Archie Prentice

BACKGROUND: The survival of older patients with acute myeloid leukemia has not improved. Few clinical trials have been available for older patients who are not considered fit for an intensive chemotherapy approach. METHODS: Between December 1998 and November 2003, as part of National Cancer Research Institute Acute Myeloid Leukemia 14 Trial, 217 patients, who were deemed unfit for intensive chemotherapy were randomized to receive low-dose cytarabine (Ara-C) (20 mg twice daily for 10 days) or hydroxyurea with or without all-trans retinoic acid (ATRA). RESULTS: Low-dose ara-C produced a better remission rate (18% vs 1%; odds ratio [OR], 0.15; 95% confidence interval [95% CI], 0.06-0.37; P = .00006) and better overall survival (OR, 0.60; 95% CI, 0.44-0.81; P = .0009), which was accounted for by the achievement of complete remission (CR) (duration of CR: 80 weeks vs 10 weeks for patients with no CR). Patients who had adverse cytogenetics did not benefit. ATRA had no effect. Toxicity scores or supportive care requirements did not differ between the treatment arms. CONCLUSIONS: Older, less fit patients have a poor outcome, and few trials have been conducted in this patient group. Low-dose ara-C treatment was superior to best supportive care and hydroxyurea because it had greater success in achieving CR, and it could represent standard care against which new treatments may be compared in this patient group.

PURPOSE: The aims of this study were to quantify the prospects of salvage treatment of patients who did not undergo transplantation in first complete remission (CR1) and to assess the contribution of allograft in second complete remission (CR2) with respect to major risk groups. This evaluation can inform the decision whether to offer a transplant in CR1. PATIENTS AND METHODS: Of 8,909 patients who entered the Medical Research Council AML10, AML12, and AML15 trials, 1,271 of 3,919 patients age 16 to 49 years who did not receive a transplant in CR1 relapsed. Of these patients, 19% are alive beyond 5 years compared with 7% of patients who relapsed after an allograft in CR1. Overall survival and the contribution of a transplant in CR2 were assessed overall and by cytogenetic risk group by using Mantel-Byar analysis. RESULTS: Fifty-five percent of patients who relapsed entered CR2. This percentage varied by risk group as follows: favorable (82%), intermediate (54%), adverse (27%), and unknown (53%), which resulted in 5-year survivals of 32%, 17%, 7%, and 23%, respectively. Sixty-seven percent of remitters received an allotransplant that delivered superior survival compared with patients who did not receive a stem-cell transplant (42% v 16%). A more-stringent assessment of a transplant by using delayed-entry (Mantel-Byar) analysis confirmed the benefit of transplant overall and within intermediate and adverse risk groups but not the favorable subgroup. CONCLUSION: Successful salvage treatment of patients who do not undergo transplantation in CR1 and relapse can be achieved in 19% of patients, which is improved by a transplant except in favorable risk disease. This result suggests that, for intermediate-risk patients in particular, equivalent overall survival can be achieved by delaying transplantation until after relapse, which would require many fewer transplants.

The major issues in the management of fungal infection are prevention, diagnosis and treatment. Our main goal must remain prevention, but for a number of reasons prophylaxis against invasive fungal infection (IFI), in the neutropenic patient, remains controversial. We consider this aim to be desirable, given the high mortality rates associated with established infection, due in part to inadequacies and substantial delays in diagnosis. In a meta-analysis Bow et al (1997) , found a mortality rate of 47% in patients with invasive fungal infection. In a more recent prospective EORTC survey, Denning et al (1999) found that the 3-month mortality rate in patients with invasive aspergillosis was 64%. However, clinical trials of antifungal prophylaxis have failed to show a reduction in overall mortality in almost every case, other than those conducted in the highest risk patients. Fuel was recently added to the fire of this debate by the publication of a further meta-analysis on the role of prophylaxis and empirical treatment ( Gotzsche & Johansen, 1997) . We have criticized this report for a number of reasons ( Kibber et al, 1997 ) but do support the contention that more studies are needed using modern diagnostic tools, including PCR, as end-points, given the inadequacy of conventional antemortem methods. In particular, we feel that consideration should be given to the study of targeted prophylaxis and treatment in defined risk groups. IFI is most commonly caused in Europe and North America by Candida spp. and Aspergillus spp.. Fusarium spp. and the mucorales are less common causes and cryptococcosis is seen almost exclusively in those with a T-cell deficiency. There has been a shift in the dominant species in most centres in Europe and N. America, where the use of prophylaxis by oral non-absorbable polyenes (amphotericin B or nystatin) has been replaced by the absorbable azoles and, more recently, the triazoles. The introduction of fluconazole led to a clear reduction in both colonization and invasive infection with Candida albicans. The reduction in candidosis due to fluconazole prophylaxis has been accompanied by an increase in invasive mould infections in some studies, e.g. 29% in bone marrow transplantation (BMT) patients given fluconazole vs. 18% in those not receiving it ( van Burik et al, 1998 ). What is consistently reported is the failure of fluconazole to prevent lower gastrointestinal tract (GIT) colonization in contrast to the unabsorbed polyenes. There have been few reports of fluconazole-resistant C. albicans infections occurring in neutropenic patients (in contrast to AIDS) but some major studies have documented an increased risk from C. krusei and C. glabrata, which are inherently resistant to this agent ( Wingard et al, 1991 ). By the use of the combination of fluconazole and oral amphotericin B, we have found a reduction in colonization with C. albicans in the absence of an increase in non-albicans species compared with our prior prophylaxis with ketoconazole and amphotericin B, in a longitudinal study, with well-documented cohorts separated by some years (C. albicans 48% to 23%, C. glabrata 18% to 18%, C. torulopsis 5% to 3%, C. krusei 5% to 1%). The overall rate of colonization by Candida species was therefore reduced (unpublished observations). The dominant species, in some bone marrow transplant centres, is now Aspergillus, for the prevention of which systemic agents such as itraconazole, voriconazole or amphotericin B lipid-associated formulations are used. Specific facilities including laminar airflow (LAF) or high-efficiency particulate air (HEPA) filtered rooms are also needed as acquisition is largely by the airborne route and > 95% of primary infections are in the lung. Studies in patients undergoing either chemotherapy or bone marrow transplantation have shown that the resulting neutropenia carries a risk for IFI of 2–40% ( Schwartz et al, 1984 ; Tollemar et al, 1989a , b; Goodrich et al, 1991 ; Verfaille et al, 1991; Goodman et al, 1992 ; McWhinney et al, 1993 ; Winston et al, 1993 ; Fetscher et al, 1999 ; Salazar et al, 1999 ). The risk is dependent on the underlying disease, the treatment given and the prophylaxis used and is higher in older patients ( Kiwan & Anaissie, 1999) . Historical data are likely to underestimate the risk of IFI during life because of inadequacies in diagnosis and, conversely, autopsy studies are likely to overestimate the risk given the likely recent antecedent immune status, e.g. very prolonged neutropenia. Within the neutropenic population, the risk of IFI ranges from ‘low’ in those with moderate (0·1–0·5 × 109/l) to ‘high’ in those with severe (< 0·1 × 109/l) neutropenia and, most substantially, with the length of neutropenia, less than 7 d carrying a minimal risk. However, this rises to a fourfold increased risk of invasive aspergillosis in those with > 21 d of profound neutropenia ( Gerson et al, 1984 ). Numerous studies have demonstrated additional factors associated with an increased risk of IFI and these are listed in Table I. We shall explore the evidence for some of these to assist in the construction of risk categories. Beyond the gastrointestinal mucosa, the upper airways and the epidermis, the next line of defence against IFI are the phagocytic elements of tissue macrophages and the blood neutrophils and monocytes. Fungal translocation is likely to be an occasional occurrence, as with transient bacteraemia from the oral flora and the lower GIT, yet those with normal numbers of functional phagocytic cells rarely suffer IFI. Macrophages are long lived and only defects in their function (see under corticosteroids) are likely to be of consequence. In contrast, the normal intravascular life span of the neutrophil is of the order of 8–12 h. Neutropenia is the dominant risk factor for IFI ( Goodrich et al, 1991; Goodman et al, 1992 ). Myelosuppressive chemotherapy, for example DAT (daunorubicin, cytarabine and 6-thioguanine) in acute myeloblastic leukaemia (AML), leads to an average 3 weeks of neutropenia, with 2 weeks being severe (< 0·1 × 109/l). Patients undergoing remission induction chemotherapy are thus in an intermediate risk group. That few infections are diagnosed probably reflects more the inadequacy of our tests and clearance of undetected IFI, with neutrophil recovery, than a true low incidence. It is reasonable to assume that those patients with relapsed and/or refractory disease who have prolonged moderate or severe neutropenia are at a greater risk. Shorter periods of neutropenia should carry minimal risk and may not require prophylaxis, such as for those with moderate (0·1–0·5 × 109/l) neutropenia for < 7 d, but even these patients can develop IFI. Fetscher et al (1999) found a 3% rate of invasive Aspergillus infection (with 85% mortality) in 683 patients receiving peripheral blood stem cell (PBSC) rescue after high-dose chemotherapy treatment for a range of solid tumours or lymphoma, although 9 of the 20 occurred at the time of disease relapse. Reliable rapid diagnostic methods targeted more than prophylaxis for such patients. Within the higher stem cell has been found to in be by the of low cell on the risk of IFI, found by Tollemar et al . myeloblastic leukaemia may be associated with a higher risk from invasive aspergillosis to the EORTC study ( Denning et al, 1998 ) , which used the was not of that risk although more of the reported vs. acute leukaemia a the more common use of in the Within the Bow et al found a higher risk in those with with the risk being in the the mortality in with undergoing remission and marrow transplant is now the of was found to be infection the fungal infection was the most of with Aspergillus spp. being the et al, 1999 ). The higher rate in be due either to an functional or to a reduction in the numbers of neutrophils at the of treatment. the conventional chemotherapy used in the treatment of most solid the risk of IFI is to be and the main risk factors of prolonged neutropenia and use of are prophylaxis is not but an should be at risk. study in to or not in to as disease prophylaxis was by et al at The was given during the d after transplant at a of and and fungal infection rates with Candida albicans being the most common fungal was an risk factor in this study, evidence that the combination of these factors is of risk. In a study of prophylaxis by the of the introduction of at a of in a increase in the of IFI ( et al, ). than are the most of line infections and a of invasive disease ( et al, 1992 ). Candida has a high for and is as a and systemic of fungal infection ( is and the of such a is not for antifungal prophylaxis ( et al, 1997 ) The risk of IFI in from to a Bow et al, ). In those with an it is as low as for those with an the risk is increased ( et al, 1997 in those with a it can be as high as In the study, profound T-cell was used as What is by the is have an increased risk of IFI in those with including Tollemar et al, 1989a ; et al, ; et al, 1997 ). It is to be that has been a shift to and an increase in the rate of IFI, associated with and and is not associated with neutropenia ( et al, 1997 ). In the of this report found a by the of neutropenia by the use of corticosteroids) this risk. In our during the the risk of invasive aspergillosis was in the of was probably an is a of gastrointestinal the for as a risk factor for a higher rate of IFI in ( Tollemar et al, 1989a ; Goodrich et al, 1991 ; Verfaille et al, 1991 ). colonization is almost a of invasive candidosis and is of infection ( et al, ; Schwartz et al, 1984 ). evidence for an increased risk of IFI more than or colonization at a is found ( et al, ). that fungal infection was in those not compared with in those at more than Bow et al only a with of show an increased risk of IFI in those patients with in a very that colonization was the most factor in fungal with of patients compared with in those not with Candida is of such that and as is an established infection on colonization at a ( Wingard et al, ; et al, ). et al (1999) found of patients at or after fungal colonization of the lung. However, this high and in of the of most other study from et al in a of patients with that of who for or IFI, compared with only of with < a is for IFI but a not such a recently, et al have demonstrated the of from the lower tract and using as a more and Bow et al more IFI in those with bacteraemia compared with those et al reported that of those patients with who a mortality rate of 21 of a of which due to fungal infection compared with of who blood and 5% bacteraemia has also been shown to be a risk factor by other ( et al, ). it that these patients are for The of the gastrointestinal flora by the of is a of fungal colonization in the risk for IFI ( Schwartz et al, 1984 ; et al, 1991 ; Bow et al, et al, ; et al, ). studies have an increased risk from IFI in patients undergoing In the study from this was found to be a factor ( Tollemar et al, ). caused by high-dose was as the factor to a increased rate of IFI by Bow et al . In this study, IFI was seen in of patients receiving in contrast to and seen in and used in induction treatment for which not reports are as yet our that the combination of with moderate to high-dose is with a high because of a combination of a T-cell and neutropenia. In patients we of as of but due to or aspergillosis and infection was seen with the an of (unpublished observations). is a with a prolonged length of on and a which that this agent even used patients in a prophylaxis for neutropenic is not the risk is probably and of for those at low risk of infection. for the of modern from a number of studies in risk groups. It is to older less such as ( et al, ) , by the route in a which is undergoing or the older azoles such as ketoconazole in an of We have therefore on the and evidence from the studies, including those in which polyenes used as the has been compared with polyenes in studies in at a ‘low’ of ( et al, 1993 ). The either amphotericin B and/or and of There was a reduction in the risk of fungal infection vs. in patients although the in invasive infection not colonization rates was more colonization with in the fluconazole in this study was a of patients ( et al, 1991 ) who fluconazole at compared with amphotericin B, given as . and invasive infection rates amphotericin B was more in colonization in the lower with compared with in the fluconazole which the almost of fluconazole in the upper tract is not the is at in these In of neutropenia, & reported that fluconazole compared with increased the time to use of amphotericin B and reduced the of by was was on invasive study an on neutropenia, not reported further recent study failed to show for fluconazole prophylaxis at in patients with ( et al, 1998 ). in undergoing chemotherapy for acute leukaemia compared fluconazole or every with ( Winston et al, 1993 ). fungal colonization of in patients compared with of at the of prophylaxis < and fungal infections with of seen in patients compared with of in the fluconazole patients fungal infections occurred in 20 of patients but in only 7 of fluconazole patients invasive fungal infections in of patients and in of fluconazole patients was in colonization and infection by Candida species other than Candida krusei with of patients at the of prophylaxis compared with of fluconazole patients < and of patients compared with 7 of fluconazole patients There was seen in Aspergillus infections and The use of amphotericin B, the of and overall mortality in both study ( Winston et al, 1993 ). was compared with oral amphotericin B in a study for the prevention of fungal infection in neutropenic patients with acute compared as a with oral amphotericin B every systemic fungal infection occurred in of fluconazole and of amphotericin B systemic fungal infection the empirical use of amphotericin B occurred in of fluconazole and of oral amphotericin B fungal infection was found in of fluconazole compared with of amphotericin B > and less with the amphotericin ( et al, ). In a substantial study in a reduction in IFI, with in the fluconazole compared with with the polyenes The of fluconazole to an for of the species is et al, in the from of the which has been by the introduction of an in studies of this have shown high with at in chemotherapy patients ( et al, in at ( et al, ) and in at 7 ( et al, 1998 ). In the study, the of was a that of and has antifungal data in patients. data in show a of at years and at years . are the for Aspergillus spp. reported as for prophylaxis by et al and for Candida species infections (C. albicans and non-albicans It is that higher blood in ( et al, 1999 although are as yet data and the of resistant trials have compared with a of In study of chemotherapy was the and was an but not of Aspergillus spp. infections with use vs. being in which was conducted in centres, used a with and reported very high rates ( et al, 1999 ). The study ( et al, 1999 ) of vs. fluconazole chemotherapy, and some of more than during neutropenic to the to d after the of prophylaxis, or Aspergillus spp. infections seen with itraconazole, with only C. albicans and fungal In with fungal Aspergillus C. Aspergillus spp. In the infections with fluconazole and with infections in chemotherapy patients. was a major because of In a study of chemotherapy amphotericin B used as a ( et al, 1998 ). There more IFI vs. Aspergillus spp. vs. and fungal vs. in the amphotericin than with itraconazole, but of these also reported with the their and trials reported a lower overall of fungal infection, a lower rate from fungal infection and a lower use of amphotericin B for IFI in patients given study that compared with fluconazole as in patients chemotherapy with or stem cell transplantation for the in of the blood of of The for of this ( et al, 1999 ). study the for and of the blood for The recent of both the oral and an of has in prophylaxis for most ( et al, 1999 ). The in by Goodman et al compared high-dose fluconazole with study, conducted during the neutropenic found a reduction in infection invasive infection vs. and vs. by of C. Aspergillus spp. and a as for IFI in the fluconazole et al conducted a of fluconazole at against in of or marrow Fungal infection in of patients was reduced compared with of seen in those receiving Candida albicans was seen in the fluconazole and infection, fungal colonization and the empirical use of amphotericin B at was in the fluconazole amphotericin B has been used at a low as a agent for prophylaxis ( et al, 1992 although study, using or a reduction in fungal infection from to in of ( et al, ). Patients peripheral blood stem cell with profound T-cell of the for prevention are at very high risk of IFI. to after amphotericin B has been used with in the mortality from to compared with an studies in support a role for this of amphotericin B, from reduced colonization to reduced IFI ( Tollemar et al, et al, 1999 ). with an has a substantial reduction in risk of invasive fungal infection in undergoing induction for leukaemia using on There was a reduction from an of of to in ( et al, ). in this study, the in was 47% in contrast to in and the in those with was Candida spp. the most common of infection, Aspergillus spp. infection was seen only in The of is the of an by the support for a role in prophylaxis from the of a reduction in the of fungal infection in a North of empirical ( et al, 1999 ). than some studies some studies have of et al conducted a study of the of high-dose fluconazole to and amphotericin B in patients. to be at high risk because of used for the treatment of disease and an prolonged of neutropenia. The of was but the rate of defined infection, including vs. was higher in the fluconazole amphotericin B was given more in the fluconazole and of IFI in There was on The study by & an on neutrophil et al in the fluconazole a higher rate of but and, in contrast, Goodman et al an increased mortality in fluconazole but this not The risk of is documented but is not due to the ( et al, ). reasonable of acute is in every the should be in those with has the but also the of being in the upper and some studies have the and colonization by resistant of non-albicans in the lower tract (see et al, ). this as we have from the introduction of the their use with non-absorbable amphotericin B or to the lower has to albicans increase in non-albicans (unpublished observations). In support of this et al found that although fluconazole the upper of patients for in contrast to in the oral amphotericin B with other for are and those most in are the which should not be given with ( and the of and is ( & than Gotzsche and (1997) only Bow et al (1999) have a meta-analysis of prophylaxis and, to this has been only in In the by Bow et al (1999) , the of the studies compared fluconazole with either or polyenes. There was less overall fungal infection for the for empirical antifungal was a reduction The azoles a reduction in IFI fungal mortality was reduced in patients undergoing stem cell was on overall The of of IFI during of neutropenia has led to the of et al amphotericin B to patients undergoing bone marrow of of invasive aspergillosis compared with a in an In an study ( et al, 1998 patients with or IFI during treatment for or an amphotericin B or or or (in of patients given prophylaxis IFI against of receiving IFI was in of patients who not We have used prophylaxis with amphotericin B or and of patients to has fungal infection (unpublished observations). such patients require prophylaxis during of neutropenia. The to be from the is that antifungal prophylaxis should not be given to patient, but this the seen in the The is to prophylaxis to those at but where do we the those with a of neutropenia are at some risk. more is to a such that the for prophylaxis as our to risk more infection and should and targeted prophylaxis or for those in the intermediate groups. our for and patients are the we have not as the patients are at for fungal infection. may with diagnostic methods. In the use of prophylaxis be for that the of an risk factor in a of risk and the introduction of prophylaxis not of these are given in Table from is as for the prevention of fungal infection as it is for infection ( et al, ). of have been to the use of ( & and other and is a of infection by and although it is to be have been documented of infection with this and are of for fungal and are the most of infection, after ( et al, 1992 ). ( et al, 1999 factor ( et al, ) and factor ( Fetscher & 1999) have the to the from and thus to the risk of fungal translocation and infection. of Aspergillus infection by the airborne route particulate air (HEPA) has been demonstrated to the rate of aspergillosis from to in a ( McWhinney et al, 1993 ; & and to in a study by Wingard et al . et al (1997) found that patients in rooms a risk of invasive aspergillosis of compared with those in laminar airflow a has recently demonstrated a substantial in for receiving for mortality and for those for in which the was in either or laminar airflow rooms ( et al, 1998 ). with under have been as the of ( . the use of such some patients fungal airways infection. and have demonstrated that the are found on and and, in the ( et al, 1997 ). studies of the use of amphotericin B have a ( et al, 1993 ). has not been in a recent prospective study by Schwartz et al (1999) . In this study, the of or invasive aspergillosis was of vs. of in those receiving the agent study patients in rooms and those receiving systemic amphotericin B or There in either mortality or In to the is substantial evidence for a from the use of and study of in the treatment of or mould infections from ( et al, 1993 ) a with some factor vs. of patients with in an study ( et al, 1993 ). In a survey, 48% of patients with or IFI, invasive Aspergillus infection, an mortality rate ( et al, 1998 ). was to our (unpublished rate using of invasive aspergillosis as the management after diagnosis and of amphotericin B (unpublished observations). However, as a of this only of patients for the management of of invasive and empirical studies are now that methods of fungal are is by substantial in data and of ( et al, ) with and of Candida with neutrophils ( et al, 1992 ). most studies of factor have demonstrated other than of neutrophil and occasional reduction in the for treatment and reduced the study ( et al, ) a lower rate of infection in those receiving further using modern diagnostic methods to the the risk of IFI, but have a of ( et al, 1993 ; of et al, ; et al, 1998 ). The of a targeted prophylaxis the of rapid diagnostic tests to that is likely to rescue patients who are this We have used in patients with for more than years ( McWhinney et al, 1993 ; . and 3 weeks on the seen in conventional ( et al, 1997 ) . et al (1999) in neutropenic patients with to for more than who a normal conventional time of d was by the use of are by ( et al, 1992 but we now to use amphotericin B or in the ( et al, 1998 ). of fungal has been used for years ( et al, ) and the is and for aspergillosis ( et al, ; et al, 1998 in our to be used to treatment. the Candida be as remains to be as a of Candida infection ( et al, ). The of a for fungal diagnosis has some years to is that of a which can be species by the use of ( et al, 1997 ). It is now to be used with for targeted ( et al, 1998 ). under is the ( ( & We that these can to a further on we use the combination of and to targeted and have empirical in in most The evidence for the of prophylaxis is in the but it is in other groups. The of invasive fungal infection is in the neutropenic patient, that we consider it to the use of both and empirical in the and, the rapid diagnostic methods are in the intermediate risk is not in the group. for failure to of prophylaxis is that the to be The should the range of likely have systemic and the of should be the upper and lower the and the upper tract by the of these we have recently to our of using a combination of oral amphotericin B or in the intermediate and who are in The use of empirical should with increased in the and be targeted to those with very invasive fungal In the we the of clinical trials of and antifungal with studies of of the and using We that this

There is increasing evidence of T cell dysfunction in B cell chronic lymphocytic leukaemia (B-CLL) which may contribute to the aetiology and progress of the disease. An absolute CD8+ lymphocytosis correlates with disease progression and low expression of CD4 and CD8 (as found in autoimmune disease) is seen with abnormal expression of other surface molecules. Although the expression of T cell surface activation markers, CD25 and CD152, may be increased on culture in B-CLL serum, response to the common mitogens, PHA and PWM, is reduced. This and the excess of CD8 cells may explain partly the variable cooperation of T cells with B cell production of immunoglobulin in B-CLL. In the context of T cell cross-talk with antigen presenting cells, B-CLL B cells are poor antigen presenters. But the T cells themselves have significant abnormalities of expression of the many antigens and ligands necessary for this process. In particular, they exhibit variable expression of the low affinity and non-specific adhesion molecules LFA-1 and ICAM-1, variable, clonally restricted and skewed expression of the TCR repertoire (implying repeated antigenic stimulation possibly by CLL antigens), reduced CD28 and CD152 expression (implying impairment of ability to start or stop an immune response) and reduced IL2 and CD25 (IL2 R) expression (critical for positive feed-back in maintenance and expansion of the T cell response to antigen presentation). Although the production of IL2 and other cytokines by the T cell in B-CLL may be impaired, production of the anti-apoptotic cytokine IL4 is not and there may be a unique and expanded subset of CD8/CD30 cells capable of releasing IL4. The relationship of this T cell subset to the malignant B cell in vivo is unknown. However, T cells which are CD4+/CD152+/CCR4+ migrate selectively in vitro in response to the chemokine CCL22 (specific for the receptor CCR4) produced by the malignant B cells and are always seen amongst the malignant cells in bone marrow and lymph nodes from B-CLL patients. Other abnormalities of cytokine secretion are described. These findings suggest that the T cell in B-CLL may be unable to start, maintain and complete an immune response to the malignant B cell and other antigens and may be involved directly in sustaining the tumour. However, autologous tumour specific cytotoxicity has been shown in vitro and T cells which recognise tumour-derived heavy chain fragments circulate in vivo. If adoptive immunotherapy of any nature is to succeed in B-CLL, manipulation to optimise these CTL responses is needed to overcome the profound and variable T cell dysfunction in this disease.

Background Multidisciplinary team (MDT) meetings have been endorsed by the Department of Health as the core model for managing chronic diseases. However, the evidence for their effectiveness is mixed and the degree to which they have been absorbed into clinical practice varies widely across conditions and settings. We aimed to identify the key characteristics of chronic disease MDT meetings that are associated with decision implementation, a measure of effectiveness, and to derive a set of feasible modifications to MDT meetings to improve decision-making. Methods We undertook a mixed-methods prospective observational study of 12 MDTs in the London and North Thames area, covering cancer, heart failure, mental health and memory clinic teams. Data were collected by observation of 370 MDT meetings, completion of the Team Climate Inventory (TCI) by 161 MDT members, interviews with 53 MDT members and 20 patients, and review of 2654 patients’ medical records. We examined the influence of patient-related factors (disease, age, sex, deprivation indicator, whether or not their preferences and other clinical/health behaviours were mentioned) and MDT features (team climate and skill mix) on the implementation of MDT treatment plans. Interview and observation data were thematically analysed and integrated to explore possible explanations for the quantitative findings, and to identify areas of diverse beliefs and practice across MDT meetings. Based on these data, we used a modified formal consensus technique involving expert stakeholders to derive a set of indications of good practice for effective MDT meetings. Results The adjusted odds of implementation were reduced by 25% for each additional professional group represented [95% confidence interval (CI) 0.66 to 0.87], though there was some evidence of a differential effect by type of disease. Implementation was more likely in MDTs with clear goals and processes and a good team climate (adjusted odds of implementation increased by 7%; 95% CI 1% to 13% for a 0.1-unit increase in TCI score). Implementation varied by disease category (with the lowest adjusted odds of implementation in mental health teams) and by patient deprivation (adjusted odds of implementation for patients in the most compared with least deprived areas were 0.60, 95% CI 0.39 to 0.91). We ascertained 16 key themes within five domains where there was substantial diversity in beliefs and practices across MDT meetings. These related to the purpose, structure, processes and content of MDT meetings, as well as to the role of the patient. We identified 68 potential recommendations for improving the effectiveness of MDT meetings. Of these, 21 engendered both strong agreement (median ≥ 7) and low variation in the extent of agreement (mean absolute deviation from the median of &lt; 1.11) among the expert consensus panel. These related to the purpose of the meetings (e.g. that agreeing treatment plans should take precedence over other objectives); meeting processes (e.g. that MDT decision implementation should be audited annually); content of the discussion (e.g. that information on comorbidities and past medical history should be routinely available); and the role of the patient (e.g. concerning the most appropriate time to discuss treatment options). Panellists from all specialties agreed that these recommendations were both desirable and feasible. We were unable to achieve consensus for 17 statements. In part, this was a result of disease-specific differences including the need to be prescriptive about MDT membership, with local flexibility deemed appropriate for heart failure and uniformity supported for cancer. In other cases, our data suggest that some processes (e.g. discussion of unrelated research topics) should be locally agreed, depending on the preferences of individual teams. Conclusions Substantial diversity exists in the purpose, structure, processes and content of MDT meetings. Greater multidisciplinarity is not necessarily associated with more effective decision-making and MDT decisions (as measured by decision implementation). Decisions were less likely to be implemented for patients living in more deprived areas. We identified 21 indications of good practice for improving the effectiveness of MDT meetings, which expert stakeholders from a range of chronic disease specialties agree are both desirable and feasible. These are important because MDT meetings are resource-intensive and they should deliver value to the NHS and patients. Priorities for future work include research to examine whether or not the 21 indications of good practice identified in this study will lead to better decision-making; for example, incorporating the indications into a modified MDT and experimentally evaluating its effectiveness in a pragmatic randomised controlled trial. Other areas for further research include exploring the value of multidisciplinarity in MDT meetings and the reasons for low implementation in community mental health teams. There is also scope to examine the underlying determinants of the inequalities demonstrated in this study, for example by exploring patient preferences in more depth. Finally, future work could examine the association between MDT decision implementation and improvements in patient outcomes. Funding The National Institute for Health Research Health Services and Delivery Research programme.