FEATURE A Prospective, Randomized, Controlled Trial of Autologous Platelet-Rich Plasma Gel for the Treatment of Diabetic Foot Ulcers Vickie R. Driver, DPM, MS, FACFAS; Jason Hanft, DPM, FACFAS; Carelyn P. Fylling, RN, MSN; Judy M. Beriou, RN, MHA; and the AutoloGel™ Diabetic Foot Ulcer Study Group Nonhealing diabetic foot ulcers are a common cause of amputation. Emerging cellular therapies such as platelet-rich plasma gel provide ulcer management options to avoid loss of limb. The purpose of this prospective, randomized, controlled, blinded, multicenter clinical study was to evaluate the safety and efficacy of autologous platelet-rich plasma gel for the treatment of nonhealing diabetic foot ulcers. One hundred, twenty-nine (129) patients were screened; 72 completed a 7-day screening period and met the study inclusion criteria.Patients were randomized into two groups — the standard care with platelet-rich plasma gel or control (saline gel) dressing group — and evaluated biweekly for 12 weeks or until healing. Healing was confirmed 1 week following closure and monitored for another 11 weeks. An independent audit led to the exclusion of 32 patients from the final per-protocol analysis because of protocol violations and failure to complete treatment. In the 40 wounds per-protocol group, 13 out of 19 (68.4%) of the platelet-rich plasma gel and nine out of 21 (42.9%) of the control wounds healed. After adjusting for wound size outliers (n = 5), significantly more platelet-rich plasma gel (13 out of 16, 81.3%) than control gel (eight out of 19, 42.1%) treated wounds healed (P = 0.036, Fisher’s exact test). Kaplan-Meier time-to-healing also was significantly different between groups (log-rank, P = 0.0177). No treatment-related serious adverse events were reported and bovine thrombin used in the preparation of PRP did not cause Factor V inhibition. When used with good standards of care, the majority of nonhealing diabetic foot ulcers treated with autologous platelet-rich plasma gel can be expected to heal. KEYWORDS: platelet-derived growth factors; platelet releasate; platelet-rich plasma; PRP; platelet gel; diabetic foot ulcers; Factor V; bovine thrombin, autologous growth factors Ostomy/Wound Management 2006;52(6):68–87 ore than 20.8 million persons in the US have diabetes mellitus; 2002 data estimates from the Cen ters for Disease Con trol and Prevention indicate that 82,000 lower limb amputations M were performed in pers ons with diabetes.1 Characteri s tic pathological changes attri buted to autonomic and sensory neuropathy, often combi n ed with vascular disease, lead to a high-risk situati on for the person with diabetes.2,3 Pers ons Dr. Driver is Director, Clinical Research, Center for Lower Extremity Ambulatory Research, Dr. William M. Scholl College of Podiatric Medicine, Rosalind Franklin University of Medicine; and Chief, National Center for Limb Preser vation, Advocate Lutheran General Hospital, Niles, Ill. Dr. Hanft is Dire c t o r, Doctors Research Network, South Miami, Fla. Ms. Fylling is Vice President, Professional Ser vices, and Ms. Beriou is Senior Research and Clinical Coordinator, Cytomedix, Inc. Please address correspondence to: Carelyn Fylling, Cytomedix, Inc., 416 Hunge r f o rd Drive, Rockville, MD 20850; email: [email protected] Carelyn Fylling and Judy Beriou disclose they are employees and/or stockholders of Cytomedix, Inc. 68 OstomyWound Management who have had such pathology and ex peri ence trauma or published.7-21 In vi tro research has shown that platel ets infection are at high risk for devel oping ulcera ti on of the contain com ponents and properties for wound healing22; 4 foot or ankle. Pecora ro et al documen ted the causal pathlikewi s e , plasma contains fibrin matri x .23 way of an amput a ti on and found that in 81% of cases In 2001, Margolis24 p u bl i s h ed a retro s pective study faulty wound healing con tri buted to amput a ti on . analyzing the treatment re sults of 26,599 patients wi t h Healthcare practi tion ers should utilize wound tre a tments diabetic neu ropathic foot ulcers who had been tre a ted that can redu ce the ra te of f a u l ty wound healing; thus, with an auto l ogous platel et releasate. The re sults su ggest preven ting amputations. that platel et rel e a s a teprovi ded with standardized care was According to the Am erican Di a betes As s ociation, m ore more ef fective than standard care alone. than 60% of n on traumatic lower-limb amputations occur The purpose of the current stu dy was to determine the in people with diabetes; the ra te of amputation for people safety and ef fectiveness of tre a ting diabetic foot ulcers with diabetes is 10 times high er than for people wi t h o ut with PRP gel versus a con trol tre a tm ent (normal saline d i a betes; Mexican Am ericans are 1.8 times as likely, n on gel ) . The primary objective of the 12-week stu dy was to Hispanic Blacks are 2.7 times as likely, and Am erican compare the safety and inciden ce of com p l ete wound cl oIndians are three to four times as likely to ex peri ence sure bet ween PRP gel- and con tro l - treated wounds at the lower-limb amputations. Amputation ra tes are 1.4 to 2.7 end of the stu dy. Secon d a ry obj ectives inclu ded com p a rtimes high er in men than women with diabetes.5 Fryk berg ing the ra te of wound healing during the 12-week study 6 et al cites a 1998 stu dy of 67,000 diabetes-rel a ted lower and inciden ce of wound rec i d ivism among healed ulcers ex trem i ty amput a ti on (LEA) and a similar stu dy that during a 3-month fo ll ow-up peri od . Sa fety va ri a bl e s resulted in a total of 984,000 hospital days, e ach length of inclu ded adverse events, s erious adverse events, and cl i n istay avera ging 15 days. Nonhealing diabetic foot ulcers cal labora tory tests. and the re su l ting po ten tial amputations pre s ent sign i f iStudy Design and Methods cant costs to the healthcare system and redu ce patient qu a l i ty of life. This pro s pective, ra n domized, con tro ll ed , do u bl e The goal of diabetic foot ulcer treatment is to obtain blinded, mu l ti cen ter trial was condu cted under the US wound cl o su re as exped i ti o u s ly as possible. Accepted therFood and Drug Ad m i n i s tration (FDA) Inve s ti ga ti onal apeutic obj ectives and standards of c a re for diabetic foot Device Exem pti on (IDE) reg u l a ti ons. Constella Clinical ulcers inclu de wound debri dement, pressu re rel i ef in the wound area, Ostomy Wound Management 2006;52(6):68–87 a ppropri a te wound managem en t KEY POINTS ( eg, moist wound healing), infection • While some foot ulcers in persons with diabetes mellitus will heal within a reasonm a n a gement, isch emia manageable period of time when optimal care is provided, others will not; thus, increasing m en t , medical managem ent of the risk of complications such as amputation. com orbidities, and su r gical man• To test the safety and efficacy of autologous platelet-rich plasma (PRP) gel in the agem ent as needed .6 Emer ging celtreatment of chronic nonhealing diabetic foot ulcers, researchers randomly lular therapies su ch as platelet - ri ch assigned persons with nonhealing diabetic foot ulcers to a 12-week treatment periplasma (PRP) can have an adjuncod of offloading and control or offloading and PRP gel. tive role in a standardized , qu a l i ty • No short-term or long-term (24-week) safety concerns were observed but a variety tre a tm ent plan. of protocol violations reduced the power of the study to detect statistically significant differences. P l a tel et releasates, i n cluding • In the total group, 68% of PRP gel and 43% of control gel treated wounds healed. mu l tiple growth factors, have been When standardized for baseline size, the proportion of wounds healed was signifiused to treat wounds since 1985. In cantly different (81% versus 42%). vivo pro s pective con tro ll ed studies • The results of this controlled study suggest that, as part of an overall program of as well as retro s pective and co s t optimal care, PRP gel is more effective than non-PRP gel. The results also confirm ef fectiveness studies documenting the appropriateness of using moisture-retentive control dressings in all chronic the ef fect of this thera py have been wound studies. June 2006 Vol. 52 Issue 6 69 Informatics (Du rh a m , NC) served as the Cl i n i c a l Research Organization (CRO) to implem ent and monitor the tri a l , gather the data into a cen tral database, and audit the data. A data safety monitoring board provi ded eva lua ti ons of the safety - rel a ted events thro u gh o ut the tre a tment phase of the study. Independent stati s ticians were con tracted to initi a lly power the stu dy, devel op the statistical plan, and analy ze the safety and ef fectiveness data. Fourteen (14) investi ga tive sites from ac ross the co u ntry participated in the stu dy. Sites inclu ded wound care physicians’ and podiatrists’ offices, o utpatient wound care cen ters, a univers i ty - b a s ed co ll ege of pod i a tric medicine clinic, Vetera n’s Ad m i n i s tration wound care clinics, and an Army hospital limb pre s erva ti on progra m . E ach site obt a i n ed IRB approval to condu ct the stu dy. Study el i gi b i l i ty. Pers ons with type 1 or type 2 diabetes bet ween the ages of 18 and 95 with an ulcer of at least 4wee k s’ durati on were el i gi ble for the stu dy if t h ey met ad d i ti onal inclusion / exclu s i on cri teria: h em oglobin A1C <12; index foot ulcer located on the plantar, medial, or lateral aspect of the foot (including all toe su rf aces); and wound area (length x width) measurem ent bet ween 0.5 cm2 and 20 cm2, inclusive. Wounds located under a Charcot deform i ty had to be free of ac ute ch a n ges and must have under gone appropriate stru ctu ral consolidation. The index ulcer had to be cl i n i c a lly noninfected ( a l t h o u gha culture was obtained, infecti on was diagn o s ed thro u gh clinical signs and sym ptoms ra t h er than cultu re results25) and full - t h i ckness wi t h o ut ex po su re of bone, mu s cl e , ligaments, or ten dons (Un iversity of Tex a s Tre a tm en t - Ba s ed Di a betic Foot Classification System : Grade 1 A26). The protocol requ i red that post debridem ent the ulcer would be free of necro tic debri s , forei gn bodies, sinus tract s , tu n n el i n g, and undermining; compri s ed of healthy vascularized tissue; and at least 4 cm from any ad d i ti onal wound. Addition a lly, the limb had to have adequ a te perfusion as shown by examination and non-inva s ive va s c ular te s ting ankle brachial index (ABI) and toe brachial index (TBI). Women of childbe a ring age could not be pregnant or lact a ti n g ; both men and women had to be wi lling to use a med i c a lly accepted form of bi rth con trol throughout the trial and for 6 months fo ll owi n g. Pa ti ent h i s tory, physical examinati on (including a SemmesWei n s tein monofilament test for neu ropathy), and bl ood for baseline labora tory studies were obtained. 70 OstomyWound Management Approved , inform ed , s i gn ed consent stipulating that the patient was able to com p ly with all spec i f i ed care and visit requ i rem ents was sec u red from the patient, caregiver, or legal repre s en t a tive before stu dy en ro ll m ent. The Investi ga tor documen ted reasonable ex pect a ti on that the patient was medically stable and capable of com p l eting the stu dy. Stu dy exclu s i on cri teria are listed in Table 1. After meeting all initial inclusion cri teria and signing the informed consent, a ll pati ents completed a 7-day screening-peri od. This included initial excision / debri dem ent, baseline wound measurements and evaluation, and application of the control saline gel to the wound. For standardization, s h a rp debri dement guidelines were provi ded as p a rt of the protocol. Patients, a family member, or other designated parties were provi ded supplies and instru cted to change the dressings once midway thro u gh the screening peri od . The patient also was requ i red to use a fixed anklefoot orthoses that could be removed for the dressing change and at night. Crutches or a walker were used for added safety. Screening data were captured on case report forms (CRFs) for data analysis. Pati ents whose wounds reduced in area by >50% du ring the screening peri od were not ra n domized to treatment and discontinued from any furt h er study participati on because they appe a red to be able to heal wi t h o ut more advanced interven ti on . Random i z a ti on and blinding procedu res. The ra ndom i z a ti on sch edule was el ectron i c a lly gen era ted, bl ocked per investi ga ti onal cen ter, and provi ded to the site by the con tract research organization (CRO). E ach el i gible stu dy participant was assign ed to one of two tre a tm en t groups, PRP or control, and received the next available consecutive ra n dom i z a ti on nu m ber. E ach site had one designated “unbl i n ded” pers on to treat the patient (also blinded) and maintain doc u m ents in a sec u re priva te area to maintain blinding of the investi ga tor, investi ga tive site staff, patient, s ponsor, and CRO staff and mon i tor. Th i s pers on did not parti c i p a te in any other aspect of the patient’s care. The bl i n ded investi ga tors and staff measured the wounds; perform ed all tests, assessments, and debri dement; and determ i n ed wound closure. A stra tegic a lly placed dra pe prohibited the pati ent from seei n g wh i ch treatment was app l i ed to the wound. Bl ood was drawn from both the treatment and con trol pati ents to maintain blinding. PRP prepara ti on process. The PRP separa ti on system utilized in the stu dy is a newer gen era ti on , point-of - c a re TABLE 1 STUDY EXCLUSION CRITERIA • • • • • • • • • • • • • • • • • • • • • Patient currently enrolled in another investigational device or d rug trial or previously enrolled (within last 30 days) in investigative research of a device or pharmaceutical agent Ulcer decreased ≥50% in area during 7-day screening period Ulcer is due to non-diabetic etiology Patient’s blood vessels are non-compressible for ABI testing Evidence of gangrene in ulcer or on any part of the foot Patient has radiographic evidence consistent with diagnosis of acute Charcot foot Patient is currently receiving or has received radiation or chemotherapy within 3 months of randomization Topical, oral, or IV antibiotic/antimicrobial agents or medications have been used within 2 days (48 hours) of randomization Patient has received growth factor therapy (eg, autologous platelet-rich plasma gel, becaplermin, bilayered cell therapy, d e rmal substitute, extracellular matrix) within 7 days of randomization. Screening serum albumin level <2.5 g/dL Screening hemoglobin <10.5 mg/dL Screening platelet count < 100 x 109/L Patient is undergoing renal dialysis, has known immune insufficiency, known abnormal platelet activation disorders — ie, gray platelet syndrome, liver disease, active cancer (except remote basal cell of the skin), eating/nutritional, hematologic, collagen vascular disease, rheumatic disease, or bleeding disorders History of peripheral vascular repair within the 30 days of randomization Patient has known or suspected osteomyelitis Surgical correction (other than debridement) required for ulcer to heal Index ulcer has exposed tendons, ligaments, muscle, or bone Patient is known to have a psychological, developmental, physical, emotional, or social disorder, or any other situation that may interfere with compliance with study requirements and/or healing of the ulcer History of alcohol or drug abuse within the last year prior to randomization Patient has inadequate venous access for blood draw Patient has a religious or cultural conflict with the use of platelet gel treatment system for processing autologous platel ets and plasma to be used for the tre a tm ent of nonhealing wounds. This system is compri s ed of two components: a small , port a bl e cen tri f u ge to separate whole bl ood into PRP and a convenience kit that inclu des items for the bl ood draw, processing, and PRP gel application. The platel et - ri chplasma gel (AutoloGel™, Cytomedix, Inc, Rock vi ll e , Md) was used to treat pati ents in the treatment gro u p. Wounds in the con trol group were tre a ted with a saline gel (Normlgel®, Mölnlycke Health Care, Norcross, Ga). Either PRP gel or saline gel was app l i ed to the prep a red wound bed . The first step of the PRP sep a ra ti on process inclu ded performing a ven i p u n cture to draw <20 mL of bl ood , depending on the wound size, from the patient. The bl ood was spun in a small , portable cen tri f u ge for 1.5 minutes to s ep a ra te the PRP from the whole bl ood. The PRP was ex tracted into a syri n ge wh ere re a gents were ad ded to activa te the platel ets and plasma as well as to achieve proper gel con s i s tency (gel con s i s tency was usu a lly a t t a i n ed within 15 to 30 seconds); the gel then was immediately app l i ed to the wound. A contact layer dressing was applied over the gel . A foam dressing (non - a b s orben t s i de) was placed over the con t act dressing layer so the PRP gel was not absorbed. This was covered with the a b s orbent side of a foam dressing (to absorb any leaking wound exudates) and sec u red . For protection, b a rrier cream was placed on intact skin su rrounding the wound. For pati ents ra n domized to the control group, normal saline gel was applied to the wound following wound bed prep a ra tion. Similar to PRP gel application, a contact layer dressing was applied over the saline gel, followed by the non-absorbent side of a foam dressing, and covered with the absorbent side of a foam dressing before being secured. Clinical evalu a ti ons and procedures. Wounds were a s s e s s ed and measu red (length, wi d t h , and depth using a metric tape measu re at each visit. The measurem ents and o t h er wound va ri a bles including undermining or tu n n eling, ch a racteri s tics of wound exudates (ie, presence, co l or, amount, and odor), necro tic tissue, and gra nu l a ti on tissue were documented .27 Re sults rel evant to ad d i ti onal va riables/findings wi ll be published in a sep a rate paper. Care and management efforts provi ded at each treatment visit included cleansing and assessing the wound and obtaining vital signs and an interim wound history, including informati on regarding adverse events, concomitant medications, nutri ti on and wei gh t - be a ring status, and June 2006 Vol. 52 Issue 6 71 TABLE 2 LABORATORY STUDIES’ SCHEDULE week later but asked to continue wearing the offloading orthosis walker. At this Laboratory Parameter Baseline/ q2 Weeks During 6 Weeks 12 Weeks 24 Weeks visit, if the wound had Screening Treatment reopen ed, the patient was re-entered into the study at Complete blood count X X X X the same timeline (coinChemistry 7 panel, X As needed X X ciding with the retu rn vi s i t serum albumin HgbA1C X X for con ti nu ed care) and Partial thromboplastin X X X (if during X (if during continued until the wound time, prothrombin treatment) treatment) ei t h er healed or until wee k time, thrombin time 13, visit 1 without healing. If the wound stayed healed Antibody for Factor V X X X X after the 1-week interval, Fibrinogen X Antigenic fibrinogen PRN positive the patient en tered the folfibrinogen low-up phase and return ed after 3, 7, and 11 weeks. o t h er aspects of c a re since the last vi s i t . A fac i l i tyde s i gn ee During this 3-month follow-up phase, the healed wound perform ed ph l ebotomy; the unbl i n ded person perform ed was evaluated for breakdown and the patient was qu eri ed all subsequ ent gel processing. The principal investi ga tor, regarding adverse events. who did not ob s erve treatment procedu res, d i rected At the end of the 12-week treatment peri od , unhealed wound care provi s i on du ring the care visit. wounds were treated per physician protocol. The patient The need for con s i s tency in produ ct app l i c a ti on and was discharged from the site for follow-up at a fac i l i ty of maintenance of the blinding process dict a ted that dresshis/her choice. All participants were asked to return for ings were app l i ed on ly at the Investi ga tor’s site except for final Factor V testing at week 24 post-ra n domizati on date. the provi s i on of a on e - time dressing change at home End-of-study occ u rred at com p l eti on of the week 24 clinishould circumstances prevent clinic atten d a n ce . Pa ti ents cal lab evaluation, wi t h d rawal of patient consent, or death retu rn ed twi ce wee k ly at 3- or 4-day interva l s ; procedu re s of the patient. Patient procedures/instru cti ons were repe a tand processes de s c ri bed were perform ed at each visit for a ed, including a complete history and physical examination maximum of 12 weeks. Tre a tm ent con ti nued until the with testing of pedal pulses, vascular testing (ABI and TBI), wound healed , the 12-week treatment phase was comSemmes-Weinstein monofilament testi n g, l a bora tory tests, plete, or patient study participation was term i n a ted by the dressing change as per the Investigator’s order, and edu c aInvesti ga tor, s ponsor, or because the pati ent withdrew ti on with a discussion of healthcare options (see Figure 1). consent or failed to retu rn for visits. Statistical Analysis To eva luate safety, clinical labora tory tests were conducted throughout the stu dy to determine the impact of Healing rate. The power of the stu dy was determ i n ed tre a tm ent interven ti ons (see Ta ble 2). Ex pected or unexb a s ed on two data sources. The PRP gel healing ra te was pected adverse events (AE) that occ u rred du ring the determ i n ed from the healing ra te in an unpubl i s h ed diaco u rse of the stu dy, wh et h er observed by the Inve s ti ga tor betic foot ulcer retro s pective stu dy28 and feed b ack from or by the patient, were reported in det a i l . The Inve s ti ga tor clinicians who had used the PRP gel .29 The control group mon i tored the patient for AEs or lab abnorm a l i ties until healing ra te was based on a met a - a n a lysis of healing ra tes the para m eter retu rn ed to normal or it was determ i n ed in the con trol groups of 10 pro s pective studies.30 that fo ll ow-up was no lon ger necessary. Sample size. To determine the sample size, the ex pectEnd-of-trea tm ent visit procedu res. When the ed proporti on of patients with com p l etely cl o s edwounds Investi ga tor pron o u n ced the wound cl o s ed — ie, 100% was determ i n ed to be 0.60 ("t) in the PRP gel arm and epithelialized — the patient was sch edu l ed for a visit 1 0.20 ("c) in the con trol arm . To calculate the sample size, 72 OstomyWound Management Assessed for Eligibility (n = 129) h e a l ed within each group of i nvestiga tive sites and all groups com bi n ed . Because of varying Excluded en ro ll m ent at each site Patient withdrawal (bet ween one and 14 Failure to meet inclusion/exclusion criteria (n = 57) patients), sites were gro u ped for analysis purposes according to provi der setIntent to Treat ting and dem ogra ph i c s Randomization (n = 72) ( Groups I to V). Due to the va rying natu re of investi gative sites and their ability to enro ll pati en t s , sites were grouped into five categories: Randomized to Control Group Randomized to Treatment Arm te aching fac i l i ties, army (n = 32) (n = 40) facility, physicians in priva te practi ce (two site s ) , and ambulatory carecl i n i c s . The goal was to en ro ll ei ght to 21 Independent Site Audit Documentation patients in each gro u p. Patients Excluded Due to: Failure to Complete Treatment Re sults from the indepen d(n = 8); Protocol Violations (n = 24) ent audit el i m i n a ted all patients in one group, leaving four groups for per proTotal Per Protocol Treatment Group Total Per Protocol Control Group tocol (PP) analysis. (n = 19) (n = 21) Odds rati o / confiden ce l evel . The odds ra tio and Healed Wounds Entered 3-Month Follow-up Healed Wounds Entered 3-Month Follow-up 95% con f i dence interval for (n = 13) (n = 9) e ach group were calculated for the proporti on of Figure 1. Trial profile. patients with healed wo u n d s . The Mantelthe fo ll owing nu ll hypothesis (H0) and the altern a tive Haen s zel (M-H) combined odds ra tio (com bining over hypothesis (HA) were formu l a ted : groups) along with the 95% con f i dence interval was calH0: "t = "c versus HA: "t ≠ "c . culated . In additi on , the M-H test for hom ogen ei ty of A Fisher’s exact test with a 0.050, two-sided signifiodds ra tios of groups was perform ed . The M-H method cance level would have 80% power to detect the differwas used to test the hypothesis wh et h er the M-H comence between PRP gel proportion, "t, of 0.60 and conbined odds ra tio was one. trol proportion, "c, of 0.20 when the sample size in each Ad d i ti onal variables. Ot h er efficacy va ri a bles were 1) arm of treatment is 27 (ie, a total of 54 patients). The percent change in wound area at en d - of - s tu dy vi s i t sample size was increased from 54 to 72 patients to ( E O S V) from baseline (BL); 2) percent change in wound accommodate drop-outs. vo lume at EOSV from BL; 3) area cl o su re ra te per day at The primary ef f i c acy va ri a ble was the proporti on of EOSV; and 4) vo lume cl o su re ra te per day at EOSV. These patients with a healed wound. Fisher’s ex act test was ef f i c acy va ri a bles were of con ti nuous type . For each of the applied to compare the two treatments for proportions va riables, the two treatm ents were com p a red using June 2006 Vol. 52 Issue 6 73 Stu den t’s t- test. The stati s tical tests were perform ed using sof t w a re pack a ge STATA, Release 8.2 (Stata Corpora ti on, Co ll ege Station, Tex ) . Kaplan-Mei er. In ad d i ti on, the Kaplan-Mei er31,32 produ ct-limit met h od was used to analy ze time to healing of the PP wounds and the majority of wound sizes dataset . Kaplan-Mei er functi ons or curves of the PRP gel and control groups were obtained for each dataset . The log-rank test was used to test the hypothesis that the Kaplan-Mei er healing functions are the same ac ross the two treatments. Labora to ry safety. To evalu a te clinical labora tory safety, observed va lues at each visit and ch a n ges from baseline at post-baseline visits and en d point were su m m a ri zed de s c ri ptively (nu m ber (n), mean, standard deviati on (SD), minimum, median, maximum) for each treatment group. Re sults bet ween treatments at the en d point were com p a red using non - p a ra m etric analysis of va ri a n ce tech n i ques (Wilcoxon Rank Sum Test, two - s i ded). Shift analyses. The nu m ber and percent of patients from each treatment group that shifted in and out of normal ra n ge from baseline to en d point were calculated for each laboratory variable. To compare treatments for important shift changes, the number of patients whose labora tory re sults shifted from NORMAL or LOW at baseline to HIGH at endpoint and those shifting from NORMAL or HIGH at baseline to LOW at en d point were compared in separate 2 x 2 tables and differen ces were tested for statistical significanceusing two - s i ded Fisher’s ex act tests. Patients whose labora tory results did not shift from baseline ra n ge or who were NORMAL at en d point were analyzed toget her with those that had out of ra n ge shifts via ex act tests to compare treatment groups (Exact test for row [R] x column [C] tables). With these three sep a ra te analyses, va riables can be eva luated either for single directi on shifts of interest or shifts in either direction, when both HIGH and LOW devi a ti ons may be of significance. Results Initially, 129 pati ents provi ded Inform ed Con s en t forms and parti c i p a ted in active screening (see Figure 1). Of these patients, 57 (44%) were dropped from the study due to redu ction in the wound size of ≥ 50% du ring the 7-day screening period or for failure to meet the inclusion / exclusion criteria. Ul ti m a tely, 72 pati ents were en ro ll ed , each patient having one wound (index ulcer ) designated for study inclusion. 74 OstomyWound Management TABLE 3 INTENT TO TREAT GROUP: PATIENT DEMOGRAPHIC AND BASELINE WOUND VARIABLES Variable Patient Age (years) Hgb A1C Wound Area (cm2) Volume (cm3) N PRP Gel (n = 19) Mean SD Min Max N Mean Control (n= 21) SD Min Max P value 40 37 56.4 8.1 10.2 1.8 31.0 5.5 75.0 13.1 32 30 57.5 8.0 9.1 1.8 45.0 5.0 86.0 11.5 NS* NS 40 40 4.0 1.7 5.3 4.1 0.4 0.1 24.0 24.8 32 32 3.2 0.9 3.5 1.2 0.5 0.1 15.8 5.4 NS NS Treatment Characteristics Total PRP Gel (n = 40)† No. of patients Percent Sex Male Female Race Caucasian Hispanic Black Other, specify Foot Right Left Location Toe Heel * † Control (n= 32)† No. of patients Percent No. of patients Percent 32 8 80.00 20.00 27 5 84.38 15.63 59 13 81.94 18.06 26 8 5 1 65.00 20.00 12.50 2.50 18 9 3 2 56.25 28.13 9.38 6.25 44 17 8 3 61.11 23.61 11.11 4.17 23 17 57.50 42.50 18 14 56.25 43.75 41 31 56.94 43.06 13 18 32.50 45.00 14 10 43.75 31.25 27 28 37.50 38.89 NS = not significant (P >0.05) Demographic / ulcer location variables not statistically significantly different between groups In the intent-to-treat (ITT) population, the mean and standard deviations (SD) for age, HgbA1C, wound area, and volume in the two treatments were not significantly different, but the wound volume in the PRP gel group was significantly more variable than in the control group (SDs 4.1 versus 1.2, P <0.0001) (see Table 3). Ulcer location information was missing for nine patients (three in PRP gel group and six in the control group). No significant differences in patient demographics, wound distribution, or ulcer location were observed between the two treatment groups (additional data to be analyzed in a future publication). For purposes of the ITT analyses, the ITT population compri s ed all active patients wh o com p l eted the stu dy as well as those who were lost to fo llow-up, f a i l edto com p l ete the treatment, or had protocol violations. In the ITT group, 13 out of 40 patients (32.5%) in the PRP gel and nine out of 32 patients (28.1%) in the control group had co mpletely healed wounds after 12 weeks (P = 0.79). Because the results of the ITT analyses did not seem to reflect previous clinical outcomes, the study sponsor commissioned an independent audit to ensure study compliance with Good Clinical Practices (GCP) at the investigative sites. During the audit, patient source documents, Case Report Forms (CRF), and other stu dy source documents were reviewed . Five obj ective cri teria were devel oped against wh i ch all audited patient records were evalu a ted. The pro tocol vi o l a ti ons that caused exclusion of patients inclu ded use of the wrong cen tri f u ge (causing the patient not to receive the ri ght tre a tm ent — ie, PRP gel ) ; lack of June 2006 Vol. 52 Issue 6 75 source documentati on to su pport case report form en tri e s ; and inclu s i on of patients and/or wounds that did not meet the inclu s i on / exclusion cri teri a . The predeterm i n ed stati s tical plan iden ti f i ed that analysis would be perform ed on patients who com p l eted treatment; thus, patients with early termination due to re a s ons unrel a ted to the index wound and patients lost to fo ll ow-up also were exclu ded from the PP analysis. Site audits revealed that 32 out of 72 patients (44%) had protocol vi o l ations or did not meet the cri teria for participation throughout the co u rse of the study. Of the 32 exclu ded patients, 24 (75%) had protocol vi o l a ti ons and ei ght (25%) failed to com p l ete treatment. The protocol vi o l a ti ons appe a red to affect outcomes; thus, the PP dataset , at audit com p l etion, became the primary dataset for analysis — 19 patients were in the PRP gel group and 21 p a ti ents were in the con trol group. These pati ent outcomes ref l ect patients/wounds treated PP. In the PP group, on ly the proporti on of Caucasian versus non-Caucasian p a rticipants was sign i f i c a n t ly different (P = 0.02). The proporti on of Caucasians was sign i f i c a n t ly high er in the PRP gel group. No stati s ti c a lly significant differen ce bet ween the PRP gel group and the con trol group rel a ted to age , HgbA1C, wound area, wound volume, s ex , or wound loc a ti on were observed . This is the same as the ITT group (see Table 4). (Additional differen ces between baseline findings for other wound va ri a bles wi ll be ad d re s s ed in a later publication.) Efficacy outcomes. In the PP dataset , 13 of 19 (68.4%) pati ents in PRP gel and nine out of 21 (42.9%) pati ents in the con trol group healed (P = 0.125, two-sided Fisher ’s ex act test). The 95% CI for the percent proportion of completely healed wounds was 47.5% to 89.3% and 21.7% to 64.0% for PRP gel and control groups, re s pectively. The Kaplan-Mei er median time to com p l ete closure was 45 days for PRP gel compared to 85 days for con trol (log-rank test, P = 0.126) (see Figure 2). Although the inclusion/exclusion criteria included wounds with an area range at least 0.5 cm2 to no larger than 20 cm2, size frequency distributions showed that the majority (35 out of 40, 88%) of wound sizes were in the range of both ≤7.0 cm 2 in area and ≤2.0 cm 3 in volume and five (three patients/wounds in the PRP and two in the control group) were outliers. The mean area of the outliers was 10.53 cm2 (SD 8.9)and 14.63 cm2 (SD 1.6) for the PRP and control group, respectively. Because the size range of this group correlates with the average wound size in multiple published diabetic foot ulcer studies,33 the analyses were repeated using ulcers in this size range only. This subset of the PP dataset will be referred to as the majority wounds group. Wh en standardized for size, (mean area PRP = 2.01 cm2 (SD 1.3) and control 2.43 cm2 (SD 1.6), the proportion of com p l etely healed wounds was 13 out of 16 (81.3%) and ei ght out of 19 (42.1%) in PRP gel and con trol treatment groups, re s pectively (P = 0.036, Fisher’s ex act test). The 95% CI for the percent proporti ons of com p l etely healed wounds was 62.1% to 100% for the PRP and 19.9% to 64.3% for the control group. 76 OstomyWound Management Figure 2. Per protocol patient group (n = 40) Kaplan-Meier time-tohealing curves. Test = log-rank; Chi square = 2.34; Degrees of freedom = 1; P value = 0.126 Figure 3. Majority wound patient group (n = 35) Kaplan-Meier timeto-healing curves. Test = log-rank; Chi square = 5.62 Degrees of freedom = 1; P value = 0.0177 TABLE 4 PER PROTOCOL PATIENT DEMOGRAPHICS AND WOUND CHARACTERISTICS AT BASELINE Variable Patient Age (years) Hgb A1C Wound Area (cm2) Volume (cm3) N PRP Gel (n = 19) Mean SD Min Max N Mean Control (n= 21) SD Min Max P value 19 18 58.3 7.8 9.7 1.5 43.0 5.5 75.0 11.1 21 20 55.9 8.1 8.1 1.8 45.0 5.0 78.0 11.4 NS* NS 19 19 3.4 0.9 4.5 1.3 0.8 0.1 20.0 4.7 21 21 3.6 1.0 4.0 1.4 0.5 0.1 15.8 5.4 NS NS Treatment Characteristics Total PRP Gel (n = 19) No. of patients Percent Sex Male Female Race Caucasian Hispanic Black Other, specify Foot Right Left Location Toe Heel * † Control (n= 21) No. of patients Percent No. of patients Percent 16 3 84.21 15.79 16 5 76.19 23.81 32 8 80.00 20.00 15 4 78.95† 21.05 9 8 3 1 42.86† 38.10 14.29 4.76 24 12 3 1 60.00 30.00 7.50 2.50 14 5 73.68 26.32 10 11 47.62 52.38 24 16 60.00 40.00 6 10 31.58 52.63 9 6 42.86 28.57 15 16 37.50 40.00 NS = not significant (P >0.05) P = 0.02, two-sided Fisher’s Exact Test June 2006 Vol. 52 Issue 6 77 The Kaplan-Mei er curves of proporti on for com p l etely healed wounds over time for the majori ty wound dataset showed that the two curves started s ep a ra ting from each other on abo ut day 28 (see Figure 3), l og - rank test, P value = 0.018. When evaluating the range of healing outcomes in the four investigative site groups groupings, wide variations in healing outcomes between the site groups were observed. For the PP dataset, the site group percent of complete healing proportion varied from 50% to 100% for PRP gel-treated wounds and from 25% to 67% for control-treated wounds. For the majority wound dataset, the site group percent of complete healing proportion varied from 60% to 100% and from 25% to 60% in the PRP gel- and control-treated wounds, respectively. A trend for increased wound healing in the PRP group compared to the control group also was observed. Despite the between-site group variations, healing outcomes in each treatment group were consistent and similar in both the total dataset of the per protocol and majority wound groups. Rate of healing. In the PP dataset , the avera ge wound area cl o su re ra te per day was 0.051 cm2 for the PRP gel group versus 0.054 cm2 for the con trol group. In the majori ty wound dataset, the wound area closure per day was 0.042 cm2 for the PRP gel group and 0.043 cm2 for the con trol group; these differen ces were not stati s ti c a lly significant. In the PP dataset , wounds in the PRP gel group healed after a mean of 42.9 days (SD 18.3) compared to 47.4 days (SD 22.0) for wounds in the con trol group. While the nu m ber of days to healing was the same in the majori ty wound group (mean 42.9 and 42.8 days), 81.3% of PRP gel - treated wounds and 42.1% of con trol gel - tre a ted wounds healed du ring that time. Follow-up. Of the 40 patients in the PP dataset , 22 with healed wounds participated in the 12-week fo ll ow-up phase; of those, one in the PRP gel group had a wound that reopen ed. None of the control-tre a ted patients’ wounds re-open ed ; this differen ce was not stati s ti c a lly significant. Safety outcomes: adverse events (AE). An AE in a clinical stu dy patient who has been administered an investi ga ti onal agent is any unu sual med i c a l occurrence that has appe a red or wors en ed after the start of stu dy wh et h er or not the occurren ce was rel a ted to the use of the investi ga ti onal produ ct . Adverse events were captured for any clinical abnorm a l i ties that appe a red or wors en ed bet ween the patient’s start of the 7-day screening peri od and 30 days after receiving the last dose of stu dy treatment. Of the 127 adverse events, five occ u rred in two pati ents before randomizati on du ring the 7-day screening peri od. Of the remaining 122 adverse events occ u rring after random i z a ti on, 60 (49%) were in the PRP gel group and 62 (51%) in the con trol group. Of these, two were iden ti f i ed as definitely rel a ted to the treatment: one case of con t act derm a ti tis occ u rred in a PRP gel tre a ted wound and one instance of macera ti on occ u rred in a con trol tre a ted wound. Safety outcomes: serious adverse events (SAE). An SAE is an adverse event that meets any of the fo ll owing outcome criteri a : 78 OstomyWound Management TABLE 5 REPORTED SERIOUS ADVERSE EVENTS (N = 23 EVENTS) PRP Gel Group (6 events; 5 patients) Medra Term Myocardial infarction Congestive heart failure Pneumonia Pneumonia Pneumonia; osteomyelitis During Treatment Post Treatment Event X Death Hospitalization Hospitalization Hospitalization Hospitalization X X X X Severity as deter- Relationship mined by to device Investigator Severe Unrelated Mild Unlikely Moderate Unlikely Mild Unrelated Mild Unrelated Control Group (17 events; 7 patients) Localized infection Chest pain (gallstones) Infected left foot; left foot ulcer Bacterial arthritis/ encephalopathy Gangrene, anemia, renal failure, cardio-respiratory arrest Diabetic foot, cellulitis, osteomyelitis Cellulitis, arthritis bacterial, atrioventricular block, elevated blood glucose X X X X Hospitalization Hospitalization Hospitalization Hospitalization Severe Mild Mild Moderate/severe Death Severe Unlikely Unlikely Unlikely Unlikely Unlikely Unlikely X Hospitalization Severe Unlikely X Hospitalization Mild/severe Unrelated X • is fatal • is life-thre a tening — ie, the patient was, in the view of the Inve s ti ga tor, at immediate risk of death from the re action as it occ u rred ; however, it does not inclu de a re action that, had it occ u rred in a more serious form , m i ght have caused de a t h • requires or prolon gs inpatient hospitalization • re sults in significant or pers i s ten t disability/incapacity • is a con genital anomaly or bi rth defect • is an important medical event, based on appropriate medical judgment, that may jeopardize the patient or require the patient to seek medical or surgical intervention to prevent one of the other outcomes above. Of the 122 adverse events after ra n domization, 23 were cl a s s i f i ed as serious adverse events; six occurred in the PRP gel group and 17 in the control group. All seri o u s adverse events were unlikely or unrel a ted to devi ce usage as def i n ed by the investi ga tors (see Table 5). Clinical labora tory re sults. To analy ze the safety of tre a ting patients with the PRP gel , labora tory tests were condu cted according to the study’s predeterm i n ed ti m eframe. Because a pro s pective trial on the use of PRP gel in patients with diabetes had not been condu cted before, questions were ra i s ed wh et h er the bl ood draws, use of bovine thrombin, or the tre a tm ent itsel f would have a systemic ef fect on persons with diabetes. These concern s determ i n ed the clinical labora tory tests that were condu cted du ring the trial (see Ta ble 6). Of the 72 parti c i p a ting pati ents, 56 (78%) retu rn ed for the day 168 labora tory tests. No statisti c a lly or clinically s i gnificant differen ces were noted bet ween the PRP gel and con trol from baseline to en d point labora tory shifts in hemato l ogy, cl o t ting factors , and Factor V tests. Al t h o u gh no stati s ti c a lly significant differen ce was noted bet ween the PRP gel and control in rel a tion to shifts of cl o t ting factors , a shift (increase) was noted in PT and PTT re sults in both tre a tm ent groups. No clinically important changes in cl o t ting factors that would cause concern abo ut the June 2006 Vol. 52 Issue 6 79 ef fect of the PRP gel or control on Factor V activi ty were found du ring an indepen dent monitor review of the medical records, including concomitant med i c a tions. No clinical or statisti c a lly significant differen ces were noted in ch emistry test results bet ween the PRP gel and con trol from baseline to en d point for sod ium, potassium, ch l ori de , bi c a rbonate, creatinine, or albumin. A statistically significant differen ce was observed bet ween treatments in the ch a n ge from baseline for BUN; the BUN of the PRP gel - tre a ted pati ents decre a s ed while the BUN of con trol pati ents increased (no explanation was determ i n ed ). Serum glucose or HbA1C re sults showed that more patients shifted to high at en dpoint in the PRP gel compared to the control group. These differen ces were not statistically significant or cl i n i c a lly meaningful; t h ey su ggested that more patients in the PRP gel group had uncon tro ll ed diabetes. These safety re sults doc u m ent the minimal occ u rren ce of adverse events. No serious adverse events were attri but a ble to PRP gel and minimal were attri but a ble to labora tory shifts. These ef fects were comparable with the con trol group. Discussion This is the first reported pro s pective, randomized , bl i n ded, con tro ll ed trial in the US on the use of PRP for the tre a tm ent of d i a betic foot ulcers. In this FDA- a pproved study, s ome of their requ i rem ents (certain inclu s i on / exclusion cri teri a , blinding system , ch oi ce of con trol treatment, exten s ive labora tory tests, and doc u m entati on ) ad ded to the ri gor and complex i ty of the study design, wh i ch in tu rn caused some difficulty en ro lling patients. This stu dy com prised two levels of screening: pre- and active screen i n g. Initi a lly, an investi ga tor eva luated wh et h er a patient was a po ten tial candidate for the study. Approximately 650 patients were 80 OstomyWound Management TABLE 6 CHANGES IN CLINICAL LABORATORY RESULTS Laboratory Test Hematology Hemoglobin (G/DL) Hematocrit (%) Platelet count (10^3/UL) White blood cells (10^3/UL) Chemistry Albumin (G/DL) Bicarbonate (MEQ/L) Blood urea nitrogen (MG/DL) Chloride (MEQ/L) Creatinine (MG/DL) Glucose, serum (MG/DL) Potassium (MEQ/L) Sodium (MEQ/L) PRP Gel (n = 40) Baseline (mean) Endpoint (mean) 13.7 13.4 40.6 40.1 264 280 8.0 7.9 Control (n = 32) Baseline (mean) Endpoint (mean) 13.1 12.8 39.2 38.5 263 262 7.8 8.0 P value* 0.740 0.644 0.076 0.877 3.8 23.9 21.1 101.5 1.1 187.2 4.6 138.0 3.7 23.7 19.6 102.2 1.1 202.7 4.5 137.9 3.7 23.9 20.7 102.1 1.1 175.6 4.4 137.8 3.6 23.4 23.1 102.4 1.1 211.5 4.5 137.3 0.4693 0.6674 0.0405 0.5721 0.4143 0.4045 0.2343 0.6172 8.0 8.5 8.0 8.0 0.1232 Factor V activity (%) 105.2 104.8 101.0 103.1 0.6113 Clotting Factors PT (seconds) PTT (seconds) TT (seconds) TT Human (seconds) 12.8 29.7 12.4 11.3 13.4 31.9 12.9 11.7 13.0 30.2 12.8 11.3 13.2 30.3 12.4 11.1 0.8545 0.3738 0.4315 0.2196 Hemoglobin A1C (%HB) * Wilcoxon rank sum test, two-sided pre - s c reened — ie, revi ewed by the investi ga tor for po s s ible inclu s i on — to sec u re the 129 active screen i n g patients. Active screening com pri s ed baseline wound assessment, physical exam, labora tory tests, wound culture, vascular tests, wound excision/debri dement, and the 7-day screening period for patients meeting the requ i remen t s. From the 129 actively screen ed patients, 72 were randomized and 40 met stu dy protocol requ i rements (the clinical data audit prompted exclusion of 32 patients, dropping the PP nu m ber to 40 patients). Thus, ultimately, on ly 6% of patients who participated in the pre - s c reening process were en ro ll ed . During the analysis of the PP gro u p, f requency distri buti on dem on s tra ted that the majori ty of t h e wounds (35 out of 40) random i zed into the stu dy met the cri teria of wound area ≤7.0 cm2 and vo lume ≤2.0 cm3. The remaining five larger wounds had areas of 9 cm2 to 20 cm2 and re sults of various studies su ggest that a wound size of <7.0 cm2 is most com m on.30-32 Efficacy va ri a bles also were analy zed for the su b s et of “majority wo u n d .” Average baseline area in the majority wound group was similar to that reported in a tissue-engineered product study (n = 208: mean wound area 2.97 – 3.1 cm2),34 another tissue-engineered product (n = 15 wounds; efficacy noted in wounds < 6 cm2),35 two recombinant growth factor studies (n = 118 patients and 132 wounds respectively; mean area 5.5 cm2 and 2.6 cm2),36,37 and one retrospective study of diabetic foot ulcers (n = 26,599 patients of which 5,320 wounds averaged 1.53 cm2, 5,320 wounds averaged 1.84 cm2, and 5,319 wounds averaged 4.41 cm2 in area24). In the largest study published to date,24 60% of patients had wounds that matched the majority group in this PRP ge l study, increasing the potential external validity of the current study results. In the majority wound group, PRP gel - tre a ted wounds were sign i f i c a n t ly more likely to heal than con trol-tre a ted wounds even though healing ra tes in the control group June 2006 Vol. 52 Issue 6 81 were high er (42% healed after 12 weeks) than most con trol group healing ra te s reported in other studies30 ( s ee Figure 4). Spec i f i c a lly, re sults of a met a - a n a lysis of healing outcomes in the con trol arm of o t h er diabetic foot pro s pective studies su ggest that 24% can be ex pected to heal after 12 weeks of providing good care.30 Most studies use wet to - m oist ga u ze saline ga u ze dressings, the recogn i zed standard treatment by the m edical com mu n i ty5 as a con trol dressing. However, wet - to - m oist dressings requ i re dressing ch a n ges Figure 4. Comparison of diabetic foot ulcer prospective trial control outcomes.30 three to four times daily by the pati ent or caregiver and may not provi de an optimal s m a ll frequ ent ex po sure to it. To evalu a te this po tential moist environ m ent for healing. Some studies have shown i m p act , Factor V tests occ u rred at 6-week intervals for all that hydroco ll oid dre s s i n gs may be more ef fective than patients regardless of random i z a ti on arm and at 24 wee k s wet - to - m oist for the tre a tm ent of certain types of ulcers.38 post ra n domization. None of the pati ents dem on s tra ted The power calculati ons were based on reported any Factor V inhibiti on thro u gh o ut the stu dy and into the results of control treatments in other studies and a samfo ll ow-up. This is the first time that this has been doc uple size of 27 in each treatment arm. The better-thanmented in a pro s pective study. expected control healing rates and large number of proLabora tory tests indicated that the majori ty of patients tocol violations caused underpowering of the PP and the had el eva ted bl ood glu cose throughout the study peri od . majority wound groups yet some statistically significant Eleva ted glu cose levels have been documented to redu ce differen ces were ob s erved, su ggesting that larger healing40,41; h owever, in this stu dy, the majori ty of the between-group differences could be expected in studies wounds healed . using a larger sample size. Numerous articles have been published regarding the The FDA’s con cern abo ut the effects of frequ en t use of PRP in the surgical setting; notably, the orthopet h o u gh small amount of bl ood co ll ecti on (30 mL or less dic, plastic surgery, and dental field.42-44 Some of the PRP each visit) on health and safety as patients underwen t separation systems used require specialized technicians periodic hemato l ogy and other labora tory tests thro u ghto perform the necessar y procedures. This is the first out the trial was ad d re s s ed . Test re sults documented that published study of an autologous PRP separation system these frequ ent bl ood draws did not redu ce the hem ogl oto heal wounds that can be used by health professionals bin, hematocrit, or platel et co u n t . Because bovine thromwithin a traditional healthcare setting. A small, compact, bin was used in the processing to activa te the PRP and evipoint-of-care system such as the one described in the den ce exists of Factor V leading to bl eeding disorders in study makes this technology available to multiple care 33 patients ex po s ed to large doses of bovine thrombin, conproviders, including physician office, hospital unit, outcerns were ra i s ed as to wh et h er patients would similarly patient clinic, long-term care facility, and home health devel op Factor V antibodies to the bovine thrombin wi t h care staff. 82 OstomyWound Management Conclusion The results of this study show that PRP gel is safe for use in the treatment of nonhealing diabetic foot ulcers. In the most common size of diabetic foot ulcers (≤7.0 cm2 in area and ≤2.0 cm 3 in volume), PRP gel-treated wounds are also significantly more likely to heal than control gel treated wounds. Treating wounds with PRP or saline gel resulted in healing in appr oximately 6 weeks, but in the most common wound sizes, almost twice as many PRP treated wounds healed in that timeframe. The number of adverse events was minimal; no adverse events were serious. Further, the study demonstrated that bovine thrombin used in the preparation of PRP does not cause Factor V inhibition; thus, it does not cause coagulopathy. In addition, withdrawal of a small amount of blood twice weekly did not affec t patient hemoglobin, hematocrit, or platelet count. Clinically meaningful shifts in laboratory values studied from baseline to endpoint were not obser ved. This type of PRP system could be utilized by healthcare providers to treat diabetic foot ulcers in multiple settings. Using PRP gel to treat diabetic foot ulc ers may not only enhance healing, but it also may prevent lower extremity amputations caused by nonhealing wounds. Implicati ons for future re s e a rch include implem en ting a trial de s i gn that would permit greater subj ect en ro llment on a larger sample size to va l i d a te these results. Th i s trial would not need to re - evaluate some of the major qu e s tions answered in this trial, su ch as Factor V inhibition, i m p act on the pati ent’s hem a to l ogy and other cl i n ical labora tory outcomes, and impact of a con trol that had not perform ed in a prospective trial previ o u s ly. Di a beti c foot ulcers with ch a ll en ging presentati ons (ie, mild to m odera te vascular disease, ex po s ed ten don or bone, patient hyper glycemia, and/or inadequ a te nutri tional status) could be studied to determine wh et h er PRP gel could assist in healing in these comprom i s edscen a ri o s . In ad d ition, studies to determine wh et h er this novel thera py is synergistic with other adva n ced wound care modalities could be conducted . - OWM Neal Mozen, DPM; Jeffrey Pa ge , DPM; Al ex a n der Rey zelman, DPM; Thomas Ro u k i s , DPM; Roger Schech ter, MD; Th omas Seren a , MD; Ji n sup Son g, DPM; and Robert Snyder, DPM. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Acknowledgments The authors acknowl edge the AutoloGel™ Di a beti c Foot Ulcer Group: Margaret Do u cet te, DO ; Mich ael Dellacorte, DPM; Robert Fryk berg, DPM; G a briel Ha l peri n , DPM; Robert Kirsner, MD; Mi ch ael Mi ll er, DO ; 13. 14. Centers for Disease Control and Prevention. Na ti onal diabetes fact sheet: gen eral information and nati onal esti m a tes on diabetes in the United States. Atlanta, Ga: United States Dep a rtm ent of Health and Human Services. Center for Disease Control and Prevention;2005. Am erican Di a betes Association. Consensus developm ent conference on diabetic foot wound care. Diabetes Care. 1999;22(8):1354–1360. Mani R, Fa l a n ga V, Shearman CP, Sa n deman D. Chronic Wound Healing: Clinical Measurem ent and Basic Science. New York, NY: WB Sa u n ders;1999:17–23. Pecora ro RE, Reiber GE, Burgess EM. Pathways to diabeti c limb amputation. Basis for preven ti on. Di a betes Care. 1990;13(5):513–521. American Diabetes Association. Available at: www.diabetes.or g.Accessed Janu a ry 25, 2006. Fryk berg RG, Armstrong DG , Giu rini J, et al. Diabetic Foot Di sorders: A Clinical Practi ce Guideline. Park Ridge, Ill : Am erican Coll ege of Foot and Ankle Surgeons;2000. Ho ll oway GA, Steed DL, DeMarco JM, et al. A ra n domized controll ed dose response trial of activa ted platel et su pernatant, topical CT-102 in chronic, nonhealing, d i a beti c wounds. WOUNDS. 1993;5:160–168. Steed DL, Gloslen JB, Ho ll oway GA, et al. CT-102 activa ted p l a tel et su pernatant topical: a ra n domized, prospective, do u ble blind trial in healing of chronic diabetic foot ulcers. Di a betes Care. 1992;15:1598–1604. Atri SC, Mi s tr J, Bisht D, et al. Use of homologous platel et factors in achieving total healing of recalcitrant skin ulcers. Su rgery. 1990;108:508–512. Knighton DR, Ciresi K, Fiegel VD, et al. Stimulation of repair in chronic nonhealing cut a n eous ulcers: a pro s pectively ra n dom i zed blinded trial using platel et - derived wound healing formula. Surg Gy necol Obstet. 1990;170:56–60. Fylling CP, Baxter CR, Bennett KG, et al. Multi-center cl i nical revi ew: using auto l ogous platel et gel for the treatment of d i a betic plantar wounds. Di a bete s. 2001;50 (su pp l 2):229A. Glover JL, Weinga rten MS, Buchbinder DS, et al. A fouryear multicenter retrospective stu dy of topical mu l tiple growth factor thera py for chronic wounds. Adv Wound Care. 1997;10(1):27–32. Keyser JE. Di a betic wound healing and limb salva ge in an outpatient wound care program. South Med J. 1993;86:311–317. Fylling CP. A five center study of diabetic wound healing in June 2006 Vol. 52 Issue 6 83 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. comprehensive wound managem ent programs. Diabetes. 1992; 41(suppl A):82A. Knighton DR, Fylling CP, Fiegel VD, et al. Amputation prevention in an independently revi ewed at-risk diabetic population using a compreh ensive wound care protocol. Am er J Su rg. 1990;160:466–472. Fylling CP, Knighton DR, G ordinier RH. The use of a comprehensive wound care protocol including topical growth factor therapy in treatment of d i a betic neuropathic ulcers. In: Ward J, Goto Y. Diabetic Neuropathy. New York, NY: John Wi l ey & Sons;1990:567–578. Doucet te MM, Fylling CP, Knighton DR. Amputation prevention in a high - risk population through a comprehensive wound healing protocol. Arch Phys Med R ehabil. 1989;70:780–785. Knighton DR, Fylling CP, Do u cette MM. Wound healing and amputation in a high-risk diabetic population. WOUNDS. 1989;2:107–114. Knighton DR, Ciresi KF, Fiegel VD, et al. Classification and tre a tm ent of chronic nonhealing wounds: su ccessful treatment with auto l ogous platel et-derived wound healing factors (PDWHF). Ann Surg. 1986;3:322–330. Bentkover JF, Champion AH. Econ omic eva luation of alternative methods of tre a tm ent for diabetic foot ulcer pati ents: cost effectiveness of platel et releasate and wound care clinics. WOUNDS. 1993;5:207–215. Fylling CP, McKeown PC. Cost and healing ef f i c acy of treating diabetic foot ulcers in a comprehensive wound managem ent program with growth factor thera py. Diabetes. 1990;39(suppl A):18A. Bennett NT, Schultz GS. Growth factors and wound healing. Pa rt II. Role in chronic wound healing. Am J Surg. 1993;166(1):74–81. Mosesson MW. Fibri n ogen and fibrin structu re and functions. J T h romb Haemost. 2005;3(8):1894–1904. Ma r golis DJ, Kantor J, Santana J, et al. Effectiveness of platel et releasate for the tre a tm ent of diabetic neuropathic foot ulcers. Diabetes Care. 2001;24:483–488. Lipsky BA, Berendt AR, Desery H. Di a gnosis and treatment of d i a betic foot infecti ons. Guidelines for Diabetic Foot Infecti ons. Clin Infe ct Dis. 2004:39:895–910. Armstrong DG , Lavery LA, Harkless LB.Validation of a diabetic classification system; the contri bution of dept h , infection, and ischemia to risk of amputation. Diabetes Care. 1998;21(5):855–859. Cooper DM. Assessment, measurement, and evaluation: their pivotal roles in wound healing. In: Bryant RA. Acute and Chronic Wou n d s. St. Louis, Mo: Mosby;2000:61. Fylling CP, Ba x ter CR, Bennett KG, et al. Multi-center clinical revi ew: using auto l ogous platel et gel for the treatment of diabetic plantar wounds. Poster and oral presentation at the American Di a betes Association Annual Meeting and Scientific Sessions. P h i l adelphia, Pa. June 22–26, 2001. Beriou J, Fylling CP, Justice P. A retrospective revi ew of healing progress of 102 chronic wounds du ring the first 28 days 84 OstomyWound Management 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. of treatment with auto l ogous platel et rich plasma gel. Po s ter presen t a ti on at: the Sym po s ium on Advanced Wound Care and Medical Research Forum. Orlando, Fla. May 5– 7, 2004. Margolis DJ, Kantor J, Berlin J. Healing of diabetic neuropathic foot ulcers receiving standard treatment. Diabetes Care. 1999;22:692–695. Lee ET. St a ti s tical Met h ods for Su rvival Data Analys i s. Belmont, Calif: Lifetime Learning Publications, Wadsworth, Inc;1980. Kalbflei s chJD, Pren ti ce RL.The Stati s tical Analysis of Failure Time Data. New York, NY: John Wi l ey and Sons;1980. Margolis DJ, Gelfand JM, Hoffs t ad O, Berlin J. Surroga te endpoints for the tre a tm ent of d i a betic neuropathic foot u l cers. Di a betes Care. 2003;26:1696–1700. Veves A, Falanga V, Armstrong DG , et al. Graftskin, a human skin equivalent, is effective in the management of noninfected neuropathic diabetic foot ulcers. Diabetes Care. 2001;24(2):290–295. Lipkin S, Chaikof E, Isseroff Z, et al. Effectiveness of bilayered cellular matrix in healing of neuropathic diabetic foot u l cers: Results of a multicenter pilot trial. WOUNDS. 2003;15(7):230–236. S teed DL. Clinical eva lu a ti on of recombinant human p l a telet-derived growth factor for the tre a tm ent of lower extrem i ty diabetic ulcers. J Va sc Su rg. 1995;21:71–81. Wieman TJ, Smiell JM, Su Y. Efficacy and safety of a topical gel formulation of recombinant human plateletderived growth factor – BB (becaplermin) in patients with ch ronic neu ropathic ulcers. Di a betes Care. 1998;21(5):822–827. Bradl ey M, Cu llum N, Nelson EA, et al. System a tic revi ews of wound care management: dressings and topical agen t s used in the healing of chronic wounds. Health Technol Assess. 1999;3(17 Part 2). Winterbottom N, Kuo JM, Nguyen K, et al. Anti genic responses to bovine throm bin ex po sure du ring surgery: a pro s pective stu dy of 309 pati ents. J Applied Res Clin Exper T h erapeu ti cs. 2002;2(1):20–29. Jeffcoate WJ, Pri ce P, Harding KG, et al. Wound healing and tre a tm ents for people with diabetic foot ulcers. Diabetes Metab Res Rev. 2004;20(su ppl 1):S78–S89. Ma rs ton WA, Derm a graft Di a betic Foot Ul cer Stu dy Group. Risk factors associated with healing chronic diabetic foot ulcers: the importance of hyper glycemia. Ostomy Wound Manage. 2006; 52(3):26–28,30,32. Kassolis, JD, Reynolds, Ma. Eva luation of the adjunctive benefits of platel et-rich plasma in subantral sinus augm entation. J Craniof a cial Surg. 2005;16(2):280–287. Grageda E. Platel et - ri ch plasma and bone graft materials: a review and a standardized research protocol. Im plant Den t. 2004;13(4):301–309. Weibrich G, Hansen T, Kl eis W, et al. Effect of platel et concentra tion in platel et - ri ch plasma on peri-implant bone regenera tion. Bo n e. 2004;34(4):665–671.
© Copyright 2018