We have performed both sibling donor and volunteer unrelated donors allografts for chronic myeloid leukemia in first chronic phase for more than 6 years, and we gained the impression that patients with volunteer unrelated donors had more viral infections. Therefore, we compared the results of the first 46 patients with chronic myeloid leukemia in first chronic phase to have allografts from volunteer unrelated donors with those of 57 consecutive patients who received marrow from HLA-identical sibling donors during the same period [3]. This study focuses on the short- and medium-term complications of bone marrow transplantation, particularly infection and GVHD.

Methods

Top Methods Results Discussion Author & Article Info References

Patients

Most of the patients were diagnosed at other institutions and referred to Hammersmith Hospital for consideration for allogeneic bone marrow transplantation. All patients had transplants between July 1985 and December 1991. The minimum follow-up was 7 months. The 46 consecutive volunteer unrelated donor patients were compared with a contemporary group of 57 consecutive sibling donor patients, 50 of whom have previously been described in detail [3]. The 57 sibling donor patients were the first 57 patients at our institution with chronic myeloid leukemia in first chronic phase who received combined cyclosporin A and short-course methotrexate GVHD prophylaxis. Because of their uniform treatment they are readily comparable with the volunteer unrelated donor patients. Of the total 103 patients, 96 patients were Philadelphia chromosome positive, 3 patients were negative, and in 4 patients the Philadelphia chromosome status was not known. One of the 3 patients who was Philadelphia chromosome negative was shown to have breakpoint cluster region gene rearrangement; the other 2 patients were not studied.

Disease Stage

In the month before transplantation, clinical, hematologic, and cytogenetic studies were done to confirm that the patients were in uncomplicated first chronic phase. The criteria for classifying patients with chronic myeloid leukemia as chronic phase have been described previously [1]. Fifty-seven consecutive patients with sibling donors and 46 with volunteer unrelated donors are reported here; patients transplanted in acceleration, second chronic phase, or blast crisis are excluded from this analysis. The pretransplant clinical and hematologic features of the two patient groups are summarized in Table 1.

Details have been reported previously [1, 4]. All patients except one received cyclophosphamide (Farmitalia Carlo Erba; St. Albans, UK), 60 mg/kg body weight per day on days –6 and –5,followed by total body irradiation on days –4 to –2.One 6-year-old girl with a sibling donor was given busulfan (Wellcome; London, UK), 16 mg/kg in place of total body irradiation. Patients with sibling donors received 12 Gy in six fractions (n = 51) or 10 Gy in five fractions (n = 5) over 3 days. Patients with volunteer donors received 12 (n = 26) or 13.2 Gy (n = 20) in six fractions over 3 days. Radiotherapy was delivered by an 8-MV linear accelerator (Philips) at a focus skin distance of 500 cm and a dose rate of 15 cGy per minute, without lung shielding, as previously described [6]. Ten patients with volunteer unrelated donors also received busulfan (8 mg/kg total dose) and 7 received a single dose of daunorubicin (60 mg/m² body surface area). Day 0 is the date of donor marrow infusion.

One patient who received marrow from a volunteer unrelated donor later had a relapse and received a second bone marrow transplant with busulfan conditioning alone (4 mg/kg for 4 days) [7]. One patient with a sibling donor received infusions of donor leukocytes as therapy for relapse [8].

Patients were admitted to single rooms with positive-pressure clean air supply during their initial hospitalization. Mesna (uromitexan, Asta Medica; Cambridge, UK) was used to prevent hemorrhagic cystitis. All patients received prophylactic oral antibiotics. Fevers during the neutropenic period were treated with appropriate broad-spectrum intravenous antibiotics. Patients were given amphotericin 100 mg four times a day and amphotericin lozenges to prevent fungal infections.

Treatment to the Spleen

One sibling donor patient and two volunteer donor patients had a splenectomy before transplant; all other patients received splenic irradiation, consisting of 10 Gy irradiation in 2 to 4 doses on days –8 and –7.

Bone Marrow Donation and Tissue Typing

In all cases, sibling donors and recipients were genotypically HLA identical for HLA-A, -B, and -DR. Unrelated donors mainly originated from the Anthony Nolan Research Centre (London) and the British Bone Marrow and Platelet Donor Panel (Bristol), but some came from international registries. All patients and donors gave informed consent. The selection of donors has been described [4]. All volunteer donors were serologically matched with the recipients for 6 HLA "broad" antigens. Further matching studies included biochemical typing of class I proteins by isoelectric focusing [9] and class II (DR and DQ) typing by analysis of restriction fragment length polymorphisms using the restriction endonuclease Taq I and short exon-specific probes [10]. Since 1989, in cases in which two or more candidate donors seemed equally matched, the youngest male patient with the lowest frequency of cytotoxic T-cell precursors in the GVH direction [5] was chosen. Mixed lymphocyte culture studies were used initially but were found not to correlate with outcome; results are not reported here. Harvesting of marrow and subsequent processing (that is, ex-vivo T-cell depletion) were described previously [1].

Graft-versus-Host Disease Prophylaxis and Treatment

The group of patients with sibling donors received modified cyclosporin A and short-course methotrexate GVHD prophylaxis (Table 2). Intravenous cyclosporin A was given at 5 mg/kg as a loading dose starting on day –3,then 2.5 mg/kg per day thereafter. Whole-blood trough levels were measured twice weekly and the dose of cyclosporin A was adjusted to produce levels of between 200 and 400 ng/mL. In the absence of chronic GVHD, cyclosporin A was withdrawn 4 to 6 months after transplantation. Methotrexate was given at 8 mg/m² intravenously on days 2, 4, 8, and 12. The day-12 dose was omitted if the patient had severe mucositis. Two sibling donor patients, aged 49 and 54 years, received intravenous Campath 1G, 5 mg twice a day on days +1 to +5.

Various GVHD prophylaxis regimens were used in the volunteer unrelated donor patients (see Table 2). Initially, the marrow was ex-vivo T-cell depleted with the monoclonal antibody Campath 1M and complement (n = 19) as described previously [1, 11]. Because of concerns about failed engraftment, these patients also received total lymphoid irradiation (3 x 2 Gy on days –8 and –7).Later, most patients received intravenous antilymphocyte therapy and, in addition, had cyclosporin A and methotrexate GVHD prophylaxis (as did patients with sibling donors). One patient with a volunteer unrelated donor had cyclosporin A and methotrexate GVHD prophylaxis and had no T-cell depletion.

Acute GVHD was graded from I to IV using established criteria [12, 13]. Moderate to severe acute GVHD was treated with high-dose methylprednisolone at 10 to 20 mg/kg daily for 3 days with a 50% dose reduction every 2 to 3 days if improvement occurred. Steroid-refractory GVHD was treated with intravenous administration of various anti-T-cell monoclonal antibodies or with antilymphocyte globulin. One patient (with a sibling donor) who had progressive liver GVHD received a liver transplant and is well 18 months later [14]. Chronic GVHD was scored as none, limited, or extensive [15] and was treated with cyclosporin A and corticosteroids.

Cytomegalovirus Infection Surveillance, Prophylaxis, and Management

A latex agglutination test for cytomegalovirus (CMV) antibodies that detects both immunoglobulin G (IgG) and IgM (CMV scan, Becton Dickinson; Oxford, UK) was performed on recipients' and donors' sera before transplantation according to the manufacturer's instructions. After 1988, patients received acyclovir (200 mg three times a day) and CMV-seronegative blood products if they and their donor were CMV negative and acyclovir (800 mg three times a day) with unscreened blood products if either they or their donor were CMV positive. Surveillance of blood, urine, and throat gargles for CMV was performed as indicated and, more recently, weekly for inpatients and regularly at outpatient follow-up. Cytomegalovirus was detected as early antigen fluorescent foci (DEAFF test) [16] or isolated in cultures of human embryonic lung fibroblasts, or both.

In general, since 1991, patients with surveillance specimens positive for CMV were given intravenous ganciclovir 5 mg/kg twice a day for 14 days and then once daily 5 days a week until day 120; treatment was prolonged if they were being given steroids. Five volunteer donor and four sibling donor patients were given ganciclovir therapy for positive surveillance cultures when they were otherwise well. Cytomegalovirus pneumonitis was treated in the same way, but intravenous CMV hyperimmune globulin (200 mg/kg) was also administered on days 1, 3, 5, and 7. Patients with refractory fever and continued excretion of CMV or progressive pneumonitis were also given intravenous foscarnet (100 mg/kg per day).

Infection rates were analyzed by prospective collection of data with confirmation by checking of retrospective microbiologic and virologic records and review of case notes. Our policy was to review outpatients regularly (every 1 to 3 weeks) and, if necessary, to hospitalize them at Hammersmith Hospital, even for relatively minor infective episodes. For the purposes of the study, a severe viral infection was defined as one that was fatal or life-threatening. For example, CMV pneumonitis was always scored as severe, but CMV viremia was only regarded as severe if it was associated with pancytopenia, fever, and malaise. Similarly, nonviral infections were scored as severe if the pathogens were isolated from a blood culture or from a bronchoalveolar lavage specimen in a patient with pneumonia.

Posttransplant Follow-up

The frequency of hospital visits and bone marrow biopsies has been previously detailed [3]. The minimum follow-up time was 8 months. Cotrimoxazole (480 mg twice a day) was given three times a week as pneumocystis pneumonia prophylaxis until 12 months after transplant or longer if the patient was on steroids. If a patient had a severe allergy to cotrimoxazole or poor graft function, he or she was given nebulized pentamidine monthly. Penicillin V, 250 mg twice a day, was administered daily for life primarily as pneumococcal prevention. Immunoglobulin levels were measured every 3 months after transplant.

Statistical Analysis

Cumulative actuarial probabilities of survival, leukemia-free survival, and acute GVHD were calculated using the method of Kaplan and Meier [17]. The log-rank test was used to compare differences between groups [18]. Continuous variables were compared using either a t-test or a Mann-Whitney test. Categorical data were analyzed using the Fisher exact test or a chi-square test with Yates correction. All P values are two-tailed. Not significant indicates a P value greater than 0.10. The possible confounding effects of variables on the relationship between donor group and outcome were investigated using either a test for homogeneity of the odds ratio [19] or proportional hazards regression analysis [20]. Stratification of continuous variables into two groups was achieved using a cut-off point at the median. Survival (overall and leukemia-free) was assessed on the date of last patient contact and analyzed on 31 August 1992. To eliminate the effect of different risk profiles associated with duration of disease, adjusted survivor functions were calculated by insertion of mean values of duration of disease into a proportional hazards model stratified by donor type.

Results

Top Methods Results Discussion Author & Article Info References

Tissue Typing in the Volunteer Unrelated Donor Group

All patients were broad 6-antigen HLA matches with their volunteer unrelated donors. Forty-three patients were studied by restriction fragment length polymorphisms for class II genes and 8 (19%) were mismatched (4, DR; 3, DQ; 1, both DR and DQ). Biochemical matching for class I antigens was performed in 32 patients (74%); 4 (13%) had a mismatch. When all tissue typing studies had been completed, nine donor-recipient pairs did not match at one locus and two did not match at two loci. Cytotoxic T-lymphocyte precursor frequencies were measured in 41 cases (89%). The median value was 1 in 144 x 10³ (range, 1 in 7 x 10³ to <1 in 1000 x 10³).

Pretransplant Clinical Characteristics

Analysis of pretransplant clinical features revealed several differences between the two groups in the prognostic factors that determine outcome after transplant for chronic myeloid leukemia (see Table 1). Patients with volunteer unrelated donors had a significantly longer interval from diagnosis to transplant and the donor was older. They also received a higher mean dose of total body irradiation. The small difference in cell dose is due to some of the volunteer unrelated donor group having ex-vivo T-cell-depleted marrow.

Engraftment

In the sibling donor group, one patient failed to engraft (neutrophils 0.5 x 10⁹/L or more for 3 consecutive days) and required infusion of autologous marrow. Two more patients died before neutrophil engraftment could be assessed. None of these three patients achieved platelet engraftment. Five patients (10%) did not reach a platelet level of 50 x 10⁹/L; four of these survived for fewer than 60 days. However, only one of these patients died of hemorrhage.

In the volunteer unrelated donor group, five patients failed to engraft (compared with one in the sibling donor group, P > 0.2). Four of these five patients had ex-vivo Campath 1M and one received intravenous antilymphocyte therapy with Campath 1G (4 of 19 versus 1 of 30, P = 0.05). Two of these patients received additional marrow from the original donor and three received autologous marrow. None of these patients survived.

The rate of neutrophil engraftment was similar in the two groups. Platelet recovery was slower in the volunteer unrelated donor group (P = 0.02), with 25% of all patients taking 60 or more days to achieve a platelet count of 50 x 10⁹/L.

The group of patients with volunteer unrelated donors took longer to achieve normal IgM levels (175 versus 101 days; see Table 3). The median difference was 14.5 (CI, –5.0 to 87.0).

The median duration of follow-up of the surviving sibling donors and volunteer unrelated donors patients was 21 and 24 months, respectively (Figure 1). Leukemia-free survival was greater in the sibling donor group (P = 0.05). At 2 years it was 72% (CI, 60% to 84%) in the sibling donor group versus 42% (CI, 26% to 48%) in the volunteer donor group. Overall survival was also increased but not significantly. At 2 years it was 73% (CI, 61% to 85%) in the sibling donor group versus 47% (CI, 31% to 63%) in the volunteer unrelated donor group (P = 0.07).

View larger version (14K): [in this window] [in a new window]   Figure 1. Overall (top) and leukemia-free (bottom) survival. Surviving patients are represented by vertical tick marks. Survival analysis was performed on 31 August 1992. SD = sibling donor; VUD = volunteer unrelated donor.

Potential confounding variables (see Table 1) were added to proportional hazards regression models containing donor type with survival (hazard ratio, 0.55; CI, 0.28 to 1.06) or leukemia-free survival (hazard ratio, 0.55; CI, 0.29 to 1.02) as the outcome. However, with addition of duration of disease, the adjusted hazard ratio for donor group type was not significant for either survival (hazard ratio, 0.81; CI, 0.38 to 1.71) or leukemia-free survival (hazard ratio, 0.89; CI, 0.44 to 1.80). Overall and disease-free survival rates adjusted for duration of disease are shown in Figure 2. None of the other variables tested produced adjusted hazard ratios for donor group type different from the original.

View larger version (14K): [in this window] [in a new window]   Figure 2. Overall (top) and leukemia-free survival in bone marrow transplant recipients according to donor group adjusted for duration of disease. SD = sibling donor; VUD = volunteer unrelated donor.

Three sibling donor patients were not evaluable for acute GVHD (one failed engraftment, two early deaths). Thirty patients (actuarial incidence of 50%; CI, 37% to 63%) with sibling donors developed acute GVHD of grade II or worse.

Among volunteer unrelated donor patients, five did not engraft and could not be evaluated for acute GVHD. The actuarial incidence of grade II to IV acute GVHD was the same in patients with volunteer unrelated donors (52%; CI, 38% to 66%; P > 0.2). Even in the group that did not have ex-vivo T-cell depletion, the incidence was not significantly increased (65% versus 50%, P = 0.16).

The distribution and incidence of chronic GVHD are shown in Table 4. Forty-seven patients with sibling donors survived 100 days and were therefore at risk for chronic GVHD. Twenty-three of these patients (49%; CI, 35% to 63%) developed some chronic GVHD; 15 of 47 (32%) of these had extensive involvement. The volunteer unrelated donor group had more chronic GVHD (77%; CI, 63% to 91%), particularly of the gut and skin. The incidence of extensive involvement was higher in the volunteer unrelated donor group (54%) than the sibling donor group (32%) (P = 0.05).

The volunteer unrelated donors group had a markedly higher incidence of severe viral infection compared with the sibling donors group (Table 5). Nineteen viral infections occurred in 16 of 46 volunteer unrelated donor patients compared with 7 viral infections in 7 of 57 sibling donor patients (P = 0.01). The donor or the recipient was seropositive for CMV in 27 volunteer unrelated donor patients and in 42 sibling donor patients. Serious CMV infections were more common in the volunteer unrelated donor group (10 of 27 versus 5 of 42, P = 0.03). Serious viral infections not caused by CMV were also more common in the volunteer unrelated donor group (9 versus 2, P = 0.02). Six of the serious viral infections in the volunteer unrelated donor group were fatal, compared with two in the sibling donor group. Most serious viral infections in the volunteer unrelated donor group occurred before day 100, although there were six late infections (Figure 3). The overall incidence of exposure to high-dose steroids was similar in the two groups because there was a uniform policy for treating acute GVHD and the incidence was equal in the two groups. Of the 19 viral infections in the volunteer unrelated donor group, 14 were associated with the use of high-dose steroids compared with 5 of 7 in the sibling donor group. None of the clinical features identified as being statistically significantly different between the two groups had a confounding effect on the observed relationship between donor type and severe viral infection (data not shown).

View larger version (14K): [in this window] [in a new window]   Figure 3. Cumulative number of viral infections versus time. Nineteen severe viral infections occurred in 16 volunteer unrelated donor patients and 7 in 7 sibling donor patients. BMT = bone marrow transplant; SD = sibling donor; VUD = volunteer unrelated donor.

The two groups did not differ in the incidence of bacterial infections. Seventeen patients with volunteer unrelated donors had 22 bacterial isolates from blood cultures or bronchoalveolar lavage specimens compared with 16 patients with 23 isolates in the sibling donor group. The bacteria most commonly isolated were coagulase-negative staphylococci, followed by viridans streptococci and Pseudomonas species. Two patients with volunteer unrelated donors died of Pneumocystis carinii pneumonia compared with none in the sibling donor group. The number of fungal infections in the two groups was similar.

Causes of Death

Fifteen patients who had transplants from sibling donors have died. Six deaths were due to GVHD; three, aspergillus; two, CMV pneumonitis; one, pseudomonas pneumonia; one, overwhelming pneumococcal infection; one, staphylococcal septicemia; and one, pneumonia of uncertain cause. Of the six patients who died of GVHD, three had gut GVHD; one, liver; and two, both gut and liver.

Twenty-one patients with volunteer unrelated donors are dead. Nine died of pneumonitis (two, CMV; one, adenovirus; four, idiopathic; and two, Pneumocystis carinii); in three patients death was due to GVHD (two, gut and one, gut and liver); three, due to hepatic venoocclusive disease; and three, of graft failure and bacterial sepsis. Three other patients died of viral infections: one, adenovirus (hepatic necrosis); one, JC virus (progressive multifocal leukoencephalopathy); and one, herpes zoster encephalitis.

Relapse

The incidence of hematologic relapse at 2 years in the sibling donor group was 4%. One of the two patients who had a relapse appeared to have donor-origin (Ph-negative) acute lymphoblastic leukemia [21]. In the volunteer unrelated donor group, three patients (6%) relapsed. Of the five patients who have relapsed, four are alive, three of whom are in cytogenetic remission. The median time to hematologic relapse was 257 days. Eleven more patients have had cytogenetic relapses (6 in the sibling donor group and 5 in the volunteer donor group) but have not relapsed hematologically [22]. Nine of these 11 patients are alive. Four of the 11 solely cytogenetic relapses were transient. One patient has been given donor leukocytes and -interferon and is in cytogenetic remission; two patients have been given -interferon alone.

Performance Status of Survivors

Thirty-two (76%) of the 42 sibling donor survivors had a Karnofsky score of 100 at last follow-up and none of the remaining had a score of less than 80. The three patients with a Karnofsky rating of 80 or less have impaired performance status because of chronic GVHD. Fourteen of the 25 surviving patients (56%) with volunteer unrelated donors have a normal performance status, whereas 7 have a performance status of 80 or less (P = 0.08). Of the 11 volunteer unrelated donors patients with an abnormal performance status, 9 have significant GVHD, and only 4 are working at their normal jobs.

Discussion

Top Methods Results Discussion Author & Article Info References

Our study shows that some patients with chronic myeloid leukemia in first chronic phase can be treated successfully with high-dose chemoradiotherapy and allogeneic bone marrow from carefully matched volunteer unrelated donors. Unadjusted disease-free survival was inferior to patients receiving sibling donor marrow. However, the difference was in part attributable to a longer duration of disease and an increased incidence of severe viral infection in the volunteer unrelated donor group. In this series, small but significant clinical differences between the volunteer unrelated donor and sibling donor groups might have contributed to the rate of infection. The volunteer unrelated donor group had a older donor age; had a longer time from diagnosis to transplant; received a higher dose of total body irradiation; and in most cases, received ex-vivo T-cell-depleted marrow or intravenous antilymphocyte therapy. None of these variables, however, was found to statistically affect the chance of infection.

Opportunistic infections, especially with viruses such as CMV, are still a major problem for patients in the first year after transplant. The volunteer unrelated donor group had more viral infections than the sibling donor group, but there was no increase in infections caused by bacteria, fungi, or P. carinii. Although most serious viral infections were caused by CMV, other viruses were also implicated. It is possible that the higher dose of total body irradiation received by the volunteer unrelated donor group may have contributed to the higher incidence of serious CMV infection. The late viral infections in the volunteer unrelated donor group may be due to chronic GVHD but most viral infections occurred before day 100. The volunteer unrelated donor group did not take longer than the sibling donor group to regain normal lymphocyte counts. The use of T-cell depletion in the volunteer unrelated donor group may also have played a role by impairing cell-mediated immunity. Further, more sophisticated studies of immune function are required to examine this possibility.

Some of these viral infections might in principle have been preventable. Evidence is now convincing that early treatment of CMV infection improves survival [23, 24]. It has been our policy to treat patients with positive surveillance cultures, but the first positive cultures may coincide with the onset of CMV pneumonitis. Thus, ganciclovir prophylaxis may be indicated for all patients with CMV-positive grafts [25], although it may depress marrow function even at low doses. There are no proven therapies for adenovirus or JC polyomavirus, but parainfluenza-type infection being treated effectively with ribavirin has been reported [26].

The fact that nonviral infections were not increased in the volunteer unrelated donor group requires explanation. Recovery of neutrophils and immunoglobulin levels was no slower than after sibling donor transplants, but a higher incidence of chronic GVHD in the volunteer unrelated donor group would be expected to predispose patients to more bacterial infections. The use of antibiotic prophylaxis may have reduced differences between two groups.

In contrast to the Seattle series [27], the incidence of acute GVHD in the volunteer unrelated donor group was not significantly increased, even in the patients whose marrow was not subjected to T-cell depletion. This difference is to some extent due to the relatively high incidence of grade II to IV GVHD in the sibling donor group (54% versus 36% in the Seattle series) and may relate to the lower dose of methotrexate that we use. It may also relate to our use of Campath 1G in the volunteer unrelated donor group. Interestingly, chronic GVHD was significantly increased in the volunteer unrelated donor group and there was a nonsignificant trend toward more patients having extensive disease. The high incidence of chronic GVHD seen by Ash and colleagues [28] is confirmed by our study. It is worth noting that all of the patients in the Milwaukee study received T-cell-depleted marrow compared to only 38% of our patients. Campath 1G appears to have been ineffective in preventing chronic GVHD in the volunteer unrelated donor group. Nineteen of the 27 volunteer unrelated donor patients (70%) who had chronic GVHD had previous acute GVHD that progressed to chronic disease. This prolonged GVHD and its therapy presumably caused profound immunosuppression and may have contributed to the high infection rate.

The incidence of GVHD in patients transplanted from volunteer unrelated donors is high, and it is still uncertain whether any form of ex-vivo T-cell depletion should be used. Results of T-cell depletion with sibling donors revealed actuarial relapse rates exceeding 60% [3], but the possible anti-graft-versus-leukemia effect of T-cell depletion may be to some extent counteracted by the extra GVHD (and presumably graft versus leukemia effect) seen in transplants with volunteer donors. It seems that intravenous antilymphocyte therapy has not increased the probability of relapse in the short term [29, 30].

It is generally agreed that graft failure is more common after volunteer unrelated donor transplants than after sibling donor transplants, but in this series only 1 of 31 of the non-T-cell depletion group failed to engraft, and the one patient was a one-antigen mismatch with his donor. There is still no effective therapy for the patient with progressive pancytopenia and "rejection" cells on the blood film; repeated grafts with reconditioning are usually unsuccessful.

In the future, efforts to prevent (or to treat more effectively) viral infection and chronic GVHD in volunteer unrelated donor transplants may improve clinical results. Nevertheless, the early results of volunteer unrelated donor transplantation are encouraging and it seems reasonable to pursue this approach for patients with chronic myeloid leukemia who lack sibling donors.