• breast neoplasms;
  • radiation therapy;
  • treatment outcome;
  • neoadjuvant therapy


  1. Top of page
  2. Abstract


The objectives of this study were to summarize a single-institution experience in treating patients with inflammatory breast carcinoma (IBC) using trimodality therapy and to identify prognostic factors for outcome.


Sixty-one women underwent radiation therapy with curative intent for IBC between 1982 and 2001. All but five women received trimodality therapy. Neoadjuvant chemotherapy was given to the majority of women (n = 43 patients), although some received “up-front” surgery as first therapy (n = 18 patients).


With a median potential observation time after diagnosis of 14 years, freedom from locoregional disease progression was 78%, freedom from distant metastasis was 45%, and the cause-specific survival rate was 47% at 5 years. Approximately 40% of the 56 patients who received trimodality therapy remained free of disease. Multivariate analysis demonstrated three factors that were found to be associated significantly with improved cause-specific survival: pathologic tumor size < 4 cm (P = 0.0001), up-front surgery (P = 0.0078), and local disease control (P = 0.0003). Factors that were found to be associated with better freedom from locoregional disease progression were pathologic tumor size (< 4 cm; P = 0.0157), age (> 55 years; P = 0.0596), and radiation dose (≥ 60 grays [Gy]; P = 0.0621).


IBC is an aggressive disease that is treated effectively in select patients by multimodality therapy. Patient outcomes may be improved with therapies that result in better local and systemic control. Further studies are warranted to address the optimal sequence of trimodality therapy and the optimal administration of each agent. Cancer 2004;100:920–8. © 2004 American Cancer Society.

Inflammatory breast carcinoma (IBC) comprises 1–6% of all invasive breast tumors.1 Over the last 20 years, its incidence has doubled.2 IBC is distinguished from other breast tumors by both clinical and pathologic features.3 Most commonly, it presents with classic findings in the skin overlying the breast, including peau d'orange, erythema, edema, and warmth, believed to be a result of tumor emboli obstructing flow in the dermal lymphatics. Nipple retraction is noted frequently; often, no mass can be detected on physical examination or mammogram, perhaps because of the diffuse involvement throughout the lymphatics. Classic IBC develops rapidly with signs and symptoms present < 3 months before diagnosis. Distant metastases are common at diagnosis and are almost universal without chemotherapy.

The treatment of IBC remains a challenge. In the era of trimodality therapy (surgery, radiotherapy [RT], and chemotherapy), IBC has become a moderately curable disease, with overall survival rates at 5 years improved from < 10% without chemotherapy to 30–55%.1, 4 However, there is significant room for improvement. Sequence of therapy (neoadjuvant chemotherapy vs. up-front mastectomy) is one treatment-related issue that to our knowledge has not been addressed thoroughly in the literature. At our institution (the University of Florida College of Medicine, Gainesville, FL), a greater proportion of women with IBC in the middle 1980s underwent up-front mastectomy as a means of expeditiously decreasing tumor burden. After the efficacy of chemotherapy was demonstrated, there was a shift toward administering neoadjuvant chemotherapy prior to performing mastectomy or RT.

The primary objective of the current study was to describe a relatively large, single-institution experience with IBC in the era of trimodality therapy. In addition, we hoped to identify clinical and treatment-related factors associated with a successful outcome, with particular attention to the use of up-front mastectomy.


  1. Top of page
  2. Abstract

Between 1961 and 2001, as part of a University of Florida Institutional Review Board (IRB)-approved radiation outcome tracking study (IRB no. 70-2003), all patients with locally advanced breast carcinoma who were treated with RT at the University of Florida were reviewed, and 84 patients with IBC were identified who were treated with curative intent. A previous report summarized outcomes for patients with IBC who were treated up to March 1989.5 The focus of the current study was a subset of 61 patients who were diagnosed after July 1, 1982. This cut-off date was used because it represented a point when chemotherapy was included routinely as part of therapy; after that date, 56 patients (91%) received trimodality therapy compared with 30% of patients prior to that date. The diagnoses of IBC were based solely on clinical grounds (n = 26 patients), pathologic grounds (n = 11 patients), or both (n = 24 patients). Clinical signs necessary for the clinical diagnosis of IBC included peau d'orange, erythema, and skin warmth; other signs that frequently were present included pain, nipple retraction or nipple discharge, and axillary lymphadenopathy. There was no imposed limit on the duration of clinical signs, and 6 women who had clinical signs that were present for > 1 year prior to diagnosis were included in the current study. Pathologic grounds for the diagnosis of IBC included the presence of dermal lymphatic invasion. Patients with dermal lymphatic invasion in the absence of clinical signs were classified as occult IBC and were included in this study. Women with distant metastases at the time of diagnosis (as demonstrated by a staging workup that included chest X-ray, bone scan, liver-spleen scan, or computed tomographic imaging of the head, chest, or abdomen) were excluded. However, women with supraclavicular metastases (three patients) at the time of diagnosis were included, consistent with the most recent edition of the American Joint Committee on Cancer (AJCC) classification system, which implies that these patients have better outcomes compared with patients who have distant, hematogenous metastases.6 A positive family history for breast carcinoma was noted if any related family member was diagnosed with breast carcinoma at any age.

All patients were treated with either up-front mastectomy (n = 18 patients) or neoadjuvant chemotherapy (n = 43 patients) as their first modality of therapy. All 18 patients who underwent up-front mastectomy were diagnosed prior to 1994, and all patients diagnosed after January 1, 1994 were treated with neoadjuvant chemotherapy. Five patients were treated with only two treatment modalities: either surgery and RT (four patients) or chemotherapy and RT (one patient) because of patient refusal of chemotherapy or medical contraindication to surgery. The decision regarding first treatment was based on current institutional guidelines rather than patient or tumor factors at the time of presentation. Surgery was comprised of modified radical mastectomy in all but two patients, who underwent only simple mastectomy with planned inclusion of the axilla in postoperative RT fields. The standard axillary dissection included Levels I and II, with a median of 16 lymph nodes (range, 0–57 lymph nodes) submitted to pathology. Pathology review showed a median of 5 lymph nodes (range, 0–49 lymph nodes) with metastatic disease. Chemotherapy was doxorubicin-based (88%), except in those patients for whom it was contraindicated medically or if it was not adopted yet as preferred chemotherapy by the medical oncologist (n = 7 patients); in these patients, methotrexate was substituted for doxorubicin. Taxanes were used in 7 of 10 patients diagnosed after 1996 and was either given as a single agent sequential to (n = 5 patients) or concomitantly with (n = 2 patients) neoadjuvant doxorubicin. Neoadjuvant chemotherapy (n = 42 patients) was comprised of a median of 3 cycles of combination chemotherapy, mostly including doxorubicin, cyclophosphamide, and/or 5-fluorouracil. Response to chemotherapy was defined either as complete (if there was resolution of all clinical signs and no residual palpable abnormality) or partial (if there was any improvement in clinical signs or reduction in a palpable abnormality). A pathologic complete response was documented in patients who had no residual invasive carcinoma identified. Neoadjuvant chemotherapy generally was followed by mastectomy; however, patients who were judged unresectable after chemotherapy were treated with preoperative RT (n = 9 patients). In these patients, the median time to surgery after completion of RT was 69 days (range, 43–113 days). Combination chemotherapy was administered concurrently with RT on a case-by-case basis (n = 27 patients), with the omission of doxorubicin during radiation treatments. Consolidation chemotherapy brought the median number of cycles to 6 (range, 3–19 cycles). The number of cycles did not differ significantly, regardless of whether the first therapy was surgery (median, 7 cycles; range, 6–19 cycles) or chemotherapy (median, 7 cycles; range, 3–14 cycles). Tamoxifen (at a dose of 20 mg per day) was administered to some patients who had positive estrogen receptor status (n = 14 patients), to some patients who had negative estrogen receptor status (n = 6 patients), and to some patients who had undetermined or unavailable receptor status (n = 3 patients). The standard course of RT included a median of 50 gray (Gy) to the chest wall with en face electrons or tangential photon fields, 50 Gy to the internal mammary lymph nodes with en face electrons with or without a minor photon component, 50 Gy to the anterior supraclavicular axillary field with 6-megavolt (MV) photons prescribed to Dmax, and additional radiation with 20-MV photons through a posterior axillary boost field to bring the midaxillary dose to 50 Gy. Bolus was applied to the chest wall each day throughout RT. All fields typically were treated in 25 fractions, once per day with 5 treatments per week. Five patients who were treated with preoperative RT received a total breast dose of 45 Gy or 51 Gy in 30 fractions twice per day with 10 treatments per week and no subsequent boost. In patients who were treated with RT once daily, the mastectomy scar was boosted to a median of 10 Gy in 5 fractions, resulting in a final median dose of 60 Gy (range, 42–70 Gy). A total of 20 patients received < 60 Gy; in the earlier years of the study, the mastectomy scar was not always boosted (n = 10 patients), and others received a lower RT dose than intended because of moist desquamation (n = 4 patients; range, 48–56 Gy) or progression of disease (n = 1 patient; 42 Gy). No patient had a break in RT. A summary of patient, tumor, and treatment characteristics of the patients is listed in Table 1.

Table 1. Patient, Tumor, and Treatment Characteristics for 61 Patients
CharacteristicMedian (range)No. of patients (%)a
  • Ca: carcinoma; RT: radiation therapy; chemo: chemotherapy.

  • a

    Percentages are based on the number of patients with data available.

Age (yrs)55 (27–82) 
Postmenopausal (n = 61) 36 (59)
Family history of breast ca (n = 47) 14 (30)
Diagnosis (n = 61)  
 Clinical alone 26 (43)
 Pathologic alone 11 (18)
 Both 24 (39)
Duration of symptoms (weeks) (n = 56) 8 (1–400) 
 > 8 weeks 20 (36)
 > 52 weeks  6 (11)
No. of presenting signs (n = 58) 3 (0–6) 
 ≥ 3 signs 38 (66)
 ≥ 5 signs 12 (21)
Clinical tumor size (n = 60)  
 No discrete mass  6 (10)
 < 5 cm 23 (38)
Clinical axillary lymphadenopathy (n = 53) 44 (83)
Clinical supraclavicular lymphadenopathy (n = 61)  3 (5)
History of hypertension (n = 60) 17 (28)
History of diabetes (n = 59)  6 (10)
Pathologic tumor size (cm) (n = 50) 4 (0–12) 
 No residual tumor identified  6 (12)
 ≤ 4 cm 25 (50)
Pathologic dermal lymphatic invasion (n = 53) 37 (70)
Pathologic vascular space invasion (n = 37) 30 (82)
Estrogen-receptor positive (n = 45) 20 (44)
Pathologic lymph node status (n = 56)  
 pN0 11 (20)
 pN1bi  6 (11)
 pN1bii 11 (20)
 pN1biii 11 (20)
 pN1biv 13 (23)
 pN2  4 (7)
Time to RT after diagnosis (days)115 (24–448) 
Timing of RT relative to surgery (n = 60)  
 Preoperative RT 10 (17)
 Time to surgery after preoperative RT (days) (n = 9)69 (43–113) 
 Time to chemo after diagnosis (days)19 (0–371) 
Type of chemo (n = 57)  
 Doxorubicin-based 50 (88)
 Tamoxifen (n = 47) 23 (49)
Clinical response to neoadjuvant chemo (n = 42)  
 Complete 13 (31)
 Partial 25 (60)
 None  3 (7)
 Progression  1 (2)
Response to preoperative RT (n = 9)  
 Pathologically complete  1 (14)
 Clinically complete  1 (14)
 Partial  5 (71)
 Progression  1 (14)
 Unknown  1 (14)
Sequence of therapy  
 Surgery [RIGHTWARDS ARROW] RT  4 (7)

Data were analyzed using both Number Cruncher Statistical System (NCSS) 20007 software and SAS software (version 8).8 Survival estimates were computed using the Kaplan–Meier method, and the two-tailed log-rank test was used to determine statistical significance between any strata of selected explanatory variables.9 Overall survival was calculated from the date of diagnosis to the date of last contact or death; all patients had at least 2 years of potential follow-up. Potential observation times were calculated from the date of diagnosis to the date of the analysis. A comprehensive set of patient, tumor, and treatment factors were evaluated for potential association with freedom from locoregional disease progression, freedom from distant metastasis, and cause-specific survival (CSS). Four patients were known to have died of disease, but the site of failure was unknown. For these patients, recurrences were censored at the time of death. This creates a known bias in the analysis; therefore, the reported recurrence rates are underestimated slightly. In survival analyses, potential continuous explanatory variables were converted to categorical data using median values as cut-off points to prevent extreme values from biasing results artificially. Factors with any binary level that contained < 10% of patients were not analyzed because of the high likelihood of spurious survival estimates. Seven factors were selected for multivariate analysis using proportional hazards regression based on both the results of univariate analysis and predetermined intention of analysis (namely, the prognostic significance of occult IBC and up-front surgery). The Fisher exact test was used as needed to identify the existence of significant differences between any two groups of comparison. A P value ≤ 0.05 was interpreted as statistically significant.


  1. Top of page
  2. Abstract

The median potential observation time for the 61 women was 14 years, and the median overall survival was 3.4 years. One patient was lost to follow-up 1.1 years after diagnosis and was censored at that time with no evidence of disease. At 5 years, the overall survival rate for the 61 women was 45%, and the CSS rate was 47%. At 10 years, the overall survival and CSS rates were 28% and 29%, respectively. Figure 1 compares CSS for all 84 women both before and after the cut-off date, when trimodality treatment became standard therapy; the CSS rate was 26% at 5 years for patients who were diagnosed with IBC prior to July 1982. All further analyses were restricted to patients who were diagnosed after July 1982.

thumbnail image

Figure 1. Cause-specific survival before and after July 1982 (n = 84 patients; P = 0.1733).

Download figure to PowerPoint

The 5-year rates of freedom from locoregional disease progression and freedom from distant metastasis were 78% and 45%, respectively, with most failures occurring within the first 2 years (Fig. 2). Patterns of failure for patients who received trimodality therapy are demonstrated in Figure 3. Of these 56 patients, 61% developed recurrent disease after treatment, with more failures accounted for by distant metastasis alone (39%) than by locoregional disease progression alone (2%) or local and distant failure together (20%). For all 61 patients, a slight improvement in the overall recurrence rate was observed in patients who underwent surgery first compared with patients who received chemotherapy first (50% vs. 65%; P = 0.3893). There was a lower rate of distant metastasis in patients who underwent surgery first, although the difference was not statistically different (44% vs. 65%; P = 0.1614). It is interesting to note that tumor size was not found to be significantly smaller in patients who underwent up-front surgery instead of receiving neoadjuvant chemotherapy first (67% vs. 43% of patients had pathologic tumor sizes < 4 cm; P = 0.2165). Age ≤ 55 years (50% vs. 53%; P = 0.9999) and RT dose < 60 Gy (17% vs. 40%; P = 0.1341) also were not found to be statistically significant variables.

thumbnail image

Figure 2. Freedom from locoregional disease progression and freedom from distant metastases (n = 61 patients).

Download figure to PowerPoint

thumbnail image

Figure 3. Patterns of failure by sequence of trimodality therapy. The numbers of patients for each subgroup are denoted to the left of the bar graph. Chemo: chemotherapy.

Download figure to PowerPoint

For patients who underwent up-front surgery rather than receiving neoadjuvant chemotherapy first, the actuarial outcome at 5 years for overall survival was 50% versus 44%, the CSS rate was 53% versus 45%, the rate of freedom from locoregional progression was 74% versus 80%, and the rate of freedom from distant metastases was 52% versus 42%.

Univariate analysis was performed on the factors listed in Table 2 with endpoints of freedom from locoregional disease progression and CSS. The only factor that was found to be associated with better freedom from locoregional disease progression was pathologic tumor size (< 4 cm; P = 0.0157). Other marginally significant factors included age (> 55 years; P = 0.0596), RT dose (≥ 60 Gy; P = 0.0621), pathologic lymph node stage (< pN1biv; P = 0.0938), and nondoxorubicin chemotherapy (P = 0.0943). Factors that were found to be associated with better CSS included pathologic tumor size (< 4 cm; P = 0.0033), no pathologic dermal lymphatic invasion (P = 0.0584), nondoxorubicin chemotherapy (P = 0.0049), freedom from locoregional disease progression (P = 0.0093), and duration of symptoms (≤ 8 weeks; P = 0.0605). Notably, the subsets of patients who received nondoxorubicin chemotherapy and had duration of symptoms ≤ 8 weeks had smaller pathologic tumor sizes than the comparison group (P = 0.0104 and P = 0.0402, respectively).

Table 2. Freedom from Locoregional Disease Progression and Cause-Specific Survival
Prognostic factors in univariate analysis
 Age ≤ 55 yrs
 Postmenopausal status
 History of hormone replacement therapy
 Duration of symptoms ≤ 8 weeks prior to diagnosis
 Clinical tumor size < 5 cm
 No. of clinical signs at presentation (< 3)
 Family history of breast cancer
 History of hypertension at time of diagnosis
 History of diabetes at time of diagnosis
 Clinically positive axillary lymphadenopathy
 Clinically occult inflammatory breast cancer
 Pathologic tumor size < 4 cm
 Pathologic tumor size < 4 cm, subset treated with neoadjuvant chemotherapy
 Pathologic complete response (no residual tumor identified)
 Pathologic vascular space invasion
 Pathologic dermal lymphatic invasion
 Estrogen receptor status
 No. of dissected axillary lymph nodes (< 17)
 Pathologic lymph node stage ≥ pN1biv
 Up-front surgery
 Time to surgery
 Radiotherapy dose < 60 grays
 Radiotherapy intended course not completed
 Time to radiotherapy
 Preoperative radiotherapy
 Neoadjuvant chemotherapy
 Complete or partial response to neoadjuvant chemotherapy
 Concomitant chemotherapy and radiotherapy
 Type of chemotherapy (doxorubicin-based)
 Type of chemotherapy (taxane use)
 Time to chemotherapy
 Tamoxifen as part of treatment
 Freedom from locoregional disease progression

A multivariate analysis was performed on factors that were identified as significant in the univariate analysis as well as the factors of occult IBC and up-front surgery. Factors that were found to be associated with better CSS were pathologic tumor size (< 4 cm; P = 0.0001), up-front surgery (P = 0.0078), and freedom from locoregional disease progression (P = 0.0003), as shown in Table 3. Survival curves for these factors that also were significant on univariate analysis are demonstrated in Figures 4 and 5. The CSS rates at 5 years for pathologic tumor size < 4 cm were 65% versus 35%, respectively; the CSS rates at 5 years for patients with and without up-front surgery were 55% versus 45%, respectively; and the corresponding CSS rates with and without freedom from locoregional disease progression were 55% versus 20%, respectively.

Table 3. Multivariate Analysis for Cause-specific Survival
VariableP value
Duration of symptoms ≤ 8 weeks prior to diagnosis0.9985
Pathologic dermal lymphatic invasion0.9062
Occult inflammatory breast carcinoma0.7312
Type of chemotherapy (doxorubicin-based)0.2926
Pathologic tumor size < 4 cm0.0001
Freedom from locoregional disease progression0.0003
Up-front surgery0.0078
thumbnail image

Figure 4. Cause-specific survival stratified by pathologic tumor size (n = 50 patients; P = 0.0033).

Download figure to PowerPoint

thumbnail image

Figure 5. Cause-specific survival stratified by freedom from locoregional disease progression (n = 61 patients; P = 0.0093).

Download figure to PowerPoint

Treatment was tolerated fairly well. Acute complications during and shortly after RT included moist desquamation in 39 patients (63.9%), pneumonitis in 4 patients (6.6%), delayed healing that required > 2 months in 3 patients (4.9%), and esophagitis in 2 patients (3.3%). No patient died from treatment-related toxicity.

Late effects, which occurred > 6 months after the completion of RT, included arm edema in 21 patients (34.4%). The asymmetry was < 2 cm in 9 patients (14.7%), between 2 cm and 4 cm in 8 patients (13.1%), > 4 cm in 4 patients (6.6%), and rib fracture in 4 patients (6.6%). There were no wound complications that required further surgical intervention, and there were no secondary malignancies.


  1. Top of page
  2. Abstract

In this series of patients with IBC, we report the efficacy of multimodality therapy and identify a number of clinical and treatment-related factors that require special attention for the appropriate diagnosis and optimal treatment of IBC. Multivariate analysis demonstrated that the following three factors were associated with improved survival: smaller pathologic tumor size, local disease control, and up-front surgery. Of particular interest is the latter, which suggests an improved outcome for a sequence of therapy that deviates from the widespread practice of neoadjuvant chemotherapy. The optimal role of surgery in the treatment of IBC is difficult to interpret from the literature. Reports that include both surgery and RT have shown either no survival benefit10–12 or a possible benefit in combination with chemotherapy.13–17 Given favorable reported outcomes for alternating chemotherapy and RT,18–23 and given similar outcomes when either RT or surgery follows chemotherapy,24, 25 even the necessity of surgery has been questioned.26 Meanwhile, the timing of surgery has not been defined. At the University of Florida, several patients underwent mastectomy at the time of initial presentation to decrease tumor burden expeditiously. For reasons that are not entirely clear, those patients had a better outcome compared with patients who were treated initially with chemotherapy, as noted on previous analysis.5 It is noteworthy that this difference could be demonstrated only on multivariate analysis, which controlled for potentially confounding factors. It is conceivable that leaving a large source of tumor cells in place during neoadjuvant chemotherapy actually permits the seeding and growth of more metastases in the face of chemotherapy, such that initial surgery is more effective. Although the addition of chemotherapy certainly has improved treatment outcomes overall for patients with breast carcinoma, it is important to note that, in patients with IBC and with locally advanced breast carcinoma, chemotherapy is only moderately effective in eliminating distant disease, because ≥ 50% of patients succumb to breast carcinoma with distant metastases. It also is important to note that randomized trials that have analyzed the potential benefit of neoadjuvant chemotherapy have failed to show a decided survival advantage27 and have suggested a slight decrease in local disease control with a delay of locoregional treatment.

The number of prognostic factors analyzed in the current series may be larger than any other study has reported. The issue of pathologic dermal lymphatic invasion merits special mention. Uncertainty exists regarding the significance of a pathologically based diagnosis of IBC. Because of sampling heterogeneity, dermal lymphatic invasion may not be evident in all cases; it is estimated that it can be identified in only up to 75% of patients.28 Consequently, the diagnosis of IBC is permitted based on clinical criteria alone. In turn, the classification of disease in which dermal lymphatic invasion is observed in the absence of clinical features (occult IBC) is not clear. Many report outcomes similar to the outcomes reported in patients with classic IBC,13, 28–31 although some have described occult IBC as a more benign process,32, 33 which may reflect the notion that noninflammatory breast carcinomas (e.g., locally advanced breast carcinoma) can have incidental tumor emboli in the dermal lymphatics.34 In the current series, roughly 70% of patients had documented pathologic dermal lymphatic invasion, and 18% of patients had occult IBC. There was no detectable difference in outcome for patients with dermal lymphatic invasion or clinically occult IBC. Based on these results, no recommendations can be made to treat these patients differently from the currently accepted standard of care for patients with IBC.

The presumed benefit of trimodality therapy is both improved local disease control and survival. It is intuitive that successful local control would be associated with improved survival, because local failures generally are not salvageable. The 20% local failure rate in all patients who received surgery and RT indicates room for improvement. With the exception of type of chemotherapy (which likely was confounded by imbalances in pathologic tumor size and delay of effective locoregional treatment), RT dose was the only other factor that was found to be subject to treatment modification. Hyperfractionated RT was not administered routinely, because it generally was reserved for patients whose response to neoadjuvant chemotherapy was insufficient to render the tumor “operable.” Theoretically, the use of hyperfractionation enables dose escalation and can be considered the radiation counterpart of “dose density” chemotherapy. Intuitively, hyperfractionated irradiation should be effective, particularly in rapidly proliferating tumors, such as classic IBC. An early report from the M. D. Anderson Cancer Center suggested potential benefits35 that were confirmed in a later study,36 leading to improved local control. IBC patients at our institution currently are treated with hyperfractionated RT up to doses of 66 Gy.

Because of the low incidence of IBC compared with other invasive breast carcinomas, prospective controlled trials to define optimal management strategies are not feasible. Thus, retrospective studies are essential to continued progress in the management of this disease. In part because of limited numbers for analysis and patient heterogeneity, some caution must be exercised in the interpretation of their results. The findings of this series compare favorably with other published reports. Table 4 includes series that have reported overall survival for > 30 patients who were treated with trimodality therapy for IBC. Overall survival was reported to be between 34–75% at 5 years.13, 17, 36–39 In the literature, several prognostic factors have been associated with better survival in patients with IBC, including better response to neoadjuvant chemotherapy,10, 19, 20, 38, 40–43 early response to neoadjuvant chemotherapy,23 more chemotherapeutic agents,23 no or limited clinical or pathologic lymph node involvement,10, 12, 19, 23, 38, 44, 45 premenopausal status10 and younger age,10 positive hormone receptor status,13, 38, 46 the absence of peau d'orange45 or diffuse erythematous involvement19, 20, 23, 42 at presentation, any palpable mass at presentation,12 smaller pathologic tumor size,38 the absence of distant metastases at presentation,42 the inclusion of mastectomy as part of treatment,13, 47 negative surgical margin status,44 local control,13 and higher RT dose.36 Both the patient heterogeneity and the small numbers in each of the studies most likely have led to the lack of uniformly reported prognostic factors. For these reasons, the current series also was unable to confirm improved survival with any reported factors other than pathologic tumor size and freedom from locoregional disease progression.

Table 4. Outcomes with Trimodality Therapy
StudyNo. of patients treated with three modalitiesPredominant sequence of therapyOverall survival at 5 yrs (%)
  • Chemo: chemotherapy; RT: radiation therapy.

  • a

    For n = 43 patients.

  • b

    For n = 46 patients (actual numbers treated with 3 modalities were not reported).

  • c

    Cause-specific survival was estimated from graphical data.

  • d

    For n = 115 patients.

  • e

    For n = 54 patients.

Maloisel et al., 19903838Chemo [RIGHTWARDS ARROW] surgery [RIGHTWARDS ARROW] chemo [RIGHTWARDS ARROW] RT [RIGHTWARDS ARROW] chemo75.0a
Pisansky et al., 19923736Chemo/RT [RIGHTWARDS ARROW] surgery [RIGHTWARDS ARROW] chemo34.0
Pierce et al., 199217< 46Chemo [RIGHTWARDS ARROW] surgery [RIGHTWARDS ARROW] RT [RIGHTWARDS ARROW] chemo36.0b
Perez et al., 19941386Chemo [RIGHTWARDS ARROW] surgery [RIGHTWARDS ARROW] RT [RIGHTWARDS ARROW] chemo40.0c
Liao et al., 200036100Chemo [RIGHTWARDS ARROW] surgery [RIGHTWARDS ARROW] chemo [RIGHTWARDS ARROW] RT40.5d
Harris et al., 20033950Chemo [RIGHTWARDS ARROW] RT [RIGHTWARDS ARROW] surgery [RIGHTWARDS ARROW] chemo56.0e
Current series56Chemo [RIGHTWARDS ARROW] surgery [RIGHTWARDS ARROW] RT [RIGHTWARDS ARROW] chemo48.0


The treatment of patients who have IBC with trimodality therapy has led to significant improvements in survival over the past 20 years. Further gains should be targeted toward both improved local control, by investigating the optimal timing of surgery and optimal administration of RT, and improved systemic control, by incorporating the use of new or higher doses of chemotherapeutic agents.


  1. Top of page
  2. Abstract