1186

A Phase Wl Clinical Trial to Evaluatea Combination of Recombinant HumanPlatelet-Derived Growth Factor-BB andRecombinant Human Insulin-LikeGrowth Factor-I in Patients withPeriodontal DiseaseT. Howard Howell* Joseph P. Fiorellini, ' David W. Paquette,f Steven Offenbacher,'*William V. Giannobile, * and Samuel E. Lynch*

The primary objective- of this study was to assess the safety of recombinant human(rh) platelet-derived growth factor-BB (PDGF-BB) and (rh) insulin-like growth factor-I(IGF-I) when applied to periodontal osseous defects in humans; a secondary objectivewas to begin to accrue data on the therapeutic dose of these growth factors (GFs) requiredto stimulate periodontal regeneration. Thirty-eight human subjects possessing bilateral os-

seous periodontal lesions were assigned to one of two treatment groups in a split-mouthdesign. Following full-thickness flap reflection, test sites received local application of thetherapeutic drug delivered in coded syringes by a "masked" investigator. Two dose levelswere tested, 50 p,g/ml each of rhPDGF-BB and rhIGF-I in a gel vehicle (LD-PDGF/IGF-I) and 150 µg/ml each of rhPDGF-BB and rhIGF-I plus vehicle (HD-PDGF/IGF-I). Con-trol treatment consisted of either conventional periodontal flap surgery or surgery plusvehicle. Safety analyses included physical examination, hematology, serum chemistry,urinalysis, antibody titers, and radiographie evaluation of bony changes. The primarytherapeutic assessment was bone fill measured at re-entry 6 to 9 months after treatment.No local or systemic safety issues were found as a result of GF administration. No patientsdeveloped antibodies to the rhGF proteins. In subjects treated with LD-PDGF/IGF-I, therewere no enhancements in periodontal regeneration compared to controls. However, inpatients treated with HD-PDGF/IGF-I, statistically significant increases in alveolar boneformation were noted as measured by surgical re-entry 9 months following drug delivery(P < 0.05). This corresponded to an increase of 2.08 mm of new vertical bone heightand 42.3% osseous defect fill in the HD-PDGF/IGF-I subjects versus only 0.75 mm and18.5% gains in new bone height and osseous fill, respectively, in the controls. Furcationlesions, although limited in number, responded most favorably to treatment, with 2.8 mm

horizontal osseous fill. The results from this study suggest that the local application ofrhPDGF-BB and rhIGF-I to periodontal lesions is safe at the dose levels studied. LD-PDGF/IGF-I did not elicit increased defect fill compared to the control; however, HD-PDGF/IGF-I resulted in a significant promotion in bone regeneration. Additional studiesare warranted to more fully characterize the effects of PDGF/IGF-I on periodontal regen-eration in humans. J Periodontol 1997;68:1186-1193.

Key Words: Periodontal diseases/surgery; periodontal diseases/regeneration; growth fac-tors/platelet-derived; growth factors/insulin-like; clinical trials.

*Department of Periodontology, Harvard School of Dental Medicine, Boston, MA.'Department of Periodontics, University of North Carolina-Chapel Hill, Chapel Hill, NC.•Previously, Department of Periodontology, Harvard School of Dental Medicine; currently, Osteohealth Company, Shirley, NY.

Volume 68Number 12 HOWELL, FIORELLINI, PAQUETTE, OFFENBACHER, GIANNOBILE, LYNCH 1187

Table 1. Inclusion Criteria

1. Males and females between 21 and 62 years of age.'

2. If female, practicing adequate birth control, negative preg-nancy test within 3 days of dosing, post-menopausal, and/orsurgically sterilized.

3. Scaling and root planing of the 4 quadrants completed within4 weeks prior to surgery.

4. At least one tooth in contralateral quadrants with similar an-

gular or Class II furcation osseous defects.5. Angular osseous defects must be &3 mm in depth as assessed

by radiographie evaluation.6. For interproximal defects, radiographie evidence of 30% to

70% bone loss around study teeth. For furcation defects, ra-

diographie evidence of furcation involvement at baseline.7. Patients demonstrate acceptable oral hygiene prior to surgical

therapy. Plaque score (Löe and Silness)9 must be less than 2;gingival index (Silness and Löe)10 must be less than 2.

8. Involved teeth are vital and asymptomatic.9. Patients are willing and able to give informed consent.

10. Patients are willing and able to return for multiple follow-upvisits.

Table 2. Exclusion Criteria

1. If female, positive pregnancy test within 3 days of dosing ornot practicing adequate birth control.

2. Patients taking drugs which are known to interfere withwound healing such as corticosteroids, anticancer agents, orimmune modulators or who have received such drugs within4 weeks prior to application of the study medication.

3. Allergies to yeast-derived products.4. Current nutritional disorders.5. Current or previous history of systemic diseases known to

interfere with adequate wound healing such as chronic liverfailure, diabetes, AIDS, chronic renal failure, malignancy,collagen vascular diseases, and malnutrition.

6. Evidence or previous history of significant and/or unstablecardiovascular, pulmonary, gastrointestinal, hematological, orendocrine disease or other disorder which would impact onthe safety evaluations.

7. Current or previous history of cancer (with the exception ofsquamous cell and basal cell carcinomas) within 5 years.

8. Lack of appropriate periodontal defects.9. Unacceptable oral hygiene habits.

10. Grade II or greater mobility of teeth.11. Endodontic involvement of study teeth.12. Smoking or other use of tobacco products.13. Alcohol or drug abuse as defined by criteria in Diagnostic

and Statistical Manual of Mental Disorders (DSM) III-R.14. Subjects' inability or unwillingness to give informed consent.15. History of juvenile or rapidly progressive Periodontitis.16. Patients who have received an investigational drug or partic-

ipated in a research study within 30 days prior to the firstapplication of study medication.

Traditional methods of treating periodontal disease havefocused on the elimination of bacterial pathogens and themodulation of the host response in an effort to arrest or

slow disease progression. With continued success of theseforms of therapy and a clearer understanding of the dis-ease process, the regeneration of the periodontium, com-

posed of alveolar bone, ligament, and tooth root surfacecementum, is one of the major goals of therapy. Examplesof current periodontal regenerative therapies include boneautografts, allografts, and guided tissue regeneration us-

ing cell-occlusive barrier membranes. While these treat-ments can, under certain conditions, lead to partial res-toration of the periodontium, there remains a need fordevelopment of new therapies with greater predictability,ease of use, and magnitude of efficacy.

Polypeptide growth factors are a class of natural bio-logical mediators which regulate key cellular events intissue repair including cell proliferation, Chemotaxis, dif-ferentiation, and matrix synthesis via binding to specificcell-surface receptors.1 Growth factors possess pleiotropiceffects on wound repair and are potent modulators of cellsresiding within the periodontium.2-3 Preclinical animalstudies by several investigational groups have demon-strated that the combination of two such growth factors,namely platelet-derived growth factor-BB (PDGF-BB)and insulin-like growth factor-I (IGF-I), has the capacityto stimulate new bone, periodontal ligament, and cemen-

tum formation in periodontal lesions in dogs and non-human primates.4-8

The primary objective of this Phase I/II clinical trialwas to determine the safety of a single application of acombination of recombinant human (rh) PDGF-BB andrhIGF-I in a gel carrier when placed into osseous defectsduring periodontal flap surgery in humans. A secondaryobjective was to evaluate the biological response ofgrowth factor therapy on the promotion of periodontaltissue regeneration.

MATERIALS AND METHODS

Enrollment and RandomizationThirty-eight patients with moderate to severe periodontaldisease and corresponding bone destruction were selectedfor this randomized, double-masked, paired design clini-cal trial. Patients were entered into the trial at two testcenters, the Harvard School of Dental Medicine and theUniversity of North Carolina-Chapel Hill School of Den-tistry. The mean age of the patients entered was 46.1years (range 38 to 62 years). Sixty-six percent (66%) ofthe patients were male and 34% female. Subjects enteredinto the study protocol met inclusion and exclusion cri-teria outlined in Tables 1 and 2. Patients signed informedconsent according to the approved human studies proto-cols at Harvard University and the University of NorthCarolina. Treatment and control sites were randomized

and recorded in sealed envelopes, which were opened bythe investigators at the time of surgery.

TherapyEach subject possessed matching contralateral interprox-imal angular and/or Class II furcation bone defects. Allparticipants received full-mouth scaling and root planingand oral hygiene instructions prior to drug delivery. A

1188 PDGF AND IGF-I THERAPY FOR PERIODONTAL DISEASEJ Periodontol

December 1997

Safety assessment(days 2, 7,14 & 28)

Scaling and root

planing; oral hygieneinstruction

Baselineassessementsand surgerywith drugapplication

Soft tissueassessments

Soft tissue andradiographieassessments; hardtissue re-entryassessments forLD-PDGF/IGF-Isubjects

Soft tissue andradiographieassessments; hardtissue re-entryassessments forHD-PDGF/IGF-Isubjects

Figure 1. Study timeline. The axis denotes treatment assessments and procedures applied to subjects in LD-PDGF/IGF-I and HD-PDGF/IGF-I groupsover a total of 10 months of observations.

paired or split-mouth design was utilized in which all pa-tients were randomly assigned to receive in one treatmentquadrant a control therapy consisting of either periodontalflap surgery alone or surgery plus 4% methylcellulose gelvehicle.5 The other contralateral bone defect within themouth received either a single dose of [low dose (LD)]50 µg/ml each of rhPDGF-BB and rhIGF-I« (LD-PDGF/IGF-I) or 150 µg/ml [high dose (HD)] each of rhPDGFand rhIGF-I (HD-PDGF/IGF-I) per lesion. At baseline,surgery was performed on all study teeth in each patientat one visit. Full-thickness mucoperiosteal flaps were el-evated, the teeth were scaled and root planed, and allgranulation tissue was removed under local infiltration an-

esthesia. The walls of the osseous defects were perforatedapproximately 3 times with a V2 round dental bur. Intra-operative hard tissue measurements were made using an

occlusal stent and calibrated periodontal probe. Approx-imately 70 µ of study medication, vehicle, or scaling androot planing alone was administered to the bone defects.The tissues were sutured primarily with a polytetra-fluoroethylene 4-0 suture material1 and positioned at thelevel of the cemento-enamel junction (CEJ). The siteswere covered with periodontal dressing. All patients re-

ceived penicillin VK (1 g per day for 7 days) and 0.12%Chlorhexidine rinses (twice daily for a period of 8 weeks).Oral hygiene assessment and supragingival scaling were

performed on days 28 and 42 and at 3, 4.5, 6, and 9months.

Treatment and Safety AssessmentsAt baseline, and periodically thereafter, safety as well as

clinical and radiographie measurements were taken (Fig.1). All assessments were made by a masked examinerunaware of the treatment rendered. Safety assessments in-cluded comprehensive physical examinations within 2weeks of dosing and on day 28. Laboratory examinationsincluded hematology, serum chemistries and urinalysis(baseline and on days 2, 7, 14, and 28), pregnancy testing(baseline), and serum antibody assays to monitor antibody

5Dow Chemical Co., Midland, MI."Institute of Molecular Biology, Inc., Worcester, MA.'Gore-Tex, W.L. Gore and Associates, Inc., Flagstaff, AZ.

titers to the PDGF/IGF-I drug formulation (baseline ver-

sus day 42; Table 3). All adverse events were rated mildto severe according to the World Health Organization(WHO) scale.

Efficacy parameters included direct bone height and fillmeasures obtained during the initial treatment surgery andre-entry surgery 6 to 9 months later." The secondary ef-ficacy outcome included relative attachment level change(Table 4). Clinical bone and soft tissue measurementswere made from a custom-fabricated occlusal stent. Toassure accurate placement of the fixed reference point,positioning of the stent was recorded at each time periodusing a measurement from stent to CEJ. In order to fur-ther assure the accuracy of the bone height measurements,a second surgical procedure was performed on all studysites to allow direct visualization of the bone margin. Theclinical hard tissue assessments performed included theeresiai bone height, defect bone height, and defect depthusing a UNC periodontal probe. The hard tissue mea-

surements were taken at baseline and at 6 months for theLD-PDGF/IGF-I subjects and at baseline and 9 monthsfor the HD-PDGF/IGF-I subjects.

Clinical soft tissue measurements included relative at-tachment level made from the stent using a manual probe,and closed horizontal furcation probing depth (furcationlesions) taken at 3, 6, and 9 months. All clinical mea-

surements in this trial were obtained by a single masked

Table 3. Safety Assessments

I. Standard medical and dental histories (baseline)II. Standard physical examination (baseline and 4 weeks post-

drug delivery)III. Laboratory examinations

A. Hematology (baseline and days 2, 7, 14, and 28)B. Serum chemistries (baseline and days 2, 7, 14, and 28)C. Urinalysis (baseline and days 2, 7, 14, and 28)D. Plasma sample for antibody titers to PDGF-BB and IGF-

I (baseline and days 28 and 84)E. Pregnancy test (baseline)

IV. Gingival inflammation (baseline and days 2, 7, 14, and 28)V. Radiographs (baseline and month 6 [LD-PDGF/IGF-I! and

month 9 [HD-PDGF/IGF-I])Baseline denotes within 3 days prior to initiation of study med-ication

Volume 68Number 12 HOWELL, FIORELLINI, PAQUETTE, OFFENBACHER, GIANNOBILE, LYNCH 1189

examiner who was calibrated prior to the study for 95%agreement (i.e. ± 1 mm) for both soft and hard tissueoutcomes.

Statistical AnalysisThe study design considered patients with comparabledisease in two quadrants to have each quadrant (contain-ing at least one eligible site) randomly assigned to thestudy medication or one of the two control treatments. Upto two sites per quadrant were studied. The responses ofsites in the same quadrant were averaged to provide one

measurement per quadrant. Thus, the statistical designwithin each dose group (n = 19) consisted of two split-mouth designs: LD-PDGF/IGF-I versus standard care

(surgery alone) and LD-PDGF/IGF-I versus standard care

plus vehicle, with a similar split-mouth design for theHD-PDGF/IGF-I patients.

Comparisons between study drugs and controls were

made within subjects, utilizing the split-mouth aspect ofthe design. Wilcoxon signed rank tests were used forpaired comparisons. The first comparison involved the de-termination of a therapeutic effect with the vehicle versus

surgery alone. Utilizing a Wilcoxon (rank sum) test, therewere no significant differences between vehicle and stan-dard care with respect to the primary endpoints of boneheight. Consequently, the subsequent comparisons weremade between the combined control groups versus eachof the study drug groups. All 38 patients were used forthe safety-related analysis under the intent-to-treat prin-ciple; the 36 patients completing the trial were used inefficacy analyses.

RESULTS

SafetyAnalysis of the safety data indicates that both the LD-PDGF/IGF-I and HD-PDGF/IGF-I drug formulationswere well tolerated. There were no serious adverse events,discontinuations, or deaths attributed to the drug or pla-cebo. Additionally, there were no signs of uncontrollabletissue formation, such as gingival hyperplasia. Two pa-tients left the study prior to completion. One patient was

non-compliant with the study protocol and the other re-

located from the study center.The most frequently observed adverse events were as-

sociated with the periodontal surgery and included painor discomfort at the operative site, sensitivity, and edema.There were 18 study drug-related adverse events with a

severity rating of mild (WHO classification). However,pain following surgery was not unexpected and no mea-

surable differences were ascertained between treatmentand control quadrants. These events included 15 reportsof postoperative pain or discomfort, two reports of softtissue dehiscence, and one report of a tooth abscess. Twoadverse events, rated with a moderate severity by the in-

vestigators, were noted. A 39-year-old female subject pre-sented with an abscess of a study tooth 8 months afterreceiving a single administration of HD-PDGF/TGF-I. Thesecond event occurred after a patient had been discontin-ued for non-compliance and involved a hospitalization foralcohol rehabilitation. It was felt that the events couldhave been attributable to the study medication or mayhave been of endodontic or gingival origin.

Treatment-emergent abnormal laboratory events were

reported for 5 patients and included elevated liver en-

zymes (SGOT and SGPT), lymphocytosis, and hematuria.All abnormalities were considered mild (WHO classifi-cation) and were found at both baseline (prior to treat-

ment) and 28-day timepoints (i.e., the patients enteredinto the study with abnormal levels prior to therapy andcontinued to display altered levels at one month). Anti-body titers were determined on 37 patients. No patientdeveloped antibodies to rhPDGF-BB or rhIGF-I after ex-

posure to the drug.

EfficacyPatient assessment. Data from all LD-PDGF/IGF-I andHD-PDGF/IGF-I-treated subjects were included in allanalyses, except as follows. In LD-PDGF/IGF-I subjects,two patients were lost to follow-up for non-drug relatedreasons; one patient was excluded from analysis becauseof an occurrence of a gingival abscess after treatment.Thus, there were a total of 42 sites from 16 patients eval-uated (the control therapies were evenly distributed [8surgery alone and 8 surgery plus placebo]).

Data from all 19 HD-PDGF/IGF-I-treated subjectswere evaluated, except for two patients who were lost tofollow-up for non-drug related reasons; one patient wasexcluded from analysis because of an occurrence of a

gingival abscess 8 months after treatment; one patientcould not be measured for the hard tissue and relativeattachment level endpoints at baseline because the dis-tance from the edge of the stent to the base of the lesionwas greater than the calibrated length of the periodontalprobe. There were a minimum of 14 évaluable subjectsand 28 sites for all endpoints.

The demographics of the re-entry and relative attach-ment level parameters of the study teeth at baseline are

depicted in Tables 5 and 6. All data were entered into a

statistical analysis program from the case report forms.Quality control was performed by independent verifica-tion of the entered data. Initial analysis revealed that thedata did not exhibit a normal distribution. Consequently,data were analyzed for statistical differences by the non-

parametric two-sided Wilcoxon signed rank test. Statisti-cal analysis yielded no statistically significant differencesbetween the two control groups for any parameter (datanot shown). Consequently, comparisons were made be-tween the test and combined surgery alone and surgeryplus placebo gel (control) groups.

1190 PDGF AND IGF-I THERAPY FOR PERIODONTAL DISEASEJ Periodontol

December 1997

Table 4. Efficacy Measurements

Clinical (re-entry) Hard Tissue Measurements•Defect depth•Defect bone height•Percent defect fill•Open horizontal furcation probing depth•Crestal bone height

Clinical Soft Tissue Assessments•Clinical attachment level•Horizontal furcation probing depth

Hard tissue endpoints. The principal efficacy end-points were the hard tissue assessments measured duringre-entry surgery at 6 months in LD-PDGF/IGF-I subjectsand 9 months for HD-PDGF/IGF-I subjects. No signifi-cant differences between LD-PDGF/IGF-I-treated sub-jects and standard care or standard care plus vehicle couldbe noted (Fig. 2A). The LD-PDGF/IGF-I subjects were

surgically re-entered prior to the high dose-treated sub-jects. The investigators noted that re-entry of the surgicalsites at 6 months may have been early and that the pro-cedure was disturbing the newly regenerated tissue. Thus,a decision was made to re-enter the remaining HD-PDGF/IGF-I-treated patients at 9 months. At surgical re-entry at

Table 5. Baseline Periodontal Lesion Distribution for Patients Treated with LD-PDGF/IGF-I

Lesion

Angular (n = 8)(Mean ± S.E.M.)*

Furcation (n = 8)(Mean ± S.E.M.)

ControlLD-PDGF/

IGF-I ControlLD-PDGF/

IGF-I

Hard tissue Mean vertical osseous defeet depth (mm)

Mean horizontal osseous

defect depth (mm)Soft tissue Relative attachment level

(mm) 6.38Closed horizontal furcation

probing depth (mm)

4.69 ± 1.12 4.63 ± 0.84

0.31 6.77 ± 0.49

5.50 ± 0.76 5.86 ± 1.01

6.25 ± 0.53 5.13 ± 0.35

*Standard error measurement.

Linear Bone Height Osseous Defect Fill

3.6

3.0

2.4

(mm) 1.8

0.6

60

50

40

30 (%)

10

Angular Furcation Combined Angular Furcation Combined

Figure 2. Periodontal bone response in patients treated with LD-PDGF/IGF-I by surgical re-entry at 6months. Panel A shows linear bone height mean values (mm apposition ± S.E.M.) for pooled control (openbars) or LD-PDGF/IGF-I-treated subjects (closed bars). There were no significant differences between treat-ed and control groups. Panel shows the percentage of the original osseous defect filled with new bonefollowing treatment with either control therapy or LD-PDGF/IGF-I. No significant differences were foundbetween treatment and control defects (n=16).

Volume 68Number 12 HOWELL, FIORELLINI, PAQUETTE, OFFENBACHER, GIANNOBILE, LYNCH 1191

Linear Bone Height Osseous Defect Fill

3.6

3.0

(mm) 1.8

1.2

0.6

I !

60

50

30 (%)

Angular Furcation Combined Angular Furcation Combined

Figure 3. Periodontal bone regeneration in patients treated with HD-PDGF/IGF-I by surgical re-entry at9 months. Panel A shows linear bone height mean values (mm apposition ± S.E.M.) for pooled control(open bars) or HD-PDGF/IGF-I (closed bars). Statistically significant differences in furcation lesions com-

pared to paired controls as well as in combined furcation and angular lesions versus pooled control groups(P < 0.05). Panel shows the percentage of the original osseous defect filled with new bone followingtreatment with either control therapy or HD-PDGF/IGF-I. Statistically significant differences were found inthe percentage of the original osseous defect filled with new alveolar bone (P < 0.05); = 16.

9 months in the HD-PDGF/IGF-I-treated subjects, themean new bone height (i.e., the change in bone heightfrom baseline to month 9) was 0.75 mm (±0.28; S.E.M.)in the control group and 2.08 mm (±0.40; S.E.M.) in theHD-PDGF/IGF-I group (P < 0.05; Fig. 3A).

Because the absolute height of new bone does not nec-

essarily reflect the degree to which the original osseous

lesion was healed, the percentage of osseous defect fillwas also calculated. There was again no significant dif-ferences found in defect fill in the LD-PDGF/IGF-I sub-jects (Fig. 2B). However, in HD-PDGF/IGF-I subjects,control defects resulted in 18.5 ± 7% (mean ± S.E.M.)bone fill, while defects in patients treated with HD-PDGF/IGF-I resulted in 42.3 ± 9% bone fill. Similar to new

bone height measures, furcation lesions responded most

favorably (Fig. 3B).Soft tissue assessments. Relative attachment level

(from occlusal Stents) measurements were obtained at 3,6, and 9 months (HD-PDGF/IGF-I subjects) using cali-brated UNC manual probes. Data on the reduction of hor-izontal furcation probing depths are shown in Tables 7and 8. Gains in attachment level were found for both testand control groups following periodontal surgery. How-ever, for both LD-PDGF/IGF-I and HD-PDGF/IGF-I-treated subjects, no statistically significant differenceswere found between treatment and controls at any timepoint. Trends of effect were found for both horizontalfurcation probing depth and relative attachment level inthe HD-PDGF/IGF-I-treated subjects.

DISCUSSIONThis Phase I/II clinical trial was the first human study toevaluate the effects of purified recombinant proteins on

periodontal regeneration. Its principal objective was to es-

tablish the safety of the growth factors; the secondaryobjective was to begin to accrue data on the biologicaleffects of the drug with respect to efficacy. Further, it was

hoped that some information might be obtained in regardto: 1) dosing; 2) identification of a subpopulation of peri-odontal lesions which represented the highest responders;and 3) identification of the most appropriate endpoints.

The aforementioned objectives were accomplished.Most importantly, there were no safety issues which re-

sulted from the use of the drug. Additionally, data dem-onstrating the biological activity of the product were ob-tained in the HD-PDGF/IGF-I group at surgical re-entry.Significantly greater bone formation occurred in osseousdefects treated with HD-PDGF/IGF-I than in sites receiv-ing full-thickness flap reflection and thorough debride-ment. These results appeared consistent with the findingsin preclinical dog and non-human primate studies. Whenthis growth factor combination was used to treat naturalperiodontal osseous defects in beagle dogs, substantialamounts of new bone and cementum were seen within 2weeks following a single application.4 Normal maturationof the bone continued for 3 to 6 months following treat-ment. A functionally oriented periodontal ligament was

present connecting the new bone to the tooth root surface.Further, ligature-induced periodontal osseous defects in

J Periodontol1192 PDGF AND IGF-I THERAPY FOR PERIODONTAL DISEASE December 1997

monkeys treated with 150 µg/ml each of rhPDGF andrhIGF-I demonstrated 42.5% defect fill, while placebogel-treated sites showed only 14.5% bone fill measuredby histomorphometry.612 The values reported here forHD-PDGF/IGF-I (Fig. 3) represent the high degree ofconsistency between the human and non-human primatemodel in response to PDGF/IGF-I-mediated periodontalregenerative therapy.

The selection of osseous defects for future studies mayrely on the distribution of the response of defect mor-

phology to treatment. This study shows, as other regen-erative trials have demonstrated, that interproximal de-fects respond more favorably to conventional methodsthan furcations.13-15 Osseous fill was found to be variablyincreased in angular lesions with or without treatment,while new bone fill was consistently low or non-existentin furcation lesions treated by control therapy. Extensivefurcation lesions may represent "critical size" defectswhich fail to respond to conventional periodontal debri-dement and flap surgery.16 By contrast, furcation lesions

treated in this study with HD-PDGF/IGF-I respondedmost robustly and consistently to the drug. The smallsample size in this study, however, does not allow forconclusive evidence for this effect in all patients.

Future expanded studies utilizing the informationgained from this trial evaluating the continued safety andefficacy of PDGF/IGF-I therapy to patients with peri-odontal disease are warranted. Higher doses may revealimproved efficacy and give insight into the scope of thisunique therapy to the field of periodontology. This firststudy in man not only produced encouraging safety andbiological activity data, but also provided information es-

sential to the refinement of future protocols.The use of PDGF alone or in combination with IGF-I

may also have many other applications as a stimulus forbone wound healing. Several preclinical animal studieshave demonstrated potent increases in Osteogenesis inbone defect models and in the treatment of osteoporo-sis.317 More recently, systemically administered PDGFhas been demonstrated to increase bone density (by Dexa

Table 6. Baseline Periodontal Lesion Distribution for Patients Treated with HD-PDGF/IGF-I

Angular (n = 11)(Mean ± S.E.M.)*

Furcation (n = 5)(Mean ± S.E.M.)

Lesion ControlHD-PDGF/

IGF-I ControlHD-PDGF/

IGF-I

Hard tissue Mean vertical osseous defeet depth (mm)

Mean horizontal osseous

defect depth (mm)Soft tissue Relative attachment level

(mm) 6.41Closed horizontal furcation

probing depth (mm)

4.18 ± 0.36 4.25 ± 0.38

± 0.47 6.63 ± 0.50

5.40 ± 0.93 5.20 ± 0.58

6.17 ± 2.17 5.67 ± 0.Í

*Standard error measurement.

Table 7. Soft Tissue Responses to HD-PDGF/IGF-I Therapy in Patients at 3, 6, and 9 Months

Time(months) Group

HorizontalProbing Depth

Reduction (mm)Relative Attachment

Value Level Gain (mm) Value

ControlPDGF/IGF-I

ControlPDGF/IGF-I

ControlPDGF/IGF-I

0.50 ± 0.292.330.503.001.302.00

0.330.760.001.401.14

NS

NS

NS

0.501.401.031.771.981.52

0.520.640.560.620.490.62

NS

NS

NS

Numbers are means ± standard error measurement.

Table 8. Soft Tissue Responses to LD-PDGF/IGF-I Therapy in Patients at 3 and 6 Months

Time(months) Group

HorizontalProbing Depth Relative Attachment

Reduction (mm) Value Level Increase (mm) Value

ControlPDGF/IGF-I

ControlPDGF/IGF-I

2.33 ± 0.621.67 ± 0.722.38 ± 0.710.69 ± 1.21

NS

NS

1.07 ± 0.471.40 ± 0.641.22 ± 0.561.77 ± 0.62

NS

NS

Numbers are means ± standard error measurement.

Volume 68Number 12 HOWELL, FIORELLINI, PAQUETTE, OFFENBACHER, GIANNOBILE, LYNCH 1193

scanning and histomorphometry) and strength in ovarec-

tomized animals.18 Expanded human clinical trials appearwarranted to assess the effects of growth factor-mediatedtherapy on the repair of orthotopic tissues.

ConclusionThe results from this Phase I/II human clinical trial showthat: 1) the product is safe when applied topically to peri-odontal osseous lesions during periodontal surgery; and2) a single application of the HD-PDGF/IGF-I is effec-tive, resulting in a significant improvement in bonegrowth and fill of periodontal defects compared to thecurrent standard practice (with or without placebo gel).Moreover, the results of this trial make possible additionalrefinements to future protocols.

AcknowledgmentsThis investigation was approved by the Harvard MedicalSchool Human Studies Committee and the University ofNorth Carolina-Chapel Hill School of Dentistry. Thisstudy was funded in part by NIDR grants K16 DE00275and 5T32 DE07010 and by a grant from the Institute ofMolecular Biology, Inc. This paper is dedicated to thememory of Professor Harry N. Antoniades (1923-1995).

REFERENCES1. Kiritsy CP, Lynch SE. Role of growth factors in cutaneous wound

healing: A review. Crit Rev Oral Biol Med 1993;4:729-760.2. Giannobile WV. Periodontal tissue engineering by growth factors.

Bone 1996;17:23S-37S.3. Lynch SE, Trippel SB, Finkelman RD, Hernandez RA, Kiritsy CP,

Antoniades HN. The combination of platelet-derived growth factorBB and insulin-like growth factor-I stimulates bone repair in adultYucatan micropigs. Wound Rep Reg 1994;2:182-190.

4. Lynch SE, Williams RC, Poison AM, et al. A combination of plate-let-derived and insulin-like growth factors enhances periodontal re-

generation. J Clin Periodontol 1989;16:545-548.5. Lynch SE, de Castilla GR, Williams RC, et al. The effects of short-

term application of a combination of platelet-derived and insulin-

like growth factors on periodontal wound healing. J Periodontol1991;62:458-467.

6. Giannobile WV, Hernandez RA, Finkelman RD, et al. Comparativeeffects of platelet-derived growth factor, insulin-like growth factor,individually and in combination on periodontal regeneration in Ma-caca fascicularis. J Periodont Res 1996;31:301-312.

7. Rutherford RB, Niekrash CE, Kennedy JE, Charette MF. Platelet-derived and insulin-like growth factors stimulate regeneration ofperiodontal attachment in monkeys. J Periodont Res 1992;27:285-290.

8. Park JB, Matsuura M, Han KY, et al. Periodontal regeneration inClass III furcation defects of beagle dogs using guided tissue regen-erative therapy with platelet-derived growth factor. J Periodontol1995;66:462-477.

9. Löe H, Silness J. Periodontal disease in pregnancy I. Prevalence andseverity. Acta Ódontól Scand 1963;21:533-551.

10. Silness J, Löe H. Periodontal disease in pregnancy II. Correlationbetween oral hygiene and periodontal condition. Acta Ódontól Scand1964;22:121-135.

11. Lynch SE. Methods for evaluation of regenerative procedures. J Per-iodontol 1992;63:1085-1092.

12. Giannobile WV, Finkelman RD, Lynch SE. Comparison of canineand non-human primate animal models for periodontal regenerativetherapy: Results following a single administration of PDGF/IGF-I.J Periodontol 1994;65:1158-1168.

13. Becker W, Becker BE, Berg L, Samsam C. Clinical and volumetricanalysis of three-wall intrabony defects following open flap debride-ment. J Periodontol 1986;57:277-285.

14. Froum SJ, Coran J, Thaller B, et al. Periodontal healing followingopen debridement flap procedures. I. Clinical assessment of soft tis-sue and osseous repair. J Periodontol 1982;53:8-14.

15. Poison AM, Heijl L. Osseous repair in intrabony periodontal defects.J Clin Periodontol 1978;5:13-22.

16. Schroer M, Wahl , Hutchens L, Moriarty J, Bergenholtz B. Closedversus open debridement of facial grade II molar furcations. J ClinPeriodontol 1991;18:323-329.

17. Nash TJ, Howlett CR, Martin C, Steele J, Johnson KA, Hicklin DJ.Effect of platelet-derived growth factor on tibial osteotomies in rab-bits. Bone 1994;15:203-298.

18. Mitlak BH, Finkelman RD, Hill EL, et al. The effect of systemicallyadministered PDGF-BB on the rodent skeleton. J Bone Min Res1996;11:238-247.

Send reprint requests to: Dr. T. Howard Howell, Department of Per-iodontology, Harvard School of Dental Medicine, 188 Longwood Ave-nue, Boston, MA 02115-5888. Fax: 617/432-4262.

Accepted for publication April 25, 1997.