Professional Documents
Culture Documents
Background: Nonhealing ulcers affect patient quality of life and trials (9 therapies) met eligibility criteria. There was moderate-
impose a substantial financial burden on the health care system. strength evidence for improved healing with keratinocyte therapy
(RR, 1.57 [CI, 1.16 to 2.11]) compared with standard care and
Purpose: To systematically evaluate benefits and harms of ad- low-strength evidence for biological dressing and a biological skin
vanced wound care therapies for nonhealing diabetic, venous, and equivalent compared with standard care. One small trial of arterial
arterial ulcers. ulcers reported improved healing with a biological skin equivalent
Data Sources: MEDLINE (1995 to June 2013), the Cochrane Li- compared with standard care. Overall, strength of evidence was
brary, and reference lists. low for ulcer healing and low or insufficient for time to complete
healing.
Study Selection: English-language randomized trials reporting ulcer
healing or time to complete healing in adults with nonhealing ulcers Limitations: Only studies of products approved by the U.S. Food
treated with advanced therapies. and Drug Administration were reviewed. Studies were predomi-
nantly of fair or poor quality. Few trials compared 2 advanced
Data Extraction: Study characteristics, outcomes, adverse events, therapies.
study quality, and strength of evidence were extracted by trained
researchers and confirmed by the principal investigator. Conclusion: Compared with standard care, some advanced wound
care therapies may improve the proportion of ulcers healed and
Data Synthesis: For diabetic ulcers, 35 trials (9 therapies) reduce time to healing, although evidence is limited.
met eligibility criteria. There was moderate-strength evidence for
Primary Funding Source: Department of Veterans Affairs, Veterans
improved healing with a biological skin equivalent (relative risk
Health Administration, Office of Research and Development, Qual-
[RR], 1.58 [95% CI, 1.20 to 2.08]) and negative pressure wound
ity Enhancement Research Initiative.
therapy (RR, 1.49 [CI, 1.11 to 2.01]) compared with standard care
and low-strength evidence for platelet-derived growth factors and Ann Intern Med. 2013;159:532-542. www.annals.org
silver cream compared with standard care. For venous ulcers, 20 For author affiliations, see end of text.
Each dot represents 1 trial. Shaded rows indicate therapies for which meta-analyses were possible (see Figure 2 and Appendix Figure 2 [available at
www.annals.org]). BSE ⫽ biological skin equivalent; HBOT ⫽ hyperbaric oxygen therapy; NPWT ⫽ negative pressure wound therapy; PDGF ⫽
platelet-derived growth factor; PRP ⫽ platelet-rich plasma.
* Shire Regenerative Medicine, San Diego, California.
† Organogenesis, Canton, Massachusetts.
‡ Compared biological dressing with BSE (Dermagraft) and found no significant difference (15).
§ Compared biological dressing with PDGF and found no significant difference (14).
储 A second study of silver cream did not report the proportion of healed ulcers (36).
¶ Found extracorporeal shock wave therapy to be more effective than HBOT (40).
Ulcers Healed: Advanced Wound Care Versus Standard Care 1; and Appendix Figure 2, available at www.annals.org).
Overall, the strength of evidence favoring advanced Although effect sizes were similar for all 3 therapies, the
therapies over standard care was low for 7 therapies and findings were statistically significant for only 2. We found
moderate for 2 therapies (Figure 1 and Appendix Table 3, moderate-strength evidence for improved healing with the
available at www.annals.org). Study design (for example, biological skin equivalent Apligraf (Organogenesis, Can-
duration of intervention and time of final outcome assess- ton, Massachusetts) (2 fair-quality trials; ARD, 21% [95%
ment) and clinical heterogeneity (for example, ulcer size, CI, 9% to 32%]; RR, 1.58 [CI, 1.20 to 2.08]; I2 ⫽ 0%)
duration, and location and comorbid conditions) pre- (19, 20) and low-strength evidence for improved healing
cluded meta-analyses of results in many situations. with platelet-derived growth factors (2 good-quality and 5
We were able to combine results from trials comparing fair-quality trials; ARD, 21% [CI, 14% to 29%]; RR, 1.45
3 advanced therapies with standard care (Figure 1; Table [CI, 1.03 to 2.05]; I2 ⫽ 85%) (23–30). The rating of low
534 15 October 2013 Annals of Internal Medicine Volume 159 • Number 8 www.annals.org
strength of evidence for the platelet-derived growth factor We found moderate-strength evidence of improved
trials was based on high heterogeneity (inconsistent results) healing with negative pressure wound therapy (1 good-
in trials of predominantly fair quality and concerns about quality trial; ARD, 14% [CI, 4% to 24%]; RR, 1.49 [CI,
lack of directness (probable enrollment of a highly selected 1.11 to 2.01]) (37) and low-strength evidence of improved
population). Low-strength evidence suggested no statisti- healing with silver cream (1 fair-quality trial; ARD, 23%
cally significant effect with the biological skin equivalent [CI, 2% to 43%]; RR, 2.45 [CI, 0.98 to 6.09]) (33).
Dermagraft (Shire Regenerative Medicine, San Diego, Cal- For the remaining therapies, either no difference was
ifornia) (3 fair-quality trials; ARD, 10% [CI, 2% to 18%]; observed between the advanced therapy (platelet-rich
RR, 1.49 [CI, 0.96 to 2.32]; I2 ⫽ 43%) (16 –18). plasma or ozone– oxygen therapy) and standard care or
Table 1. ARDs and RRs for Percentages of Healed Ulcers: Diabetic Ulcer Studies
Study, Year Treatment Control Healed Diabetic Ulcers, n/N (%) ARD* (95% CI), % RR* (95% CI)
(Reference)
Treatment Control
Blume et al, Formulated collagen Standard care 14/31 (45) 5/16 (31) 14 (⫺15 to 43) 1.45 (0.63 to 3.29)
2011 (10) gel
Reyzelman et al, Collagen (Graftskin) Standard care 32/46 (70) 18/39 (46) 23 (3 to 44) 1.51 (1.02 to 2.22)
2009 (13)
Veves et al, Collagen Standard care 51/138 (37) 39/138 (28) 9 (⫺2 to 20) 1.31 (0.93 to 1.84)
2002 (11) (Promogran†)
Donaghue et al, Collagen (Fibracol†) Standard care 24/50 (48) 9/25 (36) 12 (⫺11 to 35) 1.33 (0.73 to 2.42)
1998 (12)
Niezgoda et al, Biological dressing PDGF 18/37 (49) 10/36 (28) 21 (⫺1 to 43) 1.75 (0.94 to 3.26)
2005 (14) (OASIS‡)
Landsman et al, Biological dressing BSE (Dermagraft§) 10/13 (77) 11/13 (85) ⫺8 (⫺38 to 22) 0.91 (0.62 to 1.33)
2008 (15) (OASIS)
Pooled studies BSE (Dermagraft) Standard care 87/251 (35) 62/254 (24) 10 (2 to 18) 1.49 (0.96 to 2.32)
(16–18)
Pooled studies BSE (Apligraf㛳) Standard care 80/145 (55) 46/134 (34) 21 (9 to 32) 1.58 (1.20 to 2.08)
(19–20)
DiDomenico et al, BSE (Apligraf) Theraskin¶ 8/17 (47) 8/12 (67) ⫺20 (⫺55 to 16) 0.71 (0.37 to 1.34)
2011 (21)
Pooled studies PDGF Standard care 189/325 (58) 133/360 (37) 21 (14 to 29) 1.45 (1.03 to 2.05)
(23–30)
Aminian et al, rhPDGF Silver sulfadiazine 4/7 (57) 0/5 (0) 57 (16 to 98) 6.75 (0.44 to 102.80)
2000 (22)
d’Hemecourt et al, PDGF (becaplermin NaCMC 15/34 (44) 25/70 (36) 8 (⫺12 to 29) 1.24 (0.76 to 2.02)
1998 (30) gel)
Saad Setta et al, PRP Platelet-poor plasma 12/12 (100) 9/12 (75) 25 (⫺1 to 51) 1.32 (0.93 to 1.86)
2011 (31)
Driver et al, PRP Placebo gel ITT: 13/40 (33) ITT: 9/32 (28) ITT: 4 (⫺17 to 26) ITT: 1.16 (0.57 to 2.35)
2006 (32) PP: 13/19 (68) PP: 9/21 (43) PP: 26 (⫺4 to 55) PP: 1.60 (0.89 to 2.85)
Belcaro et al, Silver ointment Standard care 13/34 (39) 5/32 (16) 23 (2 to 43) 2.45 (0.98 to 6.09)
2010 (33)
Jacobs and Tomczak, Silver cream Advanced therapy 6/20 (30) 8/20 (40) ⫺10 (⫺39 to 19) 0.75 (0.32 to 1.77)
2010 (34)
Jude et al, 2007 (35) Silver dressing Advanced therapy 21/67 (31) 15/67 (22) 9 (⫺6 to 24) 1.40 (0.79 to 2.47)
Blume et al, NPWT Standard care 73/169 (43) 48/166 (29) 14 (4 to 24) 1.49 (1.11 to 2.01)
2008 (37)
Wang et al, HBOT Extracorporeal shock First course of First course of ⫺30 (⫺49 to ⫺10) 0.46 (0.25 to 0.84)
2011 (40) wave therapy treatment: treatment:
10/40 (25) 24/44 (55)
Löndahl et al, HBOT Sham treatment 25/48 (52) 12/42 (29) 24 (4 to 43) 1.82 (1.05 to 3.16)
2010 (41)
Duzgun et al, HBOT Standard/multidisciplinary 33/50 (66) 0/50 (0) 66 (53 to 79) 67.00 (4.22 to 1064.23)
2008 (42) wound therapy
Kessler et al, HBOT Standard/multidisciplinary 2/14 (14) 0/13 (0) 14 (⫺7 to 36) 4.67 (0.24 to 88.96)
2003 (43) wound therapy
Abidia et al, HBOT Sham treatment 5/8 (63) 0/8 (0) 63 (27 to 98) 11.00 (0.71 to 170.98)
2003 (44)
ARD ⫽ absolute risk difference; BSE ⫽ biological skin equivalent; HBOT ⫽ hyperbaric oxygen therapy; ITT ⫽ intention-to-treat population; NaCMC ⫽ sodium
carboxymethylcellulose; NPWT ⫽ negative pressure wound therapy; PDGF ⫽ platelet-derived growth factor; PP ⫽ per-protocol population; PRP ⫽ platelet-rich plasma;
rhPDGF ⫽ recombinant human platelet-derived growth factor; RR ⫽ relative risk.
* Differences in reported ARDs and given percentages for each group may vary because of rounding.
† Systagenix, Gatwick, United Kingdom.
‡ DFB Pharmaceuticals, Fort Worth, Texas.
§ Shire Regenerative Medicine, San Diego, California.
㛳 Organogenesis, Canton, Massachusetts.
¶ Soluble Systems, Newport News, Virginia.
www.annals.org 15 October 2013 Annals of Internal Medicine Volume 159 • Number 8 535
results from multiple trials of the same therapy were mixed lacked adequate power to assess the effect of an advanced
(collagen, biological skin equivalent [Dermagraft], platelet- therapy on these outcomes.
derived growth factor, or hyperbaric oxygen therapy). Venous Ulcers
Absolute risk differences typically ranged from 4% to
We identified 20 eligible trials of 9 advanced therapies
30% (Table 1). Strength of evidence was low (Appendix
for venous ulcers (33, 46 – 66). A summary of baseline
Table 3).
characteristics is presented in Appendix Table 4 (available
at www.annals.org). Enrollees were predominately older
Ulcers Healed: Comparative Effectiveness of Advanced Wound white women. Studies enrolled 16 to 309 patients, and
Care Products treatment duration ranged from 4 to 26 weeks. Mean ulcer
In 9 trials comparing one advanced therapy with an- size ranged from 1.2 to 11.1 cm2, with mean ulcer sizes of
other (Figure 1 and Table 1), ARDs ranged from 8% to greater than 10 cm2 in 4 of the 12 studies reporting. Mean
30%. One poor-quality study found that extracorporeal ulcer duration ranged from 12 to 207 weeks. Complete
shock wave therapy statistically significantly improved ul- study characteristics, including the risk-of-bias assessment,
cer healing compared with hyperbaric oxygen therapy are presented in Table 1 of the Supplement.
(ARD, 30% [CI, 10% to 49%]) (40). Strength of evidence The ulcer was described as a “leg” ulcer in 14 trials
was low for the effect on the proportion of healed ulcers (46, 47, 50, 51, 53–56, 58, 59, 61, 62, 64 – 66) and a
when one advanced therapy was compared with another. “lower-extremity” ulcer in 3 trials (52, 57, 60), whereas 3
For 6 of the 8 comparisons, data were available from only trials did not report the ulcer location but described it only
1 trial. as a “venous” ulcer (33, 48, 63). In 12 trials, diagnosis of a
venous ulcer was based on clinical signs or symptoms of
venous insufficiency (46 – 48, 51–56, 63– 65). The remain-
Other Outcomes
ing 8 trials required patients to have adequate arterial cir-
No studies reported a statistically significant improve-
culation or excluded patients with known arterial insuffi-
ment in time to ulcer healing for collagen (12, 13), biolog-
ciency (33, 50, 57– 62, 66).
ical dressings (14, 15), or silver products (35, 36) com-
pared with standard care or another advanced therapy.
Results were mixed but generally negative for biological Ulcers Healed: Advanced Wound Care Products Versus
skin equivalents compared with standard care or another Standard Care
advanced therapy (16, 18 –20), platelet-derived growth fac- Findings for our primary outcome of interest—
tor compared with standard care or another advanced ther- proportion of healed ulcers—are presented in Figure 2,
apy (14, 24 –26, 28 –30), platelet-rich plasma compared Table 2, and Appendix Table 5 (available at www
with standard care or another advanced therapy (31, 32), .annals.org). Meta-analyses (Appendix Figure 3, available
and negative pressure wound therapy compared with stan- at www.annals.org) showed an overall benefit of keratino-
dard care (38, 39). Strength of evidence was low or insuf- cyte therapy compared with standard care (2 fair-quality
ficient for all findings related to time to ulcer healing (Ap- trials; ARD, 14% [CI, 5% to 23%]; RR, 1.57 [CI, 1.16 to
pendix Table 3). One study of silver versus calcium 2.11]; I2 ⫽ 0%), with moderate strength of evidence (54,
dressings reported no difference between groups for a 55). Data from 3 studies of silver cream versus standard
global outcome of healed or improved ulcers (35). No care (1 good-quality and 2 fair-quality studies; ARD, 9%
studies reported on return to daily activities. [CI, ⫺4% to 23%]; RR, 1.65 [CI, 0.54 to 5.03]; I2 ⫽
No studies reported a statistically significant difference 84%) (33, 57, 58) and 2 studies of silver dressing versus
in ulcers infected during treatment or ulcer recurrence for nonsilver dressing (both of fair quality; ARD, 8% [CI,
any of the reviewed advanced therapies compared with ⫺4% to 20%]; RR, 1.27 [CI, 0.80 to 2.01]; I2 ⫽ 67%)
standard care or another advanced therapy. Fewer amputa- (59, 61, 62) showed no overall benefit of silver products.
tions were reported in the advanced therapy group in 3 In single trials, there was low-strength evidence of im-
studies (1 each of a biological skin equivalent [19], negative proved ulcer healing with biological dressing (1 fair-quality
pressure wound therapy [37], and hyperbaric oxygen ther- trial; ARD, 20% [CI, 3% to 38%]) (47) and the biological
apy [42], all compared with standard care), and 5 studies skin equivalent Apligraf (1 fair-quality trial; ARD, 14%
reported no difference (13, 20, 41, 44, 45). Few studies [CI, 3% to 26%]) (48) compared with placebo or standard
reported on the need for revascularization or hospitaliza- care (Figure 2 and Table 2). Absolute risk differences in
tion. Overall, few deaths were reported and the results did ulcer healing ranged from 1% to 38% for collagen (46),
not show a statistically significant benefit of advanced ther- the biological skin equivalent Dermagraft (50, 51),
apies in reducing all-cause mortality (reported in 16 stud- platelet-rich plasma (56), hyperbaric oxygen therapy (66),
ies) (11, 13, 14, 20, 24, 26, 28 –30, 32, 35–37, 39, 41, 43, or intermittent pneumatic compression therapy (63) com-
44). Similarly, there were no observed reductions in pain, pared with placebo or standard care (Table 2). Results for
withdrawals due to adverse events, or allergic reactions to electromagnetic therapy were mixed (64, 65). Strength of
treatment (Tables 2 to 4 of the Supplement). Studies evidence was low (Appendix Table 5).
536 15 October 2013 Annals of Internal Medicine Volume 159 • Number 8 www.annals.org
Collagen 73 Low
Biological dressing 120 Low
BSE (Dermagraft*) 71 Low
BSE (Apligraf†) 309 Low
Keratinocyte therapy 425 Moderate
PRP 86 Low
Silver cream 203 Low
Silver dressing 255 Low
IPC 54 Low
EMT 63 Low
HBOT 16 Low
Each dot represents 1 trial. Shaded rows indicate therapies for which meta-analyses were possible (see Appendix Figure 3, available at www.annals.org).
BSE ⫽ biological skin equivalent; EMT ⫽ electromagnetic therapy; HBOT ⫽ hyperbaric oxygen therapy; IPC ⫽ intermittent pneumatic compression;
PRP ⫽ platelet-rich plasma.
* Shire Regenerative Medicine, San Diego, California.
† Organogenesis, Canton, Massachusetts.
‡ Found silver cream to be more effective than tripeptide copper cream (57).
Table 2. ARDs and RRs for Percentages of Healed Ulcers: Venous Ulcer Studies
Study, Year Treatment Control Healed Venous Ulcers, n/N (%) ARD* (95% CI), % RR* (95% CI)
(Reference)
Treatment Control
Vin et al, Collagen (Promogran†) Standard care ITT: 18/37 (49) ITT: 12/36 (33) ITT: 15 (⫺7 to 38) ITT: 1.46 (0.83 to 2.58)
2002 (46) PP: 15/37 (41) PP: 11/36 (31) PP: 10 (⫺12 to 32) PP: 1.33 (0.71 to 2.49)
Mostow et al, Biological dressing Standard care 34/62 (55) 20/58 (34) 20 (3 to 38) 1.59 (1.04 to 2.42)
2005 (47) (OASIS‡)
Falanga et al, BSE (Apligraf§) Standard care 92/146 (63) 63/129 (49) 14 (3 to 26) 1.29 (1.04 to 1.60)
1998 (48)
Krishnamoorthy BSE (Dermagraft㛳) Standard care 5/13 (38) (12 pieces) 2/13 (15) (12 pieces) 23 (⫺10 to 56) 2.50 (0.59 to 10.64)
et al, 2003 5/13 (38) (4 pieces) 2/13 (15) (4 pieces) (both) (both)
(50)
Omar et al, BSE (Dermagraft) Standard care 5/10 (50) 1/8 (12) 38 (⫺1 to 76) 4.00 (0.58 to 27.70)
2004 (51)
Pooled studies Keratinocytes Standard care 80/211 (38) 50/207 (24) 14 (5 to 23) 1.57 (1.16 to 2.11)
(54, 55)
Lindgren et al, Keratinocytes Compression 2/15 (13) 2/12 (17) ⫺3 (⫺31 to 24) 0.80 (0.13 to 4.87)
1998 (52) (cryopreserved, bandages
allogeneic cells)
Navrátilová Cryopreserved Lyophilized 21/25 (84) 20/25 (80) 4 (⫺17 to 25) 1.05 (0.81 to 1.36)
et al, keratinocytes keratinocytes
2004 (53)
Stacey et al, PRP Placebo 33/42 (79) 34/44 (77) 1 (⫺16 to 19) 1.02 (0.81 to 1.27)
2000 (56)
Pooled studies Silver cream Standard care 44/102 (43) 33/97 (34) 9 (⫺4 to 23) 1.65 (0.54 to 5.03)
(33, 57, 58) or placebo
Pooled studies Silver dressings Nonsilver 79/125 (63) 69/125 (55) 8 (⫺4 to 20) 1.27 (0.80 to 2.01)
(59, 61, 62) dressings
Harding et al, Silver dressing Silver dressing 24/145 (17) 21/136 (15) 1 (⫺7 to 10) 1.07 (0.63 to 1.83)
2012 (60) (AQUACEL¶) (Urgotul**)
Bishop et al, Silver cream Tripeptide 6/28 (21) 0/29 (0) 21 (6 to 37) 13.45 (0.79 to 228.07)
1992 (57) cream
Schuler et al, IPC Unna boot 20/28 (71) 15/25 (60) 11 (⫺14 to 37) 1.19 (0.80 to 1.77)
1996 (63) dressing
Ieran et al, EMT Sham treatment 12/18 (67) 6/19 (32) 35 (5 to 65) 2.11 (1.01 to 4.42)
1990 (64)
Kenkre et al, EMT Sham treatment 2/10 (20) 2/9 (22) ⫺2 (⫺39 to 35) 0.90 (0.16 to 5.13)
1996 (65)
Hammarlund and HBOT Sham treatment 2/8 (25) 0/8 (0) 25 (⫺8 to 58) 5.00 (0.28 to 90.18)
Sundberg,
1994 (66)
ARD ⫽ absolute risk difference; BSE ⫽ biological skin equivalent; EMT ⫽ electromagnetic therapy; HBOT ⫽ hyperbaric oxygen therapy; IPC ⫽ intermittent pneumatic
compression; PRP ⫽ platelet-rich plasma; RR ⫽ relative risk.
* Differences in reported ARDs and given percentages for each group may vary because of rounding.
† Systagenix, Gatwick, United Kingdom.
‡ DFB Pharmaceuticals, Fort Worth, Texas.
§ Organogenesis, Canton, Massachusetts.
㛳 Shire Regenerative Medicine, San Diego, California.
¶ ConvaTec Professional Services, Skillman, New Jersey.
** Urgo Medical, Loughborough, United Kingdom.
though these were not primary outcomes in the studies No studies reported amputation, revascularization or
(Tables 2 to 4 of the Supplement). Of 8 studies reporting other surgery, time to recurrence, or need for home care.
ulcer infection during treatment, only 1, a collagen study, Two studies reported hospitalization (47, 54) and 1 re-
found fewer infected ulcers in the collagen group than in ported quality of life (61, 62) and found no difference
the standard care group (46). Similarly, of 7 studies report- between advanced therapy and standard care. No signifi-
ing ulcer recurrence, only 1, a biological dressing study, cant differences were seen in all-cause mortality, study
saw fewer recurring ulcers in the active treatment group withdrawals due to adverse events, or allergic reactions to
than in the standard care group (47). One study of elec- treatment, although there were few events and studies were
tromagnetic therapy noted a significant reduction in pain not adequately powered to assess these outcomes.
from baseline to 30 days in patients receiving electromag-
netic therapy compared with sham treatment (65). How- Arterial Ulcers
ever, 9 other studies reported no difference in pain between We identified 1 small (n ⫽ 31), fair-quality study that
an advanced therapy and standard care or between 2 ad- enrolled patients specifically identified as having arterial
vanced therapies. ulcers (67). Mean age of the patients was 70 years, and
538 15 October 2013 Annals of Internal Medicine Volume 159 • Number 8 www.annals.org
75% were men. Mean ulcer size was 4.8 cm2; ulcer dura- studies of diabetic ulcers or venous ulcers (that is, mixed
tion was not reported. There were significant differences in cause).
ulcer healing (ARD, 46% [CI, 12% to 80%]) and time to Our literature search identified recent systematic re-
healing, suggesting that a biological skin equivalent may be views on many of the advanced therapies included in our
more effective than standard care when used on ischemic review. Although we were able to use these reviews to iden-
foot ulcers or partial open-foot amputations after revascu- tify references our search may have missed, direct compar-
larization surgery. Other outcomes did not differ signifi- isons of the findings from existing reviews and those from
cantly compared with standard care. our review are difficult. Many of the reviews included stud-
ies of ulcers that were not diabetic, venous, or arterial;
many allowed studies with any outcome that reflected heal-
DISCUSSION ing (including changes in area or volume); and some were
We found moderate-strength evidence for improved not limited to randomized, controlled trials.
healing compared with standard care in patients with dia- We identified methodological limitations. Few studies
betic ulcers treated with the biological skin equivalent Ap- provided a run-in period with carefully monitored stan-
ligraf or negative pressure wound therapy. There was also dard care to exclude patients for whom this would obviate
moderate-strength evidence for improved healing in pa- the need for advanced therapy. Although failure to exclude
tients with venous ulcers treated with keratinocyte therapy these patients may bias the study of the advanced therapy,
compared with standard care. Strength of evidence was low it may, in fact, represent a more clinically relevant situa-
for all treatment comparisons reporting time to ulcer heal- tion. Much of the existing research on advanced wound
ing, but a few therapies were associated with significant care therapies has attempted to minimize the influence of
improvement. No studies reported a significant difference ulcer area, depth, duration, and location; patient comorbid
in adverse events for any treatment comparison. conditions; and patient adherence to the treatment proto-
Although a wide range of patients were enrolled in col through strict inclusion and exclusion criteria. The
studies, many were older than 60 years, male, and white broader applicability to patients seen in many clinical set-
and were probably adherent to treatment protocols and
tings is not clear (68, 69). Results from our included stud-
had ulcers with relatively small surface areas. Most studies
ies may overestimate benefits and underestimate harms in
excluded patients with infected ulcers, and few monitored
nonstudy populations. More than half of the trials in our
adherence to standard care or intervention components.
review were industry-sponsored, and the role of the spon-
Study authors rarely reported outcomes by patient demo-
sor in study design and analysis was typically not stated. To
graphic, comorbidity, adherence, or ulcer characteristics.
minimize the potential for selective outcome reporting, we
Therefore, we found insufficient evidence about whether
efficacy differs according to these factors. excluded studies that reported changes in ulcer size without
The overall low strength of evidence reflects method- providing data on complete healing. Although we at-
ological flaws; the small number of trials; and heterogeneity tempted to minimize publication bias, the possibility of
of the comparators, study durations, and how outcomes missed publications and unreported data exists. Definitions
were assessed. For each ulcer type (diabetic, venous, or of “chronic” ulcers varied widely, and few studies were of
arterial), specific advanced wound care therapies were eval- sufficient duration to assess whether healing was main-
uated in only a few studies, often in highly selected popu- tained (69, 70).
lations and frequently with conflicting findings. Several We limited our review to studies of products approved
types of interventions were used within each category of by the U.S. Food and Drug Administration. We excluded
wound care therapy, making it difficult to determine studies with wounds of multiple causes (for example, vas-
whether results were replicable in other studies or general- cular, pressure, trauma, or surgery) if they did not report
izable to broader clinical settings. results by cause. We also excluded studies if they did not
The quality of the individual studies reviewed was pre- report healed wounds or time to complete healing. Many
dominantly fair or poor. Common factors limiting study studies report change in ulcer size, but the clinical benefit
quality were inadequate allocation concealment, lack of of change in ulcer size has not been established.
blinding (including blinding of outcome assessment), fail- Our review highlights future research needs. Although
ure to use intention-to-treat analysis methods, and failure additional randomized trials that compare advanced
to adequately describe study dropouts and withdrawals. wound care therapies with standard care are needed to rep-
Most studies compared advanced therapies with stan- licate or refute current findings, comparative effectiveness
dard care or placebo; little comparative effectiveness re- research is also needed to evaluate the relative benefits and
search has evaluated one advanced therapy against another. harms of different advanced therapies. In both effectiveness
Despite the clinical importance of arterial ulcers, we iden- and comparative effectiveness research, the sample sizes
tified only 1 study of an advanced therapy (which was should be adequate for outcome reporting according to key
compared with standard care) for this type of ulcer. Pa- patient and ulcer characteristics, including age; race; sex;
tients with arterial disease may have been included in the and ulcer size, location, and depth. In addition, patients
www.annals.org 15 October 2013 Annals of Internal Medicine Volume 159 • Number 8 539
should be followed for sufficient time to assess whether 4. Ayello EA. What does the wound say? Why determining etiology is essential
for appropriate wound care. Adv Skin Wound Care. 2005;18:98-109. [PMID:
healing has been maintained.
15788915]
Cost-effectiveness analyses are needed to assess 5. McCulloch DK, de Asia RJ. Management of diabetic foot lesions. In: Basow
whether and for whom advanced wound care products are DS, ed. UpToDate. Waltham, MA; 2012.
high-value care. Advanced wound care products are expen- 6. Mayberry JC, Moneta GL, Taylor LM Jr, Porter JM. Fifteen-year results of
sive. They may be cost-effective or even cost-saving in ap- ambulatory compression therapy for chronic venous ulcers. Surgery. 1991;109:
575-81. [PMID: 2020902]
propriate patients and ulcers when their effect on medical 7. Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of
complications and other ulcer-related costs is considered. Interventions. Version 5.1.0 [updated March 2011]. The Cochrane Collabora-
However, a substantial proportion (generally a third or tion, 2011. Accessed at www.cochrane-handbook.org on 25 April 2013.
more) of patients with nonhealing ulcers who were ran- 8. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency
in meta-analyses. BMJ. 2003;327:557-60. [PMID: 12958120]
domly assigned to standard care had complete wound heal- 9. Owens DK, Lohr KN, Atkins D, Treadwell JR, Reston JT, Bass EB, et al.
ing with this approach over study durations of 4 to 12 AHRQ series paper 5: grading the strength of a body of evidence when compar-
weeks. These findings suggest that a more rigorous “stan- ing medical interventions—Agency for Healthcare Research and Quality and the
dard care approach” than might occur in clinical practice Effective Health-Care Program. J Clin Epidemiol. 2010;63:513-23. [PMID:
19595577]
may be beneficial in many patients.
10. Blume P, Driver VR, Tallis AJ, Kirsner RS, Kroeker R, Payne WG, et al.
We conclude, on the basis of our findings from pub- Formulated collagen gel accelerates healing rate immediately after application in
lished randomized, controlled trials, that in highly con- patients with diabetic neuropathic foot ulcers. Wound Repair Regen. 2011;19:
trolled settings, advanced wound care therapies may im- 302-8. [PMID: 21371164]
prove the proportion of healed diabetic and venous ulcers 11. Veves A, Sheehan P, Pham HT. A randomized, controlled trial of Pro-
mogran (a collagen/oxidized regenerated cellulose dressing) vs standard treatment
compared with standard care in adults with nonhealing in the management of diabetic foot ulcers. Arch Surg. 2002;137:822-7. [PMID:
ulcers. A few therapies may reduce the time to ulcer heal- 12093340]
ing, but evidence is limited. Limitations in the quality of 12. Donaghue VM, Chrzan JS, Rosenblum BI, Giurini JM, Habershaw GM,
evidence and the number of studies directly comparing Veves A. Evaluation of a collagen-alginate wound dressing in the management of
diabetic foot ulcers. Adv Wound Care. 1998;11:114-9. [PMID: 9729942]
different advanced therapies decrease the strength and gen- 13. Reyzelman A, Crews RT, Moore JC, Moore L, Mukker JS, Offutt S, et al.
eralizability of our findings. Clinical effectiveness of an acellular dermal regenerative tissue matrix compared to
standard wound management in healing diabetic foot ulcers: a prospective, ran-
From Minneapolis Veterans Affairs Health Care System and University domised, multicentre study. Int Wound J. 2009;6:196-208. [PMID: 19368581]
of Minnesota, Minneapolis, Minnesota. 14. Niezgoda JA, Van Gils CC, Frykberg RG, Hodde JP. Randomized clinical
trial comparing OASIS Wound Matrix to Regranex Gel for diabetic ulcers. Adv
Disclaimer: The views expressed in this article are those of the authors Skin Wound Care. 2005;18:258-66. [PMID: 15942317]
15. Landsman A, Roukis TS, DeFronzo DJ, Agnew P, Petranto RD, Surpre-
and do not necessarily reflect the position or policy of the Department of
nant M. Living cells or collagen matrix: which is more beneficial in the treatment
Veterans Affairs or the U.S. government.
of diabetic foot ulcers? Wounds. 2008;20:111-6.
16. Gentzkow GD, Iwasaki SD, Hershon KS, Mengel M, Prendergast JJ,
Financial Support: This article is based on research conducted by the Ricotta JJ, et al. Use of Dermagraft, a cultured human dermis, to treat diabetic
Minneapolis Evidence-based Synthesis Program and was supported by foot ulcers. Diabetes Care. 1996;19:350-4. [PMID: 8729158]
the Department of Veterans Affairs, Veterans Health Administration, 17. Naughton G, Mansbridge J, Gentzkow G. A metabolically active human
Office of Research and Development, Quality Enhancement Research dermal replacement for the treatment of diabetic foot ulcers. Artif Organs. 1997;
Initiative. 21:1203-10. [PMID: 9384327]
18. Marston WA, Hanft J, Norwood P, Pollak R; Dermagraft Diabetic Foot
Potential Conflicts of Interest: Disclosures can be viewed at www Ulcer Study Group. The efficacy and safety of Dermagraft in improving the
healing of chronic diabetic foot ulcers: results of a prospective randomized trial.
.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum⫽M13
Diabetes Care. 2003;26:1701-5. [PMID: 12766097]
-1027.
19. Veves A, Falanga V, Armstrong DG, Sabolinski ML; Apligraf Diabetic Foot
Ulcer Study. Graftskin, a human skin equivalent, is effective in the management
Requests for Single Reprints: Nancy Greer, PhD, Minneapolis Veter- of noninfected neuropathic diabetic foot ulcers: a prospective randomized multi-
ans Affairs Health Care System, One Veterans Drive, Mail Code 152, center clinical trial. Diabetes Care. 2001;24:290-5. [PMID: 11213881]
Minneapolis, MN 55417; e-mail, nancy.greer@va.gov. 20. Edmonds M; European and Australian Apligraf Diabetic Foot Ulcer Study
Group. Apligraf in the treatment of neuropathic diabetic foot ulcers. Int J Low
Current author addresses and author contributions are available at Extrem Wounds. 2009;8:11-8. [PMID: 19189997]
www.annals.org. 21. DiDomenico L, Emch KJ, Landsman AR, Landsman A. A prospective
comparison of diabetic foot ulcers treated with either a cryopreserved skin allo-
graft or a bioengineered skin substitute. Wounds. 2011;23:184-9.
22. Aminian B, Shams M, Soveyd M, Omrani GR. Topical autologous platelet-
References derived growth factors in the treatment of chronic diabetic ulcers. Arch Iran Med.
1. Sen CK, Gordillo GM, Roy S, Kirsner R, Lambert L, Hunt TK, et al. 2000;3:55-9.
Human skin wounds: a major and snowballing threat to public health 23. Agrawal RP, Jhajharia A, Mohta N, Dogra R, Chaudhari V, Nayak KC.
and the economy [Editorial]. Wound Repair Regen. 2009;17:763-71. [PMID: Use of a platelet-derived growth factor gel in chronic diabetic foot ulcers. Diabetic
19903300] Foot Journal. 2009;12:80-8.
2. Jones KR, Fennie K, Lenihan A. Evidence-based management of chronic 24. Hardikar JV, Reddy YC, Bung DD, Varma N, Shilotri PP, Prasad ED,
wounds. Adv Skin Wound Care. 2007;20:591-600. [PMID: 17975367] et al. Efficacy of recombinant human platelet-derived growth factor (rhPDGF)
3. Spentzouris G, Labropoulos N. The evaluation of lower-extremity ulcers. based gel in diabetic foot ulcers: a randomized, multicenter, double-blind,
Semin Intervent Radiol. 2009;26:286-95. [PMID: 21326538] placebo-controlled study in India. Wounds. 2005;17:141-52.
540 15 October 2013 Annals of Internal Medicine Volume 159 • Number 8 www.annals.org
www.annals.org 15 October 2013 Annals of Internal Medicine Volume 159 • Number 8 541
65. Kenkre JE, Hobbs FD, Carter YH, Holder RL, Holmes EP. A randomized estimating the percentage of individuals excluded from a typical wound-care
controlled trial of electromagnetic therapy in the primary care management of population in such trials. Adv Skin Wound Care. 2009;22:316-24. [PMID:
venous leg ulceration. Fam Pract. 1996;13:236-41. [PMID: 8671131] 20375969]
66. Hammarlund C, Sundberg T. Hyperbaric oxygen reduced size of chronic leg 69. Wu SC, Marston W, Armstrong DG. Wound care: the role of advanced
ulcers: a randomized double-blind study. Plast Reconstr Surg. 1994;93:829-33. wound-healing technologies. J Am Podiatr Med Assoc. 2010;100:385-94.
[PMID: 8134442] [PMID: 20847352]
67. Chang DW, Sanchez LA, Veith FJ, Wain RA, Okhi T, Suggs WD. Can a 70. Sullivan N, Snyder DL, Tipton K, Uhl S, Schoelles KM. Negative Pressure
tissue-engineered skin graft improve healing of lower extremity foot wounds after Wound Therapy Devices. Technology Assessment Report. (Prepared by the
revascularization? Ann Vasc Surg. 2000;14:44-9. [PMID: 10629263] ECRI Institute under contract 290-2007-10063.) Rockville, MD: Agency for
68. Carter MJ, Fife CE, Walker D, Thomson B. Estimating the applicability of Healthcare Research and Quality; 2009. Accessed at www.ahrq.gov/research
wound care randomized controlled trials to general wound-care populations by /findings/ta/negative-pressure-wound-therapy/index.html on 25 April 2013.
Papers published in the year following submission are eligible for the
award in the year of publication. First author status at the time of
manuscript submission will determine eligibility. Authors should indicate
that they wish to have their papers considered for an award when they
submit the manuscript, and they must be able to provide satisfactory
documentation of their eligibility if selected for an award. Announcement
of awards for a calendar year will occur in January of the subsequent
year. We will provide award winners with a framed certificate, a letter
documenting the award, and complimentary registration for the Ameri-
can College of Physicians’ annual meeting.
542 15 October 2013 Annals of Internal Medicine Volume 159 • Number 8 www.annals.org
Hand-search or
reviewer
suggestion (n = 19)
References included in
evidence review (n = 59)
* One article provided outcomes for both diabetic and venous ulcers.
Characteristic Collagen Biological BSE PDGF PRP Silver Products NPWT HBOT Ozone–
Dressings Oxygen
Therapy
Studies involving therapy, n 4 2 7 9 2 4 3 5 1
Total patients randomly 489 (52–276) 124 (26–98) 989 (26–281) 990 (9–382) 96 (24–72) 280 (40–134) 418 (10–341) 326 (18–100) 61
assigned (range), n
Mean age (range), y 57 (56–59) (3 trials*) 59 (58–63) 57 (56–63) (5 trials*) 58 (51–61) 57 (1 trial*) 59 (56–60) (3 trials*) 60 (53–67) (3 trials*) 64 (61–71) 63
Mean proportion of men 74 (72–77) (3 trials*) 62 (60–69) 77 (69–86) (5 trials*) 69 (60–100) (8 trials*) 80 (1 trial*) 59 (44–74) (2 trials*) 60 (53–67) 70 (50–81) 62
(range)
Mean proportion of white 63 (63–64) (2 trials*) NR 71 (69–72) (2 trials*) 83 (81–86) (3 trials*) 60 (1 trial*) NR 58 (1 trial*) NR NR
patients (range)
Mean BMI (range), kg/m2 NR 33 (1 trial*) 32 (31–32) (2 trials*) 27 (22–33) (4 trials*) NR NR NR 30 (1 trial*) NR
Mean HbA1c level 8.4 (7.9–8.6) (3 trials*) 8.3 (1 trial*) 8.6 (8.4–8.6) (2 trials*) 8.0 (7.5–8.8) (3 trials*) 7.9 (1 trial*) 8.6 (8.0–10.7) (2 trials*) 8.2 (1 trial*) 8.2 (7.9–8.8) (3 trials*) 8.6
(range), %
Mean ulcer size 3.1 (2.7–4.3) 3.5 (1.9–4.1) 2.6 (1.9–3.0) (6 trials*) 7.3 (2.7–41.5) 5.6 (3.5–9.4) 3.2 (2.2–3.7) (2 trials*) 15.7 (12.3–32.4) (2 trials*) 2.9 (2.6–3.0) (2 trials*) 4.2
(range), cm2
Mean ulcer duration 22.1 (13.0–65.4) NR 56.6 (49.0–93.6) (4 trials*) 45.1 (12.9–78.0) (5 trials*) NR 52.5 (2.1–67.6) (2 trials*) 25.7 (9.9–28.8) (2 trials*) 39.7 (32.5–41.2) (2 trials*) NR
(range), wk
Treatment duration, wk 8–12 12 4–12 8–20 12–20 4–8 3–16 2–8 12
BMI ⫽ body mass index; BSE ⫽ biological skin equivalent; HbA1c ⫽ glycosylated hemoglobin; HBOT ⫽ hyperbaric oxygen therapy; NPWT ⫽ negative pressure wound therapy; NR ⫽ not reported; PDGF ⫽ platelet-derived
growth factor; PRP ⫽ platelet-rich plasma.
* Number of trials reporting variable if less than number of all included studies.
BSE ⫽ biological skin equivalent; HBOT ⫽ hyperbaric oxygen therapy; NaCMC ⫽ sodium carboxymethylcellulose; NPWT ⫽ negative pressure wound therapy; PDGF ⫽
platelet-derived growth factor; PRP ⫽ platelet-rich plasma; RR ⫽ relative risk.
* Number evaluated for primary outcome of interest.
† Rated as high, which indicates that further research is very unlikely to change the confidence in the estimate of effect (i.e., the evidence reflects the true effect); moderate,
which indicates that further research may change confidence in the estimate of effect and may change the estimate; low, which indicates that further research is very
likely to have an important effect on the confidence in the estimate of effect and is likely to change the estimate (i.e., there is low confidence that the evidence reflects the true
effect); or insufficient, which indicates that the evidence is unavailable or does not permit a conclusion.
‡ Shire Regenerative Medicine, San Diego, California.
§ Organogenesis, Canton, Massachusetts.
㛳 Soluble Systems, Newport News, Virginia.
Study, Year (Reference) Events/Total, n/N Weight, % Relative Risk Relative Risk
M–H, Random (95% CI) M–H, Random (95% CI)
Treatment Control
BSE (Dermagraft*) vs. standard care
0.2 0.5 1 2 5
Test for subgroup differences: chi-square = 0.15; P = 0.93; I 2 = 0%
Favors control Favors treatment
BSE ⫽ biological skin equivalent; M–H ⫽ Mantel–Haenszel; PDGF ⫽ platelet-derived growth factor.
* Shire Regenerative Medicine, San Diego, California.
† Organogenesis, Canton, Massachusetts.
Characteristic Collagen Biological BSE Keratinocytes PRP Silver Products IPC EMT HBOT
Dressings
Studies involving therapy, n 1 1 3 4 1 6 1 2 1
Total patients randomly 73 120 380 (18–309) 502 (27–225) 86 771 (42–281) 54 63 (19–44) 16
assigned (range), n
Mean age (range), y 73 64 62 (60–69) 66 (63–67) (2 trials*) 71 66 (47–71) 57 68 (66–71) Median, 67
Mean proportion of men 35 42 51 (42–61) 38 (33–39) 42 42 (35–50) (5 trials*) 46 34 (26–38) 50
(range)
Mean proportion of white NR 81 79 (76–94) (2 trials*) 100 (1 trial*) NR 62 (1 trial*) NR NR NR
patients (range)
Mean BMI (range), kg/m2 28 32 30 (1 trial*) 29 (29–30) (2 trials*) NR 30 (1 trial*) 33 NR NR
Mean ulcer size 8.2 11.1 2.5 (1.2–10.7) 9.9 (8.5–10.7) (2 trials*) 4.9 6.0 (3.2–10.5) (3 trials*) 9.9 94 mg (1 trial*) 9.9
(range), cm2
Mean ulcer duration 39.9 NR 119.3 (1 trial*) 102.7 (1 trial*) 12.0 84.4 (39.5–201.1) (3 trials*) 43.8 141.1 (112.7–207.0) NR
(range), wk
Treatment duration, wk 12 12 8–12 8–12 39 4–12 26 4–13 6
BMI ⫽ body mass index; BSE ⫽ biological skin equivalent; EMT ⫽ electromagnetic therapy; HBOT ⫽ hyperbaric oxygen therapy; IPC ⫽ intermittent pneumatic compression; NR ⫽ not reported; PRP ⫽ platelet-rich plasma.
* Number of trials reporting variable if less than number of all included studies.
BSE ⫽ biological skin equivalent; EMT ⫽ electromagnetic therapy; HBOT ⫽ hyperbaric oxygen therapy; IPC ⫽ intermittent pneumatic compression; PRP ⫽ platelet-rich plasma; RR ⫽ relative risk.
* Number evaluated for primary outcome of interest.
† Rated as high, which indicates that further research is very unlikely to change the confidence in the estimate of effect (i.e., the evidence reflects the true effect); moderate, which indicates that further research may change
confidence in the estimate of effect and may change the estimate; low, which indicates that further research is very likely to have an important effect on the confidence in the estimate of effect and is likely to change the estimate
(i.e., there is low confidence that the evidence reflects the true effect); or insufficient, which indicates that the evidence is unavailable or does not permit a conclusion.
‡ DFB Pharmaceuticals, Fort Worth, Texas.
§ Shire Regenerative Medicine, San Diego, California.
㛳 Organogenesis, Canton, Massachusetts.
www.annals.org
Appendix Figure 3. Meta-analyses of the proportion of venous ulcers healed.
Study, Year (Reference) Events/Total, n/N Weight, % Relative Risk Relative Risk
Treatment Control M–H, Random (95% CI) M–H, Random (95% CI)
Michaels et al, 2009 (61, 62) 62/104 59/104 60.7 1.05 (0.83–1.32)
Test for subgroup differences: chi-square = 0.61; P = 0.74; I2 = 0% Favors control Favors treatment
M–H ⫽ Mantel–Haenszel.