Overcoming radiation resistance in inflammatory breast cancer

Overcoming Radiation Resistance inInflammatory Breast Cancer* Wendy A. Woodward, MD, PhD; Bisrat G. Debeb, PhD; Wei Xu, PhD; and Thomas A. Buchholz, MD The clinical-pathological features of inflammatory breast cancer include enrichment of factors that have been previ- ously associated with radioresistant disease, including negative hormone receptor status and a phenotype enriched for relatively radioresistant breast cancer stem/progenitor cells. The risks and benefits of accelerated postmastec- tomy radiation treatment regimens in the multimodality management of inflammatory breast cancer were reviewed at the first International Inflammatory Breast Cancer Conference at The University of Texas M. D. Anderson Cancer Center. The biological basis for radiation resistance and strategies to radiosensitize these tumors were also presented.
The prevalent basal phenotype of inflammatory breast cancer makes it an ideal clinical model to examine stem cell hypotheses, which the authors believe can help guide future trials to continue making incremental progress against this aggressive disease. Cancer 2010;116(11 suppl):2840–5. V KEYWORDS: cell lines, inflammatory breast cancer, radiation, surgery, chemotherapy, history, stem cells,mammospheres.
Currently, the most widely referenced definition of inflammatory breast cancer (IBC) is that of the American JointCommittee on Cancer,1 which states in part that inflammatory carcinoma is a clinicopathologic entity characterized bydiffuse erythema and edema (peau d’orange) of the breast, often without an underlying mass. These clinical findingsshould involve the majority of the breast. It is important to remember that inflammatory carcinoma is primarily a clinicaldiagnosis. Involvement of the dermal lymphatics alone does not indicate inflammatory carcinoma in the absence of clini-cal findings.
The unique features of the primary disease in IBC make optimal local/regional management a critical component of the multidisciplinary care of IBC. The current standard or care in the management of IBC includes neoadjuvantchemotherapy Æ trastuzumab for appropriate patients, modified radical mastectomy, and high-dose postmastectomy radi-ation targeting the chest wall skin and soft tissue as well as the draining lymphatics. Clearly the skin is a target in IBC, andcare must be taken to ensure adequate dose and skin reaction. Importantly, ‘‘despite the rapid progressive nature of IBC,70% of patients present with local/regional disease without distant metastases at diagnosis,’’2 highlighting the potentialfor improved cure rates with adequate local/regional disease control.
Before the introduction of effective chemotherapy in the management of solid tumors, the 5-year actuarial overall survival for IBC was <5%, with a median survival of only 15 months. Modified radical mastectomy was not technicallypossible for many patients with IBC because of the extent of disease on the chest wall at presentation. It was often difficultto achieve negative margins, and there was little or no benefit to a positive margin surgery. This led to the use of primaryradiotherapy as the sole treatment modality yielding excellent symptom palliation, with modest short-term local controlrates of 50%.
In this era, The University of Texas M. D. Anderson Cancer Center initiated a pilot study of altered-fractionation radiation therapy with the intent of intensifying the radiation dose and improving local control. Among 80 patients stud-ied, 69 patients received standard fractionation of 2 gray (Gy) per day to the chest wall and draining lymphatics. With this Corresponding author: Wendy A. Woodward, MD, Assistant Professor of Radiation Oncology, The University of Texas M. D. Anderson Cancer Center, Box 1202,1515 Holcombe Blvd, Houston, TX 77030; Fax: (713) 563-6940; [email protected] Division of Radiation Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas The articles in this supplement were presented at the First International Inflammatory Breast Cancer Conference, Houston, Texas, December 5-7, 2008.
*Proceedings of the First International Inflammatory Breast Cancer Conference, Supplement to Cancer.
DOI: 10.1002/cncr.25173, Received: December 1, 2009; Accepted: January 14, 2010, Published online May 19, 2010 in Wiley InterScience (www.interscience.wiley.
com) Radiation and Inflammatory Breast Cancer/Woodward et al regimen, patients received a total of 50 Gy þ a 10- to 16- before radiation therapy commenced. For the entire Gy boost to gross disease. Eleven patients received twice- cohort of 256 patients, the 5-year actuarial distant metas- daily fractionation treatment of 1.5 Gy per treatment to a tasis-free survival, local control, and overall survival were total of 51 Gy þ a 15- to 20-Gy boost to gross disease.
40%, 76%, and 44%, respectively.8 Not unexpectedly, Although this was a small study, the local control was 5-year overall survival was significantly improved for 54% in the once-a-day arm versus 73% in the altered frac- patients who completed all therapy, 51% versus 24% tionation arm.3 In 1974, the introduction of doxorubicin (P ¼ .0001), and the local/regional control rate was 84% chemotherapy yielded another significant gain in IBC for patients who completed all therapy. This rate of local/ outcome. Although the addition of chemotherapy had no regional control was higher than historical rates. Examina- effect on local control (73% vs 74%3), the 5-year survival tion of the factors that predicted for local recurrence after improved from <5% to 30%-50% because of the reduc- radiation among patients in Group 1 revealed that age tion in distant metastasis. Importantly, the addition of <45 years, less than partial response to chemotherapy, !4 chemotherapy rendered more patients operable, and positive lymph nodes, lack of taxane use, and positive or modified radical mastectomy was reintroduced. Several unknown surgical margins were all associated with series subsequently reported the local control benefit of increased risk of local/regional recurrence. Patients with mastectomy in the setting of systemic chemotherapy, any 1 of these factors had higher rates of local/regional re- demonstrating an absolute increase in local control of currence, ranging from 21% to 52%. Dose escalation 24% to 45%.4-6 In the postmastectomy setting, however, (60 vs 66 Gy) appeared to have the most significant bene- total radiation dose was lowered from a gross disease dose fit in these high-risk groups, with age <45, positive/ of 66 to 70 Gy to a microscopic disease dose of a standard unknown margins, and disease unresponsive to chemo- 60 Gy. Although this lowered the complication rates therapy. Given the higher rate of grade 3 toxicity at 5 years related to radiotherapy, local failure remained a challenge, with dose escalation (29% vs 15%, P ¼ .08), we now with local failure rates exceeding 20%.7 Demonstrating reserve dose escalation for patients with high-risk features that this was a problem of radiation resistance to this low- ered dose, the majority of these local failures were centeredin the radiation field on the chest wall. This provided therationale for dose escalation in the setting of postmastec-tomy radiation, and in 1987, a dose escalation regimen was introduced for IBC patients at The University of Texas M. D. Anderson Cancer Center consisting of 51 The clinical-pathological features of IBC include enrich- Gy to the chest wall and draining lymphatics, and a 15- ment of factors that have been previously associated with Gy boost to the chest wall (to 66 Gy total) delivered at 1.5 radioresistant disease, including negative receptor status Gy/fraction twice daily. Treating twice a day reduced the (estrogen and progesterone9,10) and a phenotype enriched overall treatment time from 6 to 41=2 weeks, an example of for relatively radioresistant breast cancer stem/progenitor the early dose-dense strategy in radiation therapy.
cells.11-13 In general, it is clear that although treatment In 2008, a retrospective analysis was performed of advances have been made, clinical experience demon- all patients with nonmetastatic IBC cancer treated at The strates a relative radioresistance of IBC tumors compared University of Texas M. D. Anderson Cancer Center over with non-IBC breast tumors, where local failure rates after the progression of these treatment strategies from 1977 to contemporary postmastectomy radiation are <10%.14 2004.8 Patients who were initially intended to receive Although improved local control can be achieved in IBC neoadjuvant chemotherapy, modified radical mastec- with dose escalation, the cost in terms of toxicity can be tomy, and postmastectomy radiation were included high, and alternative approaches to radiosensitize these tu- regardless of whether they were able to complete all ther- apy and grouped by those who completed all therapy or Laboratory investigations of available IBC cell lines those who did not, often because of progressive disease.
SUM-149 and SUM-190 demonstrate relative radiore- Among 256 patients analyzed, 192 completed all therapy sistance of these lines (data not shown), making them (Group 1), whereas 64 did not and received either preop- good models for investigation of radioresistance mecha- erative radiation (n ¼ 21) or definitive radiation (n ¼ 21).
nisms. Radiation resistance is classically assessed by clono- An additional 22 patients developed a local recurrence genic cell survival curves, which are generated by counting Figure 1. Flow cytometric analysis of expression of CD24 and CD44 in SUM149 inflammatory breast cancer (IBC) cells is shown.
SUM149 IBC cells were cultured under adherent (2-dimentional [2D]) or tumor-initiating cells enrichment condition (3D). Cellswere stained with CD24 and CD44 antibodies conjugated with different dyes and subjected to cytometric analysis. (A) The distri-bution of CD24 and CD44 expression in 1 representative experiment is shown: (A1) cells cultured in 2D condition stained withCD24 only; (A2) cells cultured in 3D condition stained with CD44 only; (A3) cells grown in 2D condition stained with both CD24and CD44; (A4) cells grown in 3D condition stained with both CD24 and CD44. (B) The relative increase of CD44 þ CD24lowpopulation in cells cultured in 3D condition compared with that in 2D condition is shown. The result is a summary of 5 independ-ent experiments. *P < .05.
colonies produced from cells that survive radiation and factors in the search for selective cancer stem cell radiosen- are capable of dividing clonally. The primary mechanism of radiation cell killing in solid tumors is mitotic cell deathrather than apoptosis, and as such, this long-term assess- ment of replicative potential eliminates cells that may per- Although novel targeted cancer stem cell radiosensitizers sist briefly after radiation but are destined to die after a are desirable, they are still in early in clinical development, few cell divisions and thus are not likely to be capable of and it will be some time before optimal strategies are defined. In the meantime, building on published and Clonogenicity has been commonly assessed in institutional experience with the oral prodrug of 5-fluo- monolayer culture in serum containing media that may ruracil, capecitabine,19 we have activated a new protocol diminish the stem or progenitor cells’ potential in the cul- for patients with progressive primary IBC to radiosensitize ture system.15,16 Because normal and cancer stem/progen- the primary tumors during radiation using capecitabine.
itor cells from mammary gland have been shown to be On the basis of the long history of 5-fluorouracil resistant to radiation compared with bulk, differentiated chemotherapy for breast cancer and the encouraging cells,11,12,17 we have compared clonogenicity of IBC cell safety and efficacy results for preoperative concurrent lines assessed in 2-dimensional (2D) monolayer clono- capecitabine and radiation in rectal cancer,20,21 this genic assays as well as in 3-dimensional serum-free growth approach was advocated for carefully selected breast can- factor-enriched clonogenic assays designed to support cer patients with inoperable breast cancer who had no stem/progenitor cells in in vitro culture.15 Exposure to better treatment options at The University of Texas stem cell-promoting media increases the percentage of tu- M. D. Anderson Cancer Center. In a retrospective review mor-initiating cells that have been prospectively identified of 55 patients treated with concurrent radiation and as CD44high/CD24low18 (Fig. 1) from human breast capecitabine at The University of Texas M. D. Anderson cancers. These tumor-initiating cell populations are more Cancer Center for inoperable breast cancer (IBC and resistant to radiation in clonogenic assays compared with non-IBC), concurrent chemoradiation with capecitabine the adherent population (Fig. 2). This assay system pro- demonstrated 91% of these patients converted to having vides an effective model to test pharmacologic and genetic operable disease after treatment.22 The clinical complete approaches to inhibit putative cancer stem cell survival response rate was 33%; moreover, the overall pathological Radiation and Inflammatory Breast Cancer/Woodward et al Figure 2. Tumor-initiating cells from inflammatory breast cancer (IBC) cell lines are more resistant to radiation than monolayer(2-dimensional [2D]) cells. Clonogenic survival assay of IBC cells derived from (A) SUM149 and (B) SUM190 cultured under 2D ormammosphere (3D) cultures is shown. Cells in 3D are more resistant to radiation than cells grown in 2D. Cells were grown in 2Dor 3D culture conditions and irradiated with single, increasing doses (0, 2, 4, 6 gray [Gy]) of radiation.
complete response rate was 20%. Only 1 patient had pro- To assess the efficacy and toxicity prospectively, to gressive disease. The 5-year overall survival, local recur- set the bar for upcoming novel radiosensitizers, and to rence-free survival, and distant metastasis-free survival permit the correlation of breast cancer stem cells in the rates were 48%, 85%, and 37%. Sixteen (29%) patients primary tumor pre- and post-radiation therapy to out- had a grade 3 or higher complication (acute yet resolving come, we have recently activated a single-arm phase 2 study to quantify response, conversion to operable among In a phase 2 study of radiation and capecitabine in nonoperable patients, local control of gross disease not women with locally advanced breast cancer who had failed resected at the time of surgery, and toxicity. Patients con- first-line anthracycline-based neoadjuvant therapy, Gaui senting to optional procedures will have a core biopsy per- et al. studied the concomitant use of radiation therapy formed before and at the conclusion of radiotherapy to be and capecitabine, to determine the toxicity and efficacy of assessed for cancer stem cells before and after radiother- this regimen as a second-line neoadjuvant treatment.19 apy. We believe that advances in technology for radiation Twenty-eight patients with inoperable locally advanced therapy targeting and delivery make this the right time for breast cancer refractory to first-line anthracycline-based strategies exploring aggressive local/regional management treatment were enrolled between January 2003 and May in patients with progressive IBC. Although we anticipate 2004. Patients received radiation therapy (total dose 5000 preoperative radiation with capecitabine will provide cGy) and concomitant capecitabine (850 mg/m2) twice disease control and conversion to operability in some daily for 14 days every 3 weeks. This treatment rendered patients, patient selection and consent are critical before 23 (82%) of the 28 patients operable. The 5 remaining embarking on any new therapy with significant potential patients did not undergo surgery because of disease pro- toxicity. These data will provide a baseline against which gression. The median clinical tumor size decreased from to compare new pipeline IBC targets. The prevalent basal 80 to 49 cm2. Microscopic residual disease was observed phenotype of IBC makes it an ideal clinical model to in 3 (13%) patients, and another patient achieved a com- examine stem cell hypotheses, which we believe can help plete pathologic response. The median number of guide future trials to continue making incremental pro- involved lymph nodes was 2, and treatment was well toler- gress against this profoundly aggressive disease.
ated, with no grade 3 or 4 events. These results indicatethat second-line neoadjuvant treatment with radiation therapy and capecitabine is feasible, well tolerated, and Acceleration of radiotherapy regimens, giving increased effective in patients with locally advanced breast cancer re- dose in less time, has been an effective strategy against the fractory to primary anthracycline-based treatment.
aggressive biology of IBC. Higher-risk patient subsets have been identified that gain the greatest benefit from Program and Clinic; the State of Texas Grant for Rare and these aggressive strategies, but local control has not yet Aggressive Cancers; and The University of Texas InstitutionalResearch Grants; KL2 RR024149 and R01CA138239-01.
been maximized for all patients, and safer, less toxicapproaches are needed. Greater understanding of the biol-ogy of radiation resistance in IBC is providing new insights to guide the design of new trials with novel radio- 1. Singletary SE, Allred C, Ashley P, et al. Staging system for breast cancer: revisions for the 6th edition of the AJCC sensitizers to improve tumor control in IBC.
Cancer Staging Manual. Surg Clin North Am. 2003;83:803-819.
2. Wingo PA, Jamison PM, Young JL, Gargiullo P. Popula- tion-based statistics for women diagnosed with inflammatory breast cancer (United States). Cancer Causes Control. 2004;15:321-328.
To evaluate radiosensitivity, specified numbers of 3. Barker JL, Montague ED, Peters LJ. Clinical experience cells derived from monolayer cultures were seeded into with irradiation of inflammatory carcinoma of the breast individual wells of a 6-well tissue culture plate (for 2D) or with and without elective chemotherapy. Cancer. 1980;45:625-629.
ultralow attachment plates (for 3D). Cells from both 4. Fleming RY, Asmar L, Buzdar AU, et al. Effectiveness of 2D and 3D were exposed to increasing doses of radia- mastectomy by response to induction chemotherapy for con- tion (2, 4, or 6 Gy) 4 hours after plating. 2D plates trol in inflammatory breast carcinoma. Ann Surg Oncol.
1997;4:452-461.
were incubated for 10 to 14 days, and colonies were stained 5. Panades M, Olivotto IA, Speers CH, et al. Evolving treat- with crystal violet. For 3D, cells were incubated in mammo- ment strategies for inflammatory breast cancer: a popula- sphere media for 7 days, and spheres were stained with tion-based survival analysis. J Clin Oncol. 2005;23:1941-1950.
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro- 6. Perez CA, Fields JN, Fracasso PM, et al. Management of mide to improve visualization and counted using a locally advanced carcinoma of the breast. II. Inflammatory Gelcount colony counter (Oxford Optronix, Oxford, carcinoma. Cancer. 1994;74(1 suppl):466-476.
7. Thoms WW Jr, McNeese MD, Fletcher GH, Buzdar AU, UK). Clonogenic survival curves were generated using Singletary SE, Oswald MJ. Multimodal treatment for inflammatory breast cancer. Int J Radiat Oncol Biol Phys.
1989;17:739-745.
8. Bristol IJ, Woodward WA, Strom EA, et al. Locoregional treatment outcomes after multimodality management of SUM149 and SUM190 cells cultured under mammo- inflammatory breast cancer. Int J Radiat Oncol Biol Phys.
sphere and adherent conditions were harvested with 10 mM ethylenediaminetetraacetic acid, centrifuged, 9. Kyndi M, Sorensen FB, Knudsen H, Overgaard M, Nielsen HM, Overgaard J. Estrogen receptor, progesterone receptor, and resuspended in phosphate-buffered saline (PBS) HER-2, and response to postmastectomy radiotherapy in (105 cells/mL). They were then incubated with fluorescein high-risk breast cancer: the Danish Breast Cancer Coopera- isothiocyanate (FITC)-conjugated CD44 and phycoery- tive Group. J Clin Oncol. 2008;26:1419-1426.
10. Woodward WA, Buchholz TA. The role of locoregional thrin (PE)-conjugated CD24 antibodies (BD Biosciences, therapy in inflammatory breast cancer. Semin Oncol.
San Diego, Calif) for 30 minutes at concentrations recom- mended by the manufacturer. Cells incubated in PBS, 11. Phillips TM, McBride WH, Pajonk F. The response of CD24(-/low)/CD44þ breast cancer-initiating cells to radia- FITC, or PE alone served as controls. Cell analysis for the tion. J Natl Cancer Inst. 2006;98:1777-1785.
expression of CD44 and CD24 was performed using a 12. Woodward WA, Chen MS, Behbod F, Alfaro MP, Buch- Beckman Coulter (Fullerton, Calif) machine, and the holz TA, Rosen JM. WNT/beta-catenin mediates radiationresistance of mouse mammary progenitor cells. Proc Natl data files were analyzed using FlowJo software (Treestar, 13. Xiao Y, Ye Y, Yearsley K, Jones S, Barsky SH. The lympho- vascular embolus of inflammatory breast cancer expresses astem cell-like phenotype. Am J Pathol. 2008;173:561-574.
14. Greenbaum MP, Strom EA, Allen PK, et al. Low locore- This supplement was sponsored by the Houston Affiliate of gional recurrence rates in patients treated after 2000 with Susan G. Komen for the Cure, the National Cancer Institute, doxorubicin based chemotherapy, modified radical mastec- and the State of Texas Rare and Aggressive Breast Cancer tomy, and post-mastectomy radiation. Radiother Oncol. In Research Program. The First International Inflammatory Breast Cancer Conference was supported in part by GlaxoSmithKline, 15. Dontu G, Abdallah WM, Foley JM, et al. In vitro propaga- Pfizer, Eli Lilly and Company, and Cardinal Health. Supported tion and transcriptional profiling of human mammary stem/ by the Morgan Welch Inflammatory Breast Cancer Research progenitor cells. Genes Dev. 2003;17:1253-1270.
Radiation and Inflammatory Breast Cancer/Woodward et al 16. Woodward WA, Bristow RG. Radiosensitivity of cancer-ini- 20. Dunst J, Debus J, Rudat V, et al. Neoadjuvant capecitabine tiating cells and normal stem cells (or what the Heisenberg combined with standard radiotherapy in patients with uncertainly principle has to do with biology). Semin Radiat locally advanced rectal cancer: mature results of a phase II trial. Strahlenther Onkol. 2008;184:450-456.
17. Zhang M, Behbod F, Atkinson RL, et al. Identification of 21. Krishnan S, Janjan NA, Skibber JM, et al. Phase II study of tumor-initiating cells in a p53-null mouse model of breast capecitabine (Xeloda) and concomitant boost radiotherapy cancer. Cancer Res. 2008;68:4674-4682.
in patients with locally advanced rectal cancer. Int J Radiat 18. Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. Prospective identification of tumorigenic breast 22. Perkins GH, Middleton LP, Tran R, et al. Concurrent che- cancer cells. Proc Natl Acad Sci U S A. 2003;100:3983-3988.
moradiation with capecitabine achieves meritable response 19. Gaui MF, Amorim G, Arcuri RA, et al. A phase II study of and local control for inoperable and recurrent neoadjuvant second-line neoadjuvant chemotherapy with capecitabine chemotherapy refractory breast cancer. Paper presented at: and radiation therapy for anthracycline-resistant locally San Antonio Breast Cancer Symposium; December 13-16, advanced breast cancer. Am J Clin Oncol. 2007;30:78-81.

Source: http://myibcresearch.wikispaces.com/file/view/radiation_resistance_ibc.pdf/154048391/radiation_resistance_ibc.pdf