Impact of Preoperative Versus Postoperative Chemotherapy on the Extent and Number of Surgical Procedures in Patients Treated in Randomized Clinical Trials for Breast Cancer

Abstract

Objective:

To determine the effect of preoperative chemotherapy on the volume of tissue excised and the number of breast operations in patients undergoing breast-conserving therapy (BCT).

Summary Background Data:

Preoperative chemotherapy is increasingly being used for breast cancer and increases rates of BCT. Its impact on the extent of surgery and the number of surgical procedures in BCT has never been fully defined. The extent of surgery in BCT directly affects cosmesis.

Methods:

We reviewed the records of 509 consecutive patients with T1–T3, N0–N2 breast cancer who were treated in prospective randomized clinical trials of chemotherapy between 1998 and 2005. We analyzed the final surgical procedure (BCT or mastectomy), the number of operations, and, in patients who underwent BCT, re-excision rates, and the total volume of breast tissue excised [4Π/3(width/2 × length/2 × height/2)].

Results:

A total of 241 patients underwent BCT, and 268 patients underwent mastectomy. Among BCT patients who had initial tumor size >2.0 cm, patients who received preoperative chemotherapy had significantly smaller volumes of breast tissue excised compared with patients who received postoperative chemotherapy (113 cm³ vs. 213 cm³, P = 0.004). The re-excision rate and total number of breast operations did not significantly differ between the groups. Among BCT patients who had initial tumor size ≤2 cm, preoperative chemotherapy had no impact on volume of breast tissue excised, re-excision rate, or number of breast operations (P > 0.05).

Conclusions:

Among patients treated with BCT for larger breast tumors, patients treated with preoperative chemotherapy have less extensive resection, with no change in rates of re-excision.

Volume of breast tissue resected to achieve negative margins and number of surgical procedures undertaken were compared between patients who underwent preoperative chemotherapy and those who underwent postoperative chemotherapy for breast cancer. Volume of breast tissue resected was significantly less in patients with T2 and T3 tumors who received preoperative chemotherapy than in those who received postoperative chemotherapy.

The use of preoperative chemotherapy for breast cancer is becoming a widely accepted treatment option. Survival of breast cancer patients treated with preoperative chemotherapy has been shown to be equivalent to that of breast cancer patients treated with postoperative chemotherapy.^1,2 However, preoperative therapy has advantages: it permits in vivo assessment of tumor response to chemotherapy and increases rates of breast-conserving therapy (BCT) without significantly increasing the risk of local recurrence.²

Although preoperative chemotherapy has been shown to increase the proportion of patients who are able to undergo BCT,^1,3–5 several critical questions regarding this approach have not been resolved.

First, it is not known whether preoperative chemotherapy versus the more common approach of postoperative chemotherapy has an effect on the volume of breast tissue resected. Until now, it has been assumed that if preoperative chemotherapy increases the proportion of patients able to undergo segmental resection rather than mastectomy, patients who receive preoperative chemotherapy will have less breast tissue excised compared with patients with the same tumor size at presentation who receive postoperative chemotherapy. However, in theory, the actual volume of tissue resected could be no different or even larger because surgeons remove a large amount of tissue in an attempt to obtain clear margins. To our knowledge, how preoperative versus postoperative chemotherapy affects the volume of tissue excised has not been formally studied. The volume of breast tissue resected in patients undergoing BCT is known to have a direct impact on cosmesis.⁶

Second, it has not been determined whether preoperative chemotherapy increases the number of operative procedures required to achieve negative margins. Since chemotherapy shrinks tumors in different ways and rarely concentrically, does the use of preoperative chemotherapy make obtaining clear margins more difficult and therefore result in more surgical procedures required for margin control?

We performed this study to determine the effect of preoperative versus postoperative chemotherapy on the volume of tissue excised and the number of breast operations in patients undergoing surgery for breast cancer. This type of information is critically important to patients when they are considering their surgical options and the risks and potential benefits of various approaches to treatment.

METHODS

We reviewed the records of 509 consecutive patients with T1–T3, N0–N2 (stage I–IIIA) breast carcinoma who were treated between September 1998 and September 2005 in 2 institutional review board-approved prospective randomized clinical chemotherapy trials of taxanes plus either 5-fluorouracil, doxorubicin, and cyclophosphamide or 5-fluorouracil, epirubicin, and cyclophosphamide. Patients underwent surgery at our institution either before any chemotherapy had been delivered or after all chemotherapy had been completed. Each patient was entered prospectively into the protocol database after providing written informed consent to participate in the Institutional Review Board approved protocols and was then followed up longitudinally.

Demographic, clinicopathologic, and treatment variables were collected prospectively on each patient. Patients’ medical records were reviewed to determine the size of the tumor on physical examination, sonography, and mammography, the surgical procedures undertaken, the total number of operations, the dimensions of the surgical specimens as documented in the pathology report, and any local recurrences.

Disease was staged in accordance with the 2003 American Joint Committee on Cancer guidelines.⁷ Disease status was assessed at presentation using physical examination, sonography of the breast and lymph node basins, and mammography. Tumor size was defined as the largest dimension recorded on physical examination, sonography, or mammography.

Only patients for whom surgical specimen dimensions (width, length, and height) were known were included in the analysis of tissue volume. Tissue volume was determined using the formula for an ellipsoid volume: 4Π/3 × width axis radius × length axis radius × height axis radius.⁸ Tumor volume analysis demonstrates that use of the formula for rectangular solids (width × length × height) consistently overestimates the true volume of breast lesions. In contrast, the ellipsoid formula more accurately calculates the volume of breast specimens, which are commonly ellipsoid or spherical.⁸ Re-excision specimen volumes were also calculated using the same formula to determine the final surgical volume resected. When additional sections of breast tissue were excised at the initial surgery to obtain negative margins of resection, the volume of these specimens was similarly calculated. The total volume of tissue excised was the sum of the volumes of all excised tissue.

Chemotherapy was delivered according to prospective institutional protocols. In the initial protocol, patients received either paclitaxel weekly or paclitaxel every 3 weeks, followed by 5-fluorouracil, cyclophosphamide, and doxorubicin as postoperative or preoperative therapy. Full details concerning this protocol have been documented in previous reports.⁹ In the second protocol, patients received either weekly paclitaxel or docetaxel and capecitabine every 3 weeks followed by 5-fluorouracil, epirubicin, and cyclophosphamide as postoperative or preoperative chemotherapy. Results from this study have yet to be reported.

All preoperative chemotherapy was completed 3 to 4 weeks prior to surgical resection. BCT was offered to patients who were deemed candidates for BCT at the time of surgery. Patients who were not candidates for BCT underwent mastectomy. The BCT surgical procedure consisted of excision of the residual primary tumor with a margin of normal tissue. No attempt was made to resect the prechemotherapy tumor volume. Patients with positive margins on final pathologic analysis returned to the operating room for re-excision until negative margins were obtained or had their surgical procedure converted to mastectomy. Pathologically negative margins (defined as >2 mm) were obtained in all patients. Axillary staging was done with sentinel lymph node surgery or complete axillary lymph node dissection. Patients with a positive sentinel lymph node or axillary metastasis proven by preoperative ultrasound-guided fine-needle aspiration underwent complete axillary node dissection. All patients treated with BCT were treated with postoperative external-beam whole-breast radiation therapy to the intact breast with tangential fields. In patients treated with mastectomy, radiation therapy was delivered if the patient's tumor was 5 cm or greater or if there were 4 or more axillary lymph nodes with metastases.

Statistical comparisons between groups were assessed using the χ² test and paired t test. All comparisons were two-tailed. P levels ≤0.05 were considered significant.

RESULTS

The patient and tumor characteristics of the study population are detailed in Table 1. The median patient age was 51 years (range, 25–75 years). A total of 177 patients received postoperative chemotherapy, and 332 patients received preoperative chemotherapy. As expected, because preoperative chemotherapy is usually recommended for patients with larger primary tumors, median tumor size was significantly smaller in the patients who received postoperative chemotherapy compared with those who received preoperative chemotherapy.

TABLE 1. Patient and Tumor Characteristics

The surgical procedures performed are listed in Table 2. The majority of patients (98%) were diagnosed by minimally invasive means (ultrasound-guided core needle biopsy). A total of 241 patients underwent BCT, and 268 underwent mastectomy. Of the 268 mastectomy patients, 142 underwent breast reconstruction (53%). Nine patients had an excisional biopsy as an initial procedure, and all of these patients went on to have an additional surgical procedure. Thirty-two patients had an additional axillary staging procedure after completion of their breast operation. Needle localization was required in 63% of all patients who underwent BCT.

TABLE 2. Surgical Procedures

Table 3 shows the relationship between chemotherapy timing (postoperative vs. preoperative) and volume of breast tissue excised and re-excision rates in patients who underwent BCT. In patients with T1 tumors (≤2 cm), the mean volume of breast tissue resected was not significantly different between patients who underwent postoperative and preoperative chemotherapy. As anticipated (because preoperative chemotherapy is usually recommended for patients with larger primary tumors), patients who underwent postoperative chemotherapy had smaller tumors than those who underwent preoperative chemotherapy. The re-excision rate was similar for patients who received postoperative and those who received preoperative chemotherapy.

TABLE 3. Volume of Breast Tissue Resected and Re-excision Rates in Patients Who Underwent BCT by Timing of Chemotherapy

In patients with T2 or T3 tumors (>2 cm) who underwent BCT, the mean volume of breast tissue excised was significantly greater in patients who received postoperative chemotherapy than in those who received preoperative chemotherapy (213 cm³ vs. 113 cm³, P = 0.0043). There was no difference in tumor size between these 2 groups of patients. The re-excision rate was similar for patients who received postoperative and those who received preoperative chemotherapy.

The relationship between chemotherapy timing and number of breast operations is shown in Table 4. Among patients with T1 tumors and among patients with T2 or T3 tumors, there was no difference between the patients who received postoperative chemotherapy and those who received preoperative chemotherapy in the number of operations. The average number of operations was 1.16 in patients who received postoperative chemotherapy, compared with 1.10 in patients who received preoperative chemotherapy (P = 0.09). Among patients who had BCT, the average number of breast operations was 1.19 in patients who received postoperative chemotherapy, compared with 1.13 in patients who received preoperative chemotherapy (P = 1), and there remained no significant difference on subanalysis by T category.

TABLE 4. Number of Breast Operations Performed by Timing of Chemotherapy

Table 5 shows the relationship between chemotherapy timing and the surgical procedures performed. Rates of conversion to mastectomy after a failed attempt at BCT were 9.9% (10 of 101) in the postoperative chemotherapy group and 7.4% (12 of 162) in the preoperative chemotherapy group. In other words, preoperative chemotherapy did not increase the rate of failed BCT.

TABLE 5. Surgical Procedures by Timing of Chemotherapy

At a median follow-up time of 33 months, there was no difference in local recurrence rates between patients treated with postoperative chemotherapy (1.1%) and those treated with preoperative chemotherapy (0.67%, P = 1). Among patients treated with BCT, 1 patient treated with postoperative chemotherapy and 1 patient treated with preoperative chemotherapy had an in-breast local recurrence. The patient treated with postoperative chemotherapy had the in-breast recurrence 10 months after initial surgery. She underwent mastectomy and died of metastatic disease 20 months later. The patient treated with preoperative chemotherapy for IDC had an in-breast recurrence of ductal carcinoma in situ 5 years after initial surgery. She underwent mastectomy and, as of this writing, remains free of disease 3 months later.

DISCUSSION

The use of preoperative chemotherapy is the accepted standard treatment approach for patients with inflammatory breast cancer. This treatment approach enables many patients with initially unresectable disease to undergo surgery with acceptable rates of locoregional control. Preoperative chemotherapy is also increasingly being used in women with locally advanced disease and in patients with operable earlier-stage breast cancer. The safety of preoperative chemotherapy for patients with resectable breast cancer has been confirmed by 2 large randomized trials: the National Surgical Adjuvant Breast and Bowel Project B-18 trial^1,2 and the European Organization for Research and Treatment of Cancer 10902 trial.¹⁰ These trials compared preoperative versus postoperative chemotherapy for primary operative breast cancer. Survival rates were equivalent in the 2 arms of each trial, and the tumor downsizing by preoperative chemotherapy resulted in increased rates of BCT. The rate of conversion to BCT was greatest in patients with tumors >5 cm at diagnosis. Local-regional recurrence rates of 5% at a median follow-up time of 53 months have been reported with preoperative chemotherapy for stage II and operable stage III breast carcinoma undergoing BCT,⁵ and this is not significantly different from local recurrence rates after BCT and postoperative chemotherapy for early-stage breast cancer.

One of the most important benefits of initiating systemic chemotherapy prior to local therapy is the ability to assess the tumor's in vivo response to a particular drug regimen. Preoperative chemotherapy thus provides the opportunity to optimize therapy by changing chemotherapeutic regimens if the current regimen is not producing the desired results. Another advantage of preoperative therapy is increased drug delivery since the tumor vasculature has not been disturbed. Finally, preoperative chemotherapy also provides an invaluable opportunity to study the molecular determinants of therapeutic response. The clinical and pathologic response of primary breast cancer and axillary node metastases to preoperative chemotherapy is a surrogate marker of response of occult systemic disease because a marked response is associated with a significant increase in disease-free survival compared with the survival of patients with only a minor response to therapy. These benefits of preoperative chemotherapy have motivated the implementation of various preoperative chemotherapy protocols for patients with early-stage breast cancer.

Despite the importance of preoperative chemotherapy in terms of increasing the number of women who are eligible for BCT and potentially advancing breast cancer research with preoperative chemotherapy clinical trials, some surgeons have voiced concerns regarding the use of preoperative chemotherapy in patients with earlier-stage breast cancer. Theoretically, women who undergo preoperative chemotherapy followed by BCT may have larger, smaller, or similar volumes of breast tissue resected compared with the volumes resected in women who undergo BCT followed by chemotherapy. Answers to several clinically relevant questions remain to be elucidated.

This study shows that preoperative chemotherapy reduces the volume of tissue excised in patients with T2 and T3 tumors undergoing BCT compared with patients receiving postoperative chemotherapy with no change in rates of re-excision or recurrence.

In this study, we found that, for patients with T1 tumors at presentation, the volume of breast tissue resected was not significantly different between patients who underwent preoperative and those who underwent postoperative chemotherapy. This finding is most likely due to limitations of surgical technique, ie, a similar volume of breast tissue is removed when resecting needle localized breast tumors, when they are not palpable, whether they are very small or just small in size. Previous studies have shown that patients with primary tumors ≤2 cm receiving preoperative chemotherapy are significantly more likely than patients with larger tumors to have complete eradication of the primary tumor prior to surgery.⁵ However, these patients with complete clinical response from chemotherapy still require surgical resection to evaluate the pathologic response. The current study indicates that preoperative chemotherapy in patients with T1 tumors does not reduce the volume of tissue resected.

In patients with T2 and T3 tumors, the volume of breast tissue resected was significantly smaller in patients who received preoperative chemotherapy than in those who received postoperative chemotherapy. Patients with T2 and T3 tumors have a greater benefit than T1 tumors from the preoperative chemotherapy in terms of reducing the volume of breast tissue excised.

The fact that the total number of operations to achieve negative margins and the rates of re-excision were the same in patients treated with preoperative and those treated with postoperative chemotherapy is important information to present when timing of surgery and chemotherapy is discussed with patients. Although it was previously known that recurrence rates do not differ significantly according to the timing of chemotherapy, to our knowledge ours is the first study to evaluate the impact of chemotherapy timing on the rates of re-excision to obtain negative margins. At the University of Texas M. D. Anderson Cancer Center, all breast specimens are evaluated intraoperatively for margin assessment to allow additional excision of any close margins and thus keep re-excision rates as low as possible. The fact that re-excision rates were not higher in patients who received preoperative chemotherapy is encouraging and alleviates concerns about the ability to obtain negative margins in patients who have received preoperative chemotherapy. Similarly, patients can now be reassured that the use of preoperative chemotherapy will not increase their risk of needing to undergo an unnecessary attempt at breast-conserving surgery to ultimately necessitate the need for mastectomy, compared with the risk in patients receiving postoperative chemotherapy. Clinical assessment by the surgeon to determine appropriateness for BCT was equally reliable in patients treated with preoperative and those treated with postoperative chemotherapy.

Several studies have evaluated factors affecting breast cosmesis after BCT, and 2 independent factors identified are the volume of tissue resected and the need for re-excision of the tumor bed.^11–14 Studies have shown that breast asymmetry and loss of volume are less well tolerated cosmetically than nipple and scar deformity.¹³ With the establishment of the safety of BCT for early breast cancer, cosmesis has become an increasingly important endpoint of treatment. Although cosmetic effects were not directly evaluated in this study, resection of smaller volumes of breast tissue will likely have a positive impact on the ultimate cosmetic outcome of breast conservation surgery.

As the number of patients diagnosed with early-stage breast cancer and undergoing BCT continues to increase, the cosmetic results of the surgical procedure become significantly important, especially to long-term survivors. Optimal cosmesis and minimal complication risk require careful attention to the technical details of surgery and radiotherapy. The use of preoperative chemotherapy can reduce the volume of tissue resected and therefore assist in improving cosmesis. Prospective blinded randomized trials to evaluate the cosmetic outcome of BCT in patients receiving preoperative and postoperative chemotherapy would be required to substantiate these results.

In counseling patients regarding preoperative versus postoperative chemotherapy, there are several important factors that should be discussed. Data from the current study indicate that among patients with larger tumors, those treated with preoperative chemotherapy undergo less extensive resection compared with those treated with postoperative chemotherapy. This smaller volume of tissue resected does not lead to increased risk of re-excision rates to obtain negative margins.

ACKNOWLEDGMENTS

The authors thank Martha Belmares of the Department of Surgical Oncology and Shu-Wan Kau of the Department of Medical Oncology, and Stephanie P. Deming for their assistance.

Discussions

Dr. William G. Cance (Gainesville, Florida): This is an excellent study that addresses a fundamental question about neoadjuvant chemotherapy for breast cancer; namely, are we removing more tissue in these patients with larger tumors and do they require more reexcisions in order to get to clear margins? The M.D. Anderson group has done pioneering work in neoadjuvant chemotherapy, and this is another in a group of studies where they have shown the benefits of its usage.

Intuitively, we feel that if you preoperatively downstage the patient, you would allow treatment based on the actual tumor size at the time of operation. But as you mention, it is possible that the larger volumes may have been resected just because the tumors were bigger at the beginning. This study clearly addresses that that is not the case.

Another critical point that your paper emphasized was the local recurrence rate. There was no increased local recurrence in these patients that were treated with preoperative chemotherapy who had less breast tissue removed. So you are treating local disease with better cosmesis and no change in the oncologic outcome, which is a very major finding for treating patients in this fashion.

I have several questions that relate largely to the primary tumor in your methodologies. How do you follow these tumors during neoadjuvant chemotherapy? Do you mark them before the chemotherapy begins? As chemotherapy becomes more effective, it is getting harder to find the residual disease. So how do you know exactly where to go, particularly with patients that have a complete clinical response? What is the utility of frozen section determination of margins? Do you routinely freeze your margins during surgery? Did that allow less breast tissue to be removed or did it influence your re-excision rates?

I want to congratulate the authors on their excellent results as we move toward a paradigm shift in the management of breast cancer.

Dr. Judy C. Boughey (Houston, Texas): Thank you for your kind comments and discussion of our paper. Regarding how we follow our tumors and when we mark them, anyone who has a tumor of less than 2 cm at presentation will automatically undergo marking, which is usually performed under ultrasound guidance if the tumor can visualized by ultrasound. A complex marker that attaches to the tissue is used, which has a decreased migration rate. Patients who present with larger tumors are followed clinically throughout chemotherapy. Every month they are evaluated either by the surgeon or medical oncologist. If the tumor is showing a significant decrease in size or is less than 2 cm, a marker will be placed at that time.

The plan at the time of surgery is to resect any residual tumor. Prior to the operation every patient will have a repeat mammogram and ultrasound as well as physical exam so that all three modalities are used to assess the degree of disease remaining. The marker is localized with a needle localization wire under mammographic guidance. Then at surgery, we will remove the tissue surrounding the marker as well as anything that is abnormal by palpation or imaging.

As far as your question regarding margin assessment, frozen margins, and intraoperative evaluation, all patients at M.D. Anderson undergo an intraoperative margin evaluation where the specimen is inked with 6 different color inks and then sliced, and then a radiograph of the individual slices taken.

The decision as far as frozen section evaluation of margins or resection of additional tissue is then made between the surgeon who will evaluate the tumor in the pathology lab, the pathologist who will palpate each individual slice, and the radiologist who will intraoperatively evaluate the specimen radiograph. Any area that appears close or concerning will either have additional margins excised or a frozen section performed of the area of concern and if this shows malignancy then additional tissue is excised at the initial operation.

This intraoperative evaluation is performed on every patient regardless of whether the patient received preoperative chemotherapy or not and, therefore, should not affect the outcomes of this study.

Dr. Gerard M. Doherty (Ann Arbor, Michigan): I just have a quick methodologic question about the way you presented the tumor size data. It is a little counterintuitive that you seemed to do smaller excisions for the bigger tumors that were present in the preoperative chemo patients. Are the tumor sizes that you presented the tumor size at the time of the operation or the tumor size at presentation of the patient? So they got smaller with the preoperative chemotherapy, is that why they have the smaller excisions? Did these tumors require only smaller excisions because they had diminished in size after the preoperative chemotherapy?

Dr. Judy C. Boughey (Houston, Texas): Yes. The tumor size we looked at was the tumor size at the clinical presentation when the patient first presents to the office with a breast cancer. So if a patient presents with a tumor of 3 cm, we are comparing the volume of breast tissue resected if they received surgery first compared to the volume resected if they received preoperative chemotherapy and then underwent surgical resection.

Dr. Leigh Anne Neumayer (Salt Lake City, Utah): Thank you for this presentation. I look forward to the paper because it sounds like it will provide more information about long-term follow-up as far as local recurrence, etc. I have a couple of questions.

First, it seems that, from both the abstract and your presentation, you included patients with mastectomy. If you had a 50% mastectomy rate, why did you still need wire localization in 64% of the patients? It seems to me if you are doing a mastectomy you shouldn't need a wire.

Second, what were the mastectomy rates by T-stage?

Third, did any of your analyses include some kind of regression that would control for tumor size and other things that might affect the volume of tissue resected?

I congratulate you on the amount of effort you go to in the operating room to ensure clearance of the margins. In my institution, it is clearly cheaper and less hassle to perform the lumpectomy, assess the adequacy of the margins myself in the operating room, take a little more if necessary, and then let the pathologist provide me with a final report. This results in a much higher re-excision rate than what you report, but I am not sure I could get the pathologist in the room or keep the patient in the room long enough for margin status by frozen section.

Dr. Judy C. Boughey (Houston, Texas): I believe we are very fortunate at M.D. Anderson for the intraoperative input we have from our Pathology and Radiology colleagues, and it does enable us to decrease our re-excision rate.

As far as the wire localization question goes, I apologize if I was not clear. The 64% wire localization rate is only for those patients undergoing breast conservation therapy. We did, as you know, have a 55% mastectomy rate. So as far as the actual volume of tissue resected, that was based only on those patients who underwent breast conservation therapy as their final stage of treatment.