• gene therapy;
  • p53;
  • breast cancer;
  • primary systemic therapy


  1. Top of page
  2. Abstract
  6. Acknowledgements


Primary systemic therapy (PST) is the standard approach to the management of patients with locally advanced breast cancer (LABC). The authors hypothesized that the intratumoral administration of a nonreplicating adenoviral vector (Ad5) that contains the human wild-type p53, AdCMV-p53, combined with chemotherapy, could increase the efficacy of PST as measured by pathologic complete response.


In a prospective, open-label, Phase II trial, 13 patients with LABC were treated with 6 3-week cycles of PST, which consisted of intratumoral injections of Ad5CMV-p53 for 2 consecutive days plus docetaxel and doxorubicin followed by surgery. p53 status was determined at baseline and was assessed immediately after the first injection (up to 48 hours). Clinical response was assessed by clinical and radiologic methods.


The trial was terminated early, because none of the patients achieved a pathologic complete response. The median age was 56 years (range, 39–71 years), and the median tumor size was 8 cm (range, 5–11 cm). Eight patients (73%) had a p53 mutation. Serial biopsies showed an increase in p53 messenger RNA (mRNA) and p21WAF1/Cip1 mRNA. All 12 evaluable patients achieved an objective clinical response. The surgical specimens revealed scattered tumor cells with extensive tumor-infiltrate leukocytes (predominantly T-lymphocytes). At a median follow-up of 37 months (range, 30–41 months), 4 patients (30%) developed systemic recurrence, and 2 patients died. The estimate breast cancer-specific survival rate at 3 years was 84% (95% confidence interval, 65.7–100%). There was no increase in systemic toxicity.


Ad5CMV-p53 combined with PST is safe, active, and associated with local immunomodulatory effects. The promising clinical activity of this combination deserves further investigation in randomized studies. Cancer 2006. © 2006 American Cancer Society.

Primary systemic therapy (PST) represents the use of cytotoxic chemotherapy as the first modality of treatment for a primary malignant tumor. This modality is an important component of the multidisciplinary approach to the management of locally advanced breast cancer (LABC).1–3 Complete pathologic response (pCR) of the primary breast lesion and regional lymph node metastases after PST, usually obtained in only 5% to 20% of patients, represents the most important prognostic factor.2, 3

Alterations in the p53 gene have been documented with high frequency in primary breast cancer, particularly in the setting of locally advanced disease.4 The majority of studies suggest that the of presence of p53 dysfunction correlates with more aggressive tumors, reduced chemosensitivity, early metastasis, and decreased survival rates.4–8 The reintroduction of wild-type (wt) p53 in preclinical models by using gene-replacement strategies has been associated with increased drug sensitivity, particularly to DNA-damaging agents, and, most recently, taxanes.9–13 These observations suggested that the manipulation of p53 may increase the efficacy of standard cytotoxic therapies (particularly anthracyclines and docetaxel) as PST.14

Gene therapy as a treatment modality involves transferring nucleic acids into target cells for the purpose of perturbing or correcting pathophysiologic processes.15 Significant research in this area has identified nonreplicating adenoviral vectors as the most reliable gene-delivery systems. AdCMV-p53 (ADVEXIN®; Introgen Therapeutics Inc., Houston, TX) is constituted of a nonreplicating adenoviral vector (Ad5) that contains the wt p53 transgene in an appropriate expression cassette.15 Preclinical reports have proven that adenovirus-mediated wt-p53 (Ad-p53) protein expression can be used to induce apoptosis and kill tumors, and clinical trials have demonstrated low toxicity, successful gene transfer, and evidence of tumor regression.16–21 We tested the safety, efficacy, and biologic activity of the combination of doxorubicin and docetaxel with intratumoral injection of AdCMV-p53 in patients with newly diagnosed LABC. We completed a series of correlative studies to evaluate gene-replacement efficiency by measuring activated p53 through measurement of levels of p53 messenger RNA (mRNA) and p21WAF1/Cip1 mRNA (baseline, postinjection).22 Moreover, we performed an exploratory assessment of the local immunologic effect associated with the injection of AdCMV-p53.23


  1. Top of page
  2. Abstract
  6. Acknowledgements

Patient Population

A prospective, open-label, Phase II study was conducted at The University of Texas M. D. Anderson Cancer Center between April 2002 and March 2003. Patient eligibility criteria included newly diagnosed, noninflammatory LABC (Stage IIIA-C) with histologic confirmation of invasive disease. Patients without a palpable breast mass and only lymph node disease were excluded. All patients were evaluated by a multidisciplinary team (medical oncologist, breast surgeon, radiation oncologist, and radiologist). Locoregional extension of disease was assessed by mammography and ultrasonography of the breast and locoregional lymph nodes and, when feasible, by magnetic resonance imaging. Patients were required to have adequate organ function and no evidence of distant metastasis. Standard immunohistochemical staining of the diagnostic core-needle biopsies was performed by using the modified avidin-biotin complex method in a DAKO autostainer (DAKO, Carpinteria, CA) using primary antibodies against estrogen receptor α (ERα) (ER clone 6F11; Novocastra; 1:50 dilution) and progesterone receptor (PR) (PR Ab-9, clone 1A6; Neomarker/Labvision; 1:30 dilution). Fluorescence in situ hybridization (FISH) also was performed to detect gene amplification of HER2/neu using the Path Vysion HER2 DNA probe kit.

This study was approved by the recombinant DNA Advisory Committee of the National Institutes of Health and by the Institutional Biosafety Committee and the Institutional Review Board at The University of Texas M. D. Anderson Cancer Center. All patients provided written informed consent before enrollment.

Treatment Schema

The initial treatment plan consisted of 4 to 6 3-week cycles. After the first patient completed treatment, taking into consideration the previous experience in LABC using the same chemotherapy regimen, we decided to administered 6 cycles to all subsequent patients. Each cycle consisted of the following: 1) Intratumoral injection of Ad5CMV-p53 was administered at a dose of 2.5 × 1012 viral particles on Days 1 and 2 (starting immediately before chemotherapy administration on Day 1).

2) The local injection of Ad5CMV-p53 consisted of the following steps: a) Patients had 2.5 to 5.0 grams of lidocaine/prilocaine cream (EMLA; AstraZeneca Pharmaceutical LP, Wilmington, DE) applied to the injection site at least 30 minutes before the procedure for local anesthesia. b) The procedure was performed in a negative-pressure room, and the area of skin to be injected was cleansed with betadine. After more lidocaine (1%) was applied to the skin, Ad5CMV-p53 was injected with a fine needle directly into the breast lesion(s) at multiple sites in equally divided doses approximately 1 cm apart in 3 dimensions to cover the entire lesion(s) by following a predetermined grid. Attention was paid to adequately cover tumor margins and to go from outer areas to inner areas (Fig. 1). c) No isolation was required after injections, but the patients were instructed to practice good hygiene after voiding, coughing, or sneezing and to avoid contact with former tissue-transplantation recipients and individuals known to have severe immunodeficiency disease (either congenital or acquired) during treatment and within the 28 days after the last Ad5CMV-p53 dose.

thumbnail image

Figure 1. This image demonstrates the technique of administration of a nonreplicating adenoviral vector (Ad5) containing the wild-type p53 transgene in an appropriate expression cassette (Ad5CMV-p53). The area to be injected was cleaned and marked for accurate delimitation of breast lesion.

Download figure to PowerPoint

3) Chemotherapy on Day 1 was administered according to the following schedule: Doxorubicin 50 mg/m2 as a 15-minute, intravenous infusion was followed immediately by a 1-hour, intravenous infusion of docetaxel 75 mg/m2. 4) After the completion of induction therapy, patients who demonstrated a clinical response underwent surgery (mastectomy or lumpectomy is selected patients with axillary dissection) followed by radiotherapy and adjuvant hormone treatment (in patients with hormone receptor-positive disease). Surgery usually was performed 4 to 6 weeks after completion of the last cycle of chemotherapy.

Correlative Studies

Core-needle biopsies of the primary tumor were performed before starting therapy. These tissues were used to obtain a histologic diagnosis, to evaluate standard prognostic factors, and to assess baseline p53 mutational status (BRT Laboratories, Baltimore, MD) using the Affymetrix® p53Genechip® (Affymetrix, Santa Clara, CA) for direct DNA sequencing. Patients also underwent serial fine-needle aspiration of the primary tumor to collect specimens 1) before the initiation of treatment, 2) 24 hours after the first intratumoral injection and chemotherapy administration, 3) 48 hours after chemotherapy and injection of Ad5CMV-p53 (24 hours after the second dose), and 4) before initiation of the second cycle (Day 22). These samples were used for real-time reverse transcriptase-polymerase chain reaction (RT-PCR) studies to evaluate changes in p53 and p21WAF1/Cip1 mRNA levels. Real-time RT-PCR analysis was performed using the commercially available gene expression assays for p53, p21, and cyclophilin without multiplexing (Hs 00153340 mL, Hs 00355782 mL, 4326316E; Applied Biosystems, Foster City, CA). The 7900 sequence detection system 2.2 software was used to determine the fold changes in p53 and p21 mRNA expression in patient samples relative to Stratagene's Universal Human Reference RNA (Stratagene, LaJolla, CA). In addition, for the characterization of tumor-infiltrating leucocytes (TILs), lymphocyte antibodies were used against CD3 (antihuman CD3, clone F7.2.38; Dako Cytomation; 1:700 dilution), CD20 (antihuman CD20, clone L26; Dako Cytomation; 1:7:00 dilution), CD4 (CD4 mouse monoclonal antibody, clone 4812; Novocastra Laboratories; 1:25 dilution), and CD8 (antihuman CD8, clone C8/144B; Dako Cytomation; 1:50 dilution). The number of positively stained cells was expressed as a percentage of the entire lymphoid population.

The final surgical specimen was evaluated for residual disease. The mastectomy specimen was sampled systematically to map the amount of residual tumor. Based on histopathologic findings in the tissue sections, the greatest dimension of residual tumor was calculated.

Serum samples that ere collected before the first 2 cycles of therapy were analyzed by Institut Armand-Frappier (Quebec, Canada). P53 autoantibodies were measured by using a standard enzyme-linked immunosorbent assay (ELISA) with purified wt-p53 bound to microtiter plates (anti-p53 ELISA kit; Phoenix Bio-Tech Corp., catalog no. DIA0302E). Total antiadenovirus immunoglobulin (Ig) class antibody detection was performed using the Adenovirus IgG ELISA kit (Phoenix Bio-Tech Corp., catalog no. RE56571). Neutralizing Ad-5 antibodies were detected by using a standard bioassay. Serum samples were incubated in the presence of predetermined amounts of virus at 37°C for 60 minutes; then, the mixture was placed onto virus-susceptible 293 monolayers. After 6 days, the neutralizing titer was evaluated based on observed cytopathic effect.

Response Criteria

The primary objective of this study was to evaluate efficacy. Clinical assessment of tumor size was performed using bidimensional criteria (the product of the greatest dimension and its perpendicular dimension applied at the widest portion of the tumor) according to World Health Organization criteria. A complete clinical response was defined as the disappearance of all clinical evidence of active tumor by clinical evaluation and imaging studies. A pCR was defined as no evidence of residual invasive tumor in the breast and axillary lymph nodes. A clinical partial response (PR) was defined as decrease ≥50% in the sum of the product of the greatest dimensions of measured lesions for a minimum of 4 weeks. Every measured lesion needed to regress to qualify as a PR and had be without the appearance of new lesions. No change or stable disease (SD) indicated no change in tumor size, a decrease <50% in tumor size, or an increase <25% in tumor size. Progressive disease (PD) was reported as an increase ≥25% in any measurable lesion or the appearance of new lesions. All adverse events that were encountered during the study were evaluated according to the National Cancer Institute Common Toxicity Criteria (version 3.0) grading system.

Statistical Analysis

This study was designed as a feasibility, safety, and efficacy investigation. The efficacy endpoint of this trial was the achievement of pCR. This trial was designed to enroll a maximum of 60 patients to have 82% power to detect a pCR rate of 30% and to reject a pCR rate ≤15% with a Type I error rate of 5%. Because this was the first experience in primary breast cancer, and there were concerns regarding feasibility and safety, strict monitoring was built into the protocol. Interim monitoring was to be performed after the pathologic findings were known for every 12th patient, and a recommendation would be made to stop the trial early if there was a probability <20% that the pCR rate was ≥15%.

Patient characteristics were tabulated and continuous variables were described by their medians and ranges. The levels of p53 and p21WAF1/Cip1 mRNA over the course of treatment were assessed visually with box plots. The associations between TILs (and lymphocyte subpopulations) and residual disease were assessed visually with plots that were fit with a smoothing line and quantified with Spearman rank-sum correlations.24 Recurrence-free survival (RFS) was measured from the start of treatment to the appearance of disease recurrence. Overall survival (OS) was measured from the start of treatment to death. The median follow-up was calculated as the median observation time among all patients. RFS and OS were estimated by the Kaplan–Meier method.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Clinical Efficacy

In total, 13 patients were enrolled and treated (Table 1). Nine patients (69%) had inoperable breast cancer, which was defined as Stage IIIB or Stage IIIC disease. Twelve patients were evaluable for response. One patient was considered ineligible because, at the time of follow-up multidisciplinary assessment, she was classified with inflammatory breast cancer. The trial was terminated at the first stage, because none of the 12 evaluable patients achieved a pCR (as defined in the trial) (Table 2). Nevertheless, all evaluable patients achieved a clinical PR (100%) and underwent surgery (mainly mastectomy; 92%). The radiologic assessment of response demonstrated a 79% median reduced tumor volume (range, 56–93%). Moreover, lymph node disease showed a median reduced size of 62% (range, 38–93%). The breast surgical specimens demonstrated extensive fibrosis and no macroscopic disease (Fig. 2A). The microscopic examination showed scattered residual tumor cells in all breast specimens and extensive lymphocytic infiltrate in all samples (Fig. 2B). The largest single residual, invasive component was ≤1 cm in 8 patients (67%). Furthermore, 1 patient had residual breast disease that measured 1.3 mm and a cluster of malignant cell in 1 lymph node, an additional patient had a residual breast disease that measured 0.3 mm and negative axillary lymph nodes, and another patient had residual ductal carcinoma in situ that measured <1 cm with evidence of dermal lymphatic invasion (original T4 disease). Two patients (17%) had pathologically negative lymph nodes.

thumbnail image

Figure 2. On macroscopic examination, the surgical specimens showed irregularly distributed, large areas of fibrosis without any evidence of a discrete tumor mass. (A) On microscopic examination, there was evidence of scattered, isolated tumor cells or clusters (B) (original magnification × 200).

Download figure to PowerPoint

Table 1. Patient Demographics
CharacteristicNo. of patients (%)
  • ER indicates estrogen receptor; PR, progesterone receptor; IHC, immunohistochemistry; FISH, fluorescent in situ hybridization.

  • *

    The p53 analysis was performed in 11 patients.

  • Values shown are the greatest maximum dimension.

  • Disease stage was determined according to the American Joint Committee on Cancer Staging Manual, 6th edition, 2004. All patients had palpable lymph nodes, sonographic assessment, and cytologic confirmation of metastatic involvement.

Total no. of patients13
Median age (range), y56 (39–71)
 White11 (85)
 Black1 (7.5)
 Asian1 (7.5)
 ER-positive and PR-positive7 (54)
 ER-negative and PR-negative6 (46)
HER-2 (IHC and/or FISH) 
 HER-2 (3+, and FISH+)5 (38)
 HER-2- (0, 1, 2, FISH−)8 (62)
p53 mutation*
 Yes8 (73)
 No3 (27)
Median primary breast lesion size (range), cm8 (5–11)
Clinical stage
 Stage IIIA4 (31)
 Stage IIIB, IIIC9 (69)
Table 2. Summary of Clinical Efficacy
CharacteristicNo. of patients (%)
  1. CR indicates complete clinical response; PR, partial clinical response; SD, stable disease; PD, progressive disease; 95% CI, 95% confidence interval.

No. of evaluable patients12
Median no. of cycles treated (range)6 (4–6)
Breast clinical response (%)
 CR0 (0)
 PR12 (100)
 SD0 (0)
 PD0 (0)
Pathologic residual breast disease (size range)
 No. with residual breast disease ≤1 cm8 (0.1–1.0 cm)
 No. with residual breast disease >1 cm4 (2.0–6.0 cm)
Pathologic lymph node status (%)
 No. with 0 lymph nodes2 (17)
 No. with 1–3 lymph nodes4 (33)
 No. with >4 lymph nodes6 (50)
Median follow-up (range), mo37 (13–41)
Estimated 3-year overall survival (95% CI), %77 (57–100)
Estimated 3-year breast cancer-specific survival (95% CI), %84 (67.5–100)

The median follow-up was 37 months (range, 13–41 months). Among all treated patients (n = 13 patients), there were 4 recurrences (30%), and 3 patients died (23%). The 3-year RFS rate was 68.4% (95% confidence interval [95% CI], 46.9–99.7%), and the estimate of OS at 3 years was 77% (95% CI, 57–100%) (Fig. 3A). All 3 patients who died initially had inoperable disease (Stage IIIC) and p53 mutations (1 patient had inversion at intron 4). They all achieved a PR with treatment, and 2 of them had visceral recurrences at 11 months and 7 months, respectively. It is noteworthy that both of those patients had less extensive TILs in their final surgical specimens compared with the other patients' specimens. Furthermore, the third patient developed a second primary tumor (nonsmall cell lung cancer) 12 months she underwent breast surgery and subsequently died of it without evidence of recurrent breast cancer. The estimated breast cancer-specific survival rate at 3 years was 84% (95% CI, 65.7–100%) (Fig. 3B).

thumbnail image

Figure 3. These Kaplan-Meier plots illustrate (A) estimated overall survival at 3 years and (B) estimated breast cancer-specific survival at 3 years.

Download figure to PowerPoint

Biologic Correlative Studies

Analysis of p53 status was performed in 11 patients, with mutations documented in 8 of 11 patients (73%). All but 1 of those mutations consisted of missense, single-base mutations that involved mostly exons 5 through 8. In detail, there were 3 mutations at exon 6, 3 mutations at exon 8, 1 mutation at exon 3, and 1 mutation at intron 4. Mutations were at codons 183, 192, 193, 196, 278, 281, and 282. One patient had an inversion at the splice site (IVS4-1G [RIGHTWARDS ARROW] C).

Adequate RNA was recovered in fine-needle aspirates from the primary tumor in 7 patients to evaluate the levels of p53 mRNA and p21WAF1/Cip1 mRNA. All of those patients had an increase in p53 mRNA after the first injection (Day 2). The p21WAF1/Cip1 mRNA levels also were increased, but to a lesser degree, in all but 2 patients (Fig. 4A). The increased level of p53 mRNA expression persisted up to 21 days in all but 1 patient, who subsequently developed a recurrence and died.

thumbnail image

Figure 4. These box plots show P53 messenger RNA (mRNA) and p21WAF1/Cip1 measurement on the log scale (y-axis). P53 mRNA measurements tended to increase over time: 1) at baseline; 2) 24 hours after the first dose of the nonreplicating adenoviral vector containing the wild-type p53 (AdCMV-p53) and chemotherapy; 3) 48 hours after the first dose; and 4) on or before Day 22 of the 2nd cycle, and the measurements did not appear to depend on p53 mutation status.

Download figure to PowerPoint

The extensive lymphocytic infiltrate within the posttreatment tumor bed was characterized further, and the results demonstrated that consisted predominantly of CD3-positive T-lymphocytes (median CD3 vs. median CD20, 80% vs. 20%, respectively) (Fig. 5A,B). The additional characterization of the T-lymphocyte subpopulation indicated a prevalence of CD8 suppressor/cytotoxic lymphocytes (CD4 vs. CD8, 22.5% vs. 77.5%, respectively) (Fig. 5C). Higher levels of CD3 infiltrate and lower levels of CD20 infiltrate tended to be associated with increased residual disease (Fig. 6A,B). Furthermore, a nonlinear association was noted between the levels of CD4 and CD8 and residual disease. Specifically, lower levels of CD4 (30–40%) and higher levels of CD8 (70–90%) were associated directly (but not significantly) with increased residual disease (Fig. 6C). The extent of residual disease was similar according to the p53 status of the primary tumor (median residual disease, 0.7 vs. 1.00 in p53-negative vs. p53-positive disease, respectively), whereas positive ER status tended to be associated with a greater amount of residual breast disease (median residual disease, 1.5 vs. 0.5 in ER-negative vs. ER-positive disease, respectively).

thumbnail image

Figure 5. Immunohistochemical staining demonstrated (A) large numbers of T lymphocytes with positivity for CD3 and (B) much smaller numbers of CD20-positive B lymphocytes in the peritumoral areas. Further characterization indicated large numbers of CD8-positive cytotoxic T lymphocytes (C) (original magnification × 400).

Download figure to PowerPoint

thumbnail image

Figure 6. (A-C) These are plots of pathologic residual disease (reported in cm) and subpopulation of lymphocytic infiltrate. Each plot also shows a smoothing or best-fit line. High percentages of (A) CD3 and lower (B) CD20 infiltrates were associated with minimal residual disease in the breast. Residual disease was plotted by CD4 and CD8 with a best-fit line. The Spearman rank-sum correlation was − 0.45 between residual disease and CD4 and for CD8 was 0.45 between residual disease and CD8 (C).

Download figure to PowerPoint

Safety Data

P53 serum antibodies (anti-p53 antibodies) were detected in 4 patients (33%) at baseline and did not increase with treatment (Table 3). Anti-Ad5 antibodies were detected in all patients at baseline and increased after treatment in 11 of 12 patients (median, 163%; range, 43–543%). Six patients tested negative for neutralizing Ad5 antibody at baseline, and all 6 of those patients exhibited an apparent increase in titer after treatment. There was no correlation between any increase in antibody titer and the adverse events and 2 serious adverse events, which were not related to AdCMV-p53, including was 1 episode of febrile neutropenia and 1 episode of hypoglycemia and hypotension. Adverse events that possibly or probably were related to AdCMV-p53 treatment were skin inflammation (3 patients with Grade 1), fever (1 patient with Grade 2, 1 patient with Grade 1), fatigue (1 patient with Grade 1), myalgia (1 patient with Grade 2), breast pain (1 patient with Grade 2), anemia (1 patient with Grade 1), and weight loss (1 patient with Grade 1).

Table 3. Serum Antibodies Levels
Patient no.EIA anti-p53 autoantibodiesEIA anti-Ad5 total antibodiesNeutralizing antibodies
p53 index*InterpretationU/mLU/mLInterpretation§Total U/mLTiter anti-Ad5
  • EIA indicates enzyme-linked immunosorbent assay; Ad5 adenoviral vector 5; Pre, preinjection; Post, postinjection.

  • *

    The p53 autoimmune index = E450(sample) − E450 (cut-off)/E450(calibrator) − E450 (cut-off).

  • Serum levels with a p53 index less than or equal to the cut-off point were categorized as negative (−) for anti-p53 antibodies. Serum levels with a p53 index value between the cut-off point and the cut-off point + 20% were categorized as critical. Serum levels with a p53 index greater than the cut-off point + 20% were categorized as positive (+) for anti p53 antibodies. Patients who had critical serum levels were retested.

  • Antibody titercalibration curve(U) × dilutionsample = antibody titersample undiluted(U).

  • §

    Serum samples were considered negative for antiadenovirus antibodies if the concentration value (U/mL) before multiplying the concentration value by the dilution factor was <8 U/mL; when the concentration value was between 8 U/mL and 12 U/mL, the serum was considered borderline for the presence of antiadenovirus antibodies. Serum samples were considered positive for antiadenovirus antibodies when concentration values were >12 U/mL (+ indicates over).

  • If immunoglobulin G concentrations (U/mL) were out of the linear scale of the standard curve, then serum samples were retested at a higher dilution.



  1. Top of page
  2. Abstract
  6. Acknowledgements

Gene therapy is the introduction and expression of recombinant genes in somatic cells for the purpose of treating a disease.15 This approach has been used in multiple clinical trials with different experimental gene therapy approaches for treating a variety of diseases, including cancer, genetic disorders, and atherosclerosis.25, 26 The results of those trials, however, mostly have been disappointing because of the inability to achieve an effective and durable transgene expression by using the currently available vectors, primarily adenoviruses.27–29 The p53 mutations in primary breast cancer have been associated with a worse prognosis; therefore, therapeutic interventions designed to modulate p53 function are particularly important in this disease.5, 30

Our current results indicate that the local injection of AdCMV-p53 is associated with the immediate and effective transfection of the gene, resulting in the synthesis of p53 mRNA. The elevated p53 mRNA persists up to 21 days, indicating that this phenomenon is not transient and chemotherapy-induced but, rather, is prolonged and clearly is related to effective gene transduction.31 Furthermore, the presence of functional p53 concomitant with the administration of chemotherapy is followed by the transient activation of p21WAF1/Cip1.23 These data support the hypothesis that the p53-dependent (chemotherapy-independent) apoptotic pathway is activated functionally in patients who receive AdCMV-p53 and that the efficiency of the p53 transfection and expression is not affected negatively by any immune response that occurs in patients after treatment. Furthermore, the current results indicate that the administration of AdCMV-p53 produces local activation of innate immune response and mild inflammatory changes28 followed by activation of adaptive immunity, as demonstrated by the presence of extensive TILs in all tumors, associated with the recruitment of CD8-positive T-lymphocytes, which typically do not constitute the predominant mononuclear cell infiltrate in primary breast cancer.32, 33 However, such a prominent and consistent lymphocytic infiltration has not been reported in breast tumors treated with conventional chemotherapy. Minimal and focal lymphocytic response can be noted in 25% to 30% of tumors, and marked lymphocytic response can be noted in 5% of tumors treated with chemotherapy alone (unpublished observation). It is noteworthy that the relative distribution of the T-lymphocyte subpopulation appears to predict for residual disease, suggesting a role for local immune activation. Therefore, the relative contribution of this immunomodulatory phenomenon to the overall benefit of the combined treatment is unknown; however, it is reasonable to hypothesize that the antitumor effect observed may be related to 2 main mechanisms that probably act in sequence: The first is the increased, chemotherapy-related, proapoptotic effect as a result of the immediate, effective, and persistent p53 gene transfer. The second is the activation of a local adaptive immune response directed against the transduced and immunogenic cancer cells, which, probably acting through modulation of the microenvironment, eliminate residual cancer cells. This observation provides critical evidence within the peculiar distribution of residual disease in breast specimens that revealed only isolated clusters of cells without organized nodules or masses in all patients treated. Moreover, the significant clinical response (100% objective response), which has not been demonstrated previously in any other neoadjuvant therapy trial for primary disease using chemotherapy alone, seems to support this hypothesis further.2, 34 The evidence of tumor regression in all treated patients also suggests a reduced incidence of primary chemoresistance to anthracyclines, which would be expected with functional p53.6, 13, 31, 34 To explore these findings further and to support our hypothesis, we performed an exploratory clinical and pathologic, retrospective, comparative analysis using a carefully matched (stage and tumor size), single-institution, historic population of 22 patients with LABC who were treated previously on the same neoadjuvant regimen (Table 4). Although we understand that there are limitations to this type of historic comparison, there was a clear indication of significant, improved clinical response with the combination regimen (41% vs. 100%; P = .0006). Moreover, 13 patients (59%) who received chemotherapy alone demonstrated either SD or PD during induction chemotherapy, and 4 of those patients could not undergo definitive surgery (unresectabe rate: chemotherapy vs. combination therapy, 18% vs. 0%, respectively; P = .27). A review of the surgical specimens in the historic control group also indicated that only 1 of those patients (4%) had evidence of extensive TILs.

Table 4. Retrospective Comparison with Locally Advanced Breast Cancer who Received Chemotherapy *
CharacteristicNo. of patients (%)P (Fisher exact test)
Historic groupAdvexin regimen
  • CR indicates complete response; PR, partial response; SD, stable disease; PD, progressive disease.

  •  *

    A retrospective, exploratory analysis was performed of patients who were treated on The University of Texas M. D. Anderson Cancer Center (MDA) protocol MDA ID-97099 for patients with locally advanced breast cancer. The objectives of the analysis were to describe and compare the clinical and pathologic response in size-matched and volume-matched patients who were treated with the combination of docetaxel and doxorubicin versus the group of patients who received the same regimen with the local injection of a nonreplicating adenoviral vector (Ad5) that contains the human wild-type p53 transgene (AdCMD-p53). Twenty-two consecutive patients were collected who had an initial presentation that matched more closely the initial presentation of the 12 patients who received combined-modality treatment. After this initial selection, the following characteristics were reviewed: 1) clinical response; 2) pathologic response in the breast reported as either no invasive disease, or disease that measured <1 cm, or disease that measured >1 cm, or unresectable.

Total no. of patients2212 
Tumor status
 T24 (18)1 (8) 
 T35 (23)4 (33).78
 T413 (59)7 (58) 
Median tumor size (range), cm8 (3–10)8 (5–11) 
Clinical response
 Responder (CR/PR)9 (41)12 (100%) 
 No-response (SD/PD)13 (59)0 (0%).0006
Pathologic response (breast):   Residual disease
 No invasive disease2 (9)0 (0) 
 <1 cm4 (18)8 (67) 
 >1 cm12 (55)4 (33).0240
 Unresectable4 (18)0 (0) 

The current results also indicate that pathologic response, as defined in this trial, probably is not the most appropriate surrogate marker of efficacy; instead, RFS should be considered as a primary endpoint when evaluating this treatment modality. Clearly, the local administration of Ad5CMV-p53 has an impact on the breast pathologic response, the assessment of which depends strongly on the extent of thoroughness of breast specimen sampling. Furthermore, the presence of extensive inflammatory infiltrate, which was not described adequately in previous classifications of pathologic response, makes the evaluation of residual disease cumbersome.35

Finally, it is of foremost importance to highlight that, in the current trial, local events associated with the administration of Ad5CMV-p53 were minimal, and the procedure was tolerated very well. Moreover, there was no evidence of increased systemic side effects of chemotherapy administered in combination with Ad5CMV-p53 despite evidence of systemic immunologic response to the adenoviral vector.

In conclusion, this study represents the first demonstration to our knowledge of a safe and efficacious gene-replacement strategy in patients with breast cancer. These encouraging results provide support for a prospective, randomized study comparing the current regimen with the standard chemotherapeutic approach as primary systemic therapy for patients with LABC.


  1. Top of page
  2. Abstract
  6. Acknowledgements

We thank Prof. Alastair M Thompson (Department of Surgical Oncology, University of Dundee, Ninewells Hospital and Medical School) for the critical review and for advice about article preparation.


  1. Top of page
  2. Abstract
  6. Acknowledgements