One Hundred Years of Insulin

One Hundred Years of Insulin

Angus M Brown

1

OneHundredYearsofInsulin


AngusMBrownB.Sc.,Ph.D.,FPhysiol,SFHEA

SchoolofLifeSciences

UniversityofNottingham

Nottingham,NG72UH

UK

DepartmentofNeurology

UniversityofWashington

Seattle,WA98195

USA

Email:[email protected]

2


In the aftermath of the First World War a young researcherworking in a North American laboratory demonstrated dramaticreductionsinthebloodglucoseconcentrationofdogsrendereddiabeticby pancreatectomy when injected with solutions of freshly groundpancreas. He convincingly argued that the ability of the pancreaticextracttoreducediabetichyperglycaemiawasevidenceofarolefortheinternalsecretionofthepancreasintheoriginofdiabetes,andthatthissecretioncouldactasapotentiallyusefultherapeutictoolintreatingthedisease. That this researcher was Israel Kleiner, an American nowforgotten, and not the Canadian Frederick Banting, globally lauded asthe pioneer of insulin, dispels the notion that the discovery of insulinemerged from a vacuum in the scientific backwater of Toronto, butratherwasaninevitabilityfoundedonthreedecadesofintensivestudybynumerousresearchers,someofwhocametantalisinglyclosetopre-emptingBanting.

Diabetes At the time of Banting’s discovery of insulin in 1922 diabetes mellitus wasconsideredadiscretecondition;itwasonlyin1936thatthedistinctionbetweentype1andtype2diabeteswasmade(Himsworth,1936),wheretheformerresultsfromlackofinsulinproduction(insulindependentdiabetes),andthe latterfrominsensitivityto insulin(insulinindependent diabetes). The requirement for an effective treatment for diabetes waspainfullyobviouslytothesufferers,theirfamiliesandthedoctorstreatingdiabeticpatients(Joslin, 1916). The disease occurred typically in pre-pubescent children, its victimsindiscriminately struck down. The disease onset was rapid, the predominant symptomsbeinghunger,thirst,weightlossandexcessiveurination(polyuria)withadefinitivediagnosisconfirmedbythepresenceofglucoseintheurine(glucosurea)andelevatedbloodglucoseconcentrations(hyperglycaemia;DiMeglioetal.,2018).Thepatientswoulddiewithinayearemaciated and wretched. The only effective treatment, promoted by eminentendocrinologistFrederickAllen,wasareducedcaloriedietthatrestrictedthecarbohydratethat fuelledhyperglycaemia,ketosisandcomathat inevitablyprecededdeath(Allenetal.,1919; Mazur, 2011). However, this seemingly inhumane treatment of starving emaciatedpatientsonlyextendedlifebyafewmonths.

Diabetesresearchupto1920 In1869 theGermanmedical studentPaul Langerhansproposed that thepancreascontained two distinct types of cell (Langerhans, 1869). The acinar cells were ordered inclusters, and produced the digestive enzymes secreted into the duodenum via thepancreaticductsthatexpeditefooddigestion.Thesecondcelltypewasexpressedinislandsorisletsarrangedasdistinctclumpsthroughoutthepancreas,butnofunctionwasascribedto them. The French expert Laguesse subsequently named these cells the islets ofLangerhans, and suggested that if the pancreas had a functionother than its role in fooddigestiontheseisletcellsweremostlikelyinvolved(Laguesse,1893).Amajorbreakthrough

3


indiabetesresearchoccurredin1889whenMinkowskiandvonMering,attheUniversityofStrasburg,ligatedthepancreaticductsofdogs,therebypreventingsecretionofthedigestiveenzymesandshowedthisdidnotcausediabetes.Howeveruponcompleteremovalof thepancreas (pancreatectomy) the dog became diabetic, exhibiting glucosurea (Mering &Minkowski, 1889). The absence of the pancreas caused diabetes. Thiswas followed up in1893 by, the French researcher Hédon, who removed the majority, but retained a smallamountofpancreasindogs,whichreducedthesupplyofpancreaticjuices:thedogsdidnotbecomediabetic.HoweverwhenHédon removed the remainderof thepancreas thedogsimmediately developed diabetes and died within a week (Hédon, 1893). In 1901 EugeneOpie,atJohnsHopkins,establishedthecorrelationbetweendamagetoisletsofLangerhanscellsandthedevelopmentofdiabetes,whichledtothebeliefthatthesecellsproducedaninternal secretion that was the key to diabetes (Opie, 1910). The fundamental clinicalconclusiondrawnwasthatdiabetescouldbetreatedwithextractofthepancreascontainingtheinternalsecretion.

These resultswere convincing evidence that the pancreas performed two distinctroles, producing the digestive enzymes (external secretion) and a blood borne internalsecretionthatcontrolledbloodglucose.Theseearlystudiessuggestedamodelfordiabetesresearchthatbecamethestandardforthenextthirtyyears:prepareanextractofpancreas,which was injected subcutaneously into a dog rendered diabetic by pancreatectomy todetermineitsabilitytodecreaseglucosurea.Therelativesimplicityoftheprocedureandtheurgent clinical need for an effective treatment for diabetes attracted hundreds ofresearchers, but the collective results were inconclusive and discouraging, a seeminglyunavoidable consequence of the procedure were harmful side effects such a fever,convulsions and infections. Itwas concluded thatdegradationof the internal secretionbythedigestiveenzymesoftheexternalsecretionmightexplaintheunpredictablepotencyoftheextracts.TocircumventthisissueRennieandFraser,attheUniversityofAberdeen,usedpancreaticextractsfromfish,wheretheisletcellsareanatomicallydistinctfromtheacinarcells,buttheirresultswereinconclusive(Rennie&Fraser,1907).

GeorgZuezler

In1906Zuelzer,aGermanresearcher,preparedextractsfromfarmanimalpancreasobtainedforalocalslaughterhouseanddemonstratedreducedglucosureaindiabeticdogs.Healsoinjectedtheextractintoadiabeticpatient,andalthoughhecouldnotmeasuretheglucose in theurine, thepatient’smoribundstate improved.Hereceived financial supportfromtheScheringdrugcompany,andin1907showedtheextractwaseffectiveinreducingglucosurea in diabetic humans, but side effects included fevers, vomiting and convulsions(Zuelzer,1908).In1911,fundedbyHoffman-LaRoche,heproducedanextractthatcausedconvulsions in the test animals, and the studieswerediscontinueddue to thedifficulty inisolatingtheinternalsecretionfromtheimpuritiesassumedtounderliethesideeffects. In1912ELScott,at theUniversityofChicago,consideredthe failuresof thepreviousstudieswere due to the external secretion destroying the internal secretion, and reasoned thatligating the ducts,which caused the acinar cells to atrophy,would eliminate the externalextract, butwas unable to successfully carry out this surgical procedure. Instead he usedalcohol as a solvent to isolate the internal secretion, which Zuelzer had also used. Scottoptimised his extraction technique and found the dogs’ glucosurea disappeared and thattheir temperament improved (Scott, 1912; Scott, 1913). However his advisor Carslonwassceptical of the results, thinking the experiments were not sufficiently controlled. Scottconsulted JJR Macleod about this work, who informed him of studies by the BritishresearchersKnowltonandStarling,whoseworkfailedtoshowanyeffectoftheextractondogs.Discouraged,Scottmovedontoresearchotherareas.Inthedecadeleadinguptothe1920s technical improvements (see later) made measuring blood glucose concentrations

4


easier, faster, andmore reliable,whichwas amajor advance for researchers, since bloodglucoseconcentrationwasafarmoreaccurateindicatorofglycaemiccontrolthandetectingthepresenceofglucosurea(Clarke&Foster,2012).

IsraelKleiner

In 1914 Israel Kleiner, working with Meltzer in the Rockefeller Institute,demonstratedthatinjectionofaglucosebolusintoapancreatectomiseddogincreasedthebloodglucoseconcentrationthreefoldcomparedtoanormaldog,andthatco-injectionofpancreatic extract with the glucose bolus prevented the increase. The First World Warinterrupted the research, but in 1919 Kleiner published a landmark paper in which heshowedthatthepancreaticextractreducedthebloodglucoselevels indogs,thefirstsuchmeasurement. Kleiner stated that the ability of the pancreatic extract to reduce bloodglucosewas supportive of the role for the internal secretion in diabetes, and that humandiabetic patients could be effectively treatedwith pancreatic extract, which if sufficientlypurewouldlimitthetoxicsideeffects(Kleiner,1919).

Figure1-Infusionofpancreaticextract(opencolumns)reducedbloodglucoseconcentrationindogsrendereddiabeticbypancreatectomy,whereasextractofsubmandibulargland(greycolumns)hadnosignificanteffect.AdaptedfromFriedman’sreanalysisofKleiner’soriginaldatawithglucoseexpressedinmmoll-1(Friedman,2010;Kleiner,1919).

TheDiscoveryofInsulin FrederickBanting,ashisbiographersemphasise,wasanextremelyunlikelyscientifichero(Bliss,1992).Hefailedhis1styearattheUniversityofToronto,Canada,studyingarts,andwasonlyallowedtoenrol inmedicine in1912afterpetitioning theUniversitySenate,having liedabouthisageonhisapplication.Bantingplacedaboutaverage inhisclass.ThestartoftheFirstWorldWarinterruptedBanting’smedicaleducationwithhis5-yearcoursecompressed into4years, leavingBanting feelingdeficient inhis training.He specialised in

5


surgeryandgraduatedinDecember1916,reportedformilitarydutythenextdayandsailedforEngland inMarch1917.Hewassent toFrance in June1918andservedasa front linemedical officer where he treated thewounded. However on the 28th Sept 1918, just sixweeksbeforetheArmistice,hewaswoundedbyshrapnelandawardedtheMilitaryCross,thecitationnoting,“hisenergyandpluckwereofaveryhighorder”.HewasinvalidedtotheUK and treated in Manchester for 9 weeks when the wound became infected. Heconvalesced in Scotland andwas recalled to Canada in February 1919, where he spent 6months treating wounded soldiers at the Christie Street Hospital in Toronto. He wasdischargedinsummer1919andcompletedhissurgicaltrainingattheTorontoHospitalforSickChildren,wherebyFebruary1920hehadassistedin232surgeries.HelefttheHospitalfor Sick Children in June 1920 and commenced private practice in London, Ontario. Hispractice was slow in becoming established and he was in debt. He carried out someadditionalworkasademonstratorinsurgeryandanatomyatWesternUniversityinLondonfor FR Miller, Professor of Physiology, whom he also helped in the lab. The momentouseventsof31stOctober1920,themostpropitiousdateindiabetesresearch,aroseasaresultofa lectureBantingwasdue todeliveron the roleof thepancreas,aboutwhichonly thebasics of its exocrine function were known. Banting read a recent publication by MosesBarroninthejournalSurgery,GynecologyandObstetricsentitled“Therelationoftheisletsof Langerhans to diabetes with special reference to cases of pancreatic lithiasis”, whichreported on an autopsy in which a pancreatic stone had occluded the pancreatic ducts,leading to loss of acinar cells, although the islet of Langerhans cells survived. Theseobservations supported the previously reported ligation studies on dogs. The keyinformationBantingabsorbedwasthatligationofthepancreaticductscausedthepancreastoatrophy, theacinar cellsdied,butdiabetesdidnotdevelop.Ashe tried to sleep in theearlyhoursBantingscribbledthefamouslines:

DiabetusLigatepancreaticductsofdogs.KeepdogsalivetillacinidegenerateleavingIslets.Trytoisolatetheinternalsecretionofthesetorelieveglycosurea.

Bantingwouldlaterrefertothisas‘theIdea’.Howeveritwasfarfromoriginal,butBanting’sknowledgeofthediabetesliteraturewassuperficial,andhewasunawarethatELScotthadproposed ligating ducts and that Zuezler and Kleiner had tested extracts of pancreas ondiabeticdogs.MillerputBanting intouchwithJJRMacleod,ProfessorofPhysiologyattheUniversityofToronto,aworld-renownedexpertoncarbohydratemetabolism.Macleodhadavoluminousknowledgeofmetabolismandin1913publishedabookentitledDiabetes:ItsPathologyandPhysiology,whichwasa summaryof the search for thepancreatic internalsecretion.ThefamousmeetingbetweenBantingandMacleodattheUniversityofTorontoonMonday7thNovember1920hasbeen recountedmany times (Bliss, 1983) and canbesummarised as follows. Banting suggested experiments in which the pancreatic ducts ofdogswereligated,andafterallowingforthepancreastoatrophy,itwouldberemovedandgrafted into another dog rendereddiabetic by complete pancreatectomy. Thedog’s urinewouldbetestedforthepresenceofglucose, itsabsenceindicativeoftheglucoseloweringeffects of the graft. Macleod suggested at this meeting freezing the pancreas prior totreatmentwithalcoholtoextracttheinternalsecretion.MacleodprovidedBantingwiththeuseofalaboratory,aboutadozendogs,andalaboratoryassistant,undergraduatestudentCharlesBest(Bliss,1983).

Theexperimentscommencedon17thMay1921andwerecompletedninemonthslater.Animportantfeatureofthestudieswasthattheyevolvedtowardsanidentifiedendgoal, to produce a purified pancreatic extractwith glucose lowering properties. An under

6


appreciatedaspectoftheexperiments,butonethatultimatelyledtosuccess,wasBantingandBest’swillingnesstoimmediatelyabandondeadendsandtorapidlyadapttheirprotocolbasedonthesuccessorfailureoftheirmostrecentexperiments.Thisledtorapidchangesinthe experimental design agreed with Macleod. Banting and Best realised that delays inproductionofthepancreaticextractfromligateddogswasthebottleneckintheirprogress.BantingwasawareofLaguesse’sdescriptionof(b)ovinepancreas,wheretheisletcellsarefully developed in foetal and young calves, but the acinar cells are immature (Laguesse,1896).Bantingsuggestedthattheyobtainextractfromfoetalcalfpancreas,aninspiredideathat bypassed the limitations of workingwith ligated dogs. Banting then suggested usingadult cow rather than foetal calf pancreas,whichwas successful, and at around this timethey adoptedMacleod’s suggestion of using alcohol as a solvent. Several researchers hadreachedthestageofdemonstratingextractfromdogshadglucoseloweringproperties,butwere unable to proceed any further due to the harmful side effects that accompaniedinjection of the extract. Whether by deliberate design, or most likely as a result ofserendipitous gut feeling, Banting and Best incorporated these two critical modificationsinto theirexperimentaldesign,whichallowedthemtomaintain theirmomentumtowardsthe success thathadeluded somanyother researchers.As theirproductivity increased inNovember 1921Macleod acceded to Banting’s request for additional manpower and JMCollip joined the group. James Collip was a similar age to Banting but was Professor ofBiochemistry, an experienced researcher and knowledgeable in current biochemistrytechniques.CollipstartedworkinginthefirsthalfofDecember1921onpurifyingextractsofbeefpancreas.Macleodsuggestedusingrabbitsasthetestanimalastheywerecheaperandeasiertoworkwiththandogs.Colliprealisedthatworkingwithnormalrabbits,asopposedto diabetic rabbits, was viable as they responded with decreased blood glucoseconcentrations to injection of the extract. Two additional experiments completed theiragenda by the end of 1921. The first showed that injection of the extract into a dogincreased its hepatic glycogen concentration, normally negligible in a diabetic dog. Thesecondexperimentwastheirdemonstrationthatregularinjectionsofextractkeptadiabeticdogaliveforseventydays,whereasmostpancreatectomiseddogsdiedwithinaweek(Bliss,1983).

7


Figure2-SchematicillustrationoftheTorontogroup’sexperiments.A.TheplanproposedbyBanting and agreed by Macleod at their first meeting involved removing a substantialamountofpancreasfromadog,theremainingatrophiedpartofthepancreastobegraftedinto a separate dog rendered diabetic by complete pancreatectomy, with measures ofglucoseintheurinetotesttheglucoseloweringpoweroftheprocedure.B.Theexperimentalprotocol evolved rapidly, with bovine pancreas extract injected into rabbits, whose bloodglucosewasmeasuredtodeterminetheeffectivenessofextract.

At a conference in Yale on 30th December 1921 Banting, Best and Macleodpresented their results to cautiously interested diabetologists. However Banting felt thatMacleodwas taking undue credit for hiswork and his simmering resentment ofMacleoddeveloped into irrational hatred. Banting also quickly became jealous of Collip who hadparachuted into thestudyand immediatelyprovedmoreadeptatobtainingviableextractthanhimselfandBest.Inanattempttore-establishhisprimacyBantingmadeaverypoorlyjudgedclinicaldecisionwhenhepersuadedMacleodtoallowahumandiabeticpatienttobeinjectedwithextracthehadpreparedwithBest.Thedatewas the11th January1921andthepatientwas14-year-oldLeonardThompson.Theextractproducedamodestdecreaseinbloodglucosefrom24.4mmoll-1to17.8mmoll-1,butthepresenceofketonespersisted.Asterile abscessappearedat the injection siteand theprocedurewas judgeda failure. Theextractwasdescribedas‘thickbrownmuck’andthepatientwasinjectedwith15mlintotal,an enormous volume for a subcutaneous injection. This description clearly illustrates thelimitations of Banting andBest’s extract,whichwas toodilute and too impure for clinicaluse.CollipcontinuedtoworkonpurifyingtheextractandonJanuary23rd1921ThompsonwasadministeredwiththeCollipextractasaresultofwhichhisbloodglucoseconcentrationfell from28.7mmol l-1 to6.7mmol l-1,hisdemeanour improvedandhe felt strongerwithcontinued treatment over thenext fewdays. Itwas the first convincingdemonstrationof

8


theanti-diabeticqualitiesofpancreaticextract.AtthispointhoweverBanting’smeaningfulcontributiontotheworkwascomplete.TheConnaughtAnti-ToxinLaboratories inTorontopurifiedtheextractonanindustrialscaleunderCollip’sdirectionandpatientsweretreatedbyqualifieddiabetologists.ByFebruary1922DrWalterCampbellwastreatingsixpatientsatthe diabetes clinic he founded at the University of Toronto, and as word spread of themiraculous discovery in the press patients and doctors clamoured for insulin. Macleodpresentedthegroup’sresultson3rdMay1922atameetingoftheAssociationofAmericanPhysiciansinWashingtonDC.Apublicationoftheresultspresentedatthismeetingincludedthewordinsulin(page4ofBantingetal.,1922),MacleodusingtheLatintranslationforislettodescribethepurifiedinternalsecretion.

Figure3-Themiracleofinsulintherapy.A.(i)PatientJLaged3yearssufferingfromtype1diabetes.Weighed15lbsonDec15th1922,(ii)butafterundertakinginsulintreatmentwastransformed,weighing 29 lbs by Feb 15th 1923. B. (i) Theodore (Teddy) Ryder, diagnosedwithdiabetesatage4,cametoTorontofortreatmentasa5yearoldweighing26lbsinJuly1922.Thelifechangingeffectofinsulintherapyisevidentinphotographsfrom(ii)1923and(iii)1929.Ryderispicturedatage66(iv)andlivedto76yearsofage,atthetimeofhisdeathhewaslongestsurvivingpatienttreatedwithinsulinintheworld.HoweverRyder’slongevitywas the exception for those diabetics treated with insulin before the advent of modernglucosemonitoringandaccurateinsulinadministration.

Readers interested in theworkcarriedout inTorontoaredirected to thecriticallyacclaimed (Clarfield, 2009; Matz, 2000; Trehtewey, 1988) book The Discovery of Insulin(Bliss, 1983), which describes in forensic detail the background to the studies, theexperiments,theextractionprocedureandtheaftermath.Thereisalsoashorteraccessiblesummary, which focuses on key experiments (Cardoso et al., 2017). Numerous reviewarticleshavebeenwrittenon thediscoveryof insulinover the last onehundred years, of

9


whichthefollowingcanberecommended(Barthold,2004;Bliss,1993;Bliss,2013;deLeivaetal.,2011;Feasby,1958;Fralick&Zinman,2021;Karamitsos,2011;Macleod,1978;Pratt,1954;Rosenfeld,2002;Rothetal.,2012;Stansfield,2012;Vecchioetal.,2018;Whitfordetal., 2012).Banting,Best,MacleodandCollip, in various combinations,publishedover fiftypapers between 1922 and 1924 (listed in the Bibliography section of The Discovery ofInsulin),which if read intheappropriatechronologicalorderprovideadetailedaccountoftheirprogressfromtheoriginalstudiesondogstothepurificationoftheextractanditsusein treating diabetic patients. The city of Toronto is justifiably proud of the discovery ofinsulinandseveralwebsiteslinktohistoricallyfascinatingcontent(CBC,1990;UofToronto1;UofToronto2;UofToronto3).

NobelPrize Inlate1923theNobelPrizeCommitteeannouncedthatBantingandMacleodweretobeawardedthePrizeforPhysiologyorMedicineforthediscoveryofinsulin.ThisawardrequirednojustificationasinsulinwastheninwidespreaduseinNorthAmericaandEurope.Bantingwasonly32yearsoldandacenturylaterremainstheyoungestrecipientofthePrizefor Physiology/Medicine. However Bantingwas furiouswithMacleod: in hismind he haddiscovered insulin in spiteofMacleod, not as a result of his support andencouragement.BantingseriouslyconsideredrefusingtheawardbutsettledonsharinghisprizewithBest,asentimentechoedbyMacleodwhosharedhisprizewithCollip.Thehatred, for this isnottoo strong aword, that Banting felt forMacleod lasted until his death twenty years later(Bliss, 1983; Bliss, 1992). Unlikely as it may sound Banting and Macleod werecomplementary, ifnot ideal,partnerswhose interpersonal chemistry fuelled their success.Banting,althoughuntrained in research,hadadrivingambition toundertake research,anunhappyengagementand financialproblemspromptinghim to leaveLondon forToronto.AroundthistimeBantingcommunicatedwiththegreatSherringtoninEnglandonthetopicof reflexes,evidenceofhisemerging interest in research.Banting’snaivetéandsuperficialknowledgeofthedifficultiesofhisendeavourensuredhewasnotdeterredbythefailureofso many others. Macleod was widely read on the topic proposed by Banting and ideallyinformed to offer advice. Banting’s skill as a surgeon, and his suggestion of graftingdegenerated pancreas into dogs, likely appealed to Macleod, previous attempts at thisprocedurehavingendedinfailure.Banting’sproposalclearlyintriguedMacleodsufficientlyto inducehim to provide costly facilities and resources. If Bantinghadbeen awareof thefateofKleiner,whowasforcedtoabandonhisresearchwhileofthecuspofsuccessforlackoffunding(Friedman,2010),hemighthavebeenmoreappreciativeofMacleod’ssupport.Itischaracteristicofnaïveresearcherstovaluetheideamorethantheabilitytobringtheideato fruitionviaawell-fundedandequipped laboratory.Self-awarenesswasnot inBanting’smakeup.Heprobablydidnotappreciatethathisgreatidea,ductligation,waswrong,butitwas thecatalyst that securedhimaccess toMacleod’s lab, resourcesand theexpertiseofthreeco-workers,eachcontributingvariedbutvitalskills.Thiswaswhatledtothediscoveryof insulin. The initial progress of Banting and Best piqued Macleod’s interest and hecontinuedtofundandthenexpandsupportfortheendeavourbyenlistingthehelpoftheexperiencedbiochemistCollip.BantingfailedtoappreciateMacleod’sinput,interpretingtherobustconstructivecriticismthatisthebedrockofscientificintercourseasunjustcriticism.Banting lacked Macleod’s knowledge and the technical ability of Collip, and on threeseparate occasions came very close to fistfights with Best, Macleod and Collip, probablyfrom the realisation of his limitations as a researcher in comparisonwith his co-workers.Indeed ifanyof thegroupwere justified in feelingaggrievedfor lackof recognition itwasCollip, for it was his crucial purification process that produced an extract of internal

10


secretionofsufficientpuritytouseondiabeticpatients,afeatnopreviousresearchershadachieved. It isnowwidelyacceptedthatcomplexteamwork ledtothediscoveryof insulinwithall fourmembersof theTorontogroupprovidingvital input. It is aprofound tragedythatBantingfeltunabletoenjoyhissuccess,anymentionofinsulinawakeningmemoriesofthe unmerited apportioning of credit toMacleod,withwhomhis namewould be foreverlinked.IfonlyBantinghadbeenabletoconsolehimselfwiththerealisationthattherewereveryfewpeopleinthehistoryofmedicinewhohaverelievedthesufferingandimprovedthequalityoflifeforsomanymillionsofpeople(Bliss,1983;Bliss,1992;Bliss,1993;Bliss,2013;Cardosoetal.,2017;Fralick&Zinman,2021;Trehtewey,1988;Whitfordetal.,2012).

Standardisation The discovery of insulin in Toronto led to an immediate and enormous globaldemand. The purification of insulinwas optimised by Collip, and Bestwrote a fascinatingdescriptionoftheprocessinwhichheofferedgenerouspraisetotheresearcherswhocameclose topurifying insulin (Zuelzer, Scott,RennieandFraser, KnowltonandStarling,Murlinand Kramer, Kleiner, et al.) and provided detailed descriptions of the method used byhimself and Banting in 1921 and Collip’s method, commenced in December 1921. Thepainstaking optimisation of the purification is described in detail, clarifying Collip’s vitalcontribution and justifying his inclusion as a core member of the Toronto group (Best &Scott, 1922). Several other researchers had progressed to the stage of demonstratingpancreaticextractloweredglucose,butfailedtoprogresstothecriticalstageoftestingtheextract in human diabetic patients, as the impurities in their extract caused harmful sideeffects(Bliss,1983).ItwasCollip’sforensicattentiontodetail,borneoutbyhisexperienceof internal secretions andmaking tissue extracts that accounted for his successwhere allothershadfailed.ThekeytoCollip’ssuccesswastheuseofalcoholasasolventatavarietyofconcentrations,andethertoextract insulinfromlipoidmaterial (Best&Scott,1922). InTorontoConnaughtAntiToxinLaboratoriesproceededwithlarge-scaleproduction,whilstEliLilly was granted the exclusive right to produce and distribute insulin in the USA for oneyear. The Nobel Prize winning scientist Arthur Krogh, whose wife had recently beendiagnosed with diabetes, visited Toronto in 1922 at Macleod’s invitation, and securedexclusive permission to produce insulin in his native Denmark, establishing the NordiskInsulinLaboratory,whichhadaccesstoaplentifulsupplyofporkpancreasfromthefamousDanishbaconprocessingplants.Themanufactureofinsulinbydifferentcompaniesledtoacall for standardisation, the process by which insulin is produced at equal potencyirrespective of manufacturer, to ensure equivalent doses were given to patients. As thestructure,andhencemolecularweightofinsulin,wasunknownitcouldnotbequantifiedinterms of moles, but instead was quantified by its glucose lowering ability, known as thebiological system. In Toronto, in the spring of 1922, the effectiveness of insulin wasmeasured by its ability to reduce blood glucose, where one physiological unit of insulinequalledthenumberofcubiccentimetres(ccequivalentto1ml)ofsuspensionthatcausedthebloodglucoseofa2kgrabbit,fastedfor24hours,tofallto2.5mmoll-1in4hours(Eadie& Macleod, 1922). Thus the researchers quantified the dose of insulin as a volume ofsolution.Therewereclearlyproblemsassociatedwiththismeansofstandardisation,notablythebiologicalvariationamongrabbitswheresomemightbemoresensitiveto insulinthanothers,whichrequiredacomplexmethodofnormalisingtheresponse(Fieller,1940).Wheninsulinwas injected into dogs, and corrected forweight differences, the Torontoworkersfound that insulin was far more potent, with only one third of the volume required toproduce an equivalent effect to that in rabbits (Banting et al., 1922). In addition patients

11


requiredvaryingvolumesofinsulinindicativeofdifferencesinthepurifiedinsulin’spotency,duetovariationsinmanufacture.

This standardisation was subsequently modified by the Toronto workers, whichreducedthephysiologicalunittoonethirdof itsoriginalsizeasthatvaluewasconsideredtoolargeanumber,andwasrenamedtheClinicalUnit,whichwascalculatedas:

𝐶𝑈 𝑐𝑐!! =32𝑊𝑡 𝑅𝑎𝑏𝑏𝑖𝑡𝑁𝑜 𝑐𝑐

𝐼𝐵𝑆 − 𝐹𝐵𝑆𝐼𝐵𝑆 − 0.045

whereCUcc-1 istheclinicalunitperml,Nocc isthenumberofcubiccentimetresinjected,Wtrabbitistherabbitweightinkg,IBSistheinitialbloodsugarpercentagepriortoinjectionandFBSisthefinalbloodsugarpercentageafterinsulininjection(Fieller,1940).PlottingCUcc-1versusFBSreturnsalinearrelationshipbutplottingCUcc-1versusNoccinjectedreturnsanon-linearrelationshipandrevealshowsmallvariationsintheNocc injectedcanleadtolargevariationsintheCUcc-1.

Theneed for standardisationwasbasedon the following considerations: (1) giventhelimitedproductionitwasimportantnottowasteinsulinintheearlydaysofmanufacturebyadministeringimprecisedosages,(2)excessinsulindosagecouldleadtopotentiallyfatalsideeffectsassociatedwithhypoglycaemia,(3)introductionofastandardunitwouldensurethat non-specialist doctors and nurses could safely administer insulin to patients, and (4)irrespectiveofitssourceallinsulinwouldhavethesamepotency(Fields,2011).Sinceinsulinwasineffectivewhentakenorally(Rennie&Fraser,1907)itwaspreparedasawater-solublehydrochloride salt for subcutaneous injection. HenryDalewas consulted on behalf of theMRC in the UK and travelled to Toronto where he expressed reservations about thestandardisation of insulin relative to its ability to induce hypoglycaemia (Feldberg, 1970).DaleconsultedwiththeLeagueofNationsHealthCommitteewhosponsoredaConferencein Edinburgh in 1923 to resolve the issue. Dale proposed abandoning the standardisationrelativetobiologicalsystemsi.e.theabilityofinsulintoproducehypoglycaemia,andinsteadproposed a standardisation relative to weight of purified dried insulin, with a correlationbetween weight of the powder and the experimental activity of 1 insulin unit (IUinternationalunit)(Murnaghan&Talalay,1992).Driedinsulinfrom5sourceswascombinedandampoulessenttolabsforcomparison.Eachmgofsolidcontained8IU,or1IU=0.125mg(Lacey,1967).AdetailedhistoryoftheearlydaysofinsulinstandardisationinTorontoisavailable (Sinding, 2002). This wasmodified in 1935 to 1 IU of insulin = 1/22mg of newstandard,withthefourthinternationalstandardin1959defining24IUofinsulinpermgofstandardwhere the factor7.147wasused to convertbetween IUandmoles (Bangham&Mussett, 1959). In 1986 it was further revised to 26 IU per mg (Bristow et al., 1988).Howeverthe1986standardcontainedwaterandsaltsandrequiredcorrectionto6nmolper1IU,whichisequivalentto28.8IUpermg.Thusin2010theWHOInternationalStandardsdefined1 IUas0.0347mgequivalentto28.8 IUpermg.Howevermanyonlinecalculatorsandreferencesuse6.944toconvertbetweenIUandmg,basedontheroundingupoftheMWofhumaninsulinfrom5808to6000(Knoppetal.,2019).

Insulin-themolecule BantingandBest initiallyuseddogsas theirprinciple laboratorymodel, since theywere the standardanimalsused inpreviousdiabetes studiesandwere readilyavailable inMacleod’s laboratoryattheUniversityofToronto.ThepremiseofBantingandBest’sworkwastoisolatetheinternalsecretionfromadog,thentest itsefficacyonaseparatedog.Adistantgoalwastoisolateanextractofinternalsecretionfromdogofsufficientpuritytobeusedonhumandiabeticpatients,aprocedurethathadbeenattemptedbyZueleralmost15

12


years previously (Zuelzer, 1908). Nowhere in the published literature from the Torontogrouparetheimplicationsofcross-specieseffectivenessofpancreaticextractdiscussed,i.e.theabilityofanextractfromdogtoproduceaneffectinhumans.Wenowknowthisisdueto the evolutionary conservation of the insulinmolecule. In the Toronto studies humans,rabbit, dogs, cats, pigs, cow and ox all featured, the first three to test the effect of theextract and the last five as sources of the extract. Indeed anatomical knowledge of inter-species variations in the anatomy of the pancreas ledMacleod to seriously consider fish(Bliss,1983),where theacinarand isletcellsareanatomicallydistant,asasuitablesourcefor the internal secretion. Collip hypothesised that a substance equivalent to the internalsecretionof thepancreaswaspresent inallplantsandevenshowedtheglucose loweringability of an extract from lawn grass cuttings in rabbits (Collip, 1923). Collip realised that‘where glycogen occurs in nature a related compound to insulinwill be there too’. Thesestudies in Toronto revealed that insulin is widespread in the animal and plant kingdoms,whereitplaysafundamentalroleincarbohydratemetabolism.

Insulin ismanufactured intheβcells inthe isletsofLangerhans inthepancreas inhumans.Theinsulingeneisencodedbya14kilobaseDNAfragmentsequenceontheshortarm of chromosome 11 (Owerbach et al., 1980) at position 15.5 (Mutskov & Felsenfeld,2009).Insulinissynthesisedinitiallyaspartofpreproinsulinintheβcells,apolypeptidethatcontains a 24 amino acid (AA) residue signal peptide, which directs the molecule to therough endoplasmic reticulum (RER) where it is cleaved leaving proinsulin. In the RER themolecule folds to the correct conformational 3D shape and is then transported to thetransGlogi network. Enzymes cleave themolecule to separate insulin from the C peptide(Champe&Harvey,2008).InsulinismadeupofanAandBchainconnectedby2disulphidebondsandconsistsof51aminoacids (Fig6A),with10AA residues fully conservedduringvertebrate evolution. Human insulin has a molecular mass of 5808 Da and a molecularformulaofC257H383N65O77S6.Insulinisstoredwithingranulesawaitingreleasebysignalssuchas glucose and sympathetic nervous system β2 receptor activation (Champe & Harvey,2008).Insulinisreleasedviaelevatedsubcutaneousglucose,whichistakenupbytheβcellsultimatelyproducingATP,whichclosesATPsensitiveK+channelsleadingtodepolarisationofthecellmembrane,activationofvoltagegatedCa2+channelsandCa2+influx(Ashcroftetal.,1984).This influx causesdockingof the secretoryvesicle containing the insulinwith theβcellmembraneandreleaseofinsulinintothebloodstream.

Insulinandglucagonexistasantagonistichormones,eachopposingtheeffectoftheother,wheretheformerisanabolicinnatureandreleasedfrompancreaticβcells,andthelatter is catabolic innatureand released frompancreaticα cells. In response to increasedbloodglucoseconcentrations insulin issecretedfromβcells,whichalsoactsasasignaltoinhibit α cell glucagon secretion (Cooperberg & Cryer, 2010). The insulin causes cellularuptakeandstorageofglucosefromthebloodintoliverandadiposecells,therebyreducingthebloodglucoseconcentrationanddiminishinginsulin’srelease.Inresponsetolowbloodglucosetheβcellsareinhibitedfromreleasinginsulinandthissignalstheαcellstoreleaseglucagon, which causes liver glycogenolysis and gluconeogenesis, both of which increasesystemic glucose concentrations (Cryer, 2012; Figure 4A). In non-diabetic individuals thebloodglucoseneverfallsbelowabout3.9mmoll-1(Cryer,2012)andsincethethresholdforinsulin secretion is 3.3mmol l-1 (Henquin et al., 2006) there is a constant tonic releaseofinsulin. However in type 1 diabetes the β cell response is diminished or absent thusattenuatingtheαcellactivation,whichresults ina limitedglucagonreleaseinresponsetohypoglycaemia(Cryer,2012),whichextendsthedepthanddurationofhypoglycaemia.

13


Figure 4 - The reciprocal relationship between insulin and glucagon release. A. The circuitwhereby glucagon and insulin release responds to blood glucose concentration. Elevatedbloodglucoseactivatesthepancreaticβcellstoreleaseinsulinintotheportalvein(red).Thisactivationoftheβcellsalsoactstoinhibittheαcellreleaseofglucagon.Theinsulincausescellularstorageofbloodborneglucose,therebyloweringglucoseconcentrations(blue).Thelowbloodglucosedirectly stimulates theα cells, butalso inhibitsactivationof theβ cells,which releases the tonic inhibition on theα cells, releasing glucagon into the portal veincausinghepaticglycogenolysisandgluconeogenesis,elevatingbloodglucoseconcentrations.B.Asthebloodglucoseconcentrationincreasestheinsulinsecretionincreases,thesteepestpartoftheslopecoincidingwiththeupperlimitofthenormoglycaemicrange(7.8mmoll-1).No insulin is secreted below about 3.3mmol l-1 glucose. Glucagon is releasedwhen bloodglucose concentrations are low, with little insulin released at the lower limit of thenormoglycaemicrange(3.9mmoll-1)(Guyton&Hall,2006).

14


The insulin molecule may have originated over two billion years ago, with abiologically similar molecule present in E. coli (de Souza & López, 2004). The release ofinsulin from vesicles via Ca2+ sensitive protein (Carafoli & Krebs, 2016) also points to anancientheritage.Insulinisstoredintheβcellsashexamersstabilisedbyzincmolecules,butonce released into the blood stream the zinc is diluted and the bonds holding the insulinmolecules together weaken, disassembling into insulinmonomers, which can diffuse intothebloodvessels(Hirschetal.,2020).Sinceinsulinisreleasedintotheportalveintheinsulinmolecules are cleared by the liver (about 80%) and kidney (20%) thereby limiting insulinavailabilityintheperiphery,whereitactsonmuscleandadiposecells(Hirschetal.,2020).

Insulinanalogues Theinsulinusedtotreathumanpatientsfrom1922totheearly1980swasanimalinorigin,predominantlypigandcow,whichhave1and3AAresiduesdifferentfromhumans,respectively(Pickup,1986)(Figure5A).Theproblemsassociatedwiththeseanimal insulinswere supply and allergic reaction. Introduction of animal insulin into humans led to theformation of anti-insulin antibodies, causing insulin resistance in patients (Schernthaner,1993), mandating the use of human insulin to treat human diabetic patients. The AAsequence of human insulinwas published by Sanger in the early 1950s forwhich hewasawarded the Nobel Prize for Chemistry in 1958 (Sanger & Thompson, 1953; Sanger &Thompson,1953;Sanger&Tuppy,1951;Sanger&Tuppy,1951).The3Dcrystalstructureofinsulin was revealed by Dorothy Hodgkin (Nobel Prize for Chemistry 1964) using X-raycrystallography in1969(Adamsetal.,1969),which facilitated itsmassproductionandthepotentialtoalteritsstructure.Insulinwassubsequentlypurifiedtoexcludepro-insulinandotherpeptides inthe1970s.Thediscoveryofthe insulingene(Owerbachetal.,1980)andrecombinant DNA technology (Baeshen et al., 2014) enabled large-scale production ofhuman insulin,where the insulingenewas incorporated intoE. colibyEli LillyoryeastbyNovo Nordisk and produced en masse. The first of these mass-produced recombinanthuman insulins was Humulin manufactured by Eli Lilly in 1982, followed by Novolinmanufactured by Novo Nordisk in 1991 and Insuman manufactured by Hoecht in 1997(Hirsch et al., 2020). Thismajor advance inmanufacturing allowed patients to be treatedwithhumaninsulinforthefirsttime.However,itisimportanttoappreciatethedifferencesinthekineticsofnaturalinsulinversussubcutaneousinjectionofhumaninsulinintheblood.The natural release of insulin can be viewed simplistically as a negative feedbackmechanism,whereelevatedbloodglucoseconcentrationspromotes insulin release,whichfacilitates cellular storage of glucose thereby reducing systemic blood glucoseconcentrations, which restricts insulin release. A decrease in blood glucose concentrationpromotesglucagonrelease,whichcausesbloodglucoseconcentrationstorise(Fig4).Thusbloodglucoseconcentrationsdeterminethebalancebetweeninsulinandglucagonrelease(Cryer, 2012). At normoglycaemic concentrations of glucose there is a continuous tonicreleaseofinsulin,butingestionoffoodstimulatesabolusofinsulinrelease,whichcanleadtoelevatedbloodinsulinlevelsforseveralhours(Fig5B).Thekineticsofinsulininthebloodare defined by two separate but related properties, pharmacokinetics, which is the timecourseofthecirculatinginsulin,andpharmacodynamics,whichreferstothebloodglucoseconcentration. A comparison of the pharmacokinetic profile of subcutaneous injectionversusendogenousinsulinshowsaslowerrise,laterpeakandprolongedpresence(Fig5C),whichmayresultamismatchbetweenprevailingglucoseconcentrationsandinsulindosageleading to hypoglycaemia (see later; Hirsch et al., 2020). In order to combat thesemismatches genetically engineered insulin analogues were developed, based on human

15


insulin, but incorporating structural alterations to increase the rate of absorption, calledrapidactinginsulin.Itshouldbenotedthatallhumaninsulinanaloguesweredevelopedtoaltertheirabsorptionintothebloodstreamonceinjectedsubcutaneouslyanddonotaffectinsulin’s binding to its target receptors or the resulting physiological effect (Hirsch et al.,2020).

Figure5 -Kineticsofblood insulin.A. (1) Insulin issynthesisedasapolypeptide(2)priortocleavageofthe24AAsignalpeptide(orange)then(3)excisionoftheCpeptide(black),whichis released with insulin and can be used as a measure of insulin release. B. Insulin levelsthroughoutthedayderivefromtwoseparatecomponents,atonicbasalreleasethroughoutthe day evident as a tonic release at night time, coupled with bolus release of insulin inresponsetomealingestion.Theinsulininjectionregime(triangles)viainsulinpensshowstwoinjectionsoflongactingaspartat0630and2230and3injectionsofdetemirpriortomealsat1 IUper10g carbohydrate.C.Profileof endogenousblood insulin levels in response tomealingestion(blue)comparedtosubcutaneousingestionofinsulinpriortomealingestion(red).

16


ThethreemaininsulinanaloguesdevelopedforrapidabsorptionwereInsulinlispromarketed as Humalog by Eli Lilly in 1996, Insulin aspart marketed as NovoLog andNovoRapidbyNovoNordisk in2000,and InsulinglulisinemarketedasApidrabySanofi in2004 (Hirsch et al., 2020). The AA substitutions weaken the bonds that hold the insulinhexamerstogether leadingtorapiddissociationintomonomers inthesubcutaneousspaceandabsorption,wheretherateof insulin increaseanddecreaseinthesystemiccirculationpost injection more accurately match that of endogenous insulin. These developments,along with careful coordination betweenmeal ingestion and insulin dosing, resulted in apharmacokineticprofilethatmoreaccuratelymatchednaturalinsulinreleaseinresponsetomeal ingestion. These rapid acting insulins arenot suited tomatching the sustainedbasallevel of insulin release at the lower endof the normoglycaemic range, thus slower actinginsulin analogues were developed. Conjugating protamine with insulin results incrystallisation of the hexamers, which must dissolve before the insulin can be absorbed.Insulin glargine, marketed a Lantus, was introduced by Sanofi in 2000, Insulin detemir,marketed as Levemir, was introduced by Novo Nordisk in 2005, and Insulin degludec,marketedasTresiba,was introducedbyNovoNordik in2015.Theywereallengineeredforextendedabsorption,thealterationsinstructureprolongingthelatencyforthehexamerstodissociate into monomers. Degludec acts for more than 24 hours requiring only oneinjection per day and has the considerable advantage of reducing the risk of nocturnalhypoglycaemia.However thesedevelopmentsstill requiremultiple injectionsof fastactinginsulinperday(Hirschetal.,2020)(Fig5B).

Insulinisexpressedasmoll-1orUnitsml-1(Frieretal.,2014),whereafastinglevelofinulin in non-diabetic adult humans ranges from about 70 to 150 pmol l-1 (Owens et al.,2001),equivalent toabout10 to20mUnitsml-1.A typical type1diabeticpatient requires0.5to1IUkg-1day-1,butintheearlystagesofthedisease,whenthereissomeremainingβcell function,may requireonly0.2 to0.6 IUkg-1day-1 (Janezetal., 2020). Type1diabeticpatients are typically suppliedwith insulin formulations of 100 IUml-1 (Lane et al., 2017),which limits thevolumeof solution injected toa coupleofmlsatmostperday.Howeversome patients with type 2 diabetes may require more than 200 IU day-1 due to insulininsensitivityandincreasedbodyweight(Fig8A&B),andmaybesuppliedwithformulationsof up to 500 IU ml-1 (de la Pena et al., 2011). Insulin mixtures have been developed toincludebasalinsulinandrapidlyactinginsulintoreducethenumberofdailyinjectionsandtomore accuratelymatch the kinetic profile of endogenous insulin release (Hirsch et al.,2012). However these mixes of insulin cannot be altered and suit patients with regular,routineandequivalentmeals.

17


Figure6-Structureofinsulin.A.ComparisonofhumaninsulinAAresidueswithavarietyofvertebrates.B.(1)HumaninsulincomprisesanAchain(21AAresidues)andBchain(30AAresidues),interlinkedwithtwodisulphidebondswithanintradisulphidebondontheAchain.(2). Fastactinghuman insulinanalogues showing thealterations inAA residues for Lispro,Aspart andGlulsine. (3). Slow acting analogues showing the AA substitutions for GlargineandthereplacementoftheAAresidues30and31ontheBchainwithlargefattyacidgroupsforDetemirandDegludec.

18


Glucosemonitoring Disruption to the insulin signallingpathway that regulatesbloodglucoseunderliesdiabetes, leadingtoitscardinalsymptom,hyperglycaemia.Measurementofbloodglucose,priortotheintroductionofmeasuringHbA1c,wasusedasaclinicaltesttoinitiallydiagnosediabetes, and as ameans of assessing glycaemic control in the diabetic patient. Thus thetherapy for diabetes comprises two separate, but related, components; to administer theappropriatedoseofinsulinaccordingtothebloodglucoseconcentration,andisdependentupon a rapid and accurate means of measuring blood glucose. In sufferers of type 1diabetes, if the degree of hyperglycaemia exceeds 10 mmol l-1, the renal threshold forglucoseabsorption(Triplitt,2012),glucoseappearsintheurine,roughlyinproportiontothebloodglucoseconcentration(Griffinetal.,1979).Priortotheintroductionofaccuratebloodglucose measurements (around 1920) detecting the presence of glucosurea was thestandardmethodofdiagnosingdiabetes.Theearlydiabetesresearchersusedthereductionor absence of glucosurea after pancreatic extract injection as evidence of its glucoseloweringability.Howevertheabsenceofglucosureacouldmeanthatthebloodglucosewasmildly hyperglycaemic, normoglycaemia or hypoglycaemia (Clarke& Foster, 2012), clearlyindicating the requirement for accurate measures of blood glucose. It was with thedevelopmentoftheBenedictsolution,knownasthecopperreductionreaction,intheearly20th century, that an accurate means of measuring glucose in the urine became widelyavailable (Benedict, 1908). This complex test requiredboilingof the reagentandused thepresenceofalkalinesodiumcarbonatetoconvertglucosetoastrongreducingagent,whichwasoxidisedbyCu2+toproduceCu+,visibleasaninsolubleredcopperoxide,theintensityofthe colour proportional to the glucose concentration. The colour of the sample wascompared to a standard colour chart to estimate the glucose concentration. Variations ofthis reactionwere introducedover thenext fewdecades,but theyall utilised the ‘copperreaction’,whichrequired laboratoryanalysisandtheiraccuracywasquestioned(Shaffer&Hartmann, 1921; Somogyi & Kramer, 1928; Van Slyke & Hawkins, 1928). A significantadvance occurred in 1941 with the first test that could be carried out at home by thepatient. The introduction of the Clinitest tablet, which, when added to the urine sample,inducedaboilingreaction,promotingtheformationofcopperoxide(Free&Free,1964).Inthe 1950s the Clinistix was introduced, a new technology in which the enzyme glucoseoxidasewasembeddedonastickthatwasinsertedintotheurinesample(Kohn,1957).Theglucose oxidase catalysed the oxidation of glucose to form hydrogen peroxidase, whichreactedwithpotassiumiodidecatalysedbyhydrogenperoxidasetoproduceiodine,whosebrownishcolour indicatedthepresenceofglucose,thestrengthofthecolourproportionaltotheglucoseconcentration.Theuseofurineglucoseconcentrationsasaproxyforbloodglucoseconcentrationswasfraughtwithlimitations(Milleretal.,1983),whichincluded,butwerenotlimitedtothefollowing.Urinevolumeandconcentrationcouldaffecttheaccuracyofthestrips,thepresenceofglucoseinurinewasonlydetectedwhenglucoseexceededthethreshold for renal clearance, and comparison between the colour on the strip and thestandardcolourchartintroducedtheelementofsubjectivejudgement.

Theneedforbloodglucosetestingbecameoverwhelmingwiththerealisationthatmeasuresofurineglucoseasaproxy forbloodglucosewerehugely inaccurate, leadingtopoorglycaemiccontrolandhighincidenceofmorbidityandmortality.In1965theDextrostixwasintroduced,whichwasbasedonthesameprincipleastheClinistix,butwascapableofmeasuringbloodglucose(Free&Free,1964).Thestripusedtheglucoseoxidase/peroxidasereaction,butcamewithasemi-permeableouter layer,whichoccludedtheredbloodcellsfrom the strip, while allowing access of plasma. At about the same time Boehringer

19


introducedtheChemstripbG,whichusedthesameoxidasereactionbutofferedtheuseraclearer view of the strip. The urine and glucose tests described so far are single pointestimates of glucose concentration at the time of sampling. As such they are useful fordetermining if the patient is hypo- or hyperglycaemic, but give no indication of the long-termmaintenanceofbloodglucoseconcentrations.Additional considerations includehowaccuratea representationofbloodglucose interstitialglucose is, limitations inaccuracyofmeasuresofglucoseateitherendoftherange,andtheuseoffingerstickinthecalibrationprocess (Frieret al., 2014). In the1970s the introductionof themeasurementof glycatedhaemoglobin (HbA1c) (Rahbar et al., 1969) was an enormous step forward in estimatinglong-term glycaemic control in patients (Koenig et al., 1976; Koenig et al., 1976). Theaveragelifespanofredbloodcellsisabout120daysandthereactionbetweenhaemoglobinandglucoseto formglycatedhaemoglobinhasbeenusedasanestimateof thedegreetowhich the red blood cells have been exposed to glucose during their lifetime. Valuesexceedingthethresholdof6.5%or48mmolmol-1signifydiabetes(Garberetal.,2016).

The inaccuracies of comparing the strip colourwith a standard colour chartwerebypassedin1970bytheintroductionoftheDextrostixreflectancemeterbyAmes(Hedneret al., 1974). The device was used in doctors’ offices, and required the insertion of aDextrostixstripdippedinbloodsampleintothemeter.Aspectrophotometermeasuredthereflectance from the strip and thebloodglucose concentrationwas indicatedbyaneedlesuperimposedononeof three scales,which requiredno interpretationof thepart of theuser (Schersten et al., 1974). Inaccuracies in the Dextrosticmeter at the extremes of theblood glucose range routinely encountered in diabetic patients prompted development in1973 of the Eyetone meter, which displayed the glucose concentration on one easilyreadablescale(Scherstenetal.,1974).In1980aDextrometer,whichusedtheDextrostixasthe glucose sensor, was developed for home use complete with a digital readout foraccuracy. The introductionof small battery operated glucosemeterswithdigital readoutscommenced with the GlucoMeter M in 1980, the AccuCheck in the mid 1980s and theOneTouchUltrain2000(Clarke&Foster,2012).Thisallowedthepatienttotesttheirbloodadlibitumwithoutavisittothedoctor.Theuseofglucosemetersathomeintroducedtheconcept of self-monitoring of blood glucose (SMBG), themost important development indiabetestherapysincetheintroductionofinsulinsixdecadespreviously.

In 1993 an enormously influential study, whose conclusions remain the basis forcurrent diabetes therapy, convincingly demonstrated the link between poor glycaemiccontrol and increased probability of developing microvascular complications. The studydemonstratedthatintensiveinsulintherapy,whichattemptstomaintainthebloodglucoseconcentrationsasclosetothenormalrangeaspossibleforaslongaspossible,significantlyreducedtheprincipalcomplicationsassociatedwithdiabetes:retinopathy,nephropathyandneuropathy, offset however, by an increased incidence of hypoglycaemia. Increases inHbA1cvalueswerecorrelatedwithincreasingretinopathy,whereaslowervalues,belowthediabeticthresholdof6.5%,wereassociatedwithincreasedincidenceofhypoglycaemia.Theconclusion from this study was that although no specific target value of HbA1c could berecommended, maintaining the glycaemic status as close to the normal range as safelypossible delayed progression of microvascular complications, but also increased theincidenceofhypoglycaemicepisodes (Nathanetal.,1993).Thetargetbloodglucoserangewas between 70 - 180 mg dl-1 (3.9 - 10 mmol l-1) (Nathan et al., 1993) with patients’adherence to this range quantified as the Time in Range (TiR), and expressed as apercentage(Fig7A).

20


Figure7 -Glycaemiccontrolandriskofmicrovascularcomplications.A.Theglucoseprofilethroughouta24hourperiodwherethegreyregionindicatesthedesirablerangeofglucoseconcentrations from 3.9 to 10mmol l-1, showing an episode of hypoglycaemia during thenight(red)andhyperglycaemiaduringtheday(blue).B.TherelationshipbetweenHbA1c,anindexofglycaemic control,where6.5% is thediabetic threshold,and riskofmicrovascularcomplicationsandhypoglycaemia.IncreasingvaluesofHbA1c(blue)increasetheprobabilityof microvascular complications, whereas decreases in HbA1c (red) lead to increasedprobabilityofhypoglycaemiaepisodes.

21


In order to achieve a close relationship between prevailing glucose concentrationand target glucose concentrations continual glucosemonitoring (CGM)was introduced intheearly2000s.Thefirstdevicesmeasuredglucosebypassingverylowamplitudecurrentsacrosstheskintoattractglucosemolecules.Thefirstsuchcommerciallyavailabledevicewasthe Glucosewatch, introduced in 2002 and featured in the contemporaneous film PanicRoom,whichdisplayeddigitallytheprevailingbloodglucoseconcentrationonawatch-typedevice(Chan&Hurel,2002).Continualadvancementshavebeenmadeinthedevelopmentof CGMwith the prescription-only DextComms STS, which had a sensor, transmitter andreceiver that output the data to a computer. In 2008 the FreeStyle Navigator CGM wasintroduced, but it required a complex and lengthy calibration. In 2015 the DexCom G4Platinum was a major advance as it was approved for used by under 18s. In 2016 theMedtronicMiniMed ‘iPro2’G5mobilewas introduced and in 2017 the FreeStyle Libre ProFlashwas introducedwhich requirednocalibration (Clarke&Foster,2012).Thesecurrentsystemsuseasensorcoatedwithglucoseoxidasethatpenetratestheskinandsendsregularupdatesofthecurrentglucoseconcentrationtoanappontheuser’smobilephone.Thesedevicesallowstorageanddisplayingraphicalformoflong-termdata,andquantifytheTiRfor defined periods of time. Longitudinal studies have shown CGM ismore effective thanSMBGasmeasuredbyanincreasedpercentageofTiRandlowerincidenceofhypoglycaemia(Danneetal.,2017).

Theunitsbywhichglucoseisexpressedvary.Earlyexperimentersexpressedglucoseconcentration as a%,where1%equals 1 g of glucose in 100ml of urine/blood, althoughclinicians favour mg dl-1 and lab scientists mmol l-1 (mM). The inter-conversion betweenthese units is straightforward (Sawbridge et al., 2014). A baseline fasting glucoseconcentrationinanon-diabeticadultof0.09%isequivalentto90mgdl-1or5mmoll-1.Thebloodglucoseconcentrationrangeofanon-diabeticindividualisbetween3.9and7.8mmoll-1(70-140mgdl-1;Frieretal.,2014).Itisenlighteningtoputintoperspectivetheglucoseconcentrationofbloodrelativetotheglucoseconcentrationincommondrinksandfood.Acan of coke (355ml) contains about 21.3 g of glucose (350mmol l-1), and an equivalentamount of fructose (Varsamis et al., 2017). Instantaneous absorption of this amount ofglucoseintothebloodofan80kgman(bloodvolumeof5.6litres)wouldincreasethebloodglucoseconcentrationby21mM,farinexcessofthehyperglycaemicthresholdof7.8mmoll-1.Adiabeticpatientundergoingahypoglycaemicevent(bloodglucoseof3mmoll-1)wouldonlyneedtodrink50mlofcoke(although150ml isrecommended(Frieretal.,2014)),oreat4.2gofpasta(72gofcarbohydrateper100g)toincreasebloodglucoseconcentrationby3mmoll-1.

LifeExpectancy A longitudinalexaminationreveals theeffectsofvariousdevelopments indiabetestherapyon lifeexpectancy.Abouthalfof thosediagnosedwith type1diabetesareunderthe age of 14 (Rawshani et al., 2014), hence the description of childhood onset diabetes.Priortoinsulintherapythelifespanofthesechildrenwasbetween1and2years,althougholderpatients lived longer,between4and9years (Brostoffetal., 2007).By1945 the lifeexpectancy for type 1 diabetics had dramatically increased by 45 years for childrendiagnosedundertheageof10,by30yearsforthosediagnosedatage30and16yearsforthosediagnosedat age50 (Brostoff et al., 2007). Itmustbeborne inmind that this timeperiodpre-dates the advent of SMBG, thus thepatients in these studiesweremost likelypoorly controlled. The trend to increased longevity continued with those born between1956and1980living15yearslongerthanthosebornbetween1950and1964(68.8versus53.4 years) (Miller et al., 2012). The improvements in blood glucosemonitoring underlie

22


theseincreasesinlifeexpectancyandtheintroductionofSMBGandCGM(seelater)shouldextendtheseevenfurther.

A recentSwedish study revealed that theageofdiagnosiswas strongly correlatedwithlifeexpectancy,thosediagnosedunder10yearsofagelivedapproximately6yearslessthanthosediagnosedbetween26and30yearsofage(Rawshanietal.,2018).Thelikelihoodof death and the increased risk of complications is directly related to the degree ofglycaemic control, the higher the per cent of time spent in TiR, the lower the risk ofcomplicationssuchasretinopathy,neuropathyandnephropathy.Thegreatestriskinthesepatientswasofcardiovasculardiseaseandendocrinedisorders,whichaccountedfor70%ofdeaths in those diagnosed between 0 and 10 years old, which fell to 61% for thosediagnosedbetween26and30yearsold.Howeveryoungeronsetpatientshadup toa30-foldincreasedriskofcardiovasculardiseaseoracutemyocardialinfarction(Rawshanietal.,2018).Ageofonset isconsideredacritical factor indetermining lifeexpectancy, therapiddeathofβcellsinyoungchildrennotonlyexposingthemtohighglycaemicloadattheonsetofthedisease,but leadingtoextendedexposuretohighglycaemic loads.Thevasculature,and in particular the coronary arteries, are vulnerable to increasing glycaemic load, againunderlining the importance of maintaining strict glycaemic control (Aronson & Rayfield,2002).

Prevalenceofdiabetes Theworldwidenumberofindividualsdiagnosedwithdiabetesquadrupledfrom108millionin1980to422millionin2014,aglobalprevalenceincreasefrom4.9%to8.5%(NCD-RisC, 2016), with a projection of 592 million by 2035 (Guariguata et al., 2014) and 642millionby2040. IntheUKthenumberofdiagnoseddiabeticshasrisenfrom1.4million in1996to4.7million in2019,aprevalenceof7%(Whicheretal.,2020),and isestimatedtoreach5.5millionin2030.About90%ofthesediabeticshavetype2diabetesand10%havetype1diabetes,althoughtype1diabetesincidenceisincreasinginEurope,possiblyduetolifestyle, increasingweightandheight,C-sectiondeliveriesandreducedfrequencyofearlyinfections (Pattersonetal., 2009).One thirdofpeopleat the timeof theirdiagnosishaveirreversiblemicrovascularcomplications,but fewer thathalfof thesepatients receive therecommended eight annual health checks putting themat increased risk of complications(DiabetesUK,2017).IntheUSAtherearecurrently34.2millionpeoplewithdiabetes,10.5%ofthepopulation,with1.7millionhavingtype1diabetes.Howevertheprevalenceincreaseswithage;26.8%ofpeopleover65arediagnosedasdiabetic(CDC,2020)(Fig8).

InsulinPumps Theidealdeliveryofinsulintothediabeticpatientshouldmimicthenaturalreleaseof insulin from the pancreas as closely as possible. In non-diabetic individuals there is acontinualtoniclowlevelreleaseofinsulinthroughoutthedayandnight,referredtoabasalrelease. Insulin is released as a bolus in response to meals with blood concentrationsreachingtentimeshigherthanbasallevels(CDC,2020).Intheearlydaysoftherapyinsulinwas administered subcutaneously via injection by syringe, whereas endogenous insulin isrealisedintothehepaticveinandisclearedbytheliver,withresidualinsulintravellingtotheperiphery in the systemic circulation where it acts on skeletal muscle and adipose cells.There are numerous problems associated with subcutaneous injections of insulin, whichinclude variations in absorption dependent upon injection site, patient compliancewith arepetitivepainfulprocess,alargebolusofinsulinpresentinperipherycanleadtoexcessive

23


uptake of glucose into adipose cells leading to weight gain, inaccuracies in dosage,sufficiently concentrated insulin solutions required to limit volume of injections, andsensitivity/irritation at injection site (Kesavadev et al., 2020). However subcutaneousinjectionwas theonlymeansofadministering insulin in thesixtyyearsafter itsdiscovery.This,combinedwiththelimitedabilitytoaccuratelymeasurebloodglucoseconcentrationswithany regularity, led topoorglycaemiccontrol resulting inhigh levelsofmorbidityandmortality.ThelongevityofTheodorRyder(Fig3),oneofthefirstpatientstouseinsulin,wastheexceptionratherthantherule(Jones,1983).Inthe1980stheinsulinpenwasdevelopedwhichoperated ina similarmanner to thesyringe,butcameequippedwithanadjustablenozzleforaccuratelyvaryingdosages.Theinsulinpenoffersmoreflexibilitythanthesyringe,ismorediscreet,moreaccurate,morecosteffectiveinthelongtermandleadstoimprovedtreatment compliance. In controlled studies patients treated with insulin pens achievedgreaterglycaemiccontrolandfewercomplicationsthanpatientstreatedwithconventionalsyringes(Kesavadevetal.,2020).

Inordertoaddresstheissuesrelatedtoinsulinadministrationwithsyringesorpenstheinsulinpumpwasdevelopedinthe1960s,butonlybecameaviableoptionforpatientsat the turn of the century (Alsaleh et al., 2010). The design of the insulin pump, orcontinuoussubcutaneous insulin infusion (CSII), incorporatesa reservoirof insulinsolutionconnected to a subcutaneous catheter via tubing,with the site of subcutaneous injectionchangedevery3days(McAdams&Rizvi,2016).Thepumpisprogrammableandcandeliverinsulinatvariablerates(0.01to50IUmin-1).Theinsulinpumpissuitablemainlyfortype1diabeticsbutabout10%ofpumpusersaretype2diabeticsandonlyrapidinsulinanaloguesareused(Janezetal.,2020).Theinsulinpumpcontinuouslydeliversinsulin(basaldelivery),istailoredtothepatient’sdailyglucosecycleandcanalsodeliverabolusinjectionofinsulinwith meals, although the user has to calculate the appropriate dose based on thecarbohydratecontentofthemealandmanuallyinputthisinformation(usually1IUper10gcarbohydrate). The insulin pump offers the benefit of no injections and the insulin dosemore closely mimics the physiological response, meaning TiR is greater, the variations inabsorption are reduced, and glycaemic control is improved with a decreased risk ofcomplications(Group,2017).

Theartificialpancreas The continual improvements in CGM and the advent of the insulin pump led tothese two technologies being combined as the artificial pancreas or closed loop insulindelivery,where readings from the glucose sensor are transmittedwirelessly to a receiverthat contains a computer algorithm for calculation of the insulin dosage delivered to thepatientvia the insulinpump (Ellerietal.,2011).Thereare types twoofalgorithmused tooptimise insulin delivery relative to prevailing glucose levels (Bequette, 2013). Theproportionalintegralderivative(PID)algorithmcanbeconsideredreactiveasitrespondstoobservedglucoseconcentrations,whereasthemodelpredictivecontrol (MPC)algorithmisproactiveatforecastingglucoselevelsinanticipationoftheglucoseconcentrationsandtheeffectsofadministered insulin(Pinskeretal.,2016).MPCalgorithmscalculatethedeliveryofinsulinviathepumpbyminimisingthedifferencebetweenthedesiredglucoselevelandthemeasuredglucoseleveloverafuturetimehorizon.ThePIDalgorithmalterstheinsulindelivery by measuring the glucose level deviation from three perspectives (1) differencebetween glucose level and target glucose level (proportional component), (2) area undercurve between target and measured glucose level (integral component), and (3) rate ofchange of measured glucose (derivative component) (Steil, 2017; Steil et al., 2011). Animportant early development in closed loop systems was a suspend function where

24


decreasesinsensorreadingsofglucosetemporarilystoppedinfusionofinsulin.Theprimarygoal of the closed loop system is to achieve a high level of glycaemic control, whilstminimising the risk of hypoglycaemia. Several recent trials have demonstrated thesuperiorityoftheclosed loopsystemovertheSMBGwhencomparingTiRandepisodesofhyper- and hypoglycaemia (Benhamou et al., 2019; Brown et al., 2019; Weisman et al.,2017).Asignificant factor in the improvementsofTiRwith theclosedsystem is related tothe variations in insulin delivery that occur between day and night. The twomain factorsthat affectbloodglucose concentration, andhence insulin release, are intakeof foodandexercise, which are absent at night. As such there is less variability in blood glucoseconcentrationandhencelessvariationininsulinreleaseandglycaemiccontrolismoreeasilyachieved at night. During the day the variations in composition of meals, time ofconsumptionandexercisecanleadtobloodglucoseconcentrationsstrayingfromthetargetlevels due to the rapid increase in blood glucose concentrations resulting from foodingestion,whichrequiressignificantreleaseof insulin.Postprandial timecoversabout65-70% of the daytime and consumption of meals require user input to administer theappropriate insulindosage. Ina randomisedstudy themajorityof the improvement inTiRfortheclosed loopsystemcamefromovernightandevening improvementsandhighlightstheextremelyimportantpointthatthemajorityofthetimenotinrangeoccursasaresultofmealsandphysicalactivity(Kovatchevetal.,2020). InfuturetheuseofaclosedloopwithfullyautomaticinsulindeliveryincombinationwithultrafastactinginsulinwillimproveTiR.These closed loop systems arenot entirely independent as they require theuser to inputwhen meals occur to account for the bolus delivery of insulin (Boughton et al., 2020).Howevertrueindependentclosedloopsystemshavebeendevelopedinwhichthealgorithmfor delivering the insulin is embedded within the pump as opposed to be present on asmartphonedevice(Wangetal.,2021).Abionicpancreas,whichwilldeliverbothinsulinandglucagonwas invented in 2015 and approved by FDA in 2019, and is in the developmentstage(Kesavadevetal.,2020).

Cost Diabetesisaverycostlydiseasetotreat.Thereareseveralreasonsforthis,includingpatientswithtype1diabetescansufferwiththediseaseforsixdecadesormorebasedonachildhooddiagnosis,andtype2diabeticpatientscansufferforfourdecadesormore,givenamiddleagediagnosis.There isadirectcorrelationbetweendurationsincediagnosisandprobabilityofdeveloping complications (Rawshani et al., 2018),which include retinopathyleadingtoblindness,nephropathyleadingtokidneyfailureandneuropathywhichcanleadtoamputations,allofwhicharetimeconsumingandcostlytotreat.IntheUKitisestimatedthatdiabetescoststheNHS£10billionperyear,10%ofitsentirebudget,with80%ofthesecostsgoingtotreatcomplications(Diabetes.UK,2017).IntheUnitedStatesasimilarpictureemerges,wherediabetesconsumed$237billionindirectmedicalcostsin2017,accountingfor about25%ofhealth care spending (AmericanDiabetesAssociation, 2018;CDC, 2020).Since about 10% of adult USA citizens have a diabetes diagnosis it is clearly adisproportionally costly condition to treat. The increasing prevalence of diabetes can onlyleadtoincreasesinthesecosts.

Diabetescanalsobeexpensiveforthepatientdependinguponwherethey live. InEuropethecostofinsulintothepatientsislimitedbythenationalhealthcaresystemsandintheUKpatientsreceiveinsulinandrelatedsuppliessuchasCGMequipmentfreeiftheyhavediabetesmellitus andqualify for amedical exemption certificate. In Scotland,WalesandNorthern Ireland prescriptions for insulin are free. In theUnited States the system isradicallydifferentandcanbeahugefinancialburdentothepatient,whichcancostupto

25


$900 per month, and it is estimated that 1 in 4 patients in the USA ration their insulin(Herkert et al., 2019), leading to complications and even death (Rajkumar, 2020). Eachdiabeticpatientonaverageincurscostsof$9,600peryear,about2.3timesmorethannon-diabetic individuals.Thescandalofpatientsunable toafford life saving insulin isa regularmediastoryintheUSA,butuntanglingthedetailsofwhythisoccursiscomplicated.Thereare only three manufacturers of insulin supplied to the USA, Novo Nordisk, Eli Lilly andSantori.Thereare1.7milliontype1diabetessufferersintheUSA,andabout7.5milliontype2diabeticsintheUSAuseinsulin.Althoughonly7%oftheworldsdiabeticsliveintheUSA50%ofworldwideinsulincostoccursintheUSA,whereasChinaaccountsfor25%ofworlds’diabeticsbutonly4%of insulinsales(Bliss,1983).Thepriceof insulinproducts intheUSAtripledbetween2002 and2013 and thepricepaid for insulin doubledbetween2012 and2016.Oneexplanationfortheincreasedcostsisthatinsulinanaloguesareseventimesmoreexpensive than human insulin, and the vast majority of the insulin currently in use areanalogues.Although the cost tomanufacture insulin has remained stationary for thepasttwodecades,thepriceofavialofHumalogincreasedfrom$35to$234between2001and2015,despiteonlycosting$4 toproduce (Hirsch,2016).Globally insulinwasa$7.3billionindustry in2005,$21billion in2013andwill reach$28billionby2025 (Tsai,2016). It isabitter ironythattheonlythingtheTorontogroupagreeduponbymid1922wasthattheyshouldnotbenefit financially fromtheirdiscoveryandsoldthepatenttotheUniversityofTorontoforanominalsum,toensurepatientswouldhaveaccesstolowcostinsulin(Bliss,1983).

Thecomplexpathwaybetweeninsulinmanufacturerandthepatientisopaquewithlack of clarity of the costs involved in each stage. A large amount of the cash goes topharmacybenefitmanagers(PBM),whoseroleistoagreecontractsbetweenmanufacturersandpharmacies,forwhichtheyreceivelargerebatesfromthemanufacturers(Cefaluetal.,2018).ItisestimatedthathalfofthelistpriceofinsulingoestoPBMs,a$200billionayearindustry in the USA. Even if cheap biosimilars (Gotham et al., 2018), essentially genericversionsofinsulin,areintroduceditislikelythatthePBMswillinterveneinthesupplychainsignificantlyraisingcosts.There isnoeasysolutiontotheproblem,withtheUSACongressrefusing to limit drug costs, as is the case in Europe, mandating via Medicare that freemarkettheoriesshoulddeterminethecostofinsulin.

It is clear that the pharmaceutical companies are guilty of rampant and blatantprofiteering. They routinely issue lawsuits involvingpatentextensions to zealouslyprotecttheir monopoly on supply of insulin in the USA, and spend millions of dollars lobbyingpoliticians to bar cheaper biosimilar compounds from entry into themarket place (Know,2020).Thisissueisthefocusofpoliticalattentionatthehighestlevel,withinsulincaravansroutinely crossing the border to Canada, where patients can buy insulin freely over thecounter for a fractionof the cost in theUSA. Such trips, althoughhighly visible viamediacoverage, are available to only a limited number of the USA diabetic patients (Gambino,2019).Thesepatientsarethebeneficiariesofthediscoveryofinsulin,butarecynicallyandmercilesslyexploitedbyBigPharma,whoviewthemascashcows,willingtopayexorbitantprices for the insulin that keeps them alive, that they will use daily for decades, and forwhich there is no substitute. Banting, a war hero who risked his life treating his fellowsoldiersintheFirstWorldWar,whorefusedtopersonallyprofitfromhisdiscoverydespiteaprecariouspersonalfinancialsituation,andwhodiedonasecretmissionservinghiscountryduringtheSecondWorldWar,wouldhavebeenabsolutelydisgusted.

26


Islettransplantation Sincediabetesresults fromtheautoimmunedestructionofpancreaticβcells, thusdeprivingthebodyoftheinsulinrequiredtocontrolbloodglucose,anobvioustherapeuticstrategyistointroducethemissingcomponent,thusinjectionofexogenousinsulinhasbeenthesoletherapyfortype1diabetesforthelast100years.Howevertherearedifficultiesinmatching delivery of insulin to prevailing glucose concentrations as previously described.Withinthelast20yearstransplantingadonorpancreasintodiabeticpatientshasbecomearealityandtodateabout48,000havebeencarriedoutworldwide(Bellin&Dunn,2020).Therecipientsofthesetransplantsarealmostexclusivelytype1diabeticpatients,with lifelongimmunosuppression therapy required to prevent tissue rejection. In theUSA a controlledfollowupstudy tenyearsafter transplantation found that40%of the recipients remainedinsulinindependent,animpressivestatistic,althoughthelimitednumberofdonorpancreasmakes it an unrealistic therapeutic intervention formost type 1 diabetics (Bellin&Dunn,2020). An alternative therapy is to implant pancreatic islets, which contain both insulin-secreting β cells and glucagon-secreting α cells, introducing the potential to counteracthypoglycaemia(Rheinheimeretal.,2015).Isletsareharvestedfromhumandonorpancreasandmorethanonepancreasdonormayberequiredperpatientdependinguponyield.Theisletsarepurifiedtoridthetissueoftheexocrinesecretion,anattractionofthispreparationis that it allows more scope for innovation, particularly regarding immunosuppression(Rheinheimer et al., 2015). A study of this procedure demonstrated that 87% of patientsreported no severe hypoglycaemia and 52% were insulin independent one year afterimplantation(Heringetal.,2016).Oncetheisletsarecleanedtheyare introducedintotheportalsystemwheretheybecameimplantedintheliverdevelopingtheirownbloodsupply(Bellin & Dunn, 2020). However this can lead to a rapid inflammatory reaction (IBMIR),whichcankillasizeableproportionofthe implantedcellsand lifelong immunosuppressivedrugtherapy isrequired(Bellin&Dunn,2020).There isalsotheproblemof implantinganappropriatenumberofcells.Itisestimatedintheadulthumanthereareabout1,000,000βcells. Implanting an excess of β cells in the expectation of considerable cell death post-implant may lead to hypoglycaemia, whereas implanting too few would fail to achieveinsulin independence (Matsumoto,2010). Inorder todealwiththe immuneresponse isletcellscanbeencapsulated,whichisolatesthemfromthecirculationandalleviatestheneedforimmunosuppressivetherapy,althoughthisisstillindevelopment.Anothermethodistouse pigβcells. It is known that pig insulin is effective in humans, and there is a plentifulsupplyofpigpancreas.Thiswouldobviouslyleadtoanimmunogenicresponsebutemergingtechnologiesarebeingusedtoreducethiseffect(Matsumoto,2010).

Thedemandfordonorpancreasesfaroutstripssupplyandintroducesthepotentialuseforinsulinproducingcells(IPCs)derivedfromstemcellsasthemostadvancedapproachfor a sustainable therapy (Chen et al., 2020).Many studies have been carried out in thisarea, the IPCs derived from either embryonic stem cells (Thomson et al., 1998) or fromhumanpluripotentstemcells (Pagliucaetal.,2014).Theformercomeswithethical issues,whereas the latter involves adult cells, which are programmed back into an embryonicpluripotent state (Takahashi et al., 2007). To circumvent rejection issues the inducedpluripotentstemcellsmaybetakenfromthepatient(Millmanetal.,2016),whichcanalsobeencapsulatedwithinamatrix thatallows interstitial fluidaccesstothegraftedcellsbutpreventsimmunecellreaction(Kepsutluetal.,2014).Suchcellsresembletheadultprimarycells regarding their ability to produce insulin. Trials involving newly diagnosed type 1diabetic patients demonstrate that introduction of IPCs derived from liver mesenchymalstriomal cells achieved an increase in both insulin and C peptide secretion, which fell incontroldiabeticpatients(Caietal.,2016).However,despiteenormousprogressinthefield,itmustbeconsideredatechniqueindevelopmentwithnumerousobstaclestobeovercome

27


before it becomes a viable therapy. The technique must improve upon the islet andpancreastransplantationwhereupto60%oftreatedpatientsremaininsulinindependent5yearsaftertreatment(Rickels&Robertson,2019).

Hypoglycaemia Innon-diabeticindividualsthenormalbloodglucoseconcentrationliesbetween3.9and7.8mmoll-1andneverdeviatesfromthisrange(Frieretal.,2014).Thedirectactivationofαcellsbybloodglucoseisofsecondaryimportancetothereleaseofthetonicinhibitionthat β cells exert overα cells in the presence of normoglycaemic glucose (Cryer, 2012).Whenglucosefallstothelowerendofthenormalrangethedecreasedreleaseofinsulinbyβcellsactstoreleasethetonicinhibitionofglucagonreleasefromαcells,thusinsulinandglucagon act in an antagonistic manner in determining the glucose concentration in theblood.Intype1diabeticpatientsthelossofβcellsremovesthestimulusforαcellreleaseofglucagon,thusfallsinbloodglucoseconcentrationarenotcounteractedbyglucagonrelease(Cryer,2012).Hypoglycaemiaisextremelyuncommoninnon-diabeticindividualswheretheprinciple cause is insulinoma, a rare condition (Frier et al., 2014). Thus 100 years agohypoglycaemiawasaconditionofwhichmostdoctorswereunaware.

Intheearlydiabetesstudiesondogsacommonlyreportedconsequenceofinjectionofpancreaticextractwasconvulsions,butthesewereconsideredcomparabletofeversandinfectionandwereclassifiedasharmfulsideeffectsresultingfromimpuritiesintheextract(Bliss, 1983). Indeed Zuezler reported severe convulsions in dogs injectedwith his extractprior the First World War (Zuelzer, 1908). That convulsions were a symptom ofhypoglycaemiawas recognised by Collipwhile purifying extract in thewinter of 1921/22.Measurement of the blood glucose of rabbits rendered convulsive by extract injectionshowedverylowreadings:theinjectionofglucoseprecipitatedtheanimal’srecovery(Bliss,1983). This was the first appreciation that pancreatic extract could not only renderhyperglycaemic animals normoglycaemic, but could render normoglycaemic animalshypoglycaemic.InearlyJanuary1922BantingpersuadedMacleodtoallowhisextracttobeinjectedintoatype1diabeticpatientforthefirsttime.ItisnotclearifBantingwasawareofCollip’s discovery given the breakdown in communications within the group, but inretrospectitmustbeconsideredaverypoorclinicaldecision(Bliss,1983).Asmorepatientswere treatedwith insulin in Toronto the clinicians became extremely vigilant for signs ofhypoglycaemia (Banting et al., 1923). The Toronto group had not only created a newtherapeuticstrategytoeffectively treatdiabetesviaexogenousapplicationof insulin, theyhadalsocreatedanewpathologicalcondition,iatrogenichypoglycaemia,andthosetreatingtype 1 diabetic patients with insulin became familiar with the causes, symptoms andtreatments(Fletcher&Campbell,1922).

Hypoglycaemiaisaconditioninwhichthebrainreceivesinsufficientglucosei.e.thedemandisnotmatchedbythesupply,andtheabsenceofβcellslimitstheextenttowhichglucagon can increase systemic glucose concentrations (Cryer, 2012). The principlesymptoms of the response to hypoglycaemic attack betray the systems sensitive todecreased glucose and are categorised as being either autonomic (Lin et al., 2021) orneuroglucopenic (Garcia et al., 2021), with symptoms of sweating, trembling, weakness,hunger,poundingheart,belongingtotheformercategoryandvisualdisturbances,difficultyconcentrating and confusion belonging to the latter (Frier et al., 2014). The glucosethresholdfortheautonomicwarningsignsishigherthanthatfortheneuroglucopenicsignsallowing the patients time to recognise the symptoms and take appropriate action byingesting glucose. However a complication of hypoglycaemia is that repeated exposure

28


decreases the gap between the threshold for autonomic and neuroglucopenic symptomssuchthatthepatientsmayberenderedincapableofreactingtotheautonomicsymptoms,aprocessofunknownmechanismnamedhypoglycaemiaunawareness(Cryer,2013).

Hypoglycaemiaisthemainreasonthatpatientsfailtocomplywithaninsulinregimedesigned to maintain strict glycaemic control (Cryer, 2002). The Diabetic Control Studydemonstrated the effectiveness of a strict insulin regime in reducing the microvascularcomplicationsthatresultsfromhyperglycaemia,butitcameataprice,increasedepisodesofhypoglycaemia (Fig 7B)(Cryer, 2002). The practical measures that have been installed inmodern insulin therapy to prevent hypoglycaemia include equipping closed loop systemswithastopfunctionthatpreventsinsulininjectionwhenbloodglucosereadingsaretoolowandalarmactivationinCGMsystemswhenbloodglucoselevelsfalltowardshypoglycaemicconcentrations(Boughton&Hovorka,2021).

Insulinispotentiallyfatalandhasbeenusedoccasionallyinmurderandsuicide;thelack of a counter-regulatory glucagon response makes injected insulin particularlydangerous.Thefirstsuicidebyinsulinwasrecordedin1927byadiabeticpatientandtherehavebeensporadiccasesreportedsince(Russelletal.,2009).Thefirstmurderattributedtoinsulin occurred in 1957 when Kenneth Barlow, a hospital worker who presumably hadaccess to insulin,was convicted ofmurdering his pregnant non-diabeticwife Elizabeth byinjection of insulin forwhich hewas imprisoned for 26 years (Marks& Richmond, 2008).Similarcasesareroutinelyreportedinthemedicalliterature(Marks&Richmond,2008).TheClaus von Bulow case, in which the wealthy socialite was tried twice for the attemptedmurderofhiswife,presumablybyinsulininjection,attractedheadlinesworldwideandwasmade into the filmReversal of Fortune. Bulowwas acquittedat the second trial, the caserestingonhiswife’slowbloodglucoseandhighinsulinlevels,althoughwhethervonBulowhadaccesstoinsulinwasakeyaspectofthedefencecase(Marks,2007).Giventhelimitedaccesstoinsulinittendstobeusedbypatientsorthoseinthehealthcareprofessionsandtherehavebeenseveralcasesofhealthcareworkerswhouseinsulintokillpatients(Pidd&Grierson, 2015). However insulin is considered a poor means of killing, as it can havevariableeffectsonvictims,andexogenouslyappliedinsulincanbedetectediftheratioofCpeptide to insulin is lower than six to one in living people and twenty to one in corpses(Marks,1999).

29


Figure8-Theobesityepidemicunderliestheincreaseintype2diabetes.A&B.Unrestrainedconsumerism, manifest as morbid obesity, subsidizes the fast food industry, the dietingindustryandultimatelycorporatemedicine.C.Portionsizeplaysasignificantroleinobesity.

30


FutureDirections The core principle of current diabetes therapy is to match as closely as possibleexogenously administered insulin to the physiological release of insulin. The incrementalsteps towards this goal have been documented in this review and comprise twoindependent but related processes, accurate blood glucose monitoring coupled toadministration of an appropriate dose of insulin. It is only within the last 20 years thatadvances inthesetechnologieshaveprogressedsufficientlythatCGMcoupledwith insulinpumps have resulted in desirable TiR being achievable by the majority of patients. Thedevelopmentofglucose–responsiveinsulinbypassestherequirementforcontinuousinsulininfusion,where a pool of latent insulin is only activatedwith increasing concentrationsofglucose (Wangetal.,2021).Unfortunatelyhypoglycaemiaremainsanever-presentrisk,asthisstrategydoesnotaddresstheimpairedglucagonresponseintheabsenceofβcellinput.Curing diabetes, as opposed to effectively treating it, will require the introduction of anautonomous endogenous glucose sensitive insulin release mechanism that accuratelymimicsphysiologicalinsulin(andglucagon)releaseacrossthebroadrangeofbloodglucoseconcentrations experienced by diabetic patients. To date whole pancreas and β celltransplantationareeffective inachieving insulin independence in type1diabeticpatients,butlimitsindonorpancreasavailabilityhaveprompteddevelopmentofstemcelltherapiestocreatepatient-derivedglucoseresponsiveIPCsthatwillbothsecreteinsulinandmodulateglucagonrelease.

Conclusion The current obesity epidemic (Fig 8) underlies the explosion in type 2 diabetes,where,allowingforpopulationincrease,global incidencerateshavedoubledinthelast40years.IntheUSAinthelast20yearstheincidenceoftype2diabeteshasincreasedfrom6%to over 10% of the general population, 13% if only adults are considered. This relativelyrecentdevelopmentisextremelyconcerningforavarietyofreasons.Firstly,thedevelopedworld (north America, Europe and Australia) acts as aspirational capitalist model fordevelopingcounties(southAmerica,AfricaandAsia).Thereisapositive,directrelationshipbetween incidence of obesity and Gross Domestic Product, which is an indicator of acounty’s economic development (Kinge et al., 2015). A sense of entitlement also prevails,where economic prosperity liberates individuals from a subsistence diet to one wherefreedom of choice dominates, and medical guidance to limit junk food consumption isconsidered interfering, unwelcome and generally unheeded. Such economic freedomsinvariably leada largeproportionofthepopulationtowardsobesityanddiabetes.Aprimetarget in reducing diabetes is prevention and delaying its onset, but strategies to reduceobesityonaglobalscalehavebeenunsuccessful(NCD-RisC,2016).Individualswhoarepre-diabeticorwithshortdurationdiabetes,canreversetheconditionwitharegimenofcaloricrestriction and exercise with long-term sustained weight reduction maintaining the post-diabeticstate(Limetal.,2011;Taylor,2019).Howevertheincreasingincidenceofdiabetessuggests that fewpeoplemanage to accomplish this andhighlights aworrying underlyingtrend, where patients refuse to adopt behaviour that will favour their long-term health,probablyasaresultof thedifficulties inreversinga lifetimeofbadhabits.Secondly, thereare enormous cost implications of increasing incidence of diabetes, since it is adisproportionately expensive disease to treat. In theUK one in six hospital inpatients hasdiabetes,andintheUSA,although10%ofthepopulationhavediabetes, itconsumesover25%ofthehealthcarebudget. IntheUSAabout25%ofpatientsrationtheir insulin,whichwill only lead to increases in the rates of complications. How health care systems areplanningtocopewhendiabetesincidenceapproaches20%isunknown.

31


Acknowledgements IamindebtedtoProfessorTilliTanseyfordirectingmetorelevantliteratureandtoIanWardfordiscussionsonlivingwithtype1diabetes.

32


References

AdamsMJ,BlundellTL,DodsonEJ,DodsonGG,VijayanM,BakerEN,HardingMM,HodgkinDC,etal. (1969).Structureofrhombohedral2zinc insulincrystals.Nature224,491-495.

Allen FN, Stilman E& Fitz R (1919).Total dietary regulation in the treatment of diabetes.NewYork.

Alsaleh FM, Smith FJ, Keady S & Taylor KM (2010). Insulin pumps: from inception to thepresentandtowardthefuture.JClinPharmTher35,127-138.

Aronson D & Rayfield EJ (2002). How hyperglycemia promotes atherosclerosis: molecularmechanisms.CardiovascDiabetol1,1.

AshcroftFM,HarrisonDE&AshcroftSJ(1984).Glucoseinducesclosureofsinglepotassiumchannelsinisolatedratpancreaticβ-cells.Nature312,446-448.

American Diabetes Association (2018). Economic Costs of Diabetes in the U.S. in 2017.DiabetesCare41,917-928.

BaeshenNA,BaeshenMN,SheikhA,BoraRS,AhmedMM,RamadanHA,SainiKS&RedwanEM(2014).Cellfactoriesforinsulinproduction.MicrobCellFact13,141.

BanghamDR&MussettMV(1959).TheFourthInternationalStandardforinsulin.BullWorldHealthOrgan20,1209-1220.

Banting FG, Best CH, Collip JB,Macleod JJR&Nodble EC (1922). The effect of pancreaticextract(insulin)onnormalrabbits.AmJPhysiol62,162–176.

BantingFG,BestCH,CollipJM,CampbellWR,FletcherAA,MacleodJJR&NobleEC(1922).Theeffectproducedondiabetesbyextractsofpancreas.TransAssocAmerPhys,1-11.

Banting FG, Campbell WR & Fletcher AA (1923). Further clinical experience with insulin(pancreaticextracts)inthetreatmentofdiabetesmellitus.BrMedJ1,8-12.

BartholdSW (2004). Introduction:unsungheroes in thebattleagainstdiabetes. ILAR J45,227-230.

BellinMD&DunnTB(2020).Transplantstrategiesfortype1diabetes:wholepancreas,isletandporcinebetacelltherapies.Diabetologia63,2049-2056.

BenedictSR(1908).AReagentFortheDetectionofReducingSugars.JBiolChem5,485-487.

BenhamouPY,FrancS,ReznikY,ThivoletC,SchaepelynckP,RenardE,GuerciB,ChaillousL,et al. (2019). Closed-loop insulin delivery in adults with type 1 diabetes in real-lifeconditions:a12-weekmulticentre,open-label randomisedcontrolledcrossovertrial.LancetDigitHealth1,e17-e25.

Bequette BW (2013). Algorithms for a closed-loop artificial pancreas: the case for modelpredictivecontrol.JDiabetesSciTechnol7,1632-1643.

33


BestCH&ScottDA(1922).Thepreparationofinsulin.JBiolChem3,709-773.

BlissM(1983).TheDiscoveryofInsulin.McClellandandStewartLimited.Toronto.

BlissM(1992).Banting:ABiography.UniversityofTorontoPressInc.Toronto.

BlissM(1993).Rewritingmedicalhistory:CharlesBestandtheBantingandBestmyth.JHistMedAlliedSci48,253-274.

BlissM(2013).TheeclipseandrehabilitationofJJRMacleod,Scotland'sinsulinlaureate.JRCollPhysiciansEdinb43,366-373.

Boughton CK, Hartnell S, Allen JM, Fuchs J & Hovorka R (2020). Training and support forhybridclosed-looptherapy.JDiabetesSciTechnol,1-6.

BoughtonCK&HovorkaR(2021).Newclosed-loopinsulinsystems.Diabetologia. https://link.springer.com/article/10.1007/s00125-021-05417-3Bristow AF, Das RE & Bangham DR (1988). World Health Organization International

Standardsforhighlypurifiedhuman,porcineandbovineinsulins.JBiolStand16,165-178.

Brostoff JM, Keen H & Brostoff J (2007). A diabetic life before and after the insulin era.Diabetologia50,1351-1353.

BrownSA,KovatchevBP,RaghinaruD,LumJW,BuckinghamBA,KudvaYC,LaffelLM,LevyCJ,etal.(2019).Six-monthrandomized,multicentertrialofclosed-loopcontrolintype1diabetes.NEnglJMed381,1707-1717.

Cai J, Wu Z, Xu X, Liao L, Chen J, Huang L, Wu W, Luo F, et al. (2016). Umbilical cordmesenchymal stromal cell with autologous bone marrow cell transplantation inestablishedtype1diabetes:apilotrandomizedcontrolledopen-labelclinicalstudytoassesssafetyandimpactoninsulinsecretion.DiabetesCare39,149-157.

Carafoli E&Krebs J (2016).Whycalcium?Howcalciumbecame thebest communicator. JBiolChem291,20849-20857.

CardosoL,RodriguesD,GomesL&CarrilhoF(2017).BantingandBest:Theextraordianarydiscoveryofinsulin.RevPortEndocrinolDiabetesMetab12(1),106-115.

CBC(1990).https://www.cbc.ca/archives/entry/the-long-life-of-ted-ryder-diabetic

CDC(2020).NationalDiabetesStatisticsReport2020:EstimatesofdiabetesanditsburdenintheUnitedStates.USDepartementofHealthandHumanServices:CentresforDiseaseControlandPrevention.https://www.cdc.gov/diabetes/library/features/diabetes-stat-report.html

CefaluWT,DawesDE,GavlakG,GoldmanD,HermanWH,VanNuysK,PowersAC,TaylorSI,et al. (2018). Insulin access and affordability working group: conclusions andrecommendations.DiabetesCare41,1299-1311.

ChampePC&HarveyRA(2008).Biochemistry.LippincottWilliams&Wilkins.Baltimore.

34


ChanNN&HurelSJ(2002).Potentialimpactofanewbloodglucosemonitoringdevice:theGlucoWatch®Biographer.PractDiabInt19(4),97-100.

Chen S, Du K & Zou C (2020). Current progress in stem cell therapy for type 1 diabetesmellitus.StemCellResTher11,275.

Clarfield AM (2009). The Discovery of Insulin: Twenty-fifth anniversary edition. JAMA301(15),1607-1608.

Clarke SF & Foster JR (2012). A history of blood glucose meters and their role in self-monitoringofdiabetesmellitus.BrJBiomedSci69,83-93.

CollipJB(1923).Effectsofplanytexractsonbloodsugar.Nature111,571.

Cooperberg BA & Cryer PE (2010). Insulin reciprocally regulates glucagon secretion inhumans.Diabetes59,2936-2940.

CryerPE(2002).Hypoglycaemia:ThelimitingfactorintheglycaemicmanagementofTypeIandTypeIIDiabetes.Diabetologia45,937-948.

Cryer PE (2012). Minireview: Glucagon in the pathogenesis of hypoglycemia andhyperglycemiaindiabetes.Endocrinology153,1039-1048.

CryerPE (2013).Mechanismsofhypoglycemia-associatedautonomic failure indiabetes.NEnglJMed369,362-372.

DanneT,NimriR,BattelinoT,BergenstalRM,CloseKL,DeVriesJH,GargS,HeinemannL,etal.(2017).Internationalconsensusonuseofcontinuousglucosemonitoring.DiabetesCare40,1631-1640.

delaPenaA,RiddleM,MorrowLA,JiangHH,LinnebjergH,ScottA,WinKM,HompeschM,etal.(2011).Pharmacokineticsandpharmacodynamicsofhigh-dosehumanregularU-500 insulin versus human regular U-100 insulin in healthy obese subjects.DiabetesCare34,2496-2501.

de Leiva A, Brugues E & de Leiva-Perez A (2011). The discovery of insulin: continuedcontroversiesafterninetyyears.EndocrinolNutr58,449-456.

deSouzaAMF&López JA (2004). Insulinor insulin-likestudiesonunicellularorganisms:areview.BrazArchBiolTech47(6),973-981.

Diabetes.UK(2017).Us,diabeticsandlotsoffactsandstats.DiabetesUK.

DiMeglioLA,Evans-MolinaC&OramRA(2018).Type1diabetes.Lancet391,2449-2462.

EadieGS&MacleodJJR(1922).Thephysiologicalassayofinsulinbasedonitseffectsonthehyperglycemiafollowingglucose injectionsandepinephrine.AmJPhysiol1923,285-296.

ElleriD,DungerDB&HovorkaR(2011).Closed-loopinsulindeliveryfortreatmentoftype1diabetes.BMCMed9,120.

FeasbyWR(1958).Thediscoveryofinsulin.JHistMedAlliedSci13,68-84.

35


FeldbergWS(1970).HenryHallettDale1875-1968.BiographMemFellRoyalSoc16,75-174.

FieldsB (2011).Fromrabbits to theLeagueofNations:Earlystandardizationof the insulinunit.PharosAlphaOmegaAlphaHonorMedSoc74,28-33.

FiellerEC(1940).Thebiologicalstandardizationofinsulin.SupplJRoyalStatSoc7,1-64.

FletcherAA&CampbellWR(1922).Thebloodsugarfollowinginsulinadministrationandthesymptomcomplex-hypoglycaemia.JMetabRes2(5-6),637-649.

Fralick M & Zinman B (2021). The discovery of insulin in Toronto: beginning a 100 yearjourneyofresearchandclinicalachievement.Diabetologia.

FreeAH&FreeHM(1964).Self testing,anemergingcomponentofclinicalchemistry.ClinChem30(6),829-838.

FriedmanJM(2010).Ataleoftwohormones.NatMed16,1100-1106.

FrierBM,HellerSR&McCrimmonRJ(2014).HypoglycaemiainClinicalDiabetes.WileyandSons.Chichester.

Gambino L (2019). 'Insulin is our oxygen': Bernie Sanders rides another campaign bus toCanada.https://www.theguardian.com/us-news/2019/jul/28/bernie-sanders-americans-canada-insulin-bus-caravan

GarberAJ,AbrahamsonMJ,BarzilayJI,BlondeL,BloomgardenZT,BushMA,Dagogo-JackS,DeFronzo RA, et al. (2016). Consensus statement by the American Association ofClinical Endocrinologists and American College of Endocrinology on thecomprehensive type 2 diabetesmanagement algorithm - 2016 Executive Summary.EndocrPract22,84-113.

Garcia SM, Hirschberg PR, Sarkar P, Siegel DM, Teegala SB, Vail GM & Routh VH (2021).Insulinactionsonhypothalamicglucose-sensingneurones.JNeuroendocrinol,e12937.

GothamD,BarberMJ&HillA(2018).Productioncostsandpotentialpricesforbiosimilarsofhumaninsulinandinsulinanalogues.BMJGlobHealth3,e000850.

Griffin NK, Smith MA, Jenkins PA, Werther G & Baum JD (1979). Relationship betweenurinaryandbloodglucoseindiabeticchildren.ArchDisChild54,371-374.

Group RS (2017). Relative effectiveness of insulin pump treatment over multiple dailyinjectionsandstructurededucationduringflexibleintensiveinsulintreatmentfortype1diabetes:clusterrandomisedtrial.BMJ356,j1285.

GuariguataL,WhitingDR,HambletonI,BeagleyJ,LinnenkampU&ShawJE(2014).Globalestimatesofdiabetesprevalencefor2013andprojectionsfor2035.DiabetesResClinPract103,137-149.

Guyton AC&Hall JE (2006).Textbook ofMedical Physiology (11th Ed). Elsevier Saunders.Philadelphia.

HednerP,ScherstenB&ThulinT(1974).Dextrostixreflectancemeterasanaidindiagnostichypoglycemia.ActaMedScand195,29-31.

36


HédonE(1893).SurlaconsommationdusucrechezlechienapresI'éxtirpationdupancréas.ArchPhysiolNormPathol5,154–163.

HenquinJC,DufraneD&NenquinM(2006).Nutrientcontrolofinsulinsecretioninisolatednormalhumanislets.Diabetes55,3470-3477.

Hering BJ, Clarke WR, Bridges ND, Eggerman TL, Alejandro R, Bellin MD, Chaloner K,CzarnieckiCW,etal.(2016).Phase3trialoftransplantationofhumanisletsintype1diabetescomplicatedbyseverehypoglycemia.DiabetesCare39,1230-1240.

HerkertD,VijayakumarP,LuoJ,SchwartzJI,RabinTL,DeFilippoE&LipskaKJ(2019).Cost-related insulin underuse among patientswith diabetes. Jama InternMed179, 112-114.

Himsworth HP (1936). Diabetes Mellitus: The differentiation into insulin-sensitive andinsulin-insensitivetypes.Lancet227,127-130.

HirschIB(2016).InsulininAmerica:Arightoraprivilege?DiabetesSpectr29,130-132.

HirschIB,BodeB,CourregesJP,DykielP,FranekE,HermansenK,KingA,MersebachH,etal.(2012). Insulin degludec/insulin aspart administered once daily at any meal, withinsulinaspartatothermealsversusa standardbasal-bolus regimen inpatientswithtype 1 diabetes: a 26-week, phase 3, randomized, open-label, treat-to-target trial.DiabetesCare35,2174-2181.

HirschIB,JunejaR,BealsJM,AntalisCJ&WrightEE(2020).Theevolutionofinsulinandhowitinformstherapyandtreatmentchoices.EndocrRev41,733-755.

JanezA,Guja C,MitrakouA, LalicN, Tankova T, Czupryniak L, TabakAG, PraznyM, et al.(2020). Insulin therapy in adults with type 1 diabetes mellitus: a narrative review.DiabetesTher11,387-409.

JonesF(1983).TeddyRyder:Banting’slivingmiracle.SundayStar. https://insulin.library.utoronto.ca/islandora/object/insulin%3AC10002JoslinE(1916).Thetreatmentofdiabetesmellitus.CanadMedAssnJ6,673–684.

KaramitsosDT(2011).Thestoryofinsulindiscovery.DiabetesResClinPract93Sup1,S2-8.

KepsutluB,NazliC,BalT&KizilelS (2014).Designofbioartificialpancreaswith functionalmicro/nano-basedencapsulationofislets.CurrPharmBiotechnol15,590-608.

KesavadevJ,SabooB,KrishnaMB&KrishnanG(2020).Evolutionofinsulindeliverydevices:from syringes, pens, and pumps to DIY artificial pancreas.Diabetes Ther 11, 1251-1269.

KingeJM,StrandBH,VollsetSE&SkirbekkV(2015).Educationalinequalitiesinobesityandgross domestic product: evidence from70 countries. J Epidemiol CommunityHealth69,1141-1146.

KleinerIS(1919).Theactionofintravenousinjectionsofpancreasemulsionsinexperimentaldiabetes.JBiolChem40,153-170.

37


KnoppJL,Holder-PearsonL&ChaseJG(2019).Insulinunitsandconversionfactors:astoryoftruth,boots,andfasterhalf-truths.JDiabetesSciTechnol13,597-600.

Know R (2020). Insulin insulated: barriers to competition and affordability in the UnitedStatesinsulinmarket.JLaw&Biosciences,1-25.

KoenigRJ,PetersonCM,JonesRL,SaudekC,LehrmanM&CeramiA(1976).Correlationofglucose regulationandhemoglobinAIc indiabetesmellitus.NEngl JMed295, 417-420.

Koenig RJ, Peterson CM, Kilo C, Cerami A&Williamson JR (1976). Hemoglobin A1c as anindicatorofthedegreeofglucoseintoleranceindiabetes.Diabetes25,230-232.

KohnJ(1957).Arapidmethodofestimatingblood-glucoseranges.Lancet273,119-121.

KovatchevBP,KollarL,AndersonSM,BarnettC,BretonMD,CarrK,GildersleeveR,OliveriMC,etal. (2020).Eveningandovernightclosed-loopcontrolversus24/7continuousclosed-loop control for type 1 diabetes: a randomised crossover trial. Lancet DigitHealth2,e64-e73.

LaceyAH(1967).Theunitofinsulin.Diabetes16,198-200.

Laguesse E (1893). Sur la formation des ilots de Langerhans dans le pancreas. ComptesRendusHebdomadairesdesSeancesetMemoiresdelaSocietedeBiologie5,819–820.

Laguesse E (1896). Recherches sur l’histogenie du pancreas chez lemouton, III Origine etevolutiondesilotsdeLangerhans(pseudo-follicules).JAnatPhysiol32,209–255.

LaneW,BaileyTS,GeretyG,GumprechtJ,Philis-TsimikasA,HansenCT,NielsenTSS,WarrenM,etal.(2017).EffectofinsulindegludecvsinsulinglargineU100onhypoglycemiainpatientswithtype1diabetes:TheSWITCH1randomizedclinicaltrial.JAMA318,33-44.

LangerhansP(1869).Inaugural-Dissertation.Berlin:GustavLange.

LimEL,HollingsworthKG,AribisalaBS,ChenMJ,MathersJC&TaylorR(2011).Reversaloftype 2 diabetes: normalisation of beta cell function in association with decreasedpancreasandlivertriacylglycerol.Diabetologia54,2506-2514.

Lin EE, Scott-Solomon E & Kuruvilla R (2021). Peripheral innervation in the regulation ofglucosehomeostasis.TiNS44,189-202.

Macleod JJ (1978). History of the researches leading to the discovery of insulin: with anintroductionbyLloydG.Stevenson.BullHistMed52,295-312.

MarksV(1999).MurderbyInsulin.TheMedico-legalJournal67(4),147-163.

MarksV(2007).InsulinMurders:Truelifecases.RSMPress.

MarksV&RichmondC(2008).ColinBouwer:professorofpsychiatryandmurderer.JRSocMed101,400-408.

38


Marks V& Richmond C (2008). Kenneth Barlow: the first documented case ofmurder byinsulin.JRSocMed101,19-21.

MatsumotoS(2010).IsletcelltransplantationforType1diabetes.JDiabetes2,16-22.

MatzR(2000).Abookthatchangedme:TheDiscoveryofInsulin.BritMedJ321,1418.

Mazur A (2011). Why were "starvation diets" promoted for diabetes in the pre-insulinperiod?NutrJ10:23.

McAdamsBH&RizviAA(2016).Anoverviewof insulinpumpsandglucosesensorsforthegeneralist.JClinMed5,2-17.

Mering JV & Minkowski O (1889). Diabetes mellitus nach Pankreasextirpation. Arch ExpPatholPharmakol26,371–387.

MillerPF,StrattonC&TrippJH(1983).Bloodtestingcomparedwithurinetestinginthelongtermcontrolofdiabetes.ArchDisChild58,294-297.

MillerRG,SecrestAM,SharmaRK,SongerTJ&OrchardTJ(2012).Improvementsinthelifeexpectancyoftype1diabetes:thePittsburghEpidemiologyofDiabetesComplicationsstudycohort.Diabetes61,2987-2992.

MillmanJR,XieC,VanDervortA,GurtlerM,PagliucaFW&MeltonDA(2016).Generationofstem cell-derived beta-cells from patients with type 1 diabetes. Nat Commun 7,11463.

Murnaghan JH&TalalayPHH (1992).Dale’s accountof the standardizationof insulin.BullHistMed66,440–450.

MutskovV&FelsenfeldG(2009).Thehumaninsulingeneispartofalargeopenchromatindomainspecificforhumanislets.ProcNatlAcadSciUSA106,17419-17424.

NathanDM,GenuthS,LachinJ,ClearyP,CroffordO,DavisM,RandL,SiebertC,etal.(1993).Theeffectofintensivetreatmentofdiabetesonthedevelopmentandprogressionoflong-term complications in insulin-dependent diabetes mellitus.N Engl J Med 329,977-986.

NCD-RisC (2016). Trends in adult body-mass index in 200 countries from 1975 to 2014: apooled analysis of 1698 population-based measurement studies with 19.2 millionparticipants.Lancet387,1377-1396.

NCD-RisC (2016). Worldwide trends in diabetes since 1980: a pooled analysis of 751population-basedstudieswith4.4millionparticipants.Lancet387,1513-1530.

OpieEL(1910).Diseaseofthepancreas.Philadelphia.

OwensDR,ZinmanB&BolliGB(2001).Insulinstodayandbeyond.Lancet358,739-746.

Owerbach D, Bell GI, Rutter WJ & Shows TB (1980). The insulin gene is located onchromosome11inhumans.Nature286,82-84.

39


PagliucaFW,MillmanJR,GurtlerM,SegelM,VanDervortA,RyuJH,PetersonQP,GreinerD,etal. (2014).Generationof functionalhumanpancreaticbetacells invitro.Cell159,428-439.

Patterson CC, Dahlquist GG, Gyurus E, Green A, Soltesz G & Group ES (2009). Incidencetrendsforchildhoodtype1diabetes inEuropeduring1989-2003andpredictednewcases2005-20:amulticentreprospectiveregistrationstudy.Lancet373,2027-2033.

PickupJ(1986).Humaninsulin.BrMedJ(ClinResEd)292,155-157.

Pidd H&Grierson J (2015). Stepping Hillmurders: how Victorino Chua's poisoningswereuncovered.https://www.theguardian.com/uk-news/2015/may/18/stepping-hill-hospital-poisonings-operation-roxburg-manchester-police-victorino-chua

PinskerJE,LeeJB,DassauE,SeborgDE,BradleyPK,GondhalekarR,BevierWC,HuyettL,etal. (2016). Randomized crossover comparison of personalizedMPC and PID controlalgorithmsfortheartificialpancreas.DiabetesCare39,1135-1142.

PrattJH(1954).Areappraisalofresearchesleadingtothediscoveryofinsulin.JHistMedAllSci9(3),281-289.

RahbarS,BlumenfeldO&RanneyHM(1969).Studiesofanunusualhemoglobininpatientswithdiabetesmellitus.BiochemBiophysResCommun36,838-843.

RajkumarSV (2020).Thehighcostof insulin in theUnitedStates:anurgentcall toaction.MayoClinProc95,22-28.

Rawshani A, Landin-Olsson M, Svensson AM, Nystrom L, Arnqvist HJ, Bolinder J &GudbjornsdottirS(2014).Theincidenceofdiabetesamong0-34yearoldsinSweden:newdataandbettermethods.Diabetologia57,1375-1381.

Rawshani A, Sattar N, Franzen S, Rawshani A, Hattersley AT, Svensson AM, Eliasson B &GudbjornsdottirS(2018).Excessmortalityandcardiovasculardiseaseinyoungadultswithtype1diabetes inrelationtoageatonset:anationwide,register-basedcohortstudy.Lancet392,477-486.

RennieJ&FraserT(1907).TheIsletsofLangerhansinrelationtodiabetes.BiochemJ2,1-9.

RheinheimerJ,BauerAC,SilveiroSP,EstivaletAA,BoucasAP,RosaAR,SouzaBM,OliveiraFS,etal.(2015).Humanpancreaticislettransplantation:anupdateanddescriptionoftheestablishmentofapancreaticisletisolationlaboratory.ArchEndocrinolMetab59,161-170.

Rickels MR & Robertson RP (2019). Pancreatic islet transplantation in humans: recentprogressandfuturedirections.EndocrRev40,631-668.

RosenfeldL(2002).Insulin:discoveryandcontroversy.ClinChem48,2270-2288.

Roth J, Qureshi S, Whitford I, Vranic M, Kahn CR, Fantus IG & Dirks JH (2012). Insulin'sdiscovery:newinsightsonitsninetiethbirthday.DiabetesMetabResRev28,293-304.

RussellKS,StevensJR&SternTA(2009). Insulinoverdoseamongpatientswithdiabetes:areadilyavailablemeansofsuicide.PrimCareCompanionJClinPsychiatry11,258-262.

40


SangerF&ThompsonEO(1953).Theamino-acidsequence intheglycylchainof insulin. I.Theidentificationoflowerpeptidesfrompartialhydrolysates.BiochemJ53,353-366.

SangerF&ThompsonEO(1953).Theamino-acidsequenceintheglycylchainof insulin.II.Theinvestigationofpeptidesfromenzymichydrolysates.BiochemJ53,366-374.

SangerF&TuppyH(1951).Theamino-acidsequenceinthephenylalanylchainofinsulin.I.Theidentificationoflowerpeptidesfrompartialhydrolysates.BiochemJ49,463-481.

SangerF&TuppyH(1951).Theamino-acidsequenceinthephenylalanylchainofinsulin.2.Theinvestigationofpeptidesfromenzymichydrolysates.BiochemJ49,481-490.

SawbridgeJL,QureshiHK,BoydMJ&BrownAM(2014).Revisionworkshopsinelementarymathematics enhance student performance in routine laboratory calculations. AdvPhysiolEduc38,239-245.

Schernthaner G (1993). Immunogenicity and allergenic potential of animal and humaninsulins.DiabetesCare16Suppl3,155-165.

Schersten B, Kuhl C, Hollender A & Ekman R (1974). Blood glucose measurement withDextrostixandnewreflectancemeter.BrMedJ3,384-387.

ScottEL(1912).Ontheinfluenceofintravenousinjectionsofanextractofthepancreasonexperientalpancreaticdiabetes.AmJPhysiol29,306-310.

ScottEL(1913).Therelationofpancreaticextracttothesugaroftheblood.ProcSocExpBiolMed10,101-103.

Shaffer PA&Hartmann AF (1921). The idiometric determination of copper and its use insugar analysis: IImethods for the determination of reducing sugars in blood, urine,milkandothersolutions.JBiolChem45(2),365-390.

Sinding C (2002).Making the unit of insulin: standards, clinical work, and industry, 1920-1925.BullHistMed76,231-270.

SomogyiΜ&KramerΗV(1928).Thenatureofbloodsugar.JBiolChem80,733-742.

StansfieldWD (2012).Thediscoveryof insulin:a case studyof scientificmethodology.AmBiolTeach74,10-14.

Steil GM (2017). Comment on Pinsker et al. Randomized crossover comparison ofpersonalizedMPCandPIDcontrolalgorithmsfortheartificialpancreas.DiabetesCare2016;39:1135-1142.DiabetesCare40,e3.

SteilGM,PalermCC,KurtzN,VoskanyanG,RoyA,PazS&KandeelFR(2011).Theeffectofinsulin feedback on closed loop glucose control. J Clin Endocrinol Metab 96, 1402-2408.

Takahashi K, Tanabe K, OhnukiM,NaritaM, Ichisaka T, Tomoda K& Yamanaka S (2007).Inductionofpluripotent stemcells fromadulthuman fibroblastsbydefined factors.Cell131,861-872.

41


Taylor R (2019). Calorie restriction for long-term remission of type 2 diabetes. Clin Med(Lond)19,37-42.

ThomsonJA,Itskovitz-EldorJ,ShapiroSS,WaknitzMA,SwiergielJJ,MarshallVS&JonesJM(1998).Embryonicstemcelllinesderivedfromhumanblastocysts.Science282,1145-1147.

TrehteweyJ (1988).Whoreallydiscovered insulin?TheDiscoveryof Insulin.TrendsPharmSci9,146.

TriplittCL(2012).Understandingthekidneys'roleinbloodglucoseregulation.AmJManagCare18,S11-16.

UofToronto1(2021).https://heritage.utoronto.ca/exhibits/insulin

UofToronto2(2021).https://insulin100.utoronto.ca/

UofToronto3(2021).https://insulin.library.utoronto.ca/

Van Slyke DD & Hawkins JA (1928). A gasometric method for determination of reducingsugars,anditsapplicationtoanalysisofbloodandurine.JBiolChem79(2),739-767.

VarsamisP,LarsenRN,DunstanDW,JenningsGL,OwenN&KingwellBA(2017).ThesugarcontentofsoftdrinksinAustralia,EuropeandtheUnitedStates.MedJAust206,454-455.

VecchioI,TornaliC,BragazziNL&MartiniM(2018).Thediscoveryofinsulin:animportantmilestoneinthehistoryofmedicine.FrontEndocrinol9,613.

WangZ,Wang J,KahkoskaAR,Buse JB&GuZ (2021).Developing insulindeliverydeviceswithglucoseresponsiveness.TrendsPharmSci42,31-44.

WeismanA,BaiJW,CardinezM,KramerCK&PerkinsBA(2017).Effectofartificialpancreassystems on glycaemic control in patientswith type 1 diabetes: a systematic reviewand meta-analysis of outpatient randomised controlled trials. Lancet DiabetesEndocrinol5,501-512.

WhicherCA,O'NeillS&HoltRIG(2020).DiabetesintheUK:2019.DiabetMed37,242-247.

Whitford I,QureshiS&SzulcAL (2012).Thediscoveryof insulin: is theregloryenoughforall?EJBC,13-17.

Zuelzer G (1908). Ueber Versuche einer specifischen Fermenttherapie de Diabetes.ZeitschriftfurexperimentellePathologieundTherapie5,307-318.

Documents

One Hundred Years of Insulin