Introduction

Idiopathic pulmonary fibrosis (IPF) is a lethal disease of unknown etiology, characterized by irreversible alveolar damage and lung tissue fibrosis. The progression of IPF would lead to persistent decline of lung function, attenuated exercise ability, impaired quality of life and eventually death within 3∼5 years after disease onset1. Currently there is no cure for IPF, and the clinically applied treatments, such as self-care measures, anti-fibrotic medication, and pulmonary rehabilitation, can only help relieve the symptoms and/or slow down the progression of IPF. Among them, the first-line anti-fibrotic drugs Nintedanib and Pirfenidone could only reduce the rate of decline in lung function (primarily lung forced vital capacity) rather than halt or reverse the disease progression. Furthermore, patients with IPF may discontinue Nintedanib and Pirfenidone due to adverse events such as nausea and diarrhea26. More importantly, none of these conventional medications could repair or regenerate the injured lung tissues in IPF patients. Therefore, multiple pluripotent stem cell or adult stem/progenitor cell-based regenerative strategies are being widely studied at the pre-clinical or clinical level for IPF treatment710.

Following lung injury, a number of tissue-specific stem cells, including airway basal progenitor cells, secretory progenitor cells and type 2 alveolar stem cells, are responsible for lung epithelial repair and regeneration11. Among these cell populations, airway basal progenitor cells marked by the expression of TP63 (P63), keratin 5 (KRT5) have been widely studied1214. These P63+ progenitor cells located in the airway basal layer are known to be capable of differentiating into secretory, goblet and ciliated cells and reconstituting the airway epithelium15. In the recent decade, it has been discovered that these P63+ progenitor cells in airway can be activated by multiple types of major alveolar injuries (eg. IPF, influenza infection, COVID-19), and then migrate from airway to the damaged loci in the alveolar compartment, where they play complex roles in the pathogenesis process12,13,16,17. It is proposed that the normal P63+ progenitor cells in the alveolar area are able to expand and rapidly form epithelial barriers to “band-aid” the injured lung tissue14,18. When stimulated by proper microenvironment signals, the P63+ progenitor cells could also differentiate towards alveolar epithelium lineages to facilitate lung repair and regeneration12,13,19,20. However, in the context of IPF, it has been confirmed that there are some dysplastic P63+ progenitor cells distributed near the fibrotic foci, which might be pro-fibrotic with potential to exacerbate the IPF disease2527. Therefore, in IPF patients, the function of normal or dysplastic P63+ progenitor cells still need to be investigated in details especially in clinical trials.

Previously our group had evaluated the therapeutic potential of mouse/human P63+ progenitor cell transplantation in a bleomycin-injured pulmonary fibrosis murine model. The transplanted mouse/human P63+ progenitor cells were able to engraft into the injured mouse lung tissue, re-establish the epithelial barrier, reconstitute vascularized air sac, and improve the blood oxygen levels of the mice. Transplantation of P63+ progenitor cells could also attenuate collagen deposition and myofibroblast cell proliferation in lung and eventually reduce mouse mortality9,21,22. Pre-clinical safety evaluations were also performed on both mice and monkeys following Good Laboratory Practice guidelines, showing that intrapulmonary transplantation of P63+ progenitor cells is very safe22. In a pilot clinical study we conducted in 2016, two patients with severe non-cystic fibrosis bronchiectasis were treated with autologous P63+ progenitor cell transplantation. Pulmonary functions of both patients recovered remarkably, with none aberrant cell growth or other related adverse events (AEs) during the whole follow-up time23. In our very recent report of a controlled phase 1 clinical trial, autologous P63+ progenitor cell transplantation was performed to treat 17 patients with chronic pulmonary obstructive disease (COPD). The data indicated that the cell treatment could significantly improve the gas transfer function (DLCO) of the COPD lung24. Altogether these previous pre-clinical and clinical evidences encourage us to further test the cell therapy in the pulmonary fibrosis disease at clinical stage.

Here, we report the first clinical trial treating IPF patients with autologous P63+ lung progenitor cell product (REGEND001) transplantation. The progenitor cells were isolated from a trace amount of healthy airway epithelium sample obtained via bronchoscopic brushing. After 3-5 weeks of cell expansion, up to 100 million P63+ lung progenitor cells were produced and transplanted into the lungs of IPF patients via bronchoscopy. The safety and preliminary efficacy of this novel cell therapy for treating IPF were addressed in this clinical trial.

Results

Enrollment of patients

We conducted an open-label, dose-escalation phase 1 clinical trial (CTR20210349) from July, 2021 to June, 2023, to study the safety and efficacy of autologous P63+ lung progenitor cell transplantation in patients with IPF. Supplementary table 1 showed detailed patient inclusion and exclusion information. Totally 12 patients were enrolled in the trial and assigned to one of four dose groups sequentially: 0.6 million cells/kg bodyweight (0.6 M), 1 million cells/kg bodyweight (1 M), 2 million cells/kg bodyweight (2 M) and 3.3 million cells/kg bodyweight (3.3 M) (Figure S1). After the 3rd patient of the 3.3 M group was recruited, enrollment of patients was terminated. Baseline demographic data (Table 1) showed that the trial participants in all groups were generally matched. Of note, 75% participants used anti-fibrotic drug pirfenidone as concomitant medication in this trial. None of them used Nintedanib as concomitant medication.

Baseline demographic and clinical characteristics

Clone P63+ progenitor cells from normal lung area of IPF patients

First, a small tissue sample is collected via bronchoscopy from grade 3-5 bronchi of upper lung lobes in patients with IPF. P63+ progenitor cells are selectively expanded through a pharmaceutical-grade cell cloning culture system. 50 μg/mL gentamicin sulfate is applied at the initial passage but not in the subsequent 2-5 passages to prevent microbial contamination. Each batch of cell products were confirmed to be free of bacterial contamination, or endotoxin, BSA, and antibiotic residues (Figure 1A). In IPF patients, the fibrotic foci are mainly distributed in lower or middle lobes. We selected relatively healthy lung upper lobes to collect tissue samples, which allowed us to collect the normal autologous P63+ progenitor cells instead of those pro-fibrotic cells (Figure 1B). Expression of representative progenitor cell markers KRT5 and P63 was confirmed by immunofluorescence staining and flow cytometric analysis demonstrating KRT5+ cell purity > 95% (Figure 1C and Figure S2A). For each batch of cell product, no tumorigenic potential was confirmed in the soft agar colony formation assay (Figure S2B).

Cloning and characterization of P63+ lung progenitor cells isolated from IPF patients. A, A schematic diagram illustrating the manufacturing, quality control, and clinical administration of autologous P63+ lung progenitor cell product REGEND001. B, Representative lung CT images of sampling and damaged areas of the IPF patients. Green circle: sampling area; red circle: damaged area. C, Immunostaining of clonogenic cells with basal cell markers KRT5 (green) and P63 (red). Scale bar, 30μm. D, Representative images of cell clones showing distinct morphology from different individual patients (indicated by patient numbers). Scale bar, 80μm.

We noticed that the cell cloned from different individual patients exhibited diverse morphology: some clones were regularly round while others exhibited irregular shape (Figure 1D). To further analyze the individual variances in P63+ progenitors isolated from the 12 IPF patients, we conducted cell morphological analysis on P4 passages of all subjects and correlated them with patient age and lung function. The data revealed significant negative correlation between the age of patients and P63+ progenitor cell clone roundness, which might be related to previous findings that aging can negatively impact cell properties28 (Figure 2A). Interestingly, we also observed a positive correlation between P63+ progenitor cell clone roundness and the pulmonary function indexes DLCO, FVC, and FEV1 (Figure 2A).

Cell morphology and single-cell RNA sequencing analysis of P63+ lung progenitor cells isolated from IPF patients.

A, Correlation analysis of cell colony roundness with patient age and lung functions. Each point represents one single cell colony, and each color represents an individual patient. B, Uniform manifold approximation and projection (UMAP) plot showing three subpopulations of cells cloned from a patient. C, Feature plots of representative cell markers. D, Gene ontology enrichment analysis of the differentially expressed genes identified in three subpopulations. E, Feature plots of representative variant progenitor cell markers.

To gain molecular insights into the identity composition of the cells selectively grown in the culture system, we performed single-cell sequencing on the cells. Unsupervised clustering of the cells uncovered 3 distinct subpopulations:C1, C2 and C3 (Figure 2B). The C1 population was characterized by lung progenitor classic markers KRT5 and P63 and accounts for more than half of all cells (65.03%) (Figure 2B and 2C). Gene ontology analysis of differentially expressed genes showed that C1 cells demonstrated high expression of genes involved in Wnt Signaling pathway and negative regulation of apoptotic signaling pathway, which is crucial for maintaining the identity of stem cells and the development of the lung29. Notably, we found C1 cells also have the function of positive regulation of angiogenesis and negative regulation of fibroblast proliferation (Figure 2D). The C2 populations accounts for 23.40% of the total number of cells. It also expressed KRT5, P63, and meanwhile expresses the proliferation-related marker KI67 (Figure 2B and 2C). C2 was mainly enriched in mitosis-related genes, which is consistent with its identity as a proliferative state (Figure 2D).

C3 cells accounts for 11.57% of the total number of cells, which showed very low level of KRT5 or P63 expression, but expressed high level of alveolar epithelial cell types 1 (AEC1) gene HOPX. Of note, the C3 cells also expressed a set of genes known as markers for variant progenitor cells, including CXCL17, CEACAM6, IL1RN and CLDN4 (Figure 2E)27. Further gene ontology analysis identified differentially expressed genes in C3 cells were enriched in epithelial cell development and differentiation and wound healing (Figure 2D). Altogether these data suggested that the C3 cells might represent a variant cell type undergoing differentiation, and its small percentage in the total cell population indicated that the current cell isolation and expansion system could efficiently enrich normal P63+ lung progenitor cells from IPF patients.

At final passage, P63+ progenitor cells were cultured under feeder-free conditions until reaching 85-100% confluence (Figure S2C). After harvesting P63+ progenitor cells using xeno-free TrypLE, cells were suspended in 14 mL of preservation reagent. The final product was sealed in cell preservation bags and shipped fresh to three clinical research centers via cold chain transportation (2-8°C) within 48 hours. Just before transplantation, cell suspension was warmed to room temperature, further diluted in saline, and then transplanted into bronchial segments of the middle and lower lobes by bronchoscope for 3 mL per segment. After the transplant surgery, the patients were asked to assume a supine position and breathe deeply for two hours to promote the adhesion of the transplanted P63+ progenitor cells to the injured lung area.

Safety analysis of REGEND001

Patients were followed up at baseline, 1 week, 4 weeks, 12 weeks, and 24 weeks post REGEND001 treatment for safety evaluation. Until 24 weeks after transplantation, totally 62 cases of grade I-III adverse events (AEs), but no grade IV and grade V AEs, were recorded (Table 2). The most common AE is COVID-19. COVID-19 were recorded in all 6 patients in 2 M and 3.3 M dose groups, which were not related to cell therapy but due to the outbreak of pandemic in China from December 2022 to March 2023. 22.6% (14/62) AEs were considered related to the cell therapy. Fever is one of the mostly common cell therapy-related AEs (21.4%, 1 grade II, 2 grade I), which all happened within 24 hours after REGEND001 treatment and recovered 4-9 hours afterwards. Bloody sputum is another common cell therapy-related AEs (21.4%, all grade I), which all happened within 24 hours after REGEND001 treatment and recovered within 1-3 days. The reason for bloody sputum was considered to be related to the bronchoscopic surgery but not the cell product. The grade III AEs include 2 severe COVID-19 and 2 acute bronchitis. The 2 acute bronchitis happened in two patients in 2 months and 4 months post cell transplantation respectively, and both recovered after standard treatment. Considering the medical history of the two patients, we think the acute bronchitis were not related to the cell treatment.

Adverse Eventsa

The results of blood routine and blood biochemistry in all patients were plotted in Figure S3 and S4. Overall, majority of these measures were maintained within the normal reference ranges. For the abnormal readings, most of them were clinically meaningless. The most common clinically meaningful abnormality is increasement of blood glucose level in 5 patients with diabetes. By consecutive high-resolution CT (HRCT) scanning of the chest, we did not find any signs of malignancy or new pathologies in any patient. Consistently, no clinically meaningful change of 4 tumor markers in blood sample was observed (Figure S5). 12-lead Electrocardiogram (ECG) and urine routine results were also normal after cell treatment. Altogether these data indicated that autologous P63+ progenitor cells transplantation therapy had an acceptable safety profile among IPF patients.

Change of lung functions after REGEND001 treatment

One key efficacy outcome of the current study is the change in gas transfer function after cell therapy. Lung gas transfer function is determined by effective alveolar capillary surface area, which could be quantitatively measured by diffusion capacity of lung for carbon monoxide (DLCO) and alveolar volume-corrected DLCO (DLCO/VA). In patients with IPF, progression of pulmonary fibrosis is the primary cause of decline in gas transfer function, which is a consistent and strong predictor of mortality in patients with various fibrotic lung diseases1. In current trial, we found that the 7 patients in the higher dose groups (1.0∼3.3 M) showed statistically significant increase of DLCO/VA level trend from averagely 72.06% (baseline) to 84.10% (24 weeks) of predicted value (p value=0.008). Of note, the overall change of DLCO/VA in the higher dose group was statistically significant from that in the lower dose group (p value=0.019), indicating a dose-dependent therapeutic effect (Figure 3A and S6). The results were similar when analyzing the DLCO/VA absolute value, DLCO absolute value as well as DLCO percentage (%) to its predicted values (Figure 3B and S6). Interestingly, based on previous morphological studies, this may have suggested that variations in cell roundness were not significantly associated with changes in DLCO and DLCO/VA following treatment. Altogether, above data indicated that REGEND001 treatment had potential to improve lung gas transfer function when applied at higher dose. Moreover, we also noticed that the concomitant COVID-19 events happened in 2 M and 3.3 M group post cell treatment, which might compromise some the potential benefit from REGEND001 treatment (Figure S6).

Changes of lung functions and 6MWD at different time points after REGEND001 treatment. (A-D)

The plots indicated mean (S.E.M.) changes from baseline in absolute value of DLCO/VA percentage to predicted value, DLCO absolute value, FVC percentage to predicted value and 6MWD. The low dose group included the 3 patients in the 0.6 M group, while the high dose groups included the 7 patients in the 1 M, 2 M, and 3.3 M dose groups. P values between two groups were calculated according to the area under curve (AUC).

We also analyzed the change of forced vital capacity (FVC) after REGEND001 treatment. FVC represents the airflow aspect of lung function, and its decline is the indicator of the progression of IPF30. Unlike DLCO, in the higher dose groups, we only observed minimal improvement of FVC% from averagely 76.91 % (baseline) to 77.01 % (24 weeks). The overall change of FVC in the higher dose group was statistically significant from that in the lower dose group (P=0.028) (Figure 3C and S6). In the higher dose groups, 3 of 7 patients demonstrated > 2% positive change of FVC% after REGEND001 treatment, which reached the minimal clinically important difference (MCID) for FVC% change in IPF patients31. Altogether, these data suggested that REGEND001 treatment might have potential to stabilize or slightly improve FVC when given at a high dose.

Change of exercise ability, quality of life and lung CT image

We also analyzed the 6-minute walking distance (6MWD) of patients, which is a widely used measure to assess functional exercise ability. In consistent with their improvement of lung functions, patients in higher dose groups demonstrated statistically significant improvement of 6MWD from averagely 424 m at baseline to 482 m at 24 weeks (p value=0.008) (Figure 3D and S6). The 58 m increase of 6MWD is considered clinically quite meaningful32. We also evaluated the patients’ quality of life by St. George’s Respiratory Questionnaire (SGRQ), which was originally developed and validated in obstructive airway disease and has acceptable validity and reliability in IPF patients33. Its scores range from 0 to 100, with higher scores indicating more limitations in the overall health of patient, daily life, and perceived well-being. Similar to the 6MWD result, patients in higher dose groups demonstrated decrease of SGRQ numerically from mean 32.3 at baseline to 25.8 at 24 weeks (Figure S6). The 6 points change of SGRQ score were considered clinically quite meaningful34. Altogether, these findings suggest that REGEND001 treatment, especially at higher dose groups, could help to improve the exercise ability and quality of life in IPF patients.

We also used HRCT scan to evaluate changes in the lung morphology of patients. In IPF lung, honeycombing cyst was a characteristic structure, which was developed after collapse of fibrotic alveolar wall and dilatation of terminal airways. The honeycombing pattern in CT imaging is considered permanent and independently associated with the disease prognosis1. In current study, three experts independently evaluated the lung CT images at baseline and 24 weeks after REGEND001 treatment in blinded manner. The results indicated that at baseline, all recruited IPF patients shows predominantly peripheral and lower lobe bilateral reticulation and honeycombing pattern. At 24 weeks, as most patients showed no obvious change of lung CT images post cell therapy, the Patient #005 and #006 in 1 M group demonstrated resolution of honeycombing lesion in their lower lungs (Figure 4A). We also performed three-dimensional (3D) computational visualization of reticulation and honeycombing in #006 CT images. Interestingly, the 3D data showed that the improvement of lung lesion was exclusively observed in the lower lobes (Figure 4B). One explanation could be that the cell suspension delivered through the bronchoscope tend to flow into the lower lobes due to gravity, which would result in cell concentration in lower lobes. Similar effect was also observed in the repair of mild emphysematous lesion in COPD patients in our previous study24. In consistency with the improvement of CT images in #005 and #006, these two patients demonstrated improvement of DLCO, FVC and 6MWD as well (Figure S6). Interestingly, based on the data in Figure 2, the cell cloned from #005 and #006 had vastly diverse roundness, suggesting that high clone roundness morphology might not be required for therapeutic efficacy. Altogether the above data suggested that REGEND001 treatment have potential to repair the pulmonary fibrotic structure at least in some patients in 1 M group.

Representative lung CT image before and after REGEND001 treatment.

A, Representative lung CT images of the Patient #005 and #006 at baseline and 24 weeks after REGEND001 treatment. Red circle indicated resolution of honeycomb lesion. B, Three-dimensional visualization of consecutive CT images of the Patient #006. The red zone indicated the lung damaged areas (reticulation and honeycomb) before and after cell therapy. The white circle indicated resolution of the lesion in lower lobes.

Discussion

The destruction of alveolar tissues in the lungs of IPF patients is currently irreversible. Thus, there is an urgent need for treatment options that support lung repair and/or regeneration. Previous studies have demonstrated that human P63+ lung progenitor cells can be easily isolated, expanded, and transplanted for lung regeneration23,24. In the current study, using single-cell RNA sequencing, we characterized the cells cloned from healthy upper lobe of IPF patient. Interestingly, apart from the KRT5+/P63+ basal progenitor cells, we also identified a subpopulation of “variant progenitor cells” in IPF patients. Similar variant cells have been found in healthy donors in low numbers as well as in the injured lobes of IPF patients with high abundance27. Interestingly, our cells cloned from healthy lobe have a low abundance of variant cells exactly like those cloned from healthy donors27. Moreover, unlike the previously reported profibrotic subpopulation in IPF patients, our variant progenitor cells exhibited a higher expression level of the HOPX gene, which is known to be expressed in normal type I alveolar cells. The variant cells demonstrated biological functions related to epithelial development and wound healing. These data suggest that we can clone relatively normal, but not profibrotic, progenitor cells from the healthy lobes of IPF patients, which can be used to manufacture a qualified cell therapeutic product.

Some previous studies also proposed that those abnormal, dysplastic P63+ progenitor cells located in the injured lung area of IPF patients could halt the repair process, participate in “honeycomb” structure formation and be deleterious for the lung function26,35. These findings raised “red flag” for the safe transplantation of P63+ progenitor cells in IPF patients. To address the safety issue of autologous P63+ progenitor cells, we conducted the current dose-escalation phase 1 clinical trial. Our data indicated minimal safety issues, with all events related to cell administration being mild and transient. Considering our previous pre-clinical animal data altogether22, we now know that at least the P63+ progenitor cells isolated from the healthy lobes of IPF patients are safe for intrapulmonary transplantation.

Although the trial was primarily focused on safety and not designed to assess efficacy, the data obtained also suggested meaningful improvements in lung gas transfer function DLCO, which is known to be closely related to the progression of pulmonary fibrosis. DLCO is considered to be a consistent and powerful predictor of mortality in patients with various fibrotic lung diseases including IPF1. Low DLCO level are strongly associated with diminished survival rate3640. A drop of more than 15% in DLCO was associated with a three-fold increased risk of death compared to a drop of less than 15% in IPF patients41,42. Additionally, DLCO also has a predictive effect on the exercise ability of IPF patients4346.

Earlier studies indicated that conventional anti-fibrotic drugs could not achieve improvements in DLCO in IPF patients. In a clinical trial of conventional IPF treatment, Durheim et al. analyzed data from 436 IPF patients after initiating antifibrotic therapy and reported an annual change rate of predicted DLCO% of –2.9%47. Similarly, when Zulkova et al. evaluated the effect of pirfenidone on lung function decline and survival based on 383 patients with IPF, they found that the predicted mean change in DLCO% at week 24 was –1.6%48. In contrast, our study demonstrated a +12.0% increase in DLCO/VA% and a +5.4% increase in DLCO% from baseline at 24 weeks in higher dose groups. Although there were data from only 7 patients in the higher dose groups, it still implies a potential advantage over conventional therapy, showing a statistically significant difference from patients in the lower dose group.

The current study has several limitations that need to be addressed in future research. Firstly, the sample size of the study was very small. Secondly, the absence of a placebo-controlled arm makes it difficult to draw firm conclusions about the efficacy of the treatment. Thirdly, the progenitor cells cloned from individual patients had their own morphological properties as well as genetic and epigenetic background, which might lead to distinct response to therapy in different individuals. Fourthly, the data collection and analysis of the 2 M and 3.3 M patient groups were interfered by the concomitant COVID-19 events and city lockdown policy. To overcome some of these limitations, a randomized, placebo-controlled, blind, multi-center phase 2 study with a larger sample size is currently being performed in China (Clinical Trial No. NCT06081621). Hopefully, the phase 2 trial might address the limitations of the current study and provide more reliable evidence for the efficacy of the treatment.

Methods

Study design

An open-label, dose-escalation, phase I clinical trial (CTR20210349, translated English version in Supplementary Material) was conducted at three clinical research centers (Peking Union Medical College Hospital, Shanghai Ruijin Hospital, The First Affiliated Hospital of Guangzhou Medical University) in China to evaluate the safety, tolerability, and efficacy of treating IPF patients through autologous P63+ basal progenitor cell transplantation at different dosages. This trial was approved by the National Medical Products Administration and the ethic committee of the three clinical research centers, and was conducted in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice (GCP).

Participants

Eligible patients were adults between the ages of 50 and 75 who were diagnosed with IPF based on the 2018 guidelines for diagnosis of IPF by the Pulmonary Pathology Society. Additional eligibility criteria were a DLCO that was 30 to 79% of the predicted value, a FVC that was 50% or more of the predicted value in pulmonary function tests within the preceding 3 months of patients screening. Classic radiological features of IPF were observed from their high-resolution computed tomography (HRCT) in the previous 12 months and they were tolerable with bronchofiberscopy. Full inclusion and exclusion criteria are listed in supplementary table 1. Prior to participating in the study, all participants were provided with detailed information regarding the study objectives and design, and informed consent was obtained. Patients were assigned to receive a single administration of autologous P63+ lung progenitor cells at 0.6 x 106, 1 x 106, 2 x 106, or 3.3 x 106 cells/kg bodyweight. All patients received standard IPF treatment during the study period. For more details, please see full clinical trial protocol in Supplementary Materials.

Single-cell RNA sequencing and bioinformatics

After removing feeder cells through gradient digestion, the passage 0 primary cells isolated by the pharmaceutical-grade cell cloning culture system were digested to prepare a single-cell suspension. Single cells were captured and barcoded in 10× Chromium Controller (10× Genomics). Subsequently, RNA from the barcoded cells was reverse-transcribed and sequencing libraries were prepared using Chromium Single-Cell 3’v3 Reagent Kit (10× Genomics) according to the instructions of manufacturer.

Sequencing libraries were loaded on an Illumina NovaSeq with 2 × 150 paired-end kits at Novogene, China. Raw sequencing reads were processed using the Cell Ranger v.3.1.0 pipeline from 10× Genomics. In brief, reads were demultiplexed, aligned to the human GRCh38 genome, and UMI counts were quantified per gene per cell to generate a gene-barcode matrix. Post-processing, including filtering by number of genes and mitochondrial gene content expressed per cell, was performed using the Seurat v.4.3.049. Genes were filtered out that were detected in less than three cells. A global-scaling normalization method ‘LogNormalize’ was used to normalize the data by a scale factor (10000). Next, a subset of highly variable genes was calculated for downstream analysis and a linear transformation (ScaleData) was applied as a pre-processing step. Principal component analysis (PCA) dimensionality reduction was performed with the highly variable genes as input in Seurat function RunPCA. The top 30 significant PCs were selected for two-dimensional Uniform Manifold Approximation and Projection (UMAP), implemented by the Seurat software with the default parameters. FindCluster in Seurat was used to identify cell clusters. Marker genes were identified through differential expression analysis utilizing the FindAllMarkers function in Seurat. Genes showing differential expression, observed in at least 25% of cells within the cluster, and exhibiting a fold change greater than 0.25 on a logarithmic scale, were classified as marker genes. Gene Ontology (GO) enrichment analysis of differentially expressed genes were implemented by the ClusterProfiler R package. GO terms with P-value < 0.05 were considered significantly enriched. Dot plots were used to visualize enriched terms by the enrichplot R package50.

Cell morphology analysis

P63+ progenitors were selectively expanded and and cultured to the P4 generation, and photographs of the cell clones were taken at 10x magnification. ImageJ (15.2a) was used to perform cell morphology analysis of progenitor cells from each patient, with roundness parameter employed to describe the morphology of cell clones. Pearson’s correlation scores were calculated to assess the relationship between the cell morphology and the age as well as lung function index. Statistical significance was considered at a p-value < 0.05.

Interventions

To sample autologous airway tissue, the bronchoscopy was performed by a team of certified physicians using a flexible fiberoptic bronchoscope. Before the bronchoscopy, 2% lidocaine spray anesthesia or general anesthesia was used. Peking Union Medical College Hospital and Shanghai Ruijin Hospital used spray anesthesia, while The First Affiliated Hospital of Guangzhou Medical University used general anesthesia. The total dosage of lidocaine was limited to 8.2 mg/kg bodyweight. During bronchoscopy, the patient is usually placed in a supine position. To qualify for tissue sampling, the surface of the airway should be smooth with normal color under bronchoscopy, with no new growths, ulcers, or bleeding points. Afterwards, a small amount of tissue sample was collected from the 3-5th level bronchi of IPF patients by sliding the brushes tenderly. Brushed samples were then shipped from hospital to cell factory through cold-chain transport (2-8℃) within 48 hours. In the cell factory, single-cell suspension was prepared by washing and enzymatically digestion, which were then cultured under a regenerative cell cloning (R-Clone) system24. The final product passed a series of rigorous tests, including identity, purity, sterility, residual endotoxin, viral contamination, residual BSA and residual antibiotics, etc., were qualified for this clinical trial. The qualified autologous P63+ lung progenitor cells product would be suspended in 14 mL saline, sealed in a cell preservation bag and shipped freshly to the hospital through cold-chain transport (2-8℃) within 48 hours.

To prepare for autologous P63+ lung progenitor cells transplantation, the cell suspension was re-suspended in 28 mL saline and then averagely injected into bronchial segments in middle and lower lobes (3 mL each) through the fiberoptic bronchoscope. All patients maintained a supine position for 2 hours, avoided drinking and minimized coughing for 3 hours post administration. When necessary, oral codeine was given.

In principle, all patients continued to use medications they had taken before participating this clinical trial, except Nintedanib. Nintedanib was excluded from the allowed medication list because as tyrosine kinase inhibitor, it is worried to have growth inhibition effect on the transplanted progenitor cells. Compatible concomitant IPF medications include Pirfenidone, Omeprazole, N-Acetyl Cysteine, Methylprednisolone, and Prednisone. The details of all concomitant medications were recorded on the Concomitant Drug Use page of the Case Report Form, clarifying the reason for medication/treatment, administration/treatment methods, and start and end dates.

Outcomes

The primary outcome of the study was the incidence and severity of the cell therapy-related AEs within 24 weeks after treatment. AEs were defined as abnormal laboratory results, symptoms, or signs, and were graded using the Common Terminology Criteria for Adverse Events (Version 5.0). The association between AEs and cell administration was evaluated by pulmonologists. Physician assessments, clinical laboratory evaluations (including complete blood count, serum biochemistry tests, and ECGs), and follow-up assessments were recorded for all patients with AEs, with treatment given when necessary.

The secondary outcome measures included incidence of complication related to bronchoscopy, examination of blood routine, urine routine, blood biochemistry, 12-lead Electrocardiogram (ECG) and IPF exacerbation events within 24 weeks after treatment; examination of Carcinoembryonic antigen (CEA), Neuron-specific enolase (NSE), Cytokeratin-19-fragment (CYFRA21-1) and Squamous cell carcinoma antigen (SCC) at baseline, 12 weeks and 24 weeks after treatment; evaluation of cell therapy efficacy through DLCO-SB, FVC, DLCO/VA, exercise tolerance test (6MWD), quality-of-life questionnaire SGRQ at 4 weeks, 12 weeks and 24 weeks after treatment; change of imaging of lung by HR-CT at 24 weeks after treatment. The quality of lung function test would be evaluated according to European Respiratory Society (ERS)/ American Thoracic Society (ATS) guidlines51. For more details, please see full clinical trial protocol in the Supplementary Materials and Methods.

Statistical Methods

All analyses were performed using R version 4.2.1 and GraphPad Prism version 9.0. P-value < 0.05 was considered statistically significant. The n numbers for each experiment are provided in the text or figures. Descriptive statistics were used to summarize adverse events. Continuous data were presented as mean or median, and categorical data was presented as absolute numbers and percentages of patients in each category.

The statistical significance (P-value) of area under the curve (AUC) between the low and high dose groups for % predicted DLCO/VA, % predicted FVC and 6MWD, as well as the significance between baseline and 24-week data in the high dose group for DLCO, % predicted FVC and 6MWD, were determined by an unpaired t test after normality validation with Shapiro-Wilk test, as shown in Figure 3. The AUC between the low and high dose groups for DLCO and the significance between baseline and 24-week data in high dose group for % predicted DLCO/VA were analyzed using the Mann-Whitney U test following normality validation with Shapiro-Wilk test. P values for continuous data in Table1 were calculated using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc for comparisons among four dose groups, except for Viability Rate, which was evaluated using the Kruskal-Walli’s test followed by Dunn’s post hoc test, after normality validation using the Shapiro-Wilk test. Categorical variables, such as sex, pre-study medication and concomitant medication, were compared by the Chi-square test or Fisher’s exact test. DLCO data were presented as both absolute value (mmol/min/kPa) and percentage of predicted value (%). DLCO/VA data were presented as both absolute value (mmol/min/kPa/L) and percentage of predicted value (%). FVC data were presented as both absolute value (L) and percentage of predicted value (%). DLCO, DLCO/VA, FVC, SGRQ, and 6MWD were presented as means.

Data and materials availability

Single-cell RNA sequencing generated during this study have been deposited at GEO under accession code GSE269794. The reviewer link and token for these data are as follows: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE269794 (token: cvmlwkykftgnjwx). The data will be publicly available before publication. The majority of the analysis was carried out using published and freely available software and pre-existing packages mentioned in the methods. No custom code was generated.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgements

This study was supported by National High Level Hospital Clinical Research Funding (2022-PUMCH-B-108 to Z.-J.X.), Jiangsu Province Science and Technology Special Project (Key Research and Development Plan for Social Development) Funding (BE2023727 to W.Z.), National Biopharmaceutical Technology Research Project Funding (NCTIB2023XB01011 to W.Z.) and Non-profit Central Research Institute Fund of Chinese Academy of Medical Science (2020-PT320-005 to W.Z.). Regend Therapeutics also funded the study and covered costs associated with the development of this publication. The authors also thank the patients who participated in this study and their families, and the study coordinators. We thank Jie Ren for research assistance.

Additional information

Contributions

W.Z. designed the overall clinical trial study and cell biology study. Q.L., J.-M.Q., and Z.-J.X. were the principal investigators of this trial and responsible for study design, who conducted the clinical procedures, read and commented on the full study report, contributed to the evaluation and interpretation of study data. M.Z., C.S., L.N., Z.-Y.B., Q.-R.Z., Q.L., J.-M.Q. and Z.-J.X. participated in patient enrollment, diagnosis and therapy. M.Z., C.S., Z.-Y.B., and Q.-R.Z. performed the bronchoscopy. S.-Y.Z., and Y.Z. performed the data analysis and statistical analysis. S.-Y.Z., Y.Z., T.Z. and W.Z. drafted the manuscript. All authors were involved in data interpretation and in the writing, revision and critical review of the manuscript. All authors approved the submitted version and are accountable for their contributions and the integrity of the work.

Conflict of interest

Regend Therapeutics is holding the patent of human lung progenitor cell isolation and expansion technique.

Supplementary material

Profile of current clinical trial.

In total 12 patients were recruited for the trial and assigned to 4 dose groups. All patients in the 2 M and 3.3 M groups were infected by SARS-CoV-2 during the follow-up period. Two patients (#007 and #008) in the 2 M group were excluded from the efficacy analysis because they missed the 24-week visit due to the lockdown policy during COVID-19 pandemic. The remaining 10 patients were subjected to both safety and efficacy analysis

Quality control of cultured lung progenitor cells.

A, FACS gating strategy for cell identity and purity test. KRT5 was immunostained as a marker of lung progenitor cells. SSC, side-scatter, FSC, forward scatter. B, Soft agar assay, also known as a tumorigenicity test, showing the lack of tumor formation by P63+ lung progenitor cells. Human melanoma cells were used as a positive control, while growth-arrested 3T3 cells served as a negative control. C, Representative image of feeder-free cultured lung progenitor cells before being harvested for transplantation. Scale bar, 80 μm.

Changes in clinical laboratory evaluations of blood routine following cell therapy.

The line plot displays the changes of various indicators of blood routine for patients over time, with each line representing one patient, and the patient number is marked alongside the line. The horizontal dashed line indicates the normal reference range.

Changes in clinical laboratory evaluations of blood biochemistry following cell therapy.

The line plot displays the changes of various blood biochemistry indicators over time, with each line representing a patient, and the patient number is labeled alongside the line. The horizontal dashed line indicates the normal reference range.

Changes in clinical laboratory evaluations of serum tumor markers following cell therapy.

The line plot displays the changes of various tumor markers in blood sample of patients over time, with each line representing one patient, and the patient number is labeled alongside the line. The horizontal dashed line indicates the normal reference range for that indicator.

Changes of predicted DLCO%, DLCO, predicted DLCO/VA%, DLCO/VA, predicted FVC%, FVC, 6MWD and SGRQ score at different time points after cell therapy.

The line plot displays the changes of various indicators for patients after cell therapy. Each line represents a patient, and the patient number is labeled along the line. The black triangles indicate the time when the patients were infected by SARS-CoV-2.

List of inclusion and exclusion criteria