DR. NANDA You have often been called thefather of biomechanics in orthodontics. How didyou get interested in biomechanics?

DR. BURSTONE As a young clinical ortho-dontist treating patients and supervising graduatestudents, I was very careful to take progress records,including headfilms, on my patients. In more sit-uations than I imagined, I was getting un desirableside effects. Even if the teeth were becoming wellaligned, they were not oriented in space where Iwanted them.

To begin with, I thought that I was not mak-ing the wires correctly, and then it dawned on methat I was making some basic errors in my bio-mechanics. The major problem was not poor wirefabrication or sloppy wires, but something much

more fundamental. This caused me to go back andlearn and relearn basic physics and engineering, soI could have a scientific basis for orthodontic bio-mechanics. In other words, I started as a clinicianwho saw problems and tried to understand why theyoccurred and what to do about them.

DR. NANDA You have done some classic re -search described in some seminal articles on thesubject of biomechanics. Based on this research,why do you believe that a thorough grounding inbiomechanics is necessary for the clinician?

DR. BURSTONE The bread and butter oforthodontic treatment is the application of forcesand force systems to alter tooth positions or toproduce bony changes. Therefore, the applicationof scientific biomechanics as we daily treat ourpatients can pay big dividends. The quality oftreatment improves, and we work much moreefficiently.

DR. NANDA How does “scientific biomechanics”differ from biomechanics in general?

DR. BURSTONE The phrase “biomechanics”has been commonly misused in orthodontics.Some times it refers to techniques; other times itrefers to clinical procedures such as wire bending.“Scientific biomechanics” relates forces and stress-es to our orthodontic problems. It answers suchquestions as:• What is the relationship between the applianceand the force system?• What is the optimal force system to producedifferent centers of rotation?• What is an optimal force? What is the relation-ship between stress and strain?

VOLUME XLI NUMBER 3 139© 2007 JCO, Inc.

JCO INTERVIEWS

Charles J. Burstone, DDS, MSPart 2 Biomechanics

Dr. Burstone is Professor Emer i tus and Dr. Nanda is Professor andHead, Department of Craniofacial Sciences, Alumni Endowed Chair,Division of Orthodontics, School of Dental Medicine, University ofConnecticut, Farmington, CT 06032. Dr. Nanda is also an AssociateEditor of the Journal of Clinical Orthodontics. E-mail Dr. Nanda atnanda@nso.uchc.edu.

Dr. Burstone Dr. Nanda

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 139

©2007 JCO, Inc. May not be distributed without permission. www.jco-online.com

• What are the biological changes in the perio -dontal ligament and in the bone?• What are the proper materials for optimal ap -pliance design?17

DR. NANDA Can you give us an example ofhow scientific biomechanics can be a practicaladjunct for the busy clinician?

DR. BURSTONE All of us would like to havecontrolled tooth movement. This is achieved byhaving a proper moment and force exerted from thebracket by means of a wire, using a multi-attach-ment appliance. Let’s look at space closure as anexample. The patient has incisors that we want toretract. In a loop arch for anterior retraction, lin-gual root torque is placed to prevent roots frombeing displaced forward. In early Tweed mechan-ics, labial root torque was placed. This was back-ward until we all learned that the tooth movementproduced was not desirable. Lingual root torque isrequired to prevent incisor roots from displacingforward. Have we learned our lesson? In somestraightwire appliances today, a flared incisor thatrequires retraction would produce labial roottorque. It’s the wrong force system—like goingback 60 years (Fig. 8).

There is more to controlled tooth movementthan some special bracket or bracket prescription.To gain control, we can slide teeth along the arch-wire, or we can build in the force system requiredfor non-sliding mechanics. The basic force systemis independent of technique or appliance.

DR. NANDA With all the refinements we nowhave in brackets and wires, are you saying thatstraight wires don’t always work?

DR. BURSTONE If we place a straight wireinto correctly placed brackets on crooked teeth, aresulting force system is inherently produced.Many times, this is exactly what we want, and welook like geniuses no matter what wires we use. Atother times, we can produce a secondary maloc-clusion that could be more problematic than theoriginal problem. Thus, a knowledgeable orth o -dontist will be able to differentiate those situationswhere straight wires work well and those situations

where they will not work well. There are manyexamples of undesirable side effects during level-ing. For instance, if the root of the cuspid is forward,leveling an arch will increase the deep overbite(Fig. 9). Other side effects include the developmentof open bites, canted occlusal planes, and crossbitesduring treatment.

As the old statement says, there is no freelunch. Whatever we do must be grounded in soundphysics and biomechanics. This includes the ques-tion of friction, which all clinicians should thor-oughly understand. In any event, without the cor rectforce system, we will not be happy with the result.

DR. NANDA We hear much about friction thesedays, and there is considerable interest in self-lig-ation with so-called “frictionless” brackets. Howimportant is friction, and is the elimination of fric-tion always desirable?

JCO INTERVIEWS

140 JCO/MARCH 2007

Fig. 8 A. Tipping central incisor lingually around cen-ter of rotation at its apex requires lingual force andlingual root torque. B. Straight wire produces labialroot torque, which tends to displace root forward.

A

B

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 140

DR. BURSTONE Friction between wires andbrackets can be both good and bad. During ca nineretraction, friction can reduce the force and min-imize tipping movements. This could be good.Other times, friction can be so great as to mini-mize or prevent tooth movement. Friction inorthodontic appliances is complicated and in -volves many factors. I cannot give a simple an -swer other than to emphasize the importance of

understanding all of the factors involved. Theseinclude the coefficients of friction of the materialsinvolved and, particularly, the role of the appliedforce system. The force system is the primarydeterminant of the friction force; for example, iftranslation is required using sliding mechanics,

Charles J. Burstone, DDS, MS

VOLUME XLI NUMBER 3 141

Fig. 9 A. Trying to use straight wire to level canine withits root inclined forward produces undesirable forcesand moments. Occlusal force from archwire causesincisors to erupt, increasing deep overbite (dotted lineshows that archwire would lie incisal to canine if notligated to incisors). Intrusive force and clockwisemoment on first premolar also tend to tip buccal seg-ments mesially. B. Overall effect of this straight-wireforce system tends to deepen overbite and producereverse curve of Spee in upper arch. Additional wiresand longer treatment time may be required to correctthis secondary malocclusion. C. Unwanted sideeffects can be eliminated by using bypass arch—sepa-rate continuous arch stepped around canine—whichmakes use of full arch to control anchorage. Separateroot spring is placed on canine to correct axial inclina-tion. D. TMA* root spring before and after activation.Placing bypass arch occlusal to canine bracket allowscanine extrusion. If no extrusion is needed, bypassarch should contact occlusal edge of canine bracket.

*Registered trademark of Ormco/“A” Company, 1717 W. CollinsAve., Orange, CA 92867; www.ormco.com.

A

C

B

D

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 141

friction will be very high. The idea of a friction-free bracket is a misnomer, unless the on ly typeof tooth movement required is simple tipping(Fig. 10).

DR. NANDA What about anchorage control?How do you prevent mesial drift of posterior seg-ments in extraction cases?

DR. BURSTONE There are a number of ap -proaches to minimizing anchorage loss in theextraction patient. These could include headgearsor other auxiliaries, but let’s discuss the intraoralforce system. There has been the suggestion ofusing Begg appliances for differential forces—heavy forces to slip anchorage and light forces tocontrol anchorage. For many reasons, I’m not surethis is a practical and valid ap proach. What myresearch has pointed to is another concept: a dif-ferential moment-to-force ratio on the anterior andposterior teeth. The higher moment-to-force ratioon the posterior teeth can produce translation whilethe anterior teeth exhibit controlled tipping. Themore uniform stresses of posterior teeth in theperiodontal ligament enhance the anchorage. Thisprincipal is universal for both sliding and non-sliding mechanics, but is most easily applied usingfrictionless loops (Fig. 11).

DR. NANDA What do you think of using head-gear to control anchorage? Most of our patientstoday are not well motivated to wear headgear.

DR. BURSTONE Headgears are certainly notobsolete. They can play a role, particularly inClass II patients to hold the upper teeth while themandible is growing forward. However, I think itis a mistake to use sloppy mechanics that loseanchorage and expect the patient to cover up ourmistakes by wearing a headgear. There is noquestion that scientific biomechanics can reducepatient compliance by minimizing anchorage lossand side effects.

DR. NANDA The orthodontist is bombarded withmany new appliances, techniques, and brackets; canan understanding of biomechanics be helpful inselecting the best appliance for a patient?

DR. BURSTONE As clinicians, when a new ap -pliance is presented, we have a number of al -ternatives. We can try it in our office. This istime-consuming and costly; furthermore, much oforthodontics requires long-term evaluation. It mightbe years before we recognize that the new proce-dure or appliance is not what is required. An under-standing of biomechanics allows us to rationallyanalyze the appliance or technique and come to a

JCO INTERVIEWS

142 JCO/MARCH 2007

Fig. 10 When canine is retracted with slidingmechanics, archwire exerts moments to preventtipping. These moments or couples produce verti-cal forces responsible for friction. A. With playbetween wire and bracket, tooth tips without fric-tion. B. After archwire engages, friction begins andless tipping occurs. C. Greater moment is pro-duced as tooth goes through translatory phase,when friction is highest.

Fig. 11 Posterior anchorage can be controlled byusing differential moments between anterior andposterior segments, pitting controlled tipping ofanterior teeth against mesial translation of posteri-or teeth. No differential forces are created becauseforces are equal and opposite. Intrusive forces onincisors can be helpful in deep-bite patients.

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 142

valid conclusion about its efficacy. Orthodonticcompanies by necessity introduce many of thenew appliances, and while most of these companiesare genuinely interested in the profession, sales-people obviously have a conflict of interest whenpresenting these appliances.

DR. NANDA Can biomechanics also help usbetter utilize any appliance that we might currentlybe using?

DR. BURSTONE The nice thing about scientif-ic biomechanics is that it is not dependent on anygiven appliance or technique. No matter whatappliance you use, it allows you to use it better withmore predictable results. Today, we have much toomuch commercialism in orthodontics; a healthydose of science in understanding appliances andhow they work is a good antidote. It is interestingto note that many of the new appliances that aresuggested are nothing more than reinventions of oldappliances.

DR. NANDA Do you have any advice for ortho-dontic researchers who are interested in evaluatingthe efficacy of different orthodontic appliances? Arethere biomechanical considerations they must beaware of as they plan and carry out their research?

DR. BURSTONE Good epidemiologic studiesare now being carried out to evaluate orthodonticappliances. The best ones are prospective, and theforce systems used in these orthodontic appliancesare a very important consideration. Even withwell-designed sampling procedures, however,definitive conclusions may be hard to find becauseof all the variables. Let’s suppose we want to com-pare the effects of headgear and a given function-al appliance. It is not enough to say headgearalone or, more specifically, occipital headgear.Occipital headgear has many directions of pull, andeven if the direction is given, there are many dif-ferent points of force application in respect to thecenter of resistance of an arch or a molar. It is lit-tle wonder that there is so much inconclusiveresearch appearing in our journals. Properlydesigned research must define the force systemmore completely; this should include moment and

force magnitudes, moment-to-force ratios, force-deflection rates, and force direction. In addition, wealso need good numerical methods to calibratetreatment changes.

DR. NANDA You’ve described appliances as beingforce- or shape-driven. What do you mean by that?

DR. BURSTONE Historically, orthodontic ap -pliances were developed, described, and taught asshape-driven. In the era of shape-driven appli-ances, which is still with us, you were taught howto bend or twist a wire or how to properly positiona bracket. That is all geometry—driven by shape.I think a better approach is to first determine yourorthodontic goal—what you would like to do—andthen determine the force system that is required toproduce that result. Then, and then only, can youdesign your appliance. It is important to have ashape, but it is more important that the shape pro-duces the de sired force system. And of course,many times that shape will look nothing like anideal or a straight wire. This does not mean that forevery activation we have to know exactly everyforce; most of the time knowing the relative forcesystem and the possible side effects can be morethan adequate. Thus, as we train our modern orth -odontists in graduate programs, a typodont coursenot only includes fabrication of the appliance, butalso an explanation of the force system that goeswith the appliance.

DR. NANDA I remember vividly a statementyou made when I joined you in 1972, and I oftenrepeat it in my lectures: “Orthodontics should notbe driven by personalities.” Well, it seems that no-thing has changed over the years, because our pro-fession continues to associate names withtechniques or brackets or other gadgets. What doyou think is going on?

DR. BURSTONE It is much more common in artthan in science to have different schools of thoughtassociated with the names of respected leaders. Wehave seen this in psychiatry, where many schoolsof thought developed originally. As psychiatrybecame more of a science, there was no longer aneed for different schools of thought and famous

Charles J. Burstone, DDS, MS

VOLUME XLI NUMBER 3 143

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 143

gurus. I hope that as orthodontics has matured, wecan still respect our great teachers as well as theindividual thinkers. It is a major responsibility oforthodontic departments and those of us who teachgraduate students to make sure that every studenthas a good background in the science of ortho-dontics, especially biomechanics.

DR. NANDA What is the scientific basis for thefield of orthodontic biomechanics?

DR. BURSTONE Biomechanics didn’t origi-nate with orthodontics. It is based upon the pio-neers of physics like Galileo and Newton. Morerecently, it includes research in material science,mechanics of materials, beam theory, finite-ele-ment and computer science, etc. Those of us whodo research in orthodontic biomechanics build onthe foundation from the basic science of engi-neering and physics.

DR. NANDA It is not uncommon to hear lecturesor to read articles in which the orthodontist developshis or her own system of biomechanics to explainwhat is happening. How do you feel about this?

DR. BURSTONE This is part of the problem. I’msure that we have many intelligent orthodontists;however, they are not in the same genius categoryas Newton. It is not only more valid to build onaccepted scientific principles, but it has the addedadvantage of allowing us to use terminology thatcan be understood and respected by all scientificspecialties. A perusal of our orthodontic journalscan show many unusual notations, funny words,signs and symbols, diagrams not in equilibrium, andimaginary or impossible force systems.

DR. NANDA So how do we get our fellow ortho-dontists to be more knowledgeable about scientif-ic biomechanics?

DR. BURSTONE It must start with the trainingof our graduate students. In addition to courses inbasic sciences, cephalometrics, techniques, ap -pliance fabrication, treatment planning, dentalmaterials, etc., they need a solid course in scien-tific biomechanics. This course should include

how to handle resultants and components of forcesacting on a body or a point, equivalent force sys-tems, the biomechanics of tooth movement, equi-librium, equilibrium diagrams, mech anics ofappliance design, force systems produced bystraight or bent wires, and other applications forheadgears and functional appliances. Of course, thismaterial is much harder to teach than the technicalaspects of orthodontics, which can be more quick-ly learned. The challenge is there and must bemet by our orthodontic programs.

DR. NANDA There has been a lot of interestrecently in “evidence-based orthodontics”. Do youthink we can practice orthodontics like medicine andsurgery, where clear evidence is available regard-ing the efficacy of one procedure over another?

DR. BURSTONE There seems to be generalacceptance in the field of medical epidemiology ofthe importance of a hierarchy of evidence. At thelowest level, one finds anecdotal evidence, and atthe highest level, one finds a systematic review ofmany studies or a meta-analysis. This is a goodframework to start from, since orthodontics hasalways suffered from a pluralism of individualopinions, from the individual orthodontist to theorthodontic guru. However, there are some prob-lems with evidence-based orthodontics. First,unlike medicine, we are not dealing with diseaseor diseased tissues. We all establish individualizedtreatment goals and procedures. To obtain thesegoals, our mechanical procedures can be very dif-ferent for any given patient; hence, unlike medicine,the endpoint may not be clear. For example, ifthere is an arch-length inadequacy, one orthodon-tist might decide to extract, and another ortho-dontist might solve the problem by expansion.Thus, an epidemiologic study may not be com-paring the same endpoints. Nevertheless, the basicapproach of evidence-based orthodontics is a goalthat we should aim for.

DR. NANDA But are there good enough studiesto establish meaningful protocols for treatment?

DR. BURSTONE This is the problem. Many of

JCO INTERVIEWS

144 JCO/MARCH 2007

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 144

the studies that are quoted in the literature havemajor deficiencies, either in sampling or in method.Thus, they do not reach the level of quality for ameta-analysis to be performed. Med i cine has theadvantage, as I pointed out, of dealing with moredefinitive endpoints. The pa tient lives or dies.Longevity can be measured. The disease is cured.What is the endpoint for a successfully treatedClass III? Protruded upper incisors? Retruded lowerincisors? Growth alteration?

DR. NANDA Whatever the methodology, manyorthodontists are reluctant to accept evidence-based approaches to treatment. They claim thattheir clinical experience and judgment are thebasis for their treatment, rather than a bunch ofdusty old articles.

DR. BURSTONE We all certainly learn a lotfrom our clinical experience; however, this does notmake evidence. Unlike medical practitioners, mostof us practice in isolation, and thus, there is littlefeedback from the rest of the profession on whatwe’re doing. We can harbor these beliefs even ifthere is no validity to them.

DR. NANDA Let’s take the example of Class IItreatment. By my recent count, there are at least 20different appliances available to move a maxillarymolar distally. Most of the appliances have beenintroduced without any evidence-based research.How do you feel about this?

DR. BURSTONE Most of these appliances usethe principal of the Nance button for anchoragecontrol during distalization. There is good evi-dence that these appliances can move molars dis-tally, primarily by tipping, during the initial phaseof treatment. Unfortunately, the premolars comeforward; that is anchorage loss. The best studiesshow that the centers of resistance of the molarsmove back the same amount as the center of resis-tance of the premolars moves forward. Fifty per-cent of anchorage loss is not very im pressive.During an intermediate second phase, bicuspids aresupposed to drift distally. Unfortu nately, the driftis minimal, and the molars come forward. Then

there is a question of the overall effect. Many ofthese patients are also treated with Class II elastics,and therefore the studies do not show if the ClassII correction is due to molar distalization or to hold-ing of the upper molars, with mandibular growthaccounting for the occlusal change. In other words,in successful Class II correction, has the ortho-dontist distalized the molars and held them back,or is the correction mainly due to favorable growth?

DR. NANDA The FDA requires different phas-es for verification of an appliance in medicine. Dowe need a regulatory body to validate the claimsof the various orthodontic manufacturers?

DR. BURSTONE I’m not sure that we want toask the FDA to have more stringent requirementsfor appliances that are not placed into tissues. I’mnot sure that the AAO is in a position to serve asa regulatory body. Orthodontic companies can domore; they can at least have research demonstrat-ing the efficacy of their appliances. But fun-damentally, it is the intelligent, well-trained ortho-dontist who must be responsible for evaluating theappliances and materials that he or she uses.

DR. NANDA Over the years, many orthodontistshave discussed with you some of their failuresand problematic patients. Based on what they’vetold you, what do you consider the most difficultproblems?

DR. BURSTONE Leaving out problems of pa tientcooperation, I would have to pick the treatment ofasymmetries. There are extreme skeletal discrep-ancies that can be treated with orthognathic surgery,and in these cases, the orthodontics is simple. Thedifficult patients are a subset: the nonsurgical treat-ment of asymmetries. Some of these patients are nottoo difficult because they involve mandibular shifts,and fairly simple symmetrical mechanics can leadto an excellent end result. But other patients canhave small skeletal asymmetries or even develop-mental dental asymmetries that are hard to treat.Treatment may lead to failure if symmetricalmechanics are used. Since no mandibular shift ispresent in these patients, the malocclusion can

Charles J. Burstone, DDS, MS

VOLUME XLI NUMBER 3 145

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 145

only be corrected with well-planned asymmetricalmechanics. This is much more challenging than theuse of symmetrical mechanics. The use of a lot ofintermaxillary elastics, such as Class II, Class III,or criss-cross elastics, may not lead to a stable cor-rection. There must be differential tooth move-ment between the right and left sides to obtain thecorrection. These asymmetries are challengingbecause they require good planning, mechanics, andexecution.18

DR. NANDA How is treatment planning for theasymmetrical patient different from our typicalsymmetrical situations?

DR. BURSTONE A good example is the so-called midline discrepancy associated with a smallmandibular asymmetry. These are patients wherethe asymmetry is not great enough to suggestorthognathic surgery. Typically, a midline is estab-lished first. One could connect midpoints such assoft-tissue nasion, subnasale, and pogonion. Thisis valid and easy to do if all points coincide, but withan asymmetry, they do not. Picking two pointssuch as nasion and subnasale to establish an upperfacial midline may not be the best choice. I thinkin terms not of a midline, but of a “mid-arc”,which is a reasonable fit of a curve going throughkey midpalatal plane structures. This simplifiestreatment, since less incisor root movement isrequired, and a better esthetic result is achieved. Inthis view, the concept of a facial midline is a fal-lacy. We should be thinking more in terms of cur-vatures (Fig. 12).

Another example is an apical midline dis-crepancy. If an apical base discrepancy is recog-nized, it might be better to angle the incisor bracketsfor compensation. A small axial inclination dif-ference will not be noted by the patient, and treat-ment then becomes much simpler. This avoids theuse of asymmetrical intermaxillary elastics, suchas anterior criss-cross elastics, that can alter theplane of occlusion, producing unsightly canting.

DR. NANDA You have been a role model tohundreds of orthodontic residents, and thousandsof orthodontists have benefited from your contri-

butions to the profession. Where do you see the pro-fession at present?

DR. BURSTONE I am very positive about ortho-dontics in general and about orthodontic educationtoday. The quality of our graduate students isexcellent; the best of the best are electing to go intoorthodontics. The overall quality of graduate train-ing has never been better. There is more uniformitybetween programs, so that graduate students from

JCO INTERVIEWS

146 JCO/MARCH 2007

Fig. 12 A. Establishing upper facial midlinebetween nasion and subnasale would require con-siderable upper incisor translation in any nonsur-gical treatment of this patient with markedmandibular asymmetry. B. Smooth arc formingmidline curvature is better fit for skeletal discrep-ancy, as seen at both incisors and skeletal land-marks. Whenever facial symmetry will not be cor-rected surgically, tooth positions should fit curverather than straight line.

A

B

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 146

any university can obtain an excellent education.Of course, I would like to see more biomechanicsin the curriculum. This continues to be a majorproblem that is not easily resolved. Scientific bio-mechanics requires teachers with a fundamentalknowledge of physics and engineering, which is nota common background for orthodontic graduates.In addition, it is important that graduate studentshave time and proper supervision to study theirpatients in depth. Although it is necessary fororthodontic departments to earn money to sup-port their programs, an overload of patients can becounterproductive to learning.

DR. NANDA The reason often given for the cur-rent shortage of teachers is a disparity in income,but we had this disparity when we both embarkedon our educational careers. Why do you think theproblem is so acute now?

DR. BURSTONE It is important that we do every-thing we can to increase and maintain competitivesalaries for our best faculty. The AAO, the AAOFoundation, and alumni have all been helpful andshould continue their efforts. It is also important thatthe universities maintain a happy, creative, andstimulating environment for their faculty.

In my career, I’ve always enjoyed every day.Nowadays, there can be too much pressure on fac-ulty members for increased clinical production orpromotion, which can create an unproductiveatmosphere.

In addition, part-time faculty members whoare clinicians in practice have to increase their con-tributions to orthodontic programs. We all have togive back something to a wonderful profession.This is not only valuable to the programs, the grad-uate students, and the profession, but as many part-time faculty members will tell you, it is a pos itiveexperience for them. Overall, if we, the acade-micians and the practitioners, pull together, we canlook forward to a positive future for orthodontics.

DR. NANDA On behalf of the readers of JCO, Iwant to thank you for this interview.

REFERENCES

17. Burstone, C.J.: Application of bioengineering to clinical ortho-dontics, in Orthodontics: Current Principles and Techniques,ed. T.M. Graber, R.L. Vanarsdall, Jr., and K.W.L. Vig, 4th ed.,Mosby, St. Louis, 2005, pp. 293-330.

18. Burstone, C.J. and Marcotte, M.R.: Problem Solving inOrthodontics: Goal-Oriented Treatment Strategies, Quint -essence, Chicago, 2000.

Charles J. Burstone, DDS, MS

VOLUME XLI NUMBER 3 147

P.139-147 Interv_Burstone:Interv_Burstone 3/23/07 8:59 AM Page 147