Synthetic immune checkpoint engagers protect HLA-deficient iPSCs and derivatives from innate immune cell cytotoxicity.

Immune rejection of allogeneic cell therapeutics remains a major problem for immuno-oncology and regenerative medicine. Allogeneic cell products so far have inferior persistence and efficacy when compared with autologous alternatives. Engineering of hypoimmune cells may greatly improve their therapeutic benefit. We present a new class of agonistic immune checkpoint engagers that protect human leukocyte antigen (HLA)-depleted induced pluripotent stem cell-derived endothelial cells (iECs) from innate immune cells. Engagers with agonistic functionality to their inhibitory receptors TIM3 and SIRPα effectively protect engineered iECs from natural killer (NK) cell and macrophage killing. The SIRPα engager can be combined with truncated CD64 to generate fully immune evasive iECs capable of escaping allogeneic cellular and immunoglobulin G (IgG) antibody-mediated rejection. Synthetic immune checkpoint engagers have high target specificity and lack retrograde signaling in the engineered cells. This modular design allows for the exploitation of more inhibitory immune pathways for immune evasion and could contribute to the advancement of allogeneic cell therapeutics.

Effect of Lactoferrin on Clinical Outcomes of Hospitalized Patients with COVID-19: The LAC Randomized Clinical Trial.

As lactoferrin is a nutritional supplement with proven antiviral and immunomodulatory abilities, it may be used to improve the clinical course of COVID-19. The clinical efficacy and safety of bovine lactoferrin were evaluated in the LAC randomized double-blind placebo-controlled trial. A total of 218 hospitalized adult patients with moderate-to-severe COVID-19 were randomized to receive 800 mg/die oral bovine lactoferrin (n = 113) or placebo (n = 105), both given in combination with standard COVID-19 therapy. No differences in lactoferrin vs. placebo were observed in the primary outcomes: the proportion of death or intensive care unit admission (risk ratio of 1.06 (95% CI 0.63-1.79)) or proportion of discharge or National Early Warning Score 2 (NEWS2) ≤ 2 within 14 days from enrollment (RR of 0.85 (95% CI 0.70-1.04)). Lactoferrin showed an excellent safety and tolerability profile. Even though bovine lactoferrin is safe and tolerable, our results do not support its use in hospitalized patients with moderate-to-severe COVID-19.

Protection of cell therapeutics from antibody-mediated killing by CD64 overexpression.

Allogeneic cell therapeutics for cancer therapy or regenerative medicine are susceptible to antibody-mediated killing, which diminishes their efficacy. Here we report a strategy to protect cells from antibody-mediated killing that relies on engineered overexpression of the IgG receptor CD64. We show that human and mouse iPSC-derived endothelial cells (iECs) overexpressing CD64 escape antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity from IgG antibodies in vitro and in ADCC-enabled mice. When CD64 expression was combined with hypoimmune genetic modifications known to protect against cellular immunity, B2M-/-CIITA-/- CD47/CD64-transgenic iECs were resistant to both IgG antibody-mediated and cellular immune killing in vitro and in humanized mice. Mechanistic studies demonstrated that CD64 or its intracellularly truncated analog CD64t effectively capture monomeric IgG and occupy their Fc, and the IgG bind and occupy their target antigens. In three applications of the approach, human CD64t-engineered thyroid epithelial cells, pancreatic beta cells and CAR T cells withstood clinically relevant levels of graft-directed antibodies and fully evaded antibody-mediated killing.

Hypoimmune induced pluripotent stem cell-derived cell therapeutics treat cardiovascular and pulmonary diseases in immunocompetent allogeneic mice.

The emerging field of regenerative cell therapy is still limited by the few cell types that can reliably be differentiated from pluripotent stem cells and by the immune hurdle of commercially scalable allogeneic cell therapeutics. Here, we show that gene-edited, immune-evasive cell grafts can survive and successfully treat diseases in immunocompetent, fully allogeneic recipients. Transplanted endothelial cells improved perfusion and increased the likelihood of limb preservation in mice with critical limb ischemia. Endothelial cell grafts transduced to express a transgene for alpha1-antitrypsin (A1AT) successfully restored physiologic A1AT serum levels in mice with genetic A1AT deficiency. This cell therapy prevented both structural and functional changes of emphysematous lung disease. A mixture of endothelial cells and cardiomyocytes was injected into infarcted mouse hearts, and both cell types orthotopically engrafted in the ischemic areas. Cell therapy led to an improvement in invasive hemodynamic heart failure parameters. Our study supports the development of hypoimmune, universal regenerative cell therapeutics for cost-effective treatments of major diseases.

The SIRPα-CD47 immune checkpoint in NK cells.

Here we report on the existence and functionality of the immune checkpoint signal regulatory protein α (SIRPα) in NK cells and describe how it can be modulated for cell therapy. NK cell SIRPα is up-regulated upon IL-2 stimulation, interacts with target cell CD47 in a threshold-dependent manner, and counters other stimulatory signals, including IL-2, CD16, or NKG2D. Elevated expression of CD47 protected K562 tumor cells and mouse and human MHC class I-deficient target cells against SIRPα+ primary NK cells, but not against SIRPα- NKL or NK92 cells. SIRPα deficiency or antibody blockade increased the killing capacity of NK cells. Overexpression of rhesus monkey CD47 in human MHC-deficient cells prevented cytotoxicity by rhesus NK cells in a xenogeneic setting. The SIRPα-CD47 axis was found to be highly species specific. Together, the results demonstrate that disruption of the SIRPα-CD47 immune checkpoint may augment NK cell antitumor responses and that elevated expression of CD47 may prevent NK cell-mediated killing of allogeneic and xenogeneic tissues.

The H-Y Antigen in Embryonic Stem Cells Causes Rejection in Syngeneic Female Recipients.

Pluripotent stem cells are promising candidates for cell-based regenerative therapies. To avoid rejection of transplanted cells, several approaches are being pursued to reduce immunogenicity of the cells or modulate the recipient's immune response. These include gene editing to reduce the antigenicity of cell products, immunosuppression of the host, or using major histocompatibility complex-matched cells from cell banks. In this context, we have investigated the antigenicity of H-Y antigens, a class of minor histocompatibility antigens encoded by the Y chromosome, to assess whether the gender of the donor affects the cell's antigenicity. In a murine transplant model, we show that the H-Y antigen in undifferentiated embryonic stem cells (ESCs), as well as ESC-derived endothelial cells, provokes T- and B cell responses in female recipients.

Genetic activation of Nrf2 reduces cutaneous symptoms in a murine model of Netherton syndrome.

Netherton syndrome is a monogenic autosomal recessive disorder primarily characterized by the detachment of the uppermost layer of the epidermis, the stratum corneum It results from mutations in the SPINK5 gene, which codes for a kallikrein inhibitor. Uncontrolled kallikrein activity leads to premature desquamation, resulting in a severe epidermal barrier defect and subsequent life-threatening systemic infections and chronic cutaneous inflammation. Here, we show that genetic activation of the transcription factor nuclear factor (erythroid-derived 2)-like 2 (Nfe2l2/Nrf2) in keratinocytes of Spink5 knockout mice, a model for Netherton syndrome, significantly alleviates their cutaneous phenotype. Nrf2 activation promoted attachment of the stratum corneum and concomitant epidermal barrier function, and reduced the expression of pro-inflammatory cytokines such as tumor necrosis factor α and thymic stromal lymphopoietin. Mechanistically, we show that Nrf2 activation induces overexpression of secretory leukocyte protease inhibitor (Slpi), a known inhibitor of kallikrein 7 and elastase 2, in mouse and human keratinocytes in vivo and in vitro, respectively. In the Spink5-deficient epidermis, the upregulation of Slpi is likely to promote stabilization of corneodesmosomes, thereby preventing premature desquamation. Our results suggest pharmacological NRF2 activation as a promising treatment modality for Netherton syndrome patients.This article has an associated First Person interview with the first author of the paper.

A Cryoinjury Model to Study Myocardial Infarction in the Mouse.

The use of animal models is essential for developing new therapeutic strategies for acute coronary syndrome and its complications. In this article, we demonstrate a murine cryoinjury infarct model that generates precise infarct sizes with high reproducibility and replicability. In brief, after intubation and sternotomy of the animal, the heart is lifted from the thorax. The probe of a handheld liquid nitrogen delivery system is applied onto the myocardial wall to induce cryoinjury. Impaired ventricular function and electrical conduction can be monitored with echocardiography or optical mapping. Transmural myocardial remodeling of the infarcted area is characterized by collagen deposition and loss of cardiomyocytes. Compared to other models (e.g., LAD-ligation), this model utilizes a handheld liquid nitrogen delivery system to generate more uniform infarct sizes.

De novo mutations in mitochondrial DNA of iPSCs produce immunogenic neoepitopes in mice and humans.

The utility of autologous induced pluripotent stem cell (iPSC) therapies for tissue regeneration depends on reliable production of immunologically silent functional iPSC derivatives. However, rejection of autologous iPSC-derived cells has been reported, although the mechanism underlying rejection is largely unknown. We hypothesized that de novo mutations in mitochondrial DNA (mtDNA), which has far less reliable repair mechanisms than chromosomal DNA, might produce neoantigens capable of eliciting immune recognition and rejection. Here we present evidence in mice and humans that nonsynonymous mtDNA mutations can arise and become enriched during reprogramming to the iPSC stage, long-term culture and differentiation into target cells. These mtDNA mutations encode neoantigens that provoke an immune response that is highly specific and dependent on the host major histocompatibility complex genotype. Our results reveal that autologous iPSCs and their derivatives are not inherently immunologically inert for autologous transplantation and suggest that iPSC-derived products should be screened for mtDNA mutations.

Hypoimmunogenic derivatives of induced pluripotent stem cells evade immune rejection in fully immunocompetent allogeneic recipients.

Autologous induced pluripotent stem cells (iPSCs) constitute an unlimited cell source for patient-specific cell-based organ repair strategies. However, their generation and subsequent differentiation into specific cells or tissues entail cell line-specific manufacturing challenges and form a lengthy process that precludes acute treatment modalities. These shortcomings could be overcome by using prefabricated allogeneic cell or tissue products, but the vigorous immune response against histo-incompatible cells has prevented the successful implementation of this approach. Here we show that both mouse and human iPSCs lose their immunogenicity when major histocompatibility complex (MHC) class I and II genes are inactivated and CD47 is over-expressed. These hypoimmunogenic iPSCs retain their pluripotent stem cell potential and differentiation capacity. Endothelial cells, smooth muscle cells, and cardiomyocytes derived from hypoimmunogenic mouse or human iPSCs reliably evade immune rejection in fully MHC-mismatched allogeneic recipients and survive long-term without the use of immunosuppression. These findings suggest that hypoimmunogenic cell grafts can be engineered for universal transplantation.

Effects of combination of sibutramine and L-carnitine compared with sibutramine monotherapy on inflammatory parameters in diabetic patients.

The aim of the study was to evaluate the effects of 12-month treatment with sibutramine plus L-carnitine compared with sibutramine alone on body weight, glycemic control, insulin resistance, and inflammatory state in type 2 diabetes mellitus patients. Two hundred fifty-four patients with uncontrolled type 2 diabetes mellitus (glycated hemoglobin [HbA(1c)] >8.0%) in therapy with different oral hypoglycemic agents or insulin were enrolled in this study and randomized to take sibutramine 10 mg plus L-carnitine 2 g or sibutramine 10 mg in monotherapy. We evaluated at baseline and after 3, 6, 9, and 12 months these parameters: body weight, body mass index, HbA(1c), fasting plasma glucose, postprandial plasma glucose, fasting plasma insulin, homeostasis model assessment of insulin resistance index, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, leptin, tumor necrosis factor-α, adiponectin, vaspin, and high-sensitivity C-reactive protein. Sibutramine plus L-carnitine gave a faster improvement of fasting plasma glucose, postprandial plasma glucose, lipid profile, leptin, tumor necrosis factor-α, and high-sensitivity C-reactive protein compared with sibutramine alone. Furthermore, there was a better improvement of body weight, HbA(1c), fasting plasma insulin, homeostasis model assessment of insulin resistance index, vaspin, and adiponectin with sibutramine plus L-carnitine compared with sibutramine alone. Sibutramine plus L-carnitine gave a better and faster improvement of all the analyzed parameters compared with sibutramine alone without giving any severe adverse effect.

Sibutramine and L-carnitine compared to sibutramine alone on insulin resistance in diabetic patients.

OBJECTIVE: To evaluate the effects of one year of treatment with sibutramine plus L-carnitine compared to sibutramine on body weight, glycemic control, and insulin resistance state in type 2 diabetic patients. METHODS: Two hundred and fifty-four patients with uncontrolled type 2 diabetes mellitus (T2DM) [glycated hemoglobin (HbA(1c)) >8.0%] in therapy with different oral hypoglycemic agents or insulin were enrolled in this study and randomised to take sibutramine 10 mg plus L-carnitine 2 g or sibutramine 10 mg in monotherapy. We evaluated at baseline, and after 3, 6, 9, and 12 months these parameters: body weight, body mass index (BMI), glycated hemoglobin (HbA(1c)), fasting plasma glucose (FPG), post-prandial plasma glucose (PPG), fasting plasma insulin (FPI), homeostasis model assessment insulin resistance index (HOMA-IR), total cholesterol (TC), low density lipoprotein-cholesterol (LDL-C), high density lipoprotein-cholesterol (HDL-C), triglycerides (Tg), retinol binding protein-4 (RBP-4), resistin, visfatin, high sensitivity-C reactive protein (Hs-CRP). RESULTS: There was a decrease in body weight, BMI, HbA(1c), FPI, HOMA-IR, and RBP-4 in both groups, even when the values obtained with sibutramine plus L-carnitine were lower than the values obtained in sibutramine group. There was a faster decrease of FPG, PPG, TC, LDL-C, resistin and Hs-CRP with sibutramine plus L-carnitine even when no differences between the two groups were obtained. Furthermore, only sibutramine plus L-carnitine improved Tg, and visfatin. CONCLUSION: Sibutramine plus L-carnitine gave a faster improvement of lipid profile, insulin resistance parameters, glycemic control, and body weight compared to sibutramine.

Comparison of orlistat treatment and placebo in obese type 2 diabetic patients.

AIM: To evaluate the effects of 1-year treatment with orlistat compared with placebo on different inflammatory parameters in type 2 obese diabetic patients. MATERIALS AND METHODS: Two hundred and fifty-four type 2 diabetic patients were randomized to take orlistat 120 mg three times a day or placebo for 12 months. We evaluated at baseline and after 3, 6, 9 and 12 months: leptin, tumor necrosis factor (TNF)-alpha, adiponectin (ADN), vaspin and high-sensitivity C-reactive protein (HS-CRP), body weight, waist circumference, body mass index (BMI), lipid profile, glycemic profile, fasting plasma insulin (FPI) and homeostasis model assessment insulin resistance index (HOMA-IR). RESULTS: Regarding inflammatory parameters, there was a significant improvement of ADN and TNF-alpha, and a faster decrease of leptin and HS-CRP in the orlistat group compared with the control group. We also recorded a significant reduction of body weight and BMI with orlistat, but not with placebo. A faster improvement of glycemic profile and FPI was obtained with orlistat compared with the controls. Also, there was a significant reduction of lipid profile with orlistat, not reached with placebo. CONCLUSIONS: Orlistat was more effective than placebo in ameliorating inflammatory parameters such as ADN and TNF-alpha, and anthropometric parameters.

Differential effects of candesartan and olmesartan on adipose tissue activity biomarkers in type II diabetic hypertensive patients.

The aim of this study is to compare the effects of candesartan and olmesartan on insulin sensitivity-related parameters, before and after antihypertensive therapy. After a 4-week washout placebo period, 194 hypertensive (diastolic blood pressure (DBP) > or =80 mm Hg and systolic blood pressure (SBP) > or =130 mm Hg) patients with well-controlled type II diabetes were randomized to receive either 8 mg of candesartan once a day (o.d.) or 10 mg olmesartan o.d. and titrated after 1 month to 16 mg candesartan o.d. or 20 mg olmesartan o.d., respectively; the treatment period had a 1-year duration. We evaluated body weight, body mass index, SBP, DBP, glycated hemoglobin, fasting plasma glucose, M value, adiponectin (ADN), resistin (r), retinol-binding protein 4, visfatin, vaspin and high-sensitivity C-reactive protein (Hs-CRP) at their baseline values and after 6 and 12 months of treatment. We observed no variation in body weight or glycemic profile for either treatment. SBP and DBP were significantly reduced by both treatments (from 144+/-8/88+/-6 to 126+/-5/77+/-4 mm Hg by candesartan (P<0.001) and from 145+/-9/89+/-7 to 128+/-7/79+/-5 mm Hg by olmesartan (P<0.001)) without any difference between them. Retinol binding protein-4, r, and the vaspin value decreased in the candesartan group but not in olmesartan group. The M value, visfatin and ADN increased with candesartan, whereas no significant variations were observed with olmesartan. Both treatments resulted in a similar reduction in Hs-CRP. Although both therapies resulted in similar reductions in blood pressure, candesartan therapy was more effective than olmesartan therapy in improving insulin sensitivity.

Effects of a standardized oral fat load on vascular remodelling markers in healthy subjects.

The most adequate way to experimentally reproduce the post-prandial lipemia condition appears to be the administration of a standardized oral fat load (OFL) to fasting patients. We studied the effects of a standardized OFL on markers of vascular remodelling in healthy subjects. We enrolled 286 Caucasians aged >or= 18 of either sex. The OFL was given after a 12-h fast. Blood samples were drawn before and 3, 6, 9 and 12h after the fat load. The following parameters were evaluated: body mass index (BMI), blood glucose (BG), systolic blood pressure (SBP), diastolic blood pressure (DBP), lipid profile, nitrites and nitrates, adiponectin (ADP), metalloproteinase-2 (MMP-2) and metalloproteinase-9 (MMP-9). High density lipoprotein-cholesterol (HDL-C) decrease was present in subjects after 6h. Triglycerides (Tg) change was observed after 6h. Nitrites/nitrates variation was observed after 6 and 9h during OFL. Adiponectin level was decreased after 6 and 9h during OFL. Both MMP-2 and MMP-9 levels were higher after 6h during OFL. We observed that nitrites/nitrates and ADP significantly decreased and MMP-2 and MMP-9 significantly increased after a standardized OFL. Other studies need to confirm the direct acute effects of post-prandial lipemia on vascular damage.

Modification of vascular and inflammation biomarkers after OGTT in overweight healthy and diabetic subjects.

We evaluated the effect of an oral glucose tolerance test (OGTT) on the level of biomarkers of vascular remodelling. We enrolled 256 Caucasian overweight healthy subjects (H) and 274 overweight type 2 diabetic patients (D). All patients underwent basal measurements of blood glucose (BG), nitrites/nitrates, adiponectin (ADP), matrix metalloproteinase-2 (MMP-2), and matrix metalloproteinase-9 (MMP-9) before and after OGTT. Nitrites/nitrates decrease was present after 60, 90, 120, and 180 min in both groups. Nitrite/nitrate levels were decreased at baseline, after 30 and 60 min in D group compared to H group. ADP decrease was present after 90, 120, and 180 min, in both groups. ADP levels were lower in D group than in H group during OGTT. MMP-2 increase was present after 60, 90, and 120 min in H group, while MMP-2 increase was observed after 90, 120, and 180 min in D group. MMP-2 levels were higher in D group than in H group during OGTT. MMP-9 increase was present in H group after 60, 90, 120, and 180 min, while MMP-9 increase was observed after 90, 120, and 180 min in D group. MMP-9 levels were higher in D group than in H group during OGTT. Postprandial glycemia induces an acute increase in biomarkers of vascular remodelling.

Candesartan effect on inflammation in hypertension.

The aim of this study was to evaluate the effect of candesartan on inflammatory biomarkers in hypertensive patients with and without type 2 diabetes mellitus after a standardized oral fat load (OFL). A total of 219 patients were enrolled: 106 patients were assigned to the non-diabetic hypertensive (NH) group, and 113 to the diabetic hypertensive (DH) group. All patients received candesartan therapy for 6 months and underwent a standardized OFL at baseline and after 6 months of therapy. We evaluated systolic blood pressure (SBP) and diastolic blood pressure (DBP), blood glucose (BG), triglycerides (Tg), soluble intercellular adhesion molecule-1 (sICAM-1), interleukin-6 (IL-6) and high-sensitivity C reactive protein (Hs-CRP). At baseline, glycated hemoglobin, homeostasis model assessment insulin resistance index, BG, fasting plasma insulin, Tg, sICAM-1, IL-6 and Hs-CRP in the DH group were significantly higher, whereas high-density lipoprotein-cholesterol value was significantly lower compared to NH group. After 6 months of candesartan therapy, sICAM-1, IL-6 and Hs-CRP were significantly lower compared to baseline in both groups; furthermore, there was a significant decrease of SBP and DBP values in both groups. After the OFL administered at baseline, there was an increase of Tg, sICAM-1, IL-6 and Hs-CRP in both groups. After the OFL administered after 6 months of therapy, instead, there was no significant variation of BG, Tg or sICAM-1 value in both groups, whereas there was an increase of IL-6 and Hs-CRP compared to time 0. We observed that candesartan treatment attenuated the inflammatory answer in both groups of patients, even if more efficiently in nondiabetic ones.

Modulation of adipokines and vascular remodelling markers during OGTT with acarbose or pioglitazone treatment.

Adipose tissue secretes biologically active mediators as adipokines. We evaluate the effect of pioglitazone and acarbose on adipokines and vascular remodelling markers during an oral glucose tolerance test (OGTT). Height and body weight, BMI, glycemic and lipid profile, blood pressure, Nitrites/nitrates, ADP, resistin, MMP-2, and MMP-9 were evaluated at titration beginning, after 3, 6 months, and at the study end in 473 type 2 diabetic patients. BMI and weight increased after full treatment with pioglitazone respect to acarbose. HbA(1c) decreased after titration period with pioglitazone compared to baseline and to acarbose, and after full treatment with pioglitazone compared to the end of titration period and to acarbose. FPG decreased after full treatment with pioglitazone compared to the end of titration period. PPG decreased with acarbose after titration period respect to baseline and after full treatment respect to the end of titration period. FPI and Homa index decreased after titration period with pioglitazone compared to baseline and to acarbose, and after full treatment with pioglitazone respect to the end of titration period and to acarbose. ADP increased after titration period with pioglitazone compared to baseline and to acarbose, and after full treatment with pioglitazone compared to the end of titration period and to acarbose. Resistin decreased after titration period with pioglitazone compared to baseline and to acarbose, and after full treatment with pioglitazone respect to acarbose. Pioglitazone improves glucose metabolism and insulin-resistance compared to acarbose in type 2 diabetic patients already treated with metformin and sulphonilureas.

Fenofibrate, simvastatin and their combination in the management of dyslipidaemia in type 2 diabetic patients.

OBJECTIVE: To evaluate the efficacy of fenofibrate, simvastatin or their combination in type 2 diabetic patients with combined dyslipidaemia. RESEARCH DESIGN AND METHODS: 241 patients, who had never previously taken lipid-lowering medications, received fenofibrate 145 mg/day, or simvastatin 40 mg/day, or fenofibrate 145 mg/day + simvastatin 40 mg/day combination for 12 months. We evaluated lipids, glycaemic, haemostatic, and inflammatory variables at baseline, and after 6 and 12 months. RESULTS: After 12 months total cholesterol (TC), LDL cholesterol (LDL-C) and triglycerides (Tg) decreased while HDL cholesterol (HDL-C) increased in all groups, even if the values obtained with fenofibrate + simvastatin were the best. At the end of the study apolipoprotein A-1 (Apo A-1) increased with fenofibrate + simvastatin, while apolipoprotein B (Apo B) decreased in all groups compared to baseline. Plasminogen activator inhibitor-1 (PAI-1) and high-sensitivity C reactive protein (hs-CRP) decreased after 12 months compared to baseline with simvastatin, and with fenofibrate + simvastatin even if the value obtained with fenofibrate-simvastatin was the lowest. After 12 months, fibrinogen (Fg) decreased compared to baseline with fenofibrate + simvastatin. LIMITATIONS: This study has some limitations. The first one is the relatively small sample of studied patients. The second one is the lack of an advanced lipid proteins evaluation, such as lipoprotein subfraction changes in the different treatment regimen. Finally, we have not selected patients that could show the best response to fibrate (i.e.: hypertriglyceridemics) or statins (i.e.: hypercholesterolemics) monotherapy, so the effect of these drugs administered alone may have been partly attenuated. CONCLUSIONS: Fenofibrate + simvastatin association improved lipid parameters, prothrombotic and inflammatory factors, and appeared to have a good tolerability profile over 12 months of therapy.