DNA repair contributes to the drug-resistant phenotype of primary acute myeloid leukaemia cells with FLT3 internal tandem duplications and is reversed by the FLT3 inhibitor PKC412.

The presence of internal tandem duplications (ITD) mutations in the FMS-like tyrosine kinase 3 (FLT3) receptor influences the risk of relapse in acute myeloid leukaemia (AML). We have investigated DNA repair in FLT3-ITD and wild-type (WT) cells. Using the comet assay, we have demonstrated that the FLT3 inhibitor PKC412 significantly inhibits repair of DNA damage in the MV4-11-FLT3-ITD cell line and FLT3-ITD patient samples but not in the HL-60-FLT3-WT cell line or FLT3-WT patient samples. Following the discovery that transcript levels of the DNA repair gene RAD51 are significantly correlated with FLT3 transcript levels in FLT3-ITD patients, we further investigated the role of RAD51 in FLT3-ITD-AML. The reduction in DNA repair in PKC412-treated FLT3-ITD cells was shown to be associated with downregulation of RAD51 mRNA and protein expression and correlates with the maintenance of phosphorylated H2AX levels, implying that PKC412 inhibits the homologous recombination double-strand break repair pathway in FLT3-ITD cells. Using FLT3-short interfering RNA (siRNA), we also demonstrated that genetic silencing of FLT3 results in RAD51 downregulation in FLT3-ITD cells but not in FLT3-WT cells. This work suggests that the use of FLT3 inhibitors such as PKC412 may reverse the drug-resistant phenotype of FLT3-ITD-AML cells by inhibiting repair of chemotherapy-induced genotoxic damage and thereby reduce the risk of disease relapse.

Bcl-x(L) is heterogenously expressed by acute myeloblastic leukaemia cells and is associated with autonomous growth in vitro and with P-glycoprotein expression.

The cells from approximately 70% patients with acute myeloblastic leukaemia exhibit autonomous growth characteristics in vitro, which have been associated with a poor response to therapy. We have previously shown that leukaemic cells with autonomous growth characteristics express high levels of bcl-2 and are relatively resistant to apoptosis. As bcl-x(L) is a bcl-2-related gene with anti-apoptotic activity which also confers resistance to cytotoxic drugs we have studied its expression in AML in relation to cellular growth characteristics and to the expression of P-glycoprotein. Cells from 15 patients were studied. Immunoblotting demonstrated bands at 31 kDa corresponding to bcl-x(L) from the cells of all patients. Bcl-x(S) was not detected in any sample. Using standardised, quantitative flow cytometry, bcl-x(L) expression ranged from 0.25 x 10(5) to 4.24 x 10(5) bound FITC molecules, (median 1.35 x 10[5]). AML blasts with autonomous growth in vitro expressed more bcl-x(L) (median 1.76 x 10[5]) than those which did not (median 0.86 x 10(5), P=0.01). Quantitative bcl-x(L) expression strongly correlated with that of P-glycoprotein, also measured by quantitative flow cytometry using the MRK16 antibody (r=0.95, P < 0.001), but not with MRPr1. These results provide a further explanation for the poor prognosis associated with autonomous in vitro growth of AML blasts and illustrate that these cells may coexpress different modalities of resistance to cytotoxic drug therapy involving both anti-apoptotic pathways (bcl-x(L), bcl-2) and classic multidrug resistance (MDR1). The implication of these findings is that the use of agents to reverse MDR1 function in AML may be unsuccessful in the absence of strategies to reduce resistance to apoptosis.

Blast maturity and CD34 expression determine the effects of the differentiating agents KH1060 and 9-cis-retinoic acid on the differentiation and clonogenicity of non-M3 acute myeloid leukaemia cells.

The vitamin D analogue KH1060 and the retinoids all-trans retinoic acid (ATRA), 9-cis-retinoic acid (9-cRA) and 4-hydroxyphenyl retinamide (4-HPR) induce differentiation and/or apoptosis and inhibit clonal growth of acute promyelocytic leukaemia cells. We have studied the effects of these agents in vitro on cells from 12 patients with other forms of acute myeloblastic leukaemia (AML). Treatment with KH1060 (10(-6) M) caused decreases in cell viability in suspension culture to a median of 44% of control values (P=0.02). However, retinoids had little effect. Subsequent clonal growth in semi-solid medium was inhibited to 5% (median) of control with 10(-6) M KH 1060 (P=0.03) and to 73% with 10(-6) M 9-cRA (P=0.01). Further inhibition of clonal growth by the combination of 5 x 10(-7) M 9-cRA and 5 x 10(-7) M KH1060 was only noted in one case. Following the primary suspension culture, cells from 6/6 CD34 positive samples grew in semi-solid cultures without analogues, whereas cells from 3/6 CD34 negative cultures grew. 10(-6) M KH1060 completely abolished colony growth in all three CD34 negative samples and 10(-6) M 9-cRA inhibited the number of colonies to a median of 11% of control values. In the six CD34 positive samples median colony growth was inhibited to 36% of control values by KH1060 and to 83% of control values by 9-cRA. CD11b expression was increased by 210% (median) with 9-cRA and by 90% (median) with KH1060 in early to intermediate myeloblast (M0, M1, M2) clones. A different pattern was noted in more mature (M4, M5, M6) clones: here there was little or no increase in CD11b expression induced by retinoids or KH1060, but the ratio of apoptotic to viable CD11b+ cells, measured by CD11b/7-AAD double staining, was increased in 6/6 cases treated with KH1060 or the combination of 9-cRA and KH 1060, and in 5/6 cases treated with 9-cRA. No overall significant change in bcl-2 or bax expression on G0/G1 cells was found after 3 days' suspension culture with the analogues. However bcl-x was downregulated in G0/G1 cells treated with KH1060 (median bcl-x relative fluorescence intensity = 45.3 in cells treated with KH1060, compared with 65.7 in control wells, P=0.028). We conclude that CD34+ samples are relatively resistant to the growth inhibition induced by KH1060 and 9-cRA. However, downregulation of bcl-x in cells which have survived treatment with KH1060 may increase the susceptibility of the remaining leukaemic cells to cytotoxic drugs.

Interleukin-1beta maintains an apoptosis-resistant phenotype in the blast cells of acute myeloid leukaemia via multiple pathways.

Blast cell survival in suspension culture is associated with chemoresistance in acute myeloid leukaemia (AML). Autonomous production of IL-1beta by AML blasts is linked with a proliferative response, although its role in survival and hence apoptosis-resistance has not been examined in this disease. Cells that secreted more than 19.7 pg/ml IL-1beta were significantly more resistant to spontaneous apoptosis in 48-h culture than those that produced less than 19.7 pg/ml IL-1beta (P=0.008). Exogenous rhIL-1beta significantly enhanced 48-h survival in 25/29 blast cell samples (P=0.0001). IL-1 receptor ligation is known to activate at least three survival pathways: those mediated by PI-3 kinase, IL-1 receptor-associated kinase (IRAK) and ceramidase. In apoptosis-sensitive AML blasts with a strong survival response to rhIL-1beta, inhibitors of all three pathways down-modulated an IL-1beta-mediated increase in blast survival, but only the inhibition of all three pathways totally eliminated viable blasts. In apoptosis-resistant and apoptosis-sensitive primary AML samples, the three inhibitors all increased apoptosis in vitro after 48 h. Exogenous rhIL-1beta induced the hyperphosphorylation of Bcl-2. It also increased the activation of NF-kappaB in 5/15 blast samples. IL-1beta-mediated survival pathways may be a factor in apoptosis-resistance in primary AML blasts, and may therefore contribute to chemoresistance.

Early allogeneic transplantation for refractory or relapsed acute leukaemia following remission induction with FLAG.

The prognosis for patients with secondary AML, primary resistant AML or ALL and early (<12 months) relapse of acute leukaemia remains extremely poor with conventional chemotherapy. As part of a strategy to improve the outcome for these patients we have treated 22 consecutive patients (18 AML, four ALL, median age 35 years) with either primary resistant disease (n=3), early relapsed leukaemia (n= 12) or secondary AML (n= 7, four RAEBt, two antecedant ALL and one antecedant Hodgkin's disease) with 'FLAG' induction chemotherapy with the aim of proceeding to early allogeneic transplantation either from sibling or unrelated donors. Eighteen patients achieved CR after one course of FLAG, including five patients who had documented p-glycoprotein-induced multidrug resistance and 10 patients with adverse cytogenetic abnormalities. Eight patients were consolidated with a second course of FLAG prior to transplantation and so far 16 patients have undergone allogeneic transplantation, 10 from unrelated donors and six from sibling donors (one mismatched). By the time of transplant three patients had progressed and were in early relapse and all have relapsed post BMT. Of the remaining 13 patients transplanted in remission, nine remain in CCR at a range of 4-26 months, three have died of transplant-related complications (18%) and one patient has relapsed. We conclude that the use of FLAG induction therapy followed by early allogeneic transplantation from either a sibling or unrelated donor can be an effective strategy for the treatment of this difficult group of young patients with poor risk acute leukaemia and appears to be associated with a low procedure-related risk.

Fas receptor-Fas ligand system is independent of both CD34 status and chemosensitivity in acute myeloid leukemia.

OBJECTIVE: To determine whether the Fas receptor-Fas ligand (FasR-FasL) system, which triggers apoptosis in sensitive cells, is an important mechanism of cytotoxicity in acute myeloblastic leukemia (AML). MATERIALS AND METHODS: We investigated FasR expression in primary AML cells and its upregulation by tumor necrosis factor (TNF), as well as the apoptosis induced by anti-Fas antibody and the potential interaction between the FasR-FasL system and the cytotoxic drug daunorubicin (DNR). RESULTS: FasR was expressed on all 25 AML samples and three normal bone marrow harvests. The intensity of expression was variable (range 1. 6-2.1 in normal bone marrow CD34(+) cells and 1.5-5.1 in AML cells, median 2.4) and was related to the morphologic FAB classification, with the highest expression in FAB types M4 and M5 (range 1.6-5.1, median 3.2). No relationship was found between FasR expression and expression of the CD34 antigen. FasR was heterogeneously upregulated in all AML cells on treatment with TNF-alpha. The degree of FasR upregulation induced was found to be related to the FAB subtype, with the greatest response observed in immature FAB types M1, M2, and M6 (range 11.0-207.1%, median 48.7%). Apoptosis could be induced in all AML samples, but not in normal bone marrow CD34(+)ve cells, by the CH11 anti-FasR antibody, although the response was variable (range 4.1-37.6%, median 16.5%). The monocytic differentiated M4 and M5 AML cells exhibited the greatest sensitivity to Fas-mediated apoptosis (range 4.4-37.6, median 20.65%); however, no relationship was found between sensitivity to Fas-mediated apoptosis and FasR expression or CD34 positivity. Apoptosis in response to DNR was observed in all AML cases; however, sensitivity was heterogeneous and found to be unrelated to FasR expression or sensitivity to Fas-mediated apoptosis. The blocking anti-FasR antibody ZB4 blocked anti-FasR-mediated apoptosis but had no inhibitory effect on DNR-induced apoptosis in AML blasts. No cytotoxic synergistic effect was demonstrated when anti-FasR antibody was used in combination with DNR. CONCLUSION: In AML, DNR induces apoptosis through an Fas-independent pathway. However, the induction of apoptosis through the Fas pathway might be a novel and effective approach for leukemia immunotherapy, particularly because Fas-mediated apoptosis was noted in CD34(+) and CD34(-) cases.

Reproducible flow cytometric methodology for measuring multidrug resistance in leukaemic blasts.

In order to bring MDR analysis into a clinical setting, reproducible assays with clear cut off points to define MDR positivity must be used. Sensitivity can also be increased by combining the results of more than one assay. We have used a combination of flow cytometric assays to define MDR positive and negative blasts in 47 AML patients entered into MRC trials. Our primary test is a standardised and reproducible assay for anthracycline accumulation in which we use carboxylate microspheres to bind the fluorescent drug daunorubicin (dnr). Cells and beads are incubated concurrently with dnr. Cellular dnr accumulation is quantified as a cell:bead fluorescence ratio. Confirmatory assays for MDR comprise the cyclosporin modulation assay for rhodamine 123 uptake and also measurement of lung resistance protein and multidrug resistance associated protein (with LRP-56 and MRPr1 respectively). 27/47 (57%) samples had both low and accumulation and at least one positive confirmatory test (a modulated functional assay and/or protein overexpression) and were categorised as "confirmed MDR". 15/47 patients (32%) were MDR negative in all 4 assays. 5/47 (11%) patients had unconfirmed low dnr accumulation. None of the patients in this cohort had high dnr accumulation alongside overexpressed LRP or MRP or functional P-glycoprotein. We believe that this approach to MDR analysis enhances the value of the highly reproducible functional assays. The use of a primary and confirmatory tests is also likely to improve specificity.

Analysis of the interaction of induction regimens with p-glycoprotein expression in patients with acute myeloid leukaemia: results from the MRC AML15 trial.

Retrospective analyses in non-randomised cohorts suggest that regimens containing fludarabine/Ara C and/or idarubicin/ara C may be more effective than daunorubicin/AraC (DA)-containing regimens in cases of acute myeloid leukaemia (AML) overexpressing p-glycoprotein (Pgp). We prospectively measured Pgp protein and function by flow cytometry in CD45-gated blasts from 434 AML15 trial patients randomised to remission induction therapy with two courses of FLAG-Ida or DA±etoposide (DA/ADE). In all, 34% were positive for Pgp protein and 38% for function. Pgp protein-positive cases had a higher incidence of resistant disease (14% vs 5%), adjusted odds ratio 2.67 (1.14-6.24). There was a trend towards a higher cumulative incidence of relapse at 5 years for Pgp-positive cases (46% vs 55%), adjusted hazard ratio 1.42 (0.98-2.07) (P=0.06). For patients treated with FLAG-Ida, the complete remission (CR) rate was 86% for both Pgp-positive and Pgp-negative patients. In patients treated with DA/ADE, 78% of Pgp-positive and 90% of Pgp-negative cases achieved CR (P=0.06). In analyses of overall survival, there was no interaction between treatment received and Pgp expression. Data for Pgp function followed similar trends. Our data suggest that FLAG-Ida may improve the remission rate for Pgp-positive AML, but the malignant clone is reduced rather than eradicated such that the relapse rate remains high in Pgp-positive patients.

Analysis of factors that affect in vitro chemosensitivity of leukaemic stem and progenitor cells to gemtuzumab ozogamicin (Mylotarg) in acute myeloid leukaemia.

Relapse in acute myeloid leukaemia (AML) is considered to result from the persistence of drug-resistant leukaemic stem and progenitor cells (LSPC) within a bone marrow 'niche' microenvironment. Identifying novel agents that have the potential to target these LSPC in their niche microenvironment will aid in the characterization of candidate agents for post-remission chemotherapy. Using an in vitro model, we found that 48-h culture with gemtuzumab ozogamicin (Mylotarg) resulted in a 34% reduction in CD34(+)CD38(-)CD123(+) LSPC number, whereas normal CD34(+)CD38(-) haemapoietic stem cells were insensitive to this agent. As there was considerable heterogeneity in LSPC response to Mylotarg treatment, various factors potentially underpinning the differential response were assessed. LSPC that overexpressed CD33 (P=0.01), which were P-glycoprotein-negative (P=0.008) and with internal tandem duplication (ITD) of the FLT3 gene (FLT3/ITD) status (P=0.006) responded better to Mylotarg treatment. LSPC from patient samples that have these combined characteristics as well as low LSPC burden showed significantly more chemosensitivity to Mylotarg compared with all other cases (P=0.002). In multivariate analysis, LSPC burden and FLT3 status were found to be predictors of LSPC chemosensitivity to Mylotarg treatment (P<0.0001). In conclusion, we have shown heterogeneity in the LSPC compartment of AML patients underpinning differential in vitro sensitivity to Mylotarg.

Resistance to FLT3 inhibition in an in vitro model of primary AML cells with a stem cell phenotype in a defined microenvironment.

Relapse in acute myeloid leukaemia (AML) is mediated by survival of leukaemic stem cells following remission-induction chemotherapy. It would therefore be useful to identify therapeutic agents that target leukaemic stem cells. We devised a flow cytometric chemosensitivity assay allowing 48 h culture of leukaemic blasts in a defined microenvironment followed by enumeration of viable CD34+CD38-CD123+ leukaemic stem and progenitor cells (LSPC). The assay was used to investigate the LSPC response to cytosine arabinoside (Ara-C) and to the FLT3 inhibitor AG1296. There was a 3.6-fold increase in Ara-C-treated LSPC survival under defined 'niche-like' conditions compared to culture without microenvironmental support. Nine AML samples with internal tandem duplications of FLT3 (FLT3/ITDs) were treated with AG1296. Three samples were very sensitive (>50% kill) and 4 were moderately sensitive (10-50% kill) in bulk suspension culture without microenvironmental support. However, under defined 'niche-like' conditions, the survival of LSPC was enhanced rather than inhibited by AG1296 treatment. We conclude that an interaction between LSPC and a defined in vitro microenvironment models a chemoresistant niche. Our data point to a need to investigate more novel chemotherapeutic agents under these stringent conditions to identify agents that may be suitable to target minimal residual disease in AML.

P-glycoprotein plays a drug-efflux-independent role in augmenting cell survival in acute myeloblastic leukemia and is associated with modulation of a sphingomyelin-ceramide apoptotic pathway.

P-glycoprotein (pgp), which is the product of the MDR1 (multidrug resistance-1) gene, has an established role as a mediator of cytotoxic drug resistance in acute myeloid leukemia (AML). To study the role of pgp in mediating apoptosis resistance in AML cells deprived of serum and growth factors, apoptosis was quantified by flow cytometry using uptake of the dye 7-amino-actinomycin D (7-AAD) alongside low forward scatter. In pgp+ve primary AML samples, there was a significant increase in apoptosis in the presence of the pgp-specific antibody UIC2 (mean increase: 58%; range: 11%-95%; P <. 05). Likewise, apoptosis in growth factor-deprived TF1 cells cultured for 30 hours increased 2.5-fold in the presence of 25 microg/mL UIC2. The pgp reversal agent PSC-833 (1 micromol/L) augmented in vitro apoptosis by a median of 52% in pgp+ve patient samples and to a comparable degree in 6 pgp-ve samples. To determine whether the sphingomyelin-ceramide (SM-ceramide) pathway of apoptosis occurs in AML blasts in response to cytotoxic drugs, cells were incubated with daunorubicin at the patient-specific IC(30) (the concentration of daunorubicin that caused apoptotic cell death in 30% of cells) in the presence of the ceramide synthase inhibitor fumonisin B1, which inhibited apoptosis by 18%-81% (median: 40%). Exogenous SM failed to augment apoptosis induced by growth factor withdrawal in pgp+ve TF1 cells and was significantly more effective at augmenting apoptosis in pgp-ve patient blasts (median increase in cell death: 33%; range: 19%-88%) than in pgp+ve samples (median: 7%; range: 0%-27%; P =.028). Cellular accumulation of exogenous SM was associated with apoptosis and also occurred in nonapoptotic patient cells treated with PSC-833. However, this effect was not seen following treatment with the UIC2 antibody. These results indicate that pgp is able to exert a protective effect on AML cell viability and that this is associated with a reduced effect of exogenous SM on apoptosis. The pgp reversal agent PSC-833 acts, at least in part, by a pgp independent mechanism to alter SM distribution and to augment apoptosis induced in AML cells by serum and growth factor withdrawal. (Blood. 2000;95:2897-2904)

Quantitative analysis of lymphocyte CD11a using standardized flow cytometry.

CD11a is the alpha-subunit of the leucocyte adhesion and costimulation molecule LFA-1. We have refined the measurement of lymphocyte CD11a density with a FACScan using commercially available fluorescent beads for standardization and fluorescein-conjugated antibody to CD11a of known fluorescein: protein ratio. The fluorescence intensity of CD11a on peripheral blood CD4+ and CD8+ lymphocytes was measured in 60 healthy subjects. We demonstrated linear correlation between age and mean CD11a density (r = 0.47 for CD4+ cells, r = 0.71 for CD8+ cells). We established that there is a consistent logical cut off point at 4.3 x 10(3) bound antibody molecules between low-expressing and high-expressing subsets of CD8+ cells and we then investigated whether the variation in lymphocyte CD11a expression in healthy subjects was sufficiently small for the application of this method to the detection of abnormal groups or individuals. Analysis of the CD11a high subsets has high statistical power (> 99% in 60 subjects to detect a 25% difference) and good precision (< 4% differences). The advantages of the method for comparative studies of cell surface accessory molecules are discussed. We have also evaluated a frozen cell line for quality control, and demonstrated up-regulation of CD11a density on CD4+ and CD8+ cells measured in three patients with infectious mononucleosis.

Nailfold capillary density as a possible indicator of pulmonary capillary loss in systemic lupus erythematosus but not in mixed connective tissue disease.

Nailfold capillary density was measured in 24 patients with systemic lupus erythematosus (SLE), 14 with mixed connective tissue disease (MCTD) and 21 healthy subjects. Pulmonary function tests were performed on all subjects and needle muscle biopsies on 12 patients with SLE and 9 with MCTD. A significant correlation was documented between nailfold capillary density and pulmonary gas transfer (KCO) in patients with SLE (p less than 0.001) but not in patients with MCTD. This suggests that in SLE poor gas transfer may be dependent on alveolar capillary loss and that nailfold capillary density may be a good indicator of alveolar capillary density. There was no significant correlation between skeletal muscle fiber atrophy and nailfold capillary density in SLE or MCTD. Additional studies to optimize the nailfold capillary counting method are described.

Chronic rupture of a patellar tendon: a technique for reconstruction with Achilles allograft.

Chronic ruptures of the patellar tendon fortunately are an uncommon event. These ruptures are often difficult to repair because they are generally accompanied by quadriceps muscle contracture and a great deal of scar tissue formation. We report the case of a repair of a chronic patellar tendon rupture. The patient's right patellar tendon was reconstructed approximately 10 months after the injury using quadricepsplasty and an Achilles tendon allograft with a suprapatellar wire for tension release. Four weeks postoperatively, he had attained 60 degrees of flexion and full active extension. At 8 weeks, the suprapatellar wire was removed allowing the distribution of stresses on the reconstructed patellar tendon. At 6 months, the patient had 130 degrees of flexion and full extension, but showed a persistent 40% deficit in right quad strength. The technique accomplished the preoperative goals of restoring quadriceps function, restoring the anatomic position of the patella, and allowing early mobilization after surgery. Although the use of a suprapatellar wire to reduce tension on the reconstructed tendon required a second operation for removal, it allowed early mobilization and better healing of the repair.

Functional multidrug resistance in acute myeloblastic leukaemia: a standardized flow cytometric assay for intracellular daunorubicin accumulation.

We have developed a standardized flow cytometric method for the measurement of in vitro multidrug resistance in acute myeloblastic leukaemia (AML) blasts, using carboxylate microspheres which bind the fluorescent drug daunorubicin. Cells and beads were incubated concurrently with the drug. Fluorescence was expressed as a cell:bead ratio. Bead fluorescence at a fixed cytometer voltage was consistent over at least a 3-month period (CV 5.47%), and repeat assays up to 8 months later correlated well (R = 0.86). Bead to drug binding provides a valuable measure of quality assurance as well as a standard for cellular drug accumulation assays and would therefore be suitable for reproducibly reporting the results of multidrug resistance analysis in a clinical setting.

Peripheral blood lymphocyte binding to a soluble FITC-fibronectin conjugate.

Fibronectin receptors are critical to lymphocyte-mediated inflammation, and binding between immobilised plasma fibronectin and peripheral blood lymphocytes (pbl) is enhanced in patients with inflammatory diseases. Such enhancement could be due to an overall increase in receptor capacity and/or be a feature of a subset of cells. It could also depend on the initial binding and/or on events that follow receptor occupancy (e.g., cell spreading). To address these questions, we used fluorescein isothiocyanate-conjugated soluble fibronectin (FITC-FN) in a flow cytometric assay of lymphocyte-fibronectin binding. EDTA-inhibitable adhesiveness to FITC-FN is enhanced in lymphocytes cultured with Interleukin 2 and in fresh peripheral blood lymphocytes (pbl) from untreated patients with acute exacerbations of chronic inflammatory diseases. The enhanced binding was noted on both CD4+ and CD8+ subsets, and decreased with anti-inflammatory therapy.

An electron microscopic study of muscle capillary wall thickening in systemic lupus erythematosus.

Capillary vessel wall thickening may contribute to the muscle fibre atrophy and impaired pulmonary gas transfer notable in SLE/MCTD. Using electron microscopy we measured the thickness of unilaminar capillary basement membranes (CBM) and quantified basement membrane laminae and pericyte layers in the skeletal muscle biopsies of 31 patients with SLE, including 14 with MCTD, and 11 controls. Capillary basement membrane thickness was significantly increased in SLE compared with controls (P < 0.01) and was significantly associated with the level of C3dg (P < 0.01). Patients with MCTD had significantly greater layering of pericytes and laminae compared with patients with SLE alone (P < 0.0001 for both variables). Multiple stepwise regression analysis indicated that ANA titre was a good predictor of pericyte layering (R = 0.63, P = 0.003) and of basement membrane lamination (P < 0.0001), with steroid therapy having a significant negative association with the number of laminae (P < 0.02). A negative association between KCO and the number of layers of basement membrane (R = 0.51, P = 0.01) was documented. However, multiple regression analysis revealed no association between parameters of thickened vessel walls and the presence of muscle fibre atrophy.