during which 13 patients had recurrent disease, eight of whom died. 18FDG uptake in the primary tumor was inversely related with disease-free survival (DFS) [hazard ratio (HR) 4.09; 95% confidence interval (CI) 1.26–13.31; P = 0.02] and this was superior to histopathology (HR 2.52; 95% CI 0.77–8.23; P = 0.13). Observer agreement of PET was excellent (intraclass correlation coefficient 0.88). Multivariate Cox regression revealed no added value of histopathology versus PET results.
Conclusion: 18FDG uptake in the primary tumor at PET was inversely associated with DFS and may help to guide adjuvant therapy.
Key words: breast cancer, neo-adjuvant chemotherapy, PET, prognostic value
|
introduction |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionReferences |
|---|
Locally advanced breast cancer (LABC) is defined as a primary tumor >5 cm, inflammatory breast cancer, skin or chest wall involvement or fixed axillary lymph node metastasis. It is usually treated with combined modality treatment consisting of chemotherapy, surgery and radiotherapy [1, 2], followed by hormonal therapy and trastuzumab where appropriate.
Even though neo-adjuvant chemotherapy (NC) appears to yield similar survival rates as adjuvant chemotherapy [3], the neo-adjuvant approach carries the advantages of easier clinical response measurement, postoperative pathological response measurement and increased breast conservation rates [4, 5]. However, there is little data as to the prognostic value of the extent of residual disease as found at postoperative pathological examination after NC [3, 6].
Honkoop et al. [6, 7] defined pathological response as gross residual disease (GRD) when extensive microscopic infiltration and/or macroscopic tumor is present, in contrast to minimal residual disease (MRD) when no or only small scattered foci of tumor are found at microscopy. They found a 2-year disease-free survival (DFS) of 80% for MRD patients versus 45% for GRD patients (P = 0.03) and a 2-year overall survival (OS) of 94% versus 68% (P = 0.05). The same study confirmed that clinical response rates are unreliable in predicting relapse or death.
[18F]fluorodeoxyglucose (18FDG) is the standard oncological tracer in positron emission tomography (PET), capitalizing on the property of cancer cells to show increased glucose metabolism [8]. 18FDG–PET is also the most widely used imaging technique to measure therapy effect on a metabolic rather than anatomical level [9]. Recent data suggest that many cytotoxic and cytostatic agents affect, directly or indirectly, the pathways, glucose transporters and metabolic enzymes controlling glycolysis [10]. In untreated breast cancer, 18FDG avidity has been shown to be positively correlated to higher microvessel density, Glut-1 expression and mitotic activity index of tumor tissue [11]. In locally advanced non-small–cell lung cancer, preoperative, postchemotherapy 18FDG–PET carries strong prognostic information to the extent that persisting 18FDG uptake is associated with poorer outcome [12]. The aim of the present prospective study was to investigate whether residual 18FDG uptake after neo-adjuvant therapy in LABC yields useful prognostic information complementing or surpassing standard histopathological classification of the surgical specimen.
|
materials and methods |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionReferences |
|---|
patients
From August 1997 to July 2003, we prospectively included 53 patients at the VU University Medical Center and the Amstelland community hospital who were treated with NC for LABC. The study protocol called for patients to undergo two 18FDG–PET scans: at presentation and after systemic therapy, i.e. before surgery. All patients gave their informed consent and the study protocol was approved by the Human Investigations Committee of the VU university medical center. At inclusion, 18FDG–PET identified distant metastases in four patients, and they were switched to other therapy [13]. Eight patients did not receive a preoperative PET scan (logistical problems, patient refusal, technical complications), and one patient refused surgery after chemotherapy. In total, there were 40 assessable patients with preoperative 18FDG–PET scans, surgery and clinical follow-up; all but one also underwent 18FDG–PET before chemotherapy (see Figure 1).
|
clinical procedures
At the time of inclusion, the standard treatment for LABC at our institutions consisted of six cycles of high-dose doxorubicin/cyclophosphamide (AC) with granulocyte colony-stimulating factor (G-CSF) or granulocyte–macrophage stimulating factor (GM-CSF), followed by mastectomy with axillary lymph node dissection and postoperative radiotherapy (n = 31). During part of the inclusion period, a clinical trial was held in which the standard regimen was compared with a split course with three neo-adjuvant cycles of high-dose AC with G-CSF or GM-CSF, surgery, three adjuvant cycles and radiotherapy. Eight patients were treated according to this second schedule. Finally, one patient was treated with six cycles of 5-fluorouracil, epirubicin and cyclophosphamide, followed by mastectomy with axillary lymph node dissection and radiotherapy. Patients with hormone receptor-positive disease were subsequently treated with tamoxifen for 5 years.
Thirty-nine patients underwent a modified radical mastectomy with axillary lymph node dissection according to Madden et al. [14]; one patient refused mastectomy and was treated with breast-conserving surgery and axillary lymph node dissection. She had a pathological complete remission (pCR). Methods of the histopathological examination have been described in detail elsewhere [6]. In summary, if tumor was visible at macroscopic evaluation of the mastectomy specimen or axillary lymph nodes, the pathological response was graded as macroscopic or GRD. If tumor was only present at microscopic examination or if there had been a complete pathological response, it was graded as MRD.
Follow-up consisted of check-ups once every 2 months in the first year after treatment, every 3 months in years 2–5 and every 6 months after year 5. Standard hematological and biochemical laboratory testing (including CA 15.3) was carried out every 6 months and a mammography of the contralateral breast every 12 months. Computed tomography scan or ultrasound of the liver, bone scan or chest X-ray were carried out on indication only.
PET procedures
18FDG–PET scans were carried out using a dedicated PET scanner (ECAT EXACT HR+; CTI/Siemens, Knoxville, TN). Patients fasted for at least 6 h before scanning. PET acquisition started at a mean of 75 min (range 60–90) after i.v. injection of 370 MBq 18FDG. The acquisition protocol consisted of two 10-min emission scans of the chest as well as two 10-min transmission scans for attenuation correction (in 4 of 40 preoperative scans, the transmission scan was not carried out for technical and/or logistical reasons). All scans were corrected for decay, scatter and randoms and reconstructed using ordered subset expectation maximization, with two iterations and 16 subsets followed by postsmoothing of the reconstructed image using a 5-mm full width at half maximum Gaussian filter. Preoperative scans were carried out at a mean of 16 days before surgery (range 1–41).
The PET images were read independently by two experienced PET readers using standard CTI software on a computer monitor in axial, sagittal and coronal planes. No standardized uptake values (SUVs) were calculated as this was not part of the original protocol and therefore the data necessary for obtaining reliable SUVs (height and weight, glucose level) were incomplete. The relevant PET images were randomly distributed among PET scans which had been obtained in LABC patients at baseline and after one, three or six cycles of chemotherapy. The observers were blinded for any clinical data including the site of the primary tumor. 18FDG uptake in the primary tumor was considered negative for residual disease in case of absent or faint accumulation (i.e. versus background and the contralateral breast) and positive for residual disease in case of at least moderately enhanced uptake. Discrepancies were solved by consensus. During the study, clinicians and pathologists were blinded to the PET data.
statistical analysis
Statistical analyses were carried out with SPSS statistical software (SPSS Inc, Chicago, IL). Observer agreement at PET was assessed with intraclass correlation coefficients (ICCs). To assess the relationship between DFS (time between onset of chemotherapy and recurrent disease) on the one hand, and preoperative 18FDG–PET, and postoperative pathology as defined by Honkoop on the other, the following approach was applied: first, Kaplan–Meier curves for DFS were plotted with PET and histopathology separately and tested using the log-rank test (with a two-sided P value of <0.05 considered significant). Then a Cox proportional hazard regression analysis was carried out to control for possible confounders (age, extent of disease at presentation, 18FDG uptake at presentation, histological tumor type, number of preoperative chemotherapy cycles). Finally, a Cox regression analysis was carried out with both histopathology and PET as independent variables in one model. Results are presented as hazard ratios (HRs) with 95% confidence intervals (CIs). The same procedure was followed with OS (time between onset of chemotherapy and death of any cause).
|
results |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionReferences |
|---|
The pretreatment characteristics of the 40 patients are given in Table 1. Histopathological examination of the surgical specimens revealed that 19 patients had MRD, of which eight (20%) had a complete pathological remission and 11 (27.5%) had microscopic residual disease, in the breast (two), the axilla (five) or both (four). GRD was found in 21 patients (52.5%); in two of them the axilla was clear of disease. All the surgical procedures were radical, in that microscopic examination revealed no tumor in the margins of the surgical specimen.
|
Median follow-up was 60 months for the entire group and 64 months for the patients without evidence of disease (range 29–94 months). Median time to progression was 26 months (range 14–90 months). At the end of follow-up, 13 patients had developed recurrent disease (all had distant metastases, two combined with a locoregional recurrence) and eight patients had died. Twenty-seven were alive without evidence of disease.
Thirty-nine patients underwent a baseline PET scan before chemotherapy. Untreated primary tumors had been 18FDG avid in all but one patient (with ductal carcinoma). Pathological examination of the surgical specimen after chemotherapy revealed MRD, and the patient had no evidence of disease at the conclusion of follow-up. At the onset of chemotherapy, 10 patients had a clinically negative axilla (one of which had postoperative pN0) and seven had a negative axilla at baseline PET (two of which had postoperative pN0).
Observer agreement of the preoperative PET classification was very high at an ICC of 0.88 (95% CI 0.78–0.94). Preoperative PET was positive in 16 cases (nine of whom developed metastatic disease during follow-up) and negative in 24 (with four subsequent relapses). Of the 21 patients with GRD, the disease recurred in nine during follow-up as it did in 4 of 19 patients with MRD. In 13 patients pathology and PET were discordant, in that one measure would show a favorable outcome while the other was poor. Four of these 13 patients relapsed during the course of follow-up (Table 2).
|
prognostic indicators
The only prognostic indicator found to be of significance was histopathology and preoperative PET. Kaplan–Meier plots for DFS and OS are shown in Figure 2. Pre- and postchemotherapy variables were tested for their ability to predict DFS and OS using Cox proportional hazard regression analysis. This revealed that none of the possible confounders (18FDG uptake at baseline, age, extent of disease at presentation, histological tumor type, number of preoperative chemotherapy cycles) influenced the results of the Cox proportional hazards analysis. Honkoop was chosen as the histopathological classification because in this dataset it was clearly superior to pCR or axillary nodal involvement as a means to predict OS and DFS (data not shown).
|
Preoperative PET (HR 4.09; 95% CI 1.26–13.31; P = 0.02; Table 3) was a better indicator for DFS than histopathological examination according to the Honkoop criteria (HR 2.52; 95% CI 0.77–8.23; P = 0.13). In predicting OS, both PET (HR 2.77; 95% CI 0.66–11.66; P = 0.16) and histopathology (HR 6.53; 95% CI 0.80–53.14; P = 0.08) were nonsignificant predictors, although the log-rank calculation for the Kaplan–Meier curve for pathology was statistically significant (P = 0.04; Figure 1D).
|
The results of the multivariate analysis are shown in Table 4. When adjusted for pathology, the ability of preoperative PET to predict DFS became slightly lower but remained significant (HR 3.51; 95% CI 1.04–11.82; P = 0.04). Pathology (adjusted for PET) could not predict DFS (HR 1.86; 95% CI 0.53–6.48; P = 0.33). Neither variable was an accurate factor for predicting OS, although pathology (adjusted for PET) outperformed PET (adjusted for pathology) with a reasonably high HR of 5.40 (95% CI 0.63–45.91; P = 0.12).
|
|
discussion |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionReferences |
|---|
In this prospective observational study, it was shown that preoperative 18FDG–PET better predicted DFS than postoperative pathological examination in women who have received NC for breast cancer. We found no significant complementary effect in combining the two predictors.
In measuring the success of a therapy, both OS and DFS are possible yardsticks. However, OS is also a function of other factors: time from relapse until death is dependent on the location of metastases as well as the success of secondary treatment. In this study, DFS was also superior because the greater number of adverse events (13 relapses versus eight deaths) allowed for stronger outcomes in statistical analysis, although this problem can be tackled with a longer follow-up period. One of the limitations of a study with few events is that multivariate analysis has to be interpreted with appropriate caution. In light of the relatively narrow CIs, we feel that in the present study multivariate analysis is a valid approach.
Three other studies have looked at preoperative 18FDG–PET scans after NC in breast cancer patients. However, none evaluated the potential association of 18FDG uptake with survival estimates: two [15, 16] investigated the ability of preoperative PET to predict the postoperative histopathological outcome, while the third [17] only investigated whether a preoperative PET scan could predict pCR. To the best of our knowledge, we present the first study evaluating the ability of preoperative 18FDG–PET to predict OS and DFS.
With our reconstruction algorithm, the resolution of PET is 
7 mm. Therefore, we expected that partial volume effects would render MRD negative at PET. However, four patients had a positive PET scan despite MRD on histopathological investigation. Although none of them had a complete remission at histopathology, we can only speculate about the biological substrate for the enhanced 18FDG avidity: uptake in diffuse microscopic tumor remnants, macrophage activity, regeneration of breast tissue, the presence of carcinoma in situ or a combination of factors. At the end of follow-up, in two patients the disease had relapsed (although all four were alive).
Alternatively, GRD might be either positive or negative at PET, depending on the metabolic activity of the remaining tumor. If this difference in metabolic activity reflects itself in differences in OS and/or DFS, PET might be able to discern two separate groups within patients with GRD. However, the number of GRD patients in the present sample (n = 21) was too low to perform a stratified analysis. Of the patients with GRD, nine had a negative preoperative scan versus 12 with positive scans. Cox regression analysis on these subgroups found PET unable to significantly predict DFS (HR 3.21; 95% CI 0.66–15.49; P = 0.15) or OS (HR 2.51; 95% CI 0.46–13.56; P = 0.29). If these findings are externally validated, a postchemotherapy PET scan might help to guide therapy to the extent that patients with a positive PET scan might be switched to second-line systemic therapy (rather than immediate surgery) with PET being used to monitor its effect in vivo.
In conclusion, enhanced 18FDG uptake in the primary tumor of LABC patients after NC was associated inversely with DFS. The data suggest that to predict DFS, a preoperative 18FDG–PET is at least as good as postoperative histopathology and possibly better.
Received for publication March 14, 2007. Revision received March 30, 2008. Accepted for publication March 31, 2008.
|
References |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionReferences |
|---|
1. Shenkier T, Weir L, Levine M, et al. Clinical practice guidelines for the care and treatment of breast cancer: 15. Treatment for women with stage III or locally advanced breast cancer. Can Med Assoc J (2004) 170:983–994.
2. Kaufmann M, Hortobagyi GN, Goldhirsch A, et al. Recommendations from an international expert panel on the use of neoadjuvant (primary) systemic treatment of operable breast cancer: an update. J Clin Oncol (2006) 24:1940–1949.
3. Fisher B, Brown A, Mamounas E, et al. Effect of preoperative chemotherapy on local-regional disease in women with operable breast cancer: findings from National Surgical Adjuvant Breast and Bowel Project B-18. J Clin Oncol (1997) 15:2483–2493.
4. Bauerfeind I, Himsl I, Lohrs B, et al. Surgical procedure following primary systemic chemotherapy for breast cancer. Gynakologe (2005) 38:193–200.[CrossRef]
5. Shen J, Valero V, Buchholz TA, et al. Effective local control and long-term survival in patients with T4 locally advanced breast cancer treated with breast conservation therapy. Ann Surg Oncol (2004) 11:854–860.[CrossRef][Web of Science][Medline]
6. Honkoop AH, Pinedo HM, De Jong JS, et al. Effects of chemotherapy on pathologic and biologic characteristics of locally advanced breast cancer. Am J Clin Pathol (1997) 107:211–218.[Web of Science][Medline]
7. Honkoop AH, van Diest PJ, de Jong JS, et al. Prognostic role of clinical, pathological and biological characteristics in patients with locally advanced breast cancer. Br J Cancer (1998) 77:621–626.[Web of Science][Medline]
8. Warburg O, Negelein E, Posener K. Über den Stoffwechsel der Tumoren. Biochemische Zeitschrift (1924) 319–344.
9. Juweid ME, Cheson BD. Positron-emission tomography and assessment of cancer therapy. N Engl J Med (2006) 354:496–507.
10. Kelloff GJ, Hoffman JM, Johnson B, et al. Progress and promise of FDG-PET imaging for cancer patient management and oncologic drug development. Clin Cancer Res (2005) 11(8):2785–2808.
11. Bos R, van Der Hoeven JJ, van Der Wall E, et al. Biologic correlates of (18)fluorodeoxyglucose uptake in human breast cancer measured by positron emission tomography. J Clin Oncol (2002) 20:379–387.
12. Hoekstra CJ, Stroobants SG, Smit EF, et al. Prognostic relevance of response evaluation using [18F]-2-fluoro-2-deoxy-D-glucose positron emission tomography in patients with locally advanced non-small-cell lung cancer. J Clin Oncol (2005) 23(33):8362–8370.
13. van der Hoeven JJ, Krak NC, Hoekstra OS, et al. 18F-2-fluoro-2-deoxy-d-glucose positron emission tomography in staging of locally advanced breast cancer. J Clin Oncol (2004) 22:1253–1259.
14. Madden JL, Kandalaft S, Bourque RA. Modified radical mastectomy. Annals of Surgery (1972) 175(5):624–634.[Web of Science][Medline]
15. Bassa P, Kim EE, Inoue T, et al. Evaluation of preoperative chemotherapy using PET with fluorine-18-fluorodeoxyglucose in breast cancer. J Nucl Med (1996) 37:931–938.
16. Kim SJ, Kim SK, Lee ES, et al. Predictive value of [18F]FDG PET for pathological response of breast cancer to neo-adjuvant chemotherapy. Ann Oncol (2004) 15:1352–1357.
17. Burcombe RJ, Makris A, Pittam M, et al. Evaluation of good clinical response to neoadjuvant chemotherapy in primary breast cancer using [18F]-fluorodeoxyglucose positron emission tomography. Eur J Cancer (2002) 38:375–379.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. H. Lee, E. L. Rosen, and D. A. Mankoff The Role of Radiotracer Imaging in the Diagnosis and Management of Patients with Breast Cancer: Part 2--Response to Therapy, Other Indications, and Future Directions J. Nucl. Med., May 1, 2009; 50(5): 738 - 748. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||