Patient Eligibility

Patients with measurable metastatic breast cancer that had progressed after or failed to respond to prior therapy with an anthracycline and a taxane were eligible for this study. Patients were also required to have received at least one therapy for metastatic breast cancer. Patients whose tumors expressed estrogen and/or progesterone receptors or HER2-neu were required to have received prior therapy with a hormonal agent and trastuzumab, respectively. Patients with brain metastases were eligible if the brain metastases were treated and remained stable for a minimum of 4 weeks. All patients had a human leukocyte antigen–matched sibling who consented to serve as an allogeneic blood stem-cell donor. This study, CC 00-C-0119, was approved by the NCI institutional review board, and informed written consent was obtained from each patient and her respective donor.

Treatments

Patients initially received one to two cycles of an induction regimen consisting of cyclophosphamide (600 mg/m²/d) and fludarabine (30 mg/m²/d) concomitantly for 4 consecutive days (Fig 1). The purposes of the induction regimen were to deplete host immunity and to provide tumor control before transplantation. The induction regimen was administered in 21-day cycles. Patients received a maximum of two cycles of the induction regimen until the CD4⁺ count was below 50 cells/μL. This level of host CD4⁺ T-cell depletion was selected as it is associated with clinically significant immunosuppression and a reduced ability to reject allogeneic cells. If after one cycle of the induction regimen, the target CD4⁺ count was reached or there was radiographic evidence of disease progression, patients proceeded directly to transplant. All of the remaining patients proceeded to transplant after two cycles of the induction regimen, regardless of whether or not the target CD4⁺ count had been reached. Patients proceeded to transplantation within 2 weeks of completing (day 21) their final cycle of the induction regimen. Patients then received a conditioning regimen consisting of fludarabine (30 mg/m²/d) and cyclophosphamide (1,200 mg/m²/d) administered intravenously and concomitantly on days −6, −5, −4, and −3 prior to transplantation.

Hematopoietic stem cells were mobilized in donors with filgrastim (8 μg/kg bid), and apheresis was performed starting on the day 5 of filgrastim administration using the Fenwal CS3000 Plus cell separator (Baxter Healthcare Corp, Deerfield, IL). Daily aphereses were performed until a minimum of 10 × 10⁶ CD34⁺ cells/kg (recipient weight) was collected. Cells were T-cell depleted with CD34-positive selection and T-cell negative–selection with a panel of monoclonal antibodies against CD2, CD6, and CD7 using a modified Isolex 300i magnetic cell selection device (Nexell Therapeutics Inc, Irvine, CA). This process resulted in a consistent 4 to 5 log depletion of T cells. T-cell doses were then adjusted so that each allograft uniformly contained 1 × 10⁵ CD3⁺ cells/kg. All allografts contained a minimum of 5 × 10⁶ CD34⁺ cells/kg. All apheresis products were cryopreserved in Plasmalyte A (Baxter Healthcare Corp) with 4% human serum albumin, 5% dimethyl sulfoxide (Research Industries, Salt Lake City, UT), and 6% Pentastarch (B. Braun, Irvine, CA), and stored in liquid nitrogen. Allografts were thawed and infused on day 0. Filgrastim was administered daily at 10 μg/kg from the day of allogeneic HSCT until the absolute neutrophil count was greater than 5,000 cells/μL for 3 consecutive days. All patients received fluconazole and acyclovir daily and trimethoprim-sulfamethoxazole on weekends for at least 6 months after transplant as infection prophylaxis. To prevent graft rejection and graft-versus-host disease (GVHD), patients received cyclosporine starting on the day before transplantation (day −1) and maintained at a level of approximately 200 μg/L through day +28 post–allogeneic HSCT. Cyclosporine was subsequently tapered to discontinuation by day +40 to permit a full GVT effect.

Starting on day +42 post-transplantation, patients were scheduled to receive dose-escalated, donor lymphocyte infusions (DLI). The DLI were scheduled for days +42, +70, and +98 post-transplantation, and contained 1 × 10⁶, 5 × 10⁶, and 1 × 10⁷ CD3⁺ cells/kg, respectively. The DLI were only administered if there was persistent disease and no evidence of GVHD. Patients were eligible to receive additional DLI beyond day +98, using the same criteria as the scheduled DLI. No additional therapy was permitted until 6 months after transplantation, with the exception of DLI.

Assessment of Response

Determinations of response were made by computed tomography (CT) scan after induction chemotherapy, at days +28, +42, +70, and +98 post-transplantation, and months +4, +5, +6, +9, and +12 post-transplantation. CT measurements were made independently by a National Institutes of Health Clinical Center radiologist (C.K.C.) not involved in patient care. A complete response was defined as complete resolution of all measurable and assessable disease for at least 28 days. A partial response was defined as a greater than 50% decrease in the sum of the products of the greatest (or longest) perpendicular dimensions of all measured lesions for at least 28 days, without evidence of new lesions. A minor response was defined as a greater than 25% but less than 50% decrease in all measured lesions for at least 28 days, without evidence of new lesions. Stable disease was defined as a less than 25% decrease or a no greater than 25% increase in all measured lesions for at least 28 days without evidence of new lesions. Progressive disease was defined as a greater than 25% increase in measured lesions or evidence of new lesions.

Adverse Events

Toxicities were defined using the NCI Common Toxicity Criteria Version 2.0. The severity of acute and chronic GVHD was graded according to Glucksberg criteria and International Bone Marrow Transplant Registry consensus criteria, respectively.^23,24

Assessment of Chimerism

Chimerism analysis was performed by the variable number tandem repeats–polymerase chain reaction method in a Clinical Laboratory Improvement Amendments–certified laboratory at the Blood Center of Southeastern Wisconsin. Chimerism was determined on total peripheral blood mononuclear cells every 2 weeks until 6 months post-transplantation, and then monthly thereafter. In addition, chimerism was determined intermittently post-transplant on samples enriched for T-lymphoid (CD3⁺) subsets. Lymphoid cell subset enrichment was performed by positive selection using either magnetic beads (Miltenyi Inc, Auburn, CA) or rosette technique (Stem Cell Technologies Inc, Vancouver, BC).

Statistical Analysis

The objective of this study was to determine if the administration of allogeneic lymphocytes remote from the chemotherapy in the transplant conditioning regimen would result in clinical responses in patients with advanced metastatic breast cancer. Since there was no prior experience as to what the response rate would be with this therapy, a predicted response rate of 30% was arbitrarily selected as a target, as similar response rates are observed with second-line chemotherapy in metastatic breast cancer patients who have received prior anthracyclines and taxanes.^4,25 The null hypothesis was that the rate of response in this group would be 10%, and the alternative hypothesis was that the response rate would be 30%. Using a one-tailed α = .05, and with more than 80% power, it was estimated that 21 patients who engrafted after transplantation would provide an adequate sample size to reject the null hypothesis in favor of the alternative hypothesis.²⁶

Overall survival durations were calculated from transplant date until date of death or last follow-up. The probability of overall survival as a function of time was calculated using the Kaplan-Meier method.

Patient Characteristics

Twenty-two patients were enrolled onto the study between May 2000 and April 2003. Six patients did not proceed to transplantation; four patients experienced rapid disease progression before receiving any study therapy; one was found to have a second primary malignancy (ovarian adenocarcinoma); and one voluntarily withdrew. The median age of the 16 patients who underwent allogeneic HSCT was 43 years, and median times from original diagnosis and diagnosis of metastatic disease were 67 months and 24 months, respectively (Table 1). All patients had infiltrating ductal carcinoma. The median number of metastatic sites was three; the most common sites of metastatic disease were liver, lungs, and bone. All patients had received adjuvant chemotherapy. Five patients had previously received high-dose chemotherapy with an autologous HSCT. The median number of prior therapies they had received for metastatic disease was four.

Hematopoietic Recovery and Engraftment

The median times from transplant to a neutrophil count of 1,000/mL and a platelet count of 50,000/mL were 10 and 14 days, respectively. All 15 patients, who survived beyond day +14 post-transplantation, had evidence of donor engraftment by variable number tandem repeats–polymerase chain reaction at day +14 post-transplantation. Mixed donor/recipient chimerism was observed in seven patients early (day +28) post-transplantation. Complete donor chimerism in the T-lymphoid compartment was observed in all 15 patients by 6 months post-transplantation after completion of scheduled DLI. There were no graft failures.

Toxicities

All patients completed at least one cycle of induction chemotherapy without significant complication, with the exception of febrile neutropenia, which was documented in 37% of cycles administered. The most common post-transplantation complications was GVHD. Acute GVHD was observed in 10 patients. Nine patients required treatment with systemic corticosteroids for grade 2 to 4 acute GVHD, and eight of these nine had complete resolution of clinical GVHD. Four of 13 assessable patients developed chronic GVHD that was extensive in two cases. Other serious adverse events included one case of Epstein-Barr virus–associated lymphproliferative disorder, which completely resolved after planned discontinuation of cyclosporine and the first scheduled DLI. There was one case of Gram-negative sepsis, one case of systemic aspergillosis, and one case of congestive heart failure, which were all successfully treated. One patient died early post-transplantation (day +3) from multiorgan failure. One patient died 6 months after transplantation from hemorrhage during a thoracentesis procedure to drain a malignant pleural effusion.

DLI

Three patients did not receive any of the scheduled DLI, as they all had developed GVHD before day +42. Eight patients received all three scheduled DLI. Seven of these patients received additional DLI (range, four to 11) for persistent disease. Four patients received either one or two of the scheduled DLI. Three of these patients developed GVHD before their next scheduled DLI. The other patient experienced disease progression and physical decline requiring removal from study before next scheduled DLI.

Efficacy

Following induction chemotherapy, two patients were documented to have a partial response, three patients had a minor response, six patients had stable disease, and six patients had progressive disease as compared with measurements at study entry. Response determination at day +28 post-transplantation for the 15 surviving patients demonstrated that the conditioning regimen improved or stabilized disease in five and nine patients, respectively, and one patient experienced further progression. In sum, the induction chemotherapy and the conditioning regimen chemotherapy resulted in four partial responses and two minor responses at day +28 evaluation as compared with measurements at study entry.

Using the day +28 tumor measurements as a new baseline, six patients were subsequently observed to have tumor regression (Table 2). There were two partial responses and four minor responses among these six patients. These responses were observed 42 days to 13 months post-transplantation. The median duration of these six responses was 3 months (range, 1 to 7 months). One response occurred after planned cyclosporine taper (patient 12); the other five responses occurred after DLI. There were no responses among patients who were in a state of mixed T-lymphoid chimerism after day +42; all responses occurred after the establishment of complete donor T-lymphoid chimerism.

In two cases (patients 2 and 7), tumor progression was observed at day +98 and day +70, respectively, before a subsequent tumor regression. One patient (patient 13), who achieved a partial response at day +28, had a subsequent 24% increase in tumor size at day +42. She was subsequently observed to have a 41% decrease in tumor at day +70. In the other three cases, disease had not responded to the conditioning regimen and remained stable before the observed regression.

The case (patient 7) that best exemplifies these responses is outlined in Figure 2. At day +28 post-transplantation, the patient’s disease was stable, and donor T-lymphoid chimerism was below 30%. Subsequent planned DLI on days +42 and +70 increased overall donor chimerism to greater than 80%; however, she was documented to have new hepatic lesions and ascites on day +98 evaluation. On day +126 evaluation, regression in the patient’s liver lesion and ascites was observed. As she had no clinical evidence of GVHD and persistent disease, an additional DLI was given. The patient subsequently experienced an approximate 90% reduction in liver metastases as compared with day +70 measurements when she was noted to have progressive disease. At approximately 6 months post-transplantation, she developed GVHD of the skin and gastrointestinal tract that required treatment with systemic corticosteroids. She was subsequently documented to have progressive disease at 8 months post-transplantation. Similarly, in four of the other five patients, tumor regression occurred before the onset of clinical GVHD, and tumor progression occurred within 2 months of starting systemic corticosteroid administration for treatment of GVHD. One response occurred without clinical evidence of GVHD.

Patient characteristics were analyzed in an attempt to identify factors associated with tumor regression observed beyond day +28. There was no association of receptor status, number, sites or size of metastatic disease, human leukocyte antigen type, cytomegalovirus status, or number of prior therapies. Similarly there was no correlation with response to either induction or conditioning chemotherapy.

Overall, six partial responses and four minor responses were observed among the sixteen patients who underwent transplantation. Using the most conservative evaluation, with 16 total patients, six responses (37.5%) have an associated lower bound of a one-sided 95% CI of 17.8%. Based on 15 patients, who survived beyond day +14 with engraftment of donor cells, the six responses (27%) have an associated lower 95% confidence bound of 19.1%. In either case, an approximately 35% to 40% partial response rate was obtained, and there is statistical evidence to reject the 10% null rate. In addition to the two patient deaths described above, there have been 10 additional deaths, all with progressive disease. For all 16 patients who underwent allogeneic HSCT, the median overall survival from time of transplant was 10.3 months, at a median potential follow-up of 23.4 months.