Erectile dysfunction · Randomized trial · 2025
Does adding stem cells to shockwave help erection problems in diabetes?
In this trial of men with diabetes, night-time erections went from almost none to about 11 minutes over six months with shockwave alone. Stem cells plus shockwave did better, at about 22 minutes, with higher erection-function scores than either treatment alone. Rates of successful sex were similar across groups.
Helped more than either treatment alone- men with diabetes and hard-to-treat erection problems
- 33
- shockwave sessions, one a week
- 4
- months of follow-up
- 6
Good to know: Usual treatment had not helped these men.
Ji YH, Zhang YF, Tan X, et al. Stem Cell Research & Therapy, 2025 · Read the paper ↗
How this study worked
What happened
Night-time erection time
Total minutes of erection overnight (middle value) · start and 6 months
StartAfter 6 months
Show the numbers
| Group | People | Start | After 6 months |
|---|---|---|---|
| Stem cells + shockwave | 10 | 0 | 22.2 |
| Shockwave alone | 10 | 0 | 11 |
| Stem cells alone | 11 | 0 | 18.2 |
The treatment
Stem cells + shockwave
Shockwave, once a week for 4 weeks
- Sessions
- 4, one a week
- Pulses
- 5,000 per session · 1,000 at each of five spots
- Strength
- 0.09mJ/mm² 120 pulses a minute
- Numbing
- Not mentioned
- Device
- RENOVA
- Everyone also got
- One stem-cell injection into the penis the day before shockwave began
Shockwave alone
Shockwave, once a week for 4 weeks
- Sessions
- 4, one a week
- Pulses
- 5,000 per session · 1,000 at each of five spots
- Strength
- 0.09mJ/mm² 120 pulses a minute
- Numbing
- Not mentioned
- Device
- RENOVA
Compared with: Each treatment on its own: one stem-cell injection, or shockwave alone
Settings are what each study used, not a prescription.
Who took part
33 men with diabetes whose erection problems had not responded to usual treatment; 11 per group.
hollow squares: no final score
Side effects
No serious side effects. Some men given stem cells had mild local pain that settled within a week without treatment; this was not reported with shockwave alone.
All the numbers and the full appraisal
The results, in numbers.
Men with refractory diabetes-associated erectile dysfunction; 55 screened/recruited, 22 excluded before randomization, and 33 allocated (11/group). Outcomes are reported at baseline and 1, 3, and 6 months; article arm denominators conflict across narrative and tables.
Table 1 baseline n=11. The Methods paragraph reports no withdrawal in this group, but the Results participant-flow paragraph reports one hPMSCs discontinuation for unsatisfactory efficacy; the analyzed count of 11 follows the Methods paragraph. Randomized/enrolled counts remain as recorded; percentages in Table 3 are transcribed because their timepoint denominators are not consistently reconciled.
Table 1 baseline n=10. The Methods paragraph reports one LI-ESWT withdrawal for unsatisfactory efficacy; the Results participant-flow paragraph instead places that discontinuation in the hPMSCs group. Randomized/enrolled counts remain as recorded; percentages in Table 3 are transcribed because their timepoint denominators are not consistently reconciled.
Table 1 baseline n=11; participant-flow narrative reports one loss to follow-up. Randomized/enrolled counts remain as recorded; percentages in Table 3 are transcribed because their timepoint denominators are not consistently reconciled.
IIEF-EF score3 reported outcomes
IIEF-EF score
International Index of Erectile Function–Erectile Function domain (IIEF-EF) score points
The Results text reports H + L 14.11 ± 0.61, hPMSCs 11.22 ± 0.91, and LI-ESWT 12.00 ± 0.25, with H + L higher than both monotherapy groups (p<0.05). The paper identifies the primary endpoint as median change from baseline but does not define these ± terms in the numerical Results paragraph; Table 1 gives baseline medians (IQR), so these values are not charted as means or SDs.
Analysis details and source
The reported group values and ± uncertainty are reproduced only in text because the summary statistic/variability represented by ± is not specified in that results passage.
IIEF-EF score
International Index of Erectile Function–Erectile Function domain (IIEF-EF) score points
The Results text reports H + L 16.90 ± 0.43, hPMSCs 13.11 ± 0.60, and LI-ESWT 13.75 ± 0.86, with H + L higher than both monotherapy groups (p<0.05). The paper identifies the primary endpoint as median change from baseline but does not define these ± terms in the numerical Results paragraph; Table 1 gives baseline medians (IQR), so these values are not charted as means or SDs.
Analysis details and source
The reported group values and ± uncertainty are reproduced only in text because the summary statistic/variability represented by ± is not specified in that results passage.
IIEF-EF score
International Index of Erectile Function–Erectile Function domain (IIEF-EF) score points
The Results text reports H + L 19.8 ± 0.31, hPMSCs 17.50 ± 0.53, and LI-ESWT 16.67 ± 1.39, with H + L higher than both monotherapy groups (p<0.05). The paper identifies the primary endpoint as median change from baseline but does not define these ± terms in the numerical Results paragraph; Table 1 gives baseline medians (IQR), so these values are not charted as means or SDs.
Analysis details and source
The reported group values and ± uncertainty are reproduced only in text because the summary statistic/variability represented by ± is not specified in that results passage.
Total erection time6 reported outcomes
Total erection time
Nocturnal penile tumescence Rigiscan minutes
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| hPMSCs | 11 | 0 (IQR 0–1.5) | 18.2 (IQR 13.35–20.55) |
| LI-ESWT | 10 | 0 (IQR 0–1.5) | 11 (IQR 11–12) |
| H + L | 10 | 0 (IQR 0–2.5) | 22.2 (IQR 15.2–30.25) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
At 6 months, medians (Q1,Q3) were 18.20 (13.35,20.55) min for hPMSCs, 11.00 (11.00,12.00) for LI-ESWT, and 22.20 (15.20,30.25) for H + L; omnibus χ²=13.68, p=0.001. Text reports H + L longer than both monotherapies (pairwise p<0.01).
Analysis details and source
Arm-level summaries are reported as medians (Q1,Q3) in Table 4; source does not provide an adjusted between-group effect estimate for this outcome. Table 4 marks hPMSCs and LI-ESWT with * and LI-ESWT and H + L with a (footnote: pairwise comparisons post-treatment, p < 0.05); no mark is shared by H + L and hPMSCs, whereas the Results text reports H + L longer than hPMSCs (Z = 92.0, p < 0.01). Both are kept as printed. Table 4 prints no group sizes; the stored n follow the analysed arm counts (hPMSCs 11, LI-ESWT 10, H + L 10; Table 2's baseline header gives 11, 10 and 11). The Methods paragraph places the two losses in LI-ESWT and H + L, the Results paragraph in hPMSCs and H + L.
Sufficient erection time
Nocturnal penile tumescence Rigiscan minutes
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| hPMSCs | 11 | 0 (IQR 0–0) | 11.4 (IQR 10.75–12) |
| LI-ESWT | 10 | 0 (IQR 0–0) | 10.7 (IQR 10.33–11.28) |
| H + L | 10 | 0 (IQR 0–0) | 11.9 (IQR 11.55–12.35) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
At 6 months, medians (Q1,Q3) were 11.40 (10.75,12.00) min for hPMSCs, 10.70 (10.33,11.28) for LI-ESWT, and 11.90 (11.55,12.35) for H + L; omnibus χ²=11.14, p=0.004. Text reports H + L longer than LI-ESWT (pairwise p<0.01).
Analysis details and source
Arm-level summaries are reported as medians (Q1,Q3) in Table 4; source does not provide an adjusted between-group effect estimate for this outcome. Table 4 prints no group sizes; the stored n follow the analysed arm counts (hPMSCs 11, LI-ESWT 10, H + L 10; Table 2's baseline header gives 11, 10 and 11). The Methods paragraph places the two losses in LI-ESWT and H + L, the Results paragraph in hPMSCs and H + L.
Average hardness at coronal sulcus
Nocturnal penile tumescence Rigiscan Rigidity reading; unit not stated
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| hPMSCs | 11 | 31 (IQR 23.5–33.5) | 47 (IQR 46.5–48.5) |
| LI-ESWT | 10 | 29.5 (IQR 26.25–33.75) | 47.5 (IQR 44.25–48) |
| H + L | 10 | 29 (IQR 23–36.5) | 49 (IQR 48–50.5) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Baseline and six-month medians (Q1,Q3): Table 4 reports baseline hPMSCs 31 (23.5, 33.5), LI-ESWT 29.5 (26.25, 33.75), and H + L 29 (23, 36.5). Six-month medians (Q1,Q3): hPMSCs 47.00 (46.5, 48.5); LI-ESWT 47.50 (44.25, 48); H + L 49.00 (48, 50.5). The source reports omnibus χ²=5.80, p=0.055; no significant between-group difference is reported.
Analysis details and source
Arm-level summaries are reported as medians (Q1,Q3) in Table 4; source does not provide an adjusted between-group effect estimate for this outcome. Table 4 prints no group sizes; the stored n follow the analysed arm counts (hPMSCs 11, LI-ESWT 10, H + L 10; Table 2's baseline header gives 11, 10 and 11). The Methods paragraph places the two losses in LI-ESWT and H + L, the Results paragraph in hPMSCs and H + L.
Average hardness at base
Nocturnal penile tumescence Rigiscan Rigidity reading; unit not stated
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| hPMSCs | 11 | 28 (IQR 21–30.5) | 46 (IQR 44.5–47) |
| LI-ESWT | 10 | 27 (IQR 22–32) | 44 (IQR 42–45.75) |
| H + L | 10 | 26 (IQR 21.5–32.5) | 47 (IQR 45–48) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Baseline and six-month medians (Q1,Q3): Table 4 reports baseline hPMSCs 28 (21, 30.5), LI-ESWT 27 (22, 32), and H + L 26 (21.5, 32.5). Six-month medians (Q1,Q3): hPMSCs 46.00 (44.5, 47); LI-ESWT 44.00 (42, 45.75); H + L 47.00 (45, 48). The source reports omnibus χ²=5.13, p=0.077; no significant between-group difference is reported.
Analysis details and source
Arm-level summaries are reported as medians (Q1,Q3) in Table 4; source does not provide an adjusted between-group effect estimate for this outcome. Table 4 prints no group sizes; the stored n follow the analysed arm counts (hPMSCs 11, LI-ESWT 10, H + L 10; Table 2's baseline header gives 11, 10 and 11). The Methods paragraph places the two losses in LI-ESWT and H + L, the Results paragraph in hPMSCs and H + L.
Increase in swelling at coronal sulcus
Nocturnal penile tumescence Rigiscan Rigiscan reading; unit not stated
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| hPMSCs | 11 | 0.5 (IQR 0.25–1) | 1.7 (IQR 1.65–1.8) |
| LI-ESWT | 10 | 0.5 (IQR 0–1) | 1.75 (IQR 1.7–1.87) |
| H + L | 10 | 1 (IQR 0–1) | 1.7 (IQR 1.7–1.8) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Baseline and six-month medians (Q1,Q3): Table 4 reports baseline hPMSCs 0.5 (0.25, 1), LI-ESWT 0.5 (0, 1), and H + L 1 (0, 1). Six-month medians (Q1,Q3): hPMSCs 1.70 (1.65, 1.8); LI-ESWT 1.75 (1.7, 1.87); H + L 1.70 (1.7, 1.8). The source reports omnibus χ²=0.23, p=0.893; no significant between-group difference is reported.
Analysis details and source
Arm-level summaries are reported as medians (Q1,Q3) in Table 4; source does not provide an adjusted between-group effect estimate for this outcome. Table 4 prints no group sizes; the stored n follow the analysed arm counts (hPMSCs 11, LI-ESWT 10, H + L 10; Table 2's baseline header gives 11, 10 and 11). The Methods paragraph places the two losses in LI-ESWT and H + L, the Results paragraph in hPMSCs and H + L.
Increase in swelling at base
Nocturnal penile tumescence Rigiscan Rigiscan reading; unit not stated
| Group | n | Baseline | Follow-up |
|---|---|---|---|
| hPMSCs | 11 | 0 (IQR 0–1) | 1.7 (IQR 1.65–1.8) |
| LI-ESWT | 10 | 1 (IQR 0–1) | 1.75 (IQR 1.7–1.87) |
| H + L | 10 | 1 (IQR 0–1) | 1.7 (IQR 1.7–1.8) |
SD: standard deviation. CI: confidence interval. IQR: interquartile range. —: not extracted.
Baseline and six-month medians (Q1,Q3): Table 4 reports baseline hPMSCs 0 (0, 1), LI-ESWT 1 (0, 1), and H + L 1 (0, 1). Six-month medians (Q1,Q3): hPMSCs 1.70 (1.65, 1.8); LI-ESWT 1.75 (1.7, 1.87); H + L 1.70 (1.7, 1.8). The source reports omnibus χ²=0.23, p=0.893; no significant between-group difference is reported.
Analysis details and source
Arm-level summaries are reported as medians (Q1,Q3) in Table 4; source does not provide an adjusted between-group effect estimate for this outcome. Table 4 prints no group sizes; the stored n follow the analysed arm counts (hPMSCs 11, LI-ESWT 10, H + L 10; Table 2's baseline header gives 11, 10 and 11). The Methods paragraph places the two losses in LI-ESWT and H + L, the Results paragraph in hPMSCs and H + L.
Participants with EHS >24 reported outcomes
Participants with EHS >2
Erection Hardness Score (EHS) reported count and percent
Table 3 reports hPMSCs 2 (18.18%), LI-ESWT 3 (30%), and H + L 2 (18.18%); group-comparison p=0.757. Parentheses are source-reported percentages; per-timepoint denominators are not consistently reconciled with the flow counts, so no binary denominator is charted.
Analysis details and source
Source percentages are transcribed as reported; no event denominators are inferred.
Participants with EHS >2
Erection Hardness Score (EHS) reported count and percent
Table 3 reports hPMSCs 2 (18.18%), LI-ESWT 3 (30%), and H + L 2 (18.18%); group-comparison p=0.757. Parentheses are source-reported percentages; per-timepoint denominators are not consistently reconciled with the flow counts, so no binary denominator is charted.
Analysis details and source
Source percentages are transcribed as reported; no event denominators are inferred.
Participants with EHS >2
Erection Hardness Score (EHS) reported count and percent
Table 3 reports hPMSCs 4 (36.30%), LI-ESWT 4 (40%), and H + L 3 (27.27%); group-comparison p=0.894. Parentheses are source-reported percentages; per-timepoint denominators are not consistently reconciled with the flow counts, so no binary denominator is charted.
Analysis details and source
Source percentages are transcribed as reported; no event denominators are inferred. The hPMSCs value 4 (36.30%) is as printed; 36.30% matches no whole-number group size for 4 participants (4 of 11 is 36.36%).
Participants with EHS >2
Erection Hardness Score (EHS) reported count and percent
Table 3 reports hPMSCs 5 (45.45%), LI-ESWT 4 (40%), and H + L 7 (70%); group-comparison p=0.045. Parentheses are source-reported percentages; per-timepoint denominators are not consistently reconciled with the flow counts, so no binary denominator is charted.
Analysis details and source
Source percentages are transcribed as reported; no event denominators are inferred.
Successful penetration (SEP-2)8 reported outcomes
Successful penetration (SEP-2)
Sexual Encounter Profile reported count and percent
Table 3 reports hPMSCs 3 (27.27%), LI-ESWT 4 (40%), and H + L 4 (36.36%); between-group p=0.866. Source percentages are transcribed; follow-up denominators are not consistently reconciled with participant flow.
Analysis details and source
Source percentages are transcribed as reported; no event denominators are inferred.
Successful intercourse (SEP-3)
Sexual Encounter Profile reported count and percent
Table 3 reports hPMSCs 2 (18.18%), LI-ESWT 3 (30%), and H + L 2 (18.18%); between-group p=0.789. Source percentages are transcribed; follow-up denominators are not consistently reconciled with participant flow.
Analysis details and source
Source percentages are transcribed as reported; no event denominators are inferred.
Successful penetration (SEP-2)
Sexual Encounter Profile reported count and percent
Table 3 reports hPMSCs 2 (18.18%), LI-ESWT 3 (30%), and H + L 2 (18.18%); between-group p=0.757. Source percentages are transcribed; follow-up denominators are not consistently reconciled with participant flow.
Analysis details and source
Counts and percentages are source-reported; no denominators are inferred.
Successful penetration (SEP-2)
Sexual Encounter Profile reported count and percent
Table 3 reports hPMSCs 3 (27.27%), LI-ESWT 4 (40%), and H + L 3 (27.27%); between-group p=0.869. Source percentages are transcribed; follow-up denominators are not consistently reconciled with participant flow.
Analysis details and source
Counts and percentages are source-reported; no denominators are inferred.
Successful penetration (SEP-2)
Sexual Encounter Profile reported count and percent
Table 3 reports hPMSCs 3 (27.27%), LI-ESWT 4 (40%), and H + L 4 (36.36%); between-group p=0.866. Source percentages are transcribed; follow-up denominators are not consistently reconciled with participant flow.
Analysis details and source
Counts and percentages are source-reported; no denominators are inferred.
Successful intercourse (SEP-3)
Sexual Encounter Profile reported count and percent
Table 3 reports hPMSCs 2 (18.18%), LI-ESWT 1 (10%), and H + L 1 (9.09%); between-group p=1.000. Source percentages are transcribed; follow-up denominators are not consistently reconciled with participant flow.
Analysis details and source
Counts and percentages are source-reported; no denominators are inferred.
Successful intercourse (SEP-3)
Sexual Encounter Profile reported count and percent
Table 3 reports hPMSCs 2 (18.18%), LI-ESWT 1 (10%), and H + L 1 (9.09%); between-group p=1.000. Source percentages are transcribed; follow-up denominators are not consistently reconciled with participant flow.
Analysis details and source
Counts and percentages are source-reported; no denominators are inferred.
Successful intercourse (SEP-3)
Sexual Encounter Profile reported count and percent
Table 3 reports hPMSCs 2 (18.18%), LI-ESWT 2 (20%), and H + L 1 (9.09%); between-group p=0.898. Source percentages are transcribed; follow-up denominators are not consistently reconciled with participant flow.
Analysis details and source
Counts and percentages are source-reported; no denominators are inferred.
How treatment was delivered
Computer-generated 1:1:1 allocation; ED medications stopped for ≥2 weeks before treatment and during study.
Third-passage human placenta-derived mesenchymal stem cells, 1×10⁷ cells/mL; one 1-mL intracavernosal injection, 0.5 mL per side.
Penile disinfection, constriction band at base for 30 min; injection into corpora cavernosa; local compression for 1 min.
RENOVA electromagnetic lithotripter; both corpora cavernosa and penile crura.
0.09 mJ/mm²; 120 pulses/min; five sites, 1,000 shocks/site (5,000/session); four weekly sessions.
hPMSC injection followed by LI-ESWT beginning 24 h later; four weekly ESWT sessions.
Baseline, 1, 3, and 6 months; IIEF-EF, EHS, SEP-2/SEP-3 and Rigiscan.
No severe adverse events were reported. Mild localized pain occurred in some participants in the hPMSCs and combination groups and resolved within one week without intervention; group counts were not given. The LI-ESWT-only arm is not described as having these events.
The article states 11 participants per randomized arm, but Tables 1–2 list n=10 for LI-ESWT. The Methods paragraph places the two losses in LI-ESWT and H + L, while the Results paragraph places them in hPMSCs and H + L, and binary percentages at follow-up do not consistently map to either account's denominators.
The stated primary endpoint is median change in IIEF-EF, and the Statistical analysis describes continuous variables both as mean ± standard deviation and as medians (interquartile range). The Results paragraph gives IIEF-EF values with ± and calls only the 6-month combined value a mean. These values are not treated as mean±SD or converted into a change estimate.
There was no sham or placebo control and follow-up ended at six months; the authors call for multicenter validation.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- The article states 11 participants per randomized arm, but Tables 1–2 list n=10 for LI-ESWT. The Methods paragraph places the two losses in LI-ESWT and H + L, while the Results paragraph places them in hPMSCs and H + L, and binary percentages at follow-up do not consistently map to either account's denominators.
- The stated primary endpoint is median change in IIEF-EF, and the Statistical analysis describes continuous variables both as mean ± standard deviation and as medians (interquartile range). The Results paragraph gives IIEF-EF values with ± and calls only the 6-month combined value a mean. These values are not treated as mean±SD or converted into a change estimate.
- There was no sham or placebo control and follow-up ended at six months; the authors call for multicenter validation.
Source-linked critical reading
Randomized comparative clinical trial
Technical source record
Complete article XML and all four tables/footnotes reviewed against original; figures were not available as local image assets. Tables 1–2 report medians (IQR) or n (%); Table 4 reports medians (Q1,Q3). Sample-size conflicts are retained as limitations.
In this study, combining placental stem cells with shock-wave treatment improved some erection-duration measures more than either treatment alone over six months; more men in the combined group reached an erection hardness score above 2, although penile hardness measured by Rigiscan did not differ significantly between groups. Success rates for penetration and intercourse did not differ significantly, and the study cannot establish long-term benefit or safety.
In men with refractory diabetic erectile dysfunction, does intracavernosal human placenta-derived mesenchymal stem-cell therapy, low-intensity shock-wave therapy, or their combination improve erectile-function measures over six months?
Prospective randomized controlled three-arm trial; 33 men were allocated 1:1:1 using a computer-generated sequence. Two participants discontinued after allocation. The publication contains internal differences in arm counts between its flow narrative and baseline tables.
Men with refractory diabetes-associated erectile dysfunction. The article states 33 eligible participants were randomized after 55 were recruited and 22 excluded; exclusions included 15 withdrawals, poor blood-pressure control (3), inadequate glycemic management (2), and incompatible comorbidities (2).
One intracavernosal injection of hPMSCs (1 mL at 1×10⁷ cells/mL; 0.5 mL per corpus cavernosum), or LI-ESWT with RENOVA electromagnetic lithotripter (0.09 mJ/mm², 120 pulses/min, five points and 5,000 shocks/session, four weekly sessions), or both, with shock waves beginning 24 hours after injection. Patients stopped ED medication for at least two weeks before intervention and during the study.
The three active groups were compared with one another; there was no sham procedure, placebo injection, or usual-care control. All participants discontinued other ED medications.
Primary endpoint was median change in IIEF-EF score from baseline, assessed through 1, 3, and 6 months. Secondary endpoints were EHS, SEP-2, SEP-3, and nocturnal Rigiscan measures: total and sufficient erection time, penile rigidity, and swelling.
At six months, the combined group had higher reported IIEF-EF values and longer median total and sufficient erection times than monotherapy groups. EHS >2 was reported in 7 (70%) in the combination group, 5 (45.45%) after hPMSCs, and 4 (40%) after LI-ESWT (p=.045). SEP-2 and SEP-3 differences were not significant. The article gives inconsistent sample denominators across its flow narrative and tables, so these percentages need cautious interpretation.
No severe adverse events were reported. Mild localized pain occurred in some hPMSC and combination recipients and resolved within one week without medical intervention. The number affected in each group is not provided.
The authors acknowledge the six-month follow-up and the need for multicenter studies. The report has inconsistent arm counts: it says 11 per randomized group, while baseline tables list LI-ESWT n=10; the Methods paragraph says one LI-ESWT and one combination participant left the study, while the Results paragraph says one hPMSCs and one combination participant discontinued, and Table 3 contains percentage denominators that do not consistently align with either account. The primary IIEF-EF outcome is described as median change, but the result text gives ± values, calling only the 6-month combined value a mean, and the statistical methods describe both mean ± standard deviation and medians (interquartile range). No sham or placebo control was used.
These results are preliminary and apply to a selected group of men with refractory diabetic ED. They do not establish comparative benefit over placebo or standard care, long-term safety, or routine clinical use of cell therapy.
A larger preregistered multicenter trial should reconcile randomization and follow-up denominators, report prespecified IIEF-EF changes with clearly defined uncertainty and between-group estimates, retain a credible control, and provide complete arm-specific adverse-event counts and longer follow-up.
Read the original article.
High-activity placenta-derived mesenchymal stem cells combined with low-intensity extracorporeal shock wave therapy for diabetic erectile dysfunction: a prospective randomized controlled trial.
Ji YH, Zhang YF, Tan X, Hou HZ, Yao Z, Zhang B. · Stem cell research & therapy · 2025
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →