Search

Patients in whom the standard of care has failed or who have uncommon tumors for which no standard of care exists are often treated with drugs selected based on the physician's best guess. The rate of success for this method is generally low. With the advent of fast, affordable tumor profiling technologies, and a growth in the understanding of predictive biomarkers, it is now possible to identify drugs potentially associated with clinical benefit for such patients. We present the Caris approach to evidence-based tumor profiling and two patients with advanced ovarian and prostate cancer in whom standard of care had failed and tumor profiling identified an effective treatment schedule. To establish Caris Molecular Intelligence(TM) (CMI), over 120,000 clinical publications were screened and graded to characterize the predictive value of biomarkers that form the panel of tests. CMI includes multiple technologies to measure changes in proteins, ribonucleic acid, and deoxyribonucleic acid and proprietary software that matches the test results with the published evidence. The CMI results enable physicians to select drugs that are more likely to benefit the patients, avoid drugs that are not likely to work, and find treatment options that otherwise would not be considered. Worldwide, over 60,000 cancer patients have undergone evidence-based tumor profiling with CMI. In the cases reported in this article, CMI identified treatments that would not have been routinely used in the respective clinical setting. The clinical outcomes observed help to illustrate the utility of this approach.

Background: A robotically guided linear accelerator has recently been developed which provides frameless radiosurgery with high precision. Potential advantages for the pediatric population include the avoidance of the cognitive decline associated with whole brain radiotherapy, the ability to treat young children with thin skulls unsuitable for frame-based methods, and the possible avoidance of general anesthesia. We report our experience with this system (the "Cyberknife") in the treatment of 21 children., Procedures: Cyberknife radiosurgery was performed on 38 occasions for 21 patients, age ranging from 8 months to 16 years (7.0 +/- 5.1 years), with tumors considered unresectable. Three had pilocytic astrocytomas, two had anaplastic astrocytomas, three had ependymomas (two anaplastic), four had medulloblastomas, three had atypical teratoid/rhabdoid tumors, three had craniopharyngiomas, and three had other pathologies. The mean target volume was 10.7 +/- 20 cm(3), mean marginal dose was 18.8 +/- 8.1 Gy, and mean follow-up is 18 +/- 11 months. Twenty-seven (71%) of the treatments were single-shot and eight (38%) patients did not require general anesthesia., Results: Local control was achieved in the patients with pilocytic and anaplastic astrocytoma, three of the patients with medulloblastoma, and the three with craniopharyngioma, but not for those with ependymoma. Two of the patients with rhabdoid tumors are alive 16 and 35 months after this diagnosis. There have been no procedure related deaths or complications., Conclusion: Cyberknife radiosurgery can be used to achieve local control for some children with CNS tumors without the need for rigid head fixation., ((c) 2004 Wiley-Liss, Inc.)

The standard of treatment for newly diagnosed patients with acute promyelocytic leukemia (APL) is all-trans retinoic acid (ATRA) plus anthracycline-based cytotoxic chemotherapy, a combination that is highly effective for remission induction. However, 20%-30% of patients relapse and require salvage therapy. Reports from China on the striking efficacy and safety of arsenic trioxide in patients with APL led to clinical trials in the United States, which culminated in U.S. Food and Drug Administration approval in September 2000. Trisenox (Cell Therapeutics, Inc., Seattle, WA) is an injectable formulation of arsenic trioxide indicated in the treatment of refractory or relapsed APL. The common side effects of Trisenox therapy are mostly mild and self-limiting and do not require interruption of therapy. Serious adverse effects that can occur include hyperleukocytosis, electrocardiographic abnormalities, and APL differentiation syndrome. These effects can be prevented or managed successfully with careful patient monitoring during treatment. Trisenox has no known cross-resistance with ATRA or other anticancer agents. It does not cause hair loss and is not myelosuppressive in patients with APL. Oncology nurses can play a major role in educating patients about this new drug, explaining its clinical benefits and side effects and the precautions that are necessary for its use.

Background: Indium-111 labeled antiferritin targets 95% of all Hodgkin's disease lesions with a diameter of 1 cm or more. Subsequent treatment with yttrium-90 labeled antiferritin secures a high response rate in patients with recurrent Hodgkin's disease., Methods: A total of 87 patients were entered on one of three different yttrium-90 labeled antiferritin protocols. Recurrences after yttrium-90 treatment were analyzed. Nine patients were retreated with involved external beam radiation fields, selected with the help of indium-111 labeled antiferritin., Results: In single-agent yttrium-90 antiferritin studies, a response rate of more than 60% was found, with an average response duration of 6 months. One-third of the patients had recurrences in previously uninvolved areas. Repeat indium antiferritin scintigraphy allowed for the selection of new radiation fields for recurrences. In-field disease control was obtained for a median of 8 months, but new recurrences in new areas occurred. Chemotherapy or radiation therapy given immediately before antiferritin decreased tumor targeting with indium-111 labeled antiferritin., Conclusions: Recurrences after radiolabeled antiferritin treatment are not due to radioresistant Hodgkin's disease. In contrast, Hodgkin's disease less than 1 cm in diameter is not targeted and not controlled by radiolabeled antiferritin. New multimodality regimens with a higher therapeutic ratio are needed for treatment of Hodgkin's disease with curative intent. Radiolabeled antiferritin can be incorporated in such regimens to secure better control of bulky Hodgkin's disease (>1 cm in diameter), but it should be given before chemotherapy or radiation therapy.

Treatment with busulphan and/or hydroxyurea rarely produces remission in patients with chronic myelogenous leukaemia (CML) in chronic phase. HLA-identical sibling transplants almost always produce remission, and only about 20% of patients relapse post-transplant. The increased anti-leukaemic efficacy of transplants results from intensive pretransplant treatment and immune-mediated anti-leukaemia effects. We studied 433 patients surviving > or = 2 years after diagnosis of CML to determine if patients who have relapsed after a transplant in chronic phase have longer survival from diagnosis than comparable subjects receiving chemotherapy. The chemotherapy cohort included 344 adults < 50 years of age treated on consecutive trials of the Italian Cooperative Study Group on CML between 1973 and 1986. The transplant cohort included 89 patients reported to the International Bone Marrow Transplant Registry who relapsed after an HLA-identical sibling bone marrow transplant carried out between 1978 and 1992. Survivals in the two groups were compared using Cox proportional hazards regression to adjust for prognostic variables. Median survival was 65 months in the chemotherapy cohort and 86 months in the transplant cohort. The 7-year probability (95% confidence interval) of survival was 34% (28-39%) in the chemotherapy cohort and 57% (43-70%) in the transplant cohort (P=0003). There was no difference in survival of patients relapsing after T-cell depleted and non-T-cell-depleted transplants. We conclude that patients who relapse after an HLA-identical sibling bone marrow transplant for CML in chronic phase have longer survival from diagnosis than comparable patients receiving chemotherapy. This effect is most likely to be the result of intensive chemotherapy and/or radiation given for pretransplant conditioning.

The brief administration of G-CSF to previously treated solid tumor patients has a positive impact on the overall cellularity and progenitor cell content of harvested bone marrow. Fifty-seven patients, fully recovered from therapy and growth factor support, had approximately 500 mL of steady-state marrow harvested as outpatients under local anesthesia. Each patient then received 5 micrograms/kg of G-CSF every 12 hours subcutaneously for either 24 hours (21 patients), 36 hours (20 patients), or 48 hours (16 patients) just before harvesting 500 mL of activated bone marrow. Bone marrow cellularity (x 10(6)/mL) increased from a steady-state mean of 10.7 (+/- 0.9) to 25.7 (+/- 2.8) after 24 hours, 9.3 (+/- 0.7) to 29 (+/- 2.5) after 36 hours, and 9.6 (+/- 0.7) to 28.4 (+/- 2.5) after 48 hours. Although the percentage of CD34+ cells did not significantly change in stimulated marrow, the total number of CD34+ cells (x 10(6)) collected increased from 34 (+/- 6.3) to 52 (+/- 6.6) after two injections, 28 (+/- 3.6) to 65 (+/- 8.5) after three injections, and 28 (+/- 5.4) to 75 (+/- 18) after four injections of G-CSF. Further phenotyping demonstrated significant increases in CD34+HLA-DR+ cells with all three schedules relative to steady-state marrow. There were no changes in the total number of CD34+HLA-DR- cells after two and four shots; however, this population increased from 10 x 10(6) in steady-state marrow to 23 x 10(6) (p = 0.012) after three injections. Analysis of peripheral blood indicated a statistically significant increase in the circulating white count, but more interestingly, there were significant increases in the number of CD34+ cells x 10(4)/mL, suggesting the onset of mobilization. Steady-state blood contained a mean of 0.86 x 10(4)/mL CD34+ cells, which increased to 4.37 x 10(4)/mL, 7.43 x 10(4)/mL, and 8.62 x 10(4)/mL after two, three, and four injections, respectively-levels of CD34+ cells that are comparable to steady-state marrow. Reinfusion of a median of 1.6 x 10(6) activated CD34+ cells/kg resulted in the recovery of > 100/mm3 neutrophils and > 20,000 platelets by days 9 and 19, respectively, which was faster than our previous patients who received steady-state marrow, and comparable to our patients who received mobilized peripheral stem cells.

There is controversy whether adults with acute myelogenous leukemia (AML) in first remission are best treated with chemotherapy or an HLA-identical sibling bone marrow transplant. We studied 1097 adults, 16-50 years old, with AML in first remission. Results of transplants from HLA-identical siblings reported to the International Bone Marrow Transplant Registry (IBMTR; n = 901) were compared with results of chemotherapy in comparable persons treated by the German AML Cooperative Group (GAMLCG; n = 196). Preliminary analyses identified subject- and disease-related variables differing between the cohorts and associated with treatment outcome within each cohort. We adjusted for these variables and differences in time-to-treatment in subsequent comparisons of treatment-related mortality, relapse, survival and leukemia-free survival (LFS). Five-year probability of treatment-related mortality was greater for transplants than chemotherapy (43% (95% confidence interval, 37-49%) vs 7% (3-11%); P< 0.0001). Five-year relapse probability was less for transplants than chemotherapy (24% (20-28%) vs 63% (55-71%); P< 0.0001). Five-year probability of survival was similar with transplants and chemotherapy (48% (43-53%) vs 42% (33-51%); P = 0.24). Five-year LFS probability was higher for transplants than chemotherapy (46% (42-50%) vs 35% (28-41%); P= 0.01). These data indicate that bone marrow transplants from HLA-identical siblings result in comparable survival but greater LFS than chemotherapy in adults with AML in first remission.

Radiolabeled immunoglobulin therapy (RIT) is a new cancer treatment that is more selective than its predecessors. Its dose-limiting normal tissue side effect is bone marrow toxicity, and hematopoietic stem cell damage appears to be its most significant mechanism. Platelet consumption in irradiated normal liver tissues and apoptosis of circulating peripheral blood lymphocytes are other, less important, hematologic side effects. 131I and 90Y are the radioisotopes most commonly used for RIT; in addition, animal toxicology and initial clinical studies of chelate immunoglobulins radiolabeled with 111In (for diagnosis) or 90Y (for therapy) are reviewed. The bone-seeking properties of free 90Y are not considered to be a major component of the hematologic damage caused by yttrium-labeled immunoglobulins. The microenvironment of the bone marrow system is not significantly damaged by current RIT protocols. Moreover, granulocyte colony-stimulating factor (G-CSF) can open the blood-marrow barrier. Bone marrow toxicity after RIT can be corrected by bone marrow transplantation, growth factors, blood products, or fractionation of RIT. Selection of the appropriate corrective regimen depends on the severity of the bone marrow damage and will further enhance the therapeutic ratio of RIT.

The best therapy for persons with acute myelogenous leukemia (AML) in 2nd remission is unknown. Bone marrow transplants from an HLA-identical sibling are reported to be better than chemotherapy but this is controversial. The objective of the study was to compare 3-year leukemia-free survival (LFS) in comparable subjects receiving chemotherapy or a transplant. 485 persons with AML in 2nd remission were studied. The chemotherapy cohort included 244 persons treated on trials of the British Medical Research Council, Eastern Cooperative Oncology Group and MD Anderson Hospital. The transplant cohort included 257 persons transplanted worldwide and reported to the international Bone Marrow Transplant Registry (16 were also chemotherapy subjects.) Subjects were selected for comparable age and year of treatment. Preliminary analyses identified two factors correlated with LFS: age < or = or > 30 years and 1st remission duration < or = or > 1 year; subsequent analyses were partitioned accordingly. Three-year probabilities of treatment-related mortality with chemotherapy and transplants were 7% (95% confidence interval, 3-15%) vs 56% (49-63%). Three-year leukemia relapse probabilities were 81% (74-86%) vs 41% (33-49%). Three-year probabilities of LFS were 17% (12-23%) vs 26 (20-32%). Cohort analysis showed significantly higher LFS with transplants vs chemotherapy in persons < or = 30 years and 1st remissions > 1 year (41% (29-53%) vs 17% (7-32%); P = 0.017) and those in > 30 years with 1st remissions < or = 1 year (18% (9-29%) vs 7% (2-16%); P = 0.046). Others had comparable LFS with both treatments. These data indicate better LFS with HLA-identical sibling transplants than chemotherapy in some persons with AML in 2nd remission.

Despite progress over the past three decades, most patients with acute myeloid leukemia (AML) treated with conventional chemotherapy alone relapse and die of recurrent leukemia. Treatments used to improve outcome include allogeneic (alloBMT) and autologous bone marrow transplantation (ABMT). Indications for transplantation and the relative merits of alloBMT and ABMT remain unclear. In this review, we evaluate evidence supporting a role of ABMT in AML and compare the results with outcomes after alloBMT. In addition, we discuss areas of controversy including the optimal timing for ABMT, the role of bone marrow purging, the place of peripheral blood stem cell collection, the high dose regimen, and post-transplant immunotherapy to reduce relapse.

This report details a bone marrow harvest procedure performed outside the hospital setting under local anesthesia, thereby avoiding many of the risks associated with the traditional surgical procedure. In approximately 30 minutes, 450 milliliters of marrow can be collected from eight bone punctures, containing a median of 4.18 x 10(9) cells and 33 x 10(6) progenitor cells as defined by CD34 expression. Reinfusion of a median 1.2 x 10(6) CD34+ cells/kg in 10 breast cancer and lung cancer patients after dose-intensive chemotherapy resulted in the recovery of granulocytes > 100/mm3 by day 14 and platelets > 20,000 by day 21. Without progenitor cell support, such recoveries could take 30 and 40 days, respectively. Collection of marrow using this protocol does not compromise the engraftment capability of the progenitor cells, seldom necessitates blood product support, is safer for the patient, and reduces the cost of harvesting by 75% compared to inpatient or day surgery procedures.

This study analyzed the impact of cytogenetic abnormalities on outcome of 1516 HLA-identical sibling bone marrow transplants for acute myelogenous leukemia (AML) in first remission reported to the International Bone Marrow Transplant Registry by 188 centers. 708 patients (47%) had cytogenetic studies performed. Transplant outcome in these subjects was similar to the 808 in whom cytogenetic studies were not performed. One or more cytogenetic abnormalities were detected in 284 (40%) of subjects studied. Relapse rates were higher and leukemia-free survival lower in patients with poor prognosis abnormalities vs those with no abnormality or with good or intermediate prognosis abnormalities (relative risk of relapse 2.40, P < 0.01; relative risk of treatment failure 1.68, P < 0.03). We conclude that cytogenetic abnormalities correlated with increased relapse in patients treated with chemotherapy. HLA-identical sibling transplants are similar.

Between 1980 and 1988, 126 patients with leukemia were treated with piperazinedione and fractionated total body irradiation (TBI) followed by allogeneic bone marrow transplantation from HLA matched siblings. Sixty-one patients had acute myelogenous leukemia, 46 acute lymphoblastic leukemia, and 19 chronic myelogenous leukemia. Patients with acute leukemia in first complete remission were transplanted only if perceived to have a low probability of remaining in remission with conventional therapy. The toxicity from the preparative regimen was similar to that of cyclophosphamide and TBI except that none of the patients in the study had hemorrhagic cystitis or veno-occlusive disease. After a median follow up of 114 months, 29 patients (23%) are still alive without relapse. The survival of patients with acute myelogenous or lymphoblastic leukemia transplanted in their first remission were 35% and 43%, respectively. The survival of patients transplanted in their first chronic phase of chronic myelogenous leukemia was 60%. The results of this preparative regimen are comparable to those of cyclophosphamide and TBI.

Peripheral blood has become an alternative to bone marrow as a source of stem cells for transplantation. One of the major disadvantages of peripheral blood as a source is the low concentration of stem cells. For successful engraftment, the infusion of at least 6 x 10(8) nucleated cells per kg is required, a cell number obtained by 6-8 cell pheresis sessions. This cell number contains approximately 0.1% CD34 cells equivalent to 600,000 CD34 cells. It is known that chemotherapy and hematopoietic growth factors increase the concentration and total number of progenitor cells in the peripheral blood. In breast and lung cancer patients we are using two cytoreductive regimens: cytoxan 2 g/m2 + platinol 90 mg/m2, and VP-16 600-900 mg/m2+ platinol 90 mg/m2, respectively, in conjunction with G-CSF for stem cell mobilization. At the time of hematopoietic recovery, between day 13 and 16, the absolute number of CD34+ cells increases in 75% of the patients more than 20-fold, from 5,000 to at least 100,000/ml blood, and to more than 40-fold, to 200,000/ml, in 54% of the patients. Therefore, 3.4-7.5 ml blood contains up to 750,000 CD34+ cells, the minimum number of CD34 cells/kg body weight infused when steady-state collected peripheral blood cells are used. Since we use a minimum of 3 x 10(6) CD34+ cells/kg body weight, 15,000 ml of mobilized blood are required. This amount can often be obtained by collecting 500 ml blood by phlebotomy in 2-3 separate sessions, which is an easy and cost-effective method for the patient.

Background: Poor prognosis of Stage IV breast cancer patients have at best a 10% 3-year survival rate with conventional chemotherapy. Dose-intensive chemotherapy improved survival rates for some of these patients., Methods: All patients were Stage IV estrogen receptor-negative or estrogen receptor-positive hormonal refractory and received conventional chemotherapy (induction phase) to the point of achieving maximal response; if disease was stable or the patients responded, they entered high-dose chemotherapy (intensive phase). Seventy-six percent of the patients received two high-dose treatments with cyclophosphamide (4.5-6.0 g/m2), etoposide (750-1500 mg/m2), and cisplatin (120-180 mg/m2). Patients were randomized to receive or not receive autologous marrow. To identify prognostic factors for survival, univariate statistical analysis and multivariate models were applied to patient subsets., Results: Univariate analysis identified a number of factors whose presence indicates improvement in overall survival rates. These include: (1) absence of liver relapse (P = 0.001); (2) absence of soft tissue relapse (P = 0.001); (3) a smaller number of metastatic sites at the time of detecting Stage IV disease (P = 0.026); and (4) disease-free interval greater than 1 year from initial diagnosis to Stage IV disease (P = 0.011). Multivariate models were fitted to the data, and three variables were identified as independent negative predictors for overall survival: (1) liver site (P = 0.001); (2) soft tissue site (P = 0.039); and (3) prior adjuvant chemotherapy (P = 0.028)., Conclusions: Shorter survival after high-dose chemotherapy is predicted independently by patients pretreated with adjuvant chemotherapy, by disease distributed to the liver or the soft tissue.

Background: There is little long-term follow-up information after autologous transplantation for Hodgkin's disease. We evaluated the influence of various prognostic factors and examined the outcome in 128 such patients., Patients and Methods: Patients received high dose cyclophosphamide, carmustine, and etoposide followed by autologous hematopoietic rescue., Results: Patients have been observed between 50-130 months (median 77 months) following transplantation. Overall survival at four years is estimated as 45 percent, and failure-free survival as 25 percent. The best results were seen in patients with a good performance status, who had failed at most one prior chemotherapy regimen. Failure-free survival at four years is estimated as 53 percent for this group. Relapses more than 24 months after transplantation were seen in 11 patients. Five patients developed myelodysplastic syndromes. Three patients became pregnant after the transplant., Conclusions: Prolonged failure-free survival may be observed following high dose chemotherapy and autologous hematopoietic rescue in patients with Hodgkin's disease. Superior results were seen in patients without extensive prior chemotherapy and in those with a good performance status. Late relapses and deaths from secondary myelodysplastic syndromes mandate prolonged follow-up after autologous transplantation for Hodgkin's disease.

Purpose: This study was undertaken to examine the feasibility of combining radiolabeled antibody therapy with high-dose chemotherapy followed by autologous bone marrow transplantation in patients with poor-prognosis Hodgkin's disease., Patients and Methods: Patients were entered onto this protocol if they had chemotherapy-resistant disease, bulky disease, or extensive prior therapy. Patients received yttrium-labeled antiferritin on day -13, -12, or -11, followed by high-dose cyclophosphamide, carmustine, and etoposide (CBV) on days -6 to -3, and then bone marrow infusion on day 0., Results: Twelve patients received both radiolabeled antibody and high-dose chemotherapy followed by autologous transplantation. Two additional patients started the study, but were unable to complete all therapy. Four of 12 patients experienced early transplant-related mortality. Four patients are alive more than 2 years following transplantation and three are free from disease progression at 24+, 25+, and 28+ months following transplantation. The progression-free survival rate at 1 year is estimated to be 21%. Considering the poor prognostic characteristics of these patients, toxicity on this protocol was not necessarily greater than that observed with high-dose chemotherapy alone., Conclusion: This report demonstrates the feasibility of combining radiolabeled antibody therapy with high-dose chemotherapy and autologous bone marrow transplantation.

The use of high-dose cyclophosphamide, carmustine, and etoposide (CBV) with autologous bone marrow transplantation (ABMT) results in long-term disease-free survival of about 30% in patients with relapsed Hodgkin's disease. Laboratory and clinical data show that cisplatin is synergistic with etoposide and carmustine, with non-overlapping extramedullary toxicity. Twenty-one patients with relapsed Hodgkin's disease that had progressed after both MOPP-like and ABVD-like regimens were treated with CBV plus cisplatin (90 mg/m2) and ABMT. The CR rate was 55%; the three-year disease-free and overall survival were 29% and 38% respectively; these results are comparable to prior experience with CBV. Performance status was strongly correlated with achievement of CR, survival, and time to treatment failure. Nephrotoxicity was seen in 3 patients, and ototoxicity in 1 patient. Although cisplatin could be added to CBV with minimal additional toxicity, the results obtained in this small patient population were not better than those of the earlier regimen. A larger trial in patients not previously exposed to cisplatin may better define the role of its addition to CBV.

Objective: To identify changes in practice and outcome of bone marrow transplants for leukemia in the 1980s., Design: Comparison of key explanatory and outcome variables in five 2-year cohorts, from 1980 through 1981 to 1988 through 1989, using a large database of detailed clinical information., Patients: Recipients (7788) of bone marrow transplants for acute lymphoblastic, acute myelogenous, or chronic myelogenous leukemia reported to the International Bone Marrow Transplant Registry, Milwaukee, Wis, by 185 transplant teams worldwide., Results: Linear increases occurred during the periods 1980 through 1981 to 1988 through 1989 as follows with 95% confidence intervals: (1) transplants for chronic myelogenous leukemia from 14% +/- 2% to 35% +/- 2%; (2) transplants from unrelated donors from 1% +/- 1% to 7% +/- 1%; (3) preparative regimens without radiation from 3% +/- 1% to 30% +/- 2%; and (4) use of methotrexate plus cyclosporine to prevent graft-vs-host disease from 2% +/- 1% to 55% +/- 2%. Among recipients of human lymphocyte antigen-identical sibling bone marrow, the 2-year probability of treatment-related mortality decreased by 6% to 22%. The probability of relapse decreased from 46% +/- 6% to 38% +/- 6% in intermediate leukemia but did not change appreciably in early or advanced leukemia. Probabilities of leukemia-free survival improved from 51% +/- 4% to 57% +/- 3% in early leukemia, from 28% +/- 4% to 36% +/- 5% in intermediate leukemia, and from 12% +/- 4% to 18% +/- 5% in advanced leukemia. A separate analysis of a homogenous population of patients indicated that improvements in outcome in the 1980s were due to improvements in transplant practice rather than improved patient selection., Conclusions: Modest increases in leukemia-free survival rates occurred after human lymphocyte antigen-identical sibling bone marrow transplants in the 1980s. Improvements were due primarily to reductions in treatment-related mortality with little or no change in relapse risk. More effective antileukemia strategies and continued reductions in treatment-related toxic effects are needed.

Seventeen patients who had a relapse at a median of 9 months after marrow transplant (14 allogeneic and three syngeneic) received second transplants. Eight patients were in remission when transplanted. Of the nine patients with active disease at the time of transplant, six had complete remissions, and one converted from blastic to chronic phase of chronic myelogenous leukemia. The median survival was 9 months (95% confidence interval, 4 to 17 months). Four patients died within 100 days of transplantation, and three were disease-free. Ten patients died after 100 days, all except two of disease relapse. Five patients had remissions that were greater than 12 months and longer than the remission after their first transplant (inversions). Three patients remain alive and disease-free at 37+, 55+, and 61+ months, the former two despite remissions of less than 1 year after their first transplant. Second transplants with a different cytoreductive regimen can eradicate disease resistant to prior myeloablative treatment; some patients may benefit from second transplants, even if the first transplant only achieves a short remission.