Increased Expression of CD25 and HLA-DR on Lymphocytes Recruited into the Peritoneal Cavity in Non-Infected CAPD Patients

The impact of uremia per se, peritoneal dialysis (PD) and hemodialysis (HD) treatment was evaluated on characteristics of lymphocytes. CD4, CD8, CD25 and HLA-DR were analyzed, with flow cytometry, in lymphocytes prepared from peripheral blood of uremic (n = 10) and hemodialysis patients (n = 10). Peritoneal dialysate was also obtained from patients on CAPD (n = 12). A decreased relative and absolute lymphocyte count was observed in peripheral blood from uremic, HD and CAPD patients compared to healthy controls (p < 0.03, p < 0.03 and p < 0.02, respectively). On the other hand, the relative distribution of lymphocytes was significantly higher in peritoneal dialysate compared to peripheral blood of CAPD patients (p < 0.02). Likewise, the absolute CD4 positive lymphocyte count was lower in the peripheral blood from uremic, HD and CAPD patients as compared to healthy controls (p < 0.001, respectively). In CAPD patients the relative distribution of CD4 positive cells (p < 0.001) was lower, while quantitative CD25 level (p < 0.01) and the relative count of HLA-DR (p < 0.0001) was increased in the peritoneal dialysate compared to blood. Taken together a selective activation of lymphocytes in peritoneal dialysate as compared to peripheral blood from uremic, HD and CAPD patients was observed. The altered biological function of the inflammatory cells may therefore explain the increased susceptibility to infectious diseases.     

Myofibroblastic differentiation in simple peritoneal sclerosis

OBJECTIVE: To analyze the presence of myofibroblasts in a series of peritoneal dialysis (PD) patients with simple sclerosis and non-PD, uremic patients. Since there is a close correlation between active fibrosis and myofibroblastic differentiation we wanted to test if myofibroblasts are present in uremic, non-PD peritoneal samples. To determine if there are correlations between myofibroblastic presence and other functional and morphologic peritoneal parameters. METHODS: Biopsies were collected from three patient groups: 1) Normal control samples (n = 15) of parietal and visceral peritoneum 2) non-PD uremic patients (n = 16); and 3) uremic patients on PD (n = 32). Peritoneal morphologic and functional parameters and immunohistochemical expression of alfa-smooth muscle actin was analyzed in each case. Vascular endothelial growth factor (VEGF), bcl-2 anti-apoptotic protein, and progesterone receptor was evaluated in a subset of cases. RESULTS: Myofibroblasts were present in 56.3% of the patients with PD-related simple sclerosis. In most cases they were distributed in the upper submesothelial area. None of the biopsies from normal controls and uremic, non-PD patients showed myofibroblasts. Within the group of PD patients, myofibroblasts showed no correlation with time on dialysis, urea/creatinine MTAC, episodes of peritonitis, submesothelial thickening, hyalinizing vasculopathy or mesothelial status. In a subset of PD patients VEGF expression was observed in submesothelial fibroblastic cells. No expression of progesterone receptor or bcl-2 was observed. CONCLUSIONS: Myofibroblasts are a reliable and simple indicator of fibrosis since they appear in early stages of PD treatment and in patients with minor morphologic anomalies. They are not exclusive of patients with sclerosing peritonitis, ultrafiltration loss or long standing treatment. Their absence in non-PD, uremic patients suggest that uremia-related fibrosis takes place without a significant participation of myofibroblasts.     

Poly(ADP-ribose) polymerase-1 in high glucose-induced epithelial-mesenchymal transition during peritoneal fibrosis

Peritoneal fibrosis is a major complication of continuous ambulatory peritoneal dialysis (CAPD). The present study tested the hypothesis that ADP-ribose polymerase-1 (PARP-1) may play a role in peritoneal epithelial-mesenchymal transition and fibrosis under high glucose conditions. High glucose (126 mmol/l)-induced peritoneal EMT and fibrosis via the PARP-1 mechanism was examined in the primary culture of rat peritoneal mesothelial cells (PMCs) and in the human peritoneal mesothelial cell line (HMrSv5) in the presence or absence of a PARP-1 inhibitor PJ34 (3x10-6 M) or by knocking down PARP-1 with the PARP-1 siRNA technique. High glucose significantly increased PARP-1 expression and EMT as demonstrated by de novo expression of a mesenchymal marker α-SMA and loss of epithelial phenotype E-cadherin by both rat and human PMC, resulting in peritoneal fibrosis including up-regulation of plasminogen activator inhibitor-1 (PAI-1), collagen I, and fibronectin mRNA and protein expression. All these fibrotic responses induced by high glucose were significantly inhibited by the PARP-1 inhibitor PJ34 (all P<0.05) or by knocking down PARP-1 with the siRNA technique. Results from this study suggested that high glucose stimulates peritoneal EMT and fibrosis via a PARP-1-dependent mechanism, and targeting the PARP-1 may represent an alternative therapeutic potential for CAPD-related peritoneal fibrosis.     

Limited influence of the mesothelium on the influx of monocytes into the peritoneal cavity

We have investigated the role of human mesothelium in an in vitro model of peritonitis on the monocyte adherence to and migration across monolayers of peritoneal mesothelial cells. Monocytes adhere avidly to non-activated mesothelial cell monolayers; however, migration in this situation was minimal. Prestimulation of the monolayers with II-1 did not alter these results. Anti-CD 18 and anti-VLA-4 mAbs used in combination had an additive inhibitory effect on monocytes adherence to resting or IL-1-pretreated mesothelial cells. MCP-1 and TGF- are secreted by mesothelial cells. Both have a modest role in mesothelium-induced monocyte chemotaxis: mAbs against these cytokines had an additive inhibitory effect on the chemotaxis induced by supernatant from 24-h prestimulated mesothelial cells. Our results indicate that the mesothelium itself has a limited role in the influx of monocytes into the peritoneal cavity during the onset of peritonitis.     

Tissue models of peritoneal fibrosis

OBJECTIVE: To evaluate the utility of peritoneal pathologic samples, unrelated to peritoneal dialysis (PD) treatment, for the study of peritoneal fibrosis and inflammation. METHODS: Comparative morphologic and immunohistochemical study of peritoneal pathologic samples unrelated to PD with peritoneal biopsies from PD patients with special emphasis on the expression of myofibroblastic and epithelial-to-mesenchymal transition markers. RESULTS: Regarding morphology, PD-related simple fibrosis was less cellular, with greater stromal hyalinization, determining a homogeneous, hypocellular aspect of the submesothelium. In contrast, non-PD fibrosis was more cellular with an extracellular matrix showing a dense and fibrillar quality with wide bundles of collagen. Hylinazing vasculopathy was only present in PD samples. Myofibroblastic differentiation and epithelial-to-mesenchymal transition were common findings in all situations of peritoneal fibrosis. Calponin and calretinin are useful cellular markers to study such fibrogenic mechanisms and correlate with other well-known markers such as a -SMA and cytokeratins. Their expression was much more intense in those samples showing acute inflammation (peritonitis). CONCLUSIONS: Non-PD models of peritoneal fibrosis seem very useful to evaluate important features of human peritoneal pathology such us fibrogenesis, and inflammation. Fibrogenic events such as myofibroblastic differentiation and epithelial-to-mesenchymal transition are evident in these tissue samples allowing us to use them as an accessible source for in vivo and ex vivo studies. Both events show their maximal expression in situations of acute inflammation supporting the important role that peritonitis episodes play in the progression of fibrosis.     

Hepatocyte growth factor/scatter factor released during peritonitis is active on mesothelial cells

Peritonitis causes mesothelial detachment that may result in persistent peritoneal denudation and fibrosis. We investigated whether hepatocyte growth factor (HGF), a scatter factor that induces detachment from substrate and fibroblastic transformation of several cell types, is produced during peritonitis and is active on mesothelial cells. We studied 18 patients on peritoneal dialysis, 9 uncomplicated, 9 with peritonitis. HGF was measured in serum, peritoneal fluid, and supernatant of peripheral blood mononuclear cells and peritoneal mononuclear cells. Primary culture of human peritoneal mesothelial cells and the human mesothelial cell line MeT-5A were conditioned with recombinant HGF, serum, and peritoneal fluid. HGF levels were significantly higher in serum and peritoneal fluid of peritonitic than uncomplicated patients. Mononuclear cells of peritonitic patients produced more HGF than cells of uncomplicated patients. Recombinant HGF, serum, and peritoneal fluid of peritonitic patients caused mesothelial cell growth, detachment, transformation from epithelial to fibroblast-like shape, overexpression of vimentin, and synthesis of type I and III collagen. In conclusion, HGF released during peritonitis causes a change in mesothelial cell phenotype and function. HGF may affect the healing process facilitating repair through mesothelial cell growth, but may contribute to peritoneal fibrosis inducing cell detachment with mesothelial denudation and collagen synthesis.     

Antigen-presenting capacity of macrophages and dendritic cells in the peritoneal cavity of patients treated with peritoneal dialysis

In this study the antigen-presenting capacity of human peritoneal cells and the influence of continuous ambulant peritoneal dialysis (CAPD) were studied. On average 6% of the peritoneal cells were dendritic cells (DC), with no difference between CAPD and control peritoneal cells. DC were enriched by selecting for non-adherent, Fc receptor-negative, low density cells. A typical spot-like CD68 positivity was seen in DC, in contrast to the pancytoplasmic staining pattern in macrophages. Peritoneal DC morphologically and functionally showed features of cells belonging to the DC lineage. Peritoneal DC were superior antigen-presenting cells for both allo-antigen, and Candida albicans antigen or purified protein derivative. CAPD peritoneal macrophages were two- to threefold better stimulator cells for allogeneic T cells compared with control macrophages. The level of integrins/adhesins or MHC class I or II, as measured semi-quantitatively on the FACS, could not account for this phenomenon. In addition, a double chamber system showed that dialysate-activated macrophages produced soluble factors that could enhance DC-induced allogeneic T cell proliferation. In conclusion, human peritoneal cells contain a relatively high percentage of classical DC. CAPD treatment does not impair the antigen-presenting capacity of peritoneal cells, but instead up-regulates the allo-antigcn-presenting capacity of peritoneal macrophages.     

Cytologic features of atypical mesothelial cells in peritoneal dialysis fluid

From February 1988 through October 1988, 23 samples of peritoneal dialysis fluid from 20 patients with end-stage renal disease were cytologically analyzed in an attempt to determine the effect of the dialysate on the mesothelial cells lining the peritoneal cavity. The patients, five female and 15 male, ranging in age from 26 to 75 yr, had been on continuous ambulatory peritoneal dialysis (CAPD) from 1 mo to 6 yr, 4 mo. The patients had no history of cirrhosis, neoplastic disease, radiation and/or chemotherapy, or current findings of infection. Smears and cytosieve filter preparations were made. Smear analysis included the mesothelial cell pattern, the degree of mesothelial cell atypia, and the presence of atypical multinucleated cells and mitoses. In the majority of the fluid samples, reactive mesothelial cells were arranged singly and in sheets. Moderately and severely atypical mesothelial cells were glandular and papillary in configuration. All samples contained at least a few reactive mesothelial cells; in six, the highest degree of cellular atypia was moderate; in 17, it was severe. The development of severe cellular atypia did not appear to be time dependent (a finding noted in samples from patients on dialysis for 6 mo up to 6 yr). When present, multinucleated mesothelial cells showed moderate to severe atypia. In four cases mitotic figures were present. On the basis of these findings, it is proposed that peritoneal dialysis plays a role in the development of mesothelial cell atypia.     

Sclerosing encapsulating peritonitis and non-occlusive mesenteric infarction found at autopsy in a man who had undergone continuous ambulatory peritoneal dialysis: a histochemical and immunohistochemical study

This is a report of a post-mortem histological, histochemical, and immunohistochemical examination of a rare case of sclerosing encapsulating peritonitis (SEP) and non-occlusive mesenteric infarction (NOMI), two serious complications of continuous ambulatory peritoneal dialysis (CAPD), with which a man suffering hepatitis C virus (HCV)-induced liver cirrhosis for 7 years and trauma-induced paraplegia for 50 years had been treated for 1 year. The direct cause of death was encephalopathy caused by extreme hyperammonemia (11 250 microg/dL in serum). The autopsy revealed that the SEP had drastically reduced the length of the small intestine to 210 cm, 180 cm of which presented acute ischemic enteritis with Gram-negative bacterial infection. Histological examination of the SEP revealed that the exterior was composed of normal serosal elastic lamina, but with a cocoon-like appearance remarkably thickened by fibrosis to 3-8 times that of the normal subserosal layer and consisting of spindle cells and blood vessels, with some infiltration of mast cells and lymphocytes. The immunohistochemical examination of the spindle cells revealed few AE1/AE3(+) cells, HHF35(+) cells, and CD34(+) cells, many CD117(+) cells with slight proliferative activity based on MIB-1 positivity (proliferation index <1%), but no CD44(+) cells. It was concluded that either the few CD34(+) and/or the many CD117(+) cells were mesenteric stem cells that had originated from the serosa, proliferated, then differentiated into myofibroblasts or fibroblasts, producing collagen and hyaluronic acid in the matrix, leading to the gradual formation of the SEP, which was induced by the continual irritation of CAPD.     

Muscle water and electrolytes in patients undergoing continuous ambulatory peritoneal dialysis

Muscle water and electrolytes were determined in percutaneous muscle biopsy material from m. quadriceps femoris in 33 uremic patients undergoing continuous ambulatory peritoneal dialysis (CAPD) for 1-38 months, and in 34 normal subjects. The patients showed increased muscle contents of water, sodium, and chloride relative to fat-free solids (FFS); both intra- and extracellular water contents were increased. The total water content was inversely correlated with the duration of CAPD. The muscle potassium content was increased, both relative to FFS and to magnesium, whereas the intracellular potassium concentration was normal. Despite hypermagnesemia, the muscle content of magnesium was normal and the intracellular concentration was even slightly decreased due to the increase in intracellular water. We conclude that muscle water and electrolyte status is abnormal in CAPD patients, but the alterations appear to be less marked than in uremic patients undergoing other forms of therapy.     

Pathogenesis of peritoneal sclerosis

Peritoneal sclerosis is an almost invariable consequence of peritoneal dialysis. In most circumstances it is "simple" sclerosis, manifesting clinically with an increasing peritoneal transport rate and loss of ultrafiltration capacity. In contrast, encapsulating peritoneal sclerosis is a life threatening and usually irreversible condition, associated with bowel obstruction, malnutrition and death. It is unknown whether common etiological factors underlie the development of these 2 clinically and pathologically distinct forms of peritoneal sclerosis. The majority of studies to date have investigated factors that contribute to "simple" sclerosis, although it remains possible that similar mechanisms are amplified in patients who develop encapsulated peritoneal sclerosis. The cellular elements that promote peritoneal sclerosis include the mesothelial cells, peritoneal fibroblasts and inflammatory cells. Factors that stimulate these cells to promote peritoneal fibrosis and neoangiogenesis, both inherent in the development of peritoneal sclerosis, include cytokines that are induced by exposure of the peritoneal membrane to high concentrations of glucose, advanced glycation of the peritoneal membrane and oxidative stress. The cumulative exposure to bioincompatible dialysate is likely to have an etiological role as the duration of dialysis correlates with the likelihood of developing peritoneal sclerosis. Indeed peritoneal dialysis using more biocompatible fluids has been shown to reduce the development of peritoneal sclerosis. The individual contribution of the factors implicated in the development of peritoneal sclerosis will only be determined by large scale peritoneal biopsy registries, which will be able to prospectively incorporate clinical and histological data and support clinical decision making.     

Influence of local inflammation of the peritoneal membrane on diuresis and residual renal function in patients treated with peritoneal dialysis

The aim of this study was to investigate the connection between local inflammation of the peritoneal membrane and diuresis, as well as the residual renal function (RRF) in patients treated with continuous ambulatory peritoneal dialysis (CAPD). Twenty patients treated with CAPD participated in this cross-sectional study. To determine the influence of local inflammation of the peritoneal membrane, effluent interleukin-6 (IL-6) and soluble interleukin-6 receptor (sIL-6R) levels were measured. The level of IL-6, in the group as a whole, was significantly higher in effluent (7.87 pg/mL) than in serum (1.29 pg/mL). There was a significant correlation between effluent and serum IL-6 (r = 0.608; P = 0.002). There was also a significant relationship between effluent and serum IL-6 and duration of CAPD treatment, respectively (r = 0.577; P = 0.004; r = 0.528; P = 0.008). Further, there was a significant negative correlation between effluent IL-6 and daily diuresis (r = −0.533; P = 0.008), but there was no significant correlation between effluent IL-6 and RRF (r = −0.339, P = 0.072). On the other hand, the concentrations of effluent IL-6 were significantly higher in patients with RRF <2 mL/min than in those with RRF ≥2 mL/min (P = 0.039). In conclusion, local inflammation has a significant impact on the amount of diuresis and probably on RRF in patients on CAPD.     

Pentraxin 3 as a new biomarker of peritoneal injury in peritoneal dialysis patients

It is well known that bioincompatible peritoneal dialysate plays a central role in the development of peritoneal fibrosis. Peritoneal inflammation continues even after the cessation of peritoneal dialysate stimulation. It is important to establish the definition of persistent inflammation in the peritoneal cavity at the cessation of peritoneal dialysis (PD). The objective of the present study was to determine whether pentraxin 3 (PTX3) in peritoneal effluent (PE) may be a new biomarker in PD patients. Serum, PE, and peritoneal specimens were obtained from 50 patients with end-stage kidney disease at Juntendo University Hospital. Samples of 19 patients were obtained at the initiation of PD and those of 31 patients at the cessation of PD. PTX3, high-sensitivity CRP, and MMP-2 and IL-6 were analyzed. An immunohistological examination using an anti-PTX3 antibody was performed. Expressions of PTX3 were observed in endothelial cells, fibroblasts, and mesothelial cells in the peritoneum. The PTX3 level in PE at the cessation of PD was significantly higher than that at the initiation of PD. Effluent PTX3 levels in patients with a history of peritonitis or a PD duration of more than 8 years were significantly higher than those in patients without peritonitis or patients with a PD duration of <8 years. The PTX3 level was significantly correlated with MMP-2 and IL-6 levels in PE, as well as the thickness of the submesothelial compact zone and the vasculopathy. It appears that PTX3 may be a new biomarker of peritoneal inflammation and progressive fibrosis.     

Increased matrix metalloproteinase-9 activity in human ovarian cancer cells cultured with conditioned medium from human peritoneal tissue

Ovarian cancer cells disseminate by attachment to the peritoneal mesothelial cell surface of the abdominal cavity. We therefore investigated the influence of conditioned medium (CM) from human peritoneal tissues and mesothelial cells on the secretion of matrix metalloproteinases (MMPs) by ovarian cancer cells. The molecular weights of MMPs stimulating factors derived from human peritoneal tissues and mesothelial cells were estimated using microconcentrators with various cut-off membranes. Human peritoneal tissues were obtained from 12 surgical patients, and mesothelial cells were isolated from three peritoneal specimens. Exposure to CM from peritoneal tissue caused a concentration-dependent increase of the MMP-2 and MMP-9 bands in CM from NOM1 ovarian cancer cells, as shown by zymography. There was a significant difference in the increase of MMP-2 and MMP-9 (2.46-fold and 7.14-fold, respectively, at 0.4mg/ml protein; P < 0.005). CM from mesothelial cells also significantly increased the secretion of MMP-9 by NOM1 cells. The molecular size of possible MMP-9-stimulating factors secreted by peritoneal tissues and mesothelial cells was above M 100000. Further, CM of peritoneal tissues and mesothelial cells also induced the invasiveness of NOM1 cells. These findings suggest that mesothelial cells may secrete some factors which predominantly induce the MMP-9 production and increase invading cell numbers.     

Schistosomiasis and in vitro transdifferentiation of murine peritoneal macrophages into fibroblastic cells

We developed a method for avoiding contamination by fibroblasts when cultures of peritoneal cells are initiated. Macrophages were identified by immunogold detection [light microscope, transmission (TEM) and scanning (SEM) electron microscopes] of membrane antigens (Mac-1⁺, Thy-1,2^–), non-specific esterase activity and ultrastructural features (TEM). As compared with controls, the yield of peritoneal macrophages was 2- and 12-fold higher, respectively, in acutely and chronically infected mice. In all, 30 chronic, 18 acute and 18 control cultures were followed up. At a given cell-density seeding, the decline of control, acute and chronic cultures starts at about day 10, 15, and 27, respectively. In chronic cultures only, fibroblast-like cells appear from day 6 onwards; their number increases with time. Cells showing characters intermediary between macrophages and fibroblasts were observed. We suggest that fibroblast-like cells result from the in vitro transdifferentiation of a limited number of in vivo committed macrophages.     

Nitric oxide production by human peritoneal mesothelial cells

Nitric oxide (NO) has multiple actions, ranging from immunomodulation to regulation of vascular tone and capillary flow. Thus NO generation within the peritoneum could potentially affect peritoneal transport by increasing capillary vasodilatation, and regulate the response to bacterial invasion. Peritoneal mesothelial cells have a common embryological derivation with endothelial cells. As mesothelial cells are the predominant cell type lining the peritoneal cavity, they could potentially be a major source of locally produced nitric oxide. Nitric oxide was measured using the Griess reaction, as total nitrite and nitrate, in fresh unused and spent dialysate effluent (SPDE) from both healthy peritoneal dialysis patients, and during episodes of bacterial peritonitis. Whereas fresh CAPD dialysate was nitrite free (5 +/- 0.1 microM), SPDE from a standard 4 h day time exchange contained 10.2 +/- 0.6 microM/L/h, and that from the overnight dwell 9.1 +/- 0.7 microM/L/h. During an episode of peritonitis, dialysate nitrite and nitrate increased significantly from 9.0 +/- 1.0 microM/L/h, when not infected to 17.5 +/- 2.4, from the first CAPD bag drained at presentation, and 15.2 +/- 1.8 for the second and 16.0 +/- 2.5 for the third exchange (p<0.01). By the following day nitrite levels had returned to baseline, 7.0 +/- 1.0 microM/L/h. Human peritoneal mesothelial cells (HPMC) were cultured and found to produce nitric oxide (261 nmol/mg cell protein), which increased in a dose dependent manner with the addition of spent uninfected CAPD dialysate. The addition of L-arginine, a NO substrate resulted in a 10% increase in nitric oxide production, whereas the addition of the blocker L-NMMA produced a 10% reduction. RNA for inducible nitric oxide synthase (iNOS) was sought using northern blotting technique following combination stimulation with lipopolysaccharide and cytokines (IL-1beta, TNFalpha and gamma-INF, and/or spent dialysate from patients with bacterial peritonitis). However, we could not demonstrate RNA production for iNOS. Peritoneal mesothelial cells may be an important source of locally generated nitric oxide within the peritoneal cavity under basal conditions, but as they do not contain iNOS, the markedly increased NO production observed with episodes of acute bacterial peritonitis is more likely due to a combination of increased NO production by peritoneal macrophages and endothelial cells.     

Control of uremic neuropathy by equilibrium peritoneal dialysis

Motor nerve conduction velocity (MNCV), sensory nerve conduction velocity (SNVC) and distal motor latencies times (DMLT) were evaluated both in upper and lower limbs in three groups of 15 patients of comparable age, treated respectively by extracorporeal dialysis (HD), continuous ambulatory peritoneal dialysis (CAPD) and combined peritoneal dialysis (CPD) for comparable sufficiently long periods. Moreover, MNCV was monitored longitudinally in two groups of patients shifted from CAPD to HD and vice versa. The results show a significant superiority of peritoneal dialysis and particularly of CAPD with respect to HD in controlling uremic neuropathy.     

Serum lipids and lipoproteins during continuous ambulatory peritoneal dialysis

The effects of continuous ambulatory peritoneal dialysis (CAPD) on serum lipids and lipoproteins over the initial year of therapy were studied in 23 uremic patients. Lipoprotein abnormalities typical for the uremic dyslipoproteinemia were present at the start of CAPD. During the first months of CAPD these abnormalities were accentuated. The concentrations of very low density lipoprotein (VLDL)-cholesterol (CHOL), low density lipoprotein (LDL)-CHOL, serum CHOL and serum triglycerides (TG) increased significantly. However, after one year of CAPD only the VLDL-CHOL and serum CHOL levels remained significantly higher than the baseline values. VLDL-TG, VLDL-CHOL and serum TG, and the changes of these variables over the study period, correlated with the amount of glucose in the dialysates. We conclude that the continuous peritoneal absorption of glucose (100-200 g/24 h) during CAPD contributes to potentially atherogenic changes in serum lipids and lipoproteins. However, some of the changes are transitory, indicating an adaptation to the peritoneal glucose load.     

Carcinoma‐associated fibroblasts derive from mesothelial cells via mesothelial‐to‐mesenchymal transition in peritoneal metastasis

Peritoneal dissemination is a frequent metastatic route for cancers of the ovary and gastrointestinal tract. Tumour cells metastasize by attaching to and invading through the mesothelial cell (MC) monolayer that lines the peritoneal cavity. Metastases are influenced by carcinoma‐associated fibroblasts (CAFs), a cell population that derives from different sources. Hence, we investigated whether MCs, through mesothelial–mesenchymal transition (MMT), were a source of CAFs during peritoneal carcinomatosis and whether MMT affected the adhesion and invasion of tumour cells. Biopsies from patients with peritoneal dissemination revealed the presence of myofibroblasts expressing mesothelial markers in the proximity of carcinoma implants. Prominent new vessel formation was observed in the peritoneal areas harbouring tumour cells when compared with tumour‐free regions. The use of a mouse model of peritoneal dissemination confirmed the myofibroblast conversion of MCs and the increase in angiogenesis at places of tumour implants. Treatment of omentum MCs with conditioned media from carcinoma cell cultures resulted in phenotype changes reminiscent of MMT. Adhesion experiments demonstrated that MMT enhanced the binding of cancer cells to MCs in a β1‐integrin‐dependent manner. Scanning electron microscopy imaging showed that the enhanced adhesion was mostly due to increased cell–cell interaction and not to a mere matrix exposure. Invasion assays suggested a reciprocal stimulation of the invasive capacity of tumour cells and MCs. Our results demonstrate that CAFs can derive from mesothelial cells during peritoneal metastasis. We suggest that MMT renders the peritoneum more receptive for tumour cell attachment/invasion and contributes to secondary tumour growth by promoting its vascularization. Copyright © 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.