the aetiology and diagnosis of haemorrhagic diseases …

453

THE AETIOLOGY AND DIAGNOSIS OFHAEMORRHAGIC DISEASES IN CHILDREN

By A. D. M. JACKSON, M.D., M.R.C.P., D.C.H.Assistant Medical Registrar, The Hospital for Sick Children, Great Ormond Street, London

The subjects of haemostasis and of the haemor-rhagic diseases are the cause of much speculationand controversy. Many theories have been putforward to explain both the physiological and thepathological processes concerned. It is proposedin this review to include an account of thosetheories which are most widely accepted at thepresent time.

The Physiology of HaemostasisHaemostasis may be defined as the physico-

chemical process by which the flow of blood frominjured vessels is arrested. For the sake ofsimplicity it may be described in a series of stages:

i. The walls of the capillaries in the injuredarea contract and considerably diminish the flow ofblood. Small amounts of thromboplastin (throm-bokinase) are liberated by the injured tissues, butnot in sufficient quantity to affect coagulationsignificantly.

2. Platelets collect in and around the breach inthe vessel wall, partly as a mechanical stop-gap,but mainly so that they may, by clumping and dis-integrating in this abnormal environment, release asubstance known as the platelet factor, which,because of its theoretical role in the clottingmechanism, has also been called thromboplastino-genase (Quick, 1949).

3. The platelet factor activates a plasma factor,to which Quick (I947) has given the name thrombo-plastinogen and thromboplastin is formed. Throm-boplastin from this source effects the formatitn ofthrombin from the plasma prothrombin complex,which includes prothrombin, calcium, and at leastone other substance variously termed the ' labilefactor,' 'factor V' and 'accelerator factor'(Stefanini, I95I). A chain reaction is now set upby thrombin, which is thought to facilitate furtherdisintegration of platelets: more thromboplastinis therefore formed, more prothrombin is con-verted into thrombin, and so on. The whole

purpose of this chain reaction is to producemaximal amounts of thrombin as rapidly aspossible.

4. In the presence of thrombin, circulatingfibrinogen is converted into fibrin. With the ap-pearance of the fibrin clot in the wound and in thedamaged vessels the chain reaction is halted owingto the fact that fibrin inhibits further formation ofthrombin.

5. The capillary walls are now beginning todilate after their initial constriction (Macfarlane,1941), but the presence of the fibrin clot and itsattachment to the vessel wall, brought about by theadhesive properties of the platelets, prevent anyfurther flow of blood through the gap. The sub-sequent retraction of the clot-again the result ofsome physical property of the platelets-serves tocomplete the process by bringing the capillarywalls together and holding them there while thegap is repaired and the debris of coagulation isremoved (Quick, 1942).

It is difficult to understand why intravascularcoagulation does not normally take place in intactvessels. The fundamental reason for this is prob-ably that, where there is no break or other abnor-mality of the capillary wall, the platelets will notdisintegrate and no platelet factor will be released.Other factors, such as heparin, which acts as ananti-thrombin, may also play some part in theprevention of thrombosis.

It appears, then, that two main mechanisms areinvolved in haemostasis: the vascular mechanismand the coagulation mechanism. Normally thesetwo systems are complementary and efficienthaemostasis cannot be achieved if either breaksdown. In certain circumstances, however, thecoagulation mechanism is redundant. This is thecase with fine puncture wounds where there islittle trauma and the break in the capillary wall isso fine that vascular contraction, aided only bythe physical action of the platelets, is sufficient toclose it. On the other hand, efficient coagulation

Protected by copyright.

on Decem

ber 6, 2021 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

454 POSTGRADUATE MEDICAL JOURNAL September 1951

is by itself incapable of arresting haemorrhage ifthe vascular mechanism is defective.

Classification of the Haemorrhagic DiseasesFrom the preceding account it will be observed

that several individual factors are intimately con-cerned with haemostasis. They may be enumer-ated thus:

I. The vascular mechanism:Capillaries.

2. The coagulation mechanism:Platelets and platelet factor.Plasma thromboplastinogen.Plasma prothrombin.Plasma fibrinogen.

The haemorrhagic diseases occurring in childhoodcan be classified according to which of these factorsis considered to be at fault.

DISORDERS OF THE VASCULAR MECHANISM(a) Congenital defect of the capillaries:

Hereditary haemorrhagic telangiectasia.(b) Mechanical rupture of the capillaries, e.g.

pertussis.(c) Avitaminosis:

Scurvy.(d) Non-thrombocytopenic purpura:

I. Anaphylactoid.2. Symptomatic, e.g. infections, acute neph-

ritis.

DISORDERS OF THE COAGULATION MECHANISM(a) Platelet or platelet factor deficiency:

I. Thrombocytopenic purpura:(i) Symptomatic, e.g. bone marrow dis-

ease, infections.(ii) Idiopathic.

(In idiopathic purpura there is also a capillarydefect.)

2. Hereditary haemorrhagic thrombasthenia.(b) Thromboplastinogen deficiency:

Haemophilia.(c) Prothrombin deficiency:

I. Haemorrhagic disease of the newborn.2. Idiopathic.

(d) Fibrinogen deficiency:Congenital afibrinogenaemia.

The term purpura simplex has been omitted fromthis classification deliberately, because it is vagueand is often loosely used to describe mild types ofpurpura whatever their aetiology.Diagnostic Tests

Diagnosis in the haemorrhagic diseases dependsto a great extent on the results of certain investiga-tions, the principles and application of which willnow be discussed.

Tourniquet test (Hess's capillary resistance test).

This clinical test is designed to demonstrate alowered resistance of the capillary wall to internalpressure. 'It is performed by applying a bloodpressure cuff to the upper arm and maintaining apressure midway between the systolic and diastolicblood pressures for at least five minutes. Theappearance, under these conditions, of more thanten petechial haemorrhages in a circle of skin5 cm. in diameter, with its centre 4 cm. below thebend of the elbow, implies a weakness of thecapillary wall. Various modifications of the testhave been suggested with the idea of improving itsaccuracy and obtaining a quantitative measure ofthe capillary defect, but these methods are notreliable and in practice the test remains crude andinaccurate. It is often, but not invariably, positivein the diseases due to a vascular defect, but in thecoagulation defects it is always negative. Inidiopathic thrombocytopenic purpura it is almostalways positive.

Platelet count. Accuracy in platelet counts isdifficult to achieve, whether the direct or indirectmethod is used, but fortunately thrombocytopenia,when it is present, is usually considerable. Thenormal platelet count is from 250,000 to 500,000per cmm. with a wide daily variation. In the new-born the count is normally much lower (150,000to 250,000 per cmm.) and rises to the higher figuresduring the first three months of life (Tocantins,I938). A repeatedly normal count excludes thediagnosis of thrombocytopenia, which is said toexist when the platelets fall below the arbitraryfigure of Ioo,ooo per cmm., although most casesof idiopathic thrombocytopenic purpura havecounts below 50,000 per cmm. In children spon-taneous haemorrhage usually arises when the plate-let count falls to 70,000 per cmm., but haemorrhagemay also occur with higher counts and, conversely,there may be no bleeding even though the plateletcount is below 70,000 per cmm. Therefore inhaemorrhagic conditions associated with thrombo-cytopenia the bleeding is not necessarily related tothe level of the platelet count.

Clot retraction. It was pointed out earlier thatclot retraction depends on the physical activity ofthe platelets. A quantitative deficiency (belowabout 70,000 per cmm.), and possibly also a quali-tative defect of the platelets, will, therefore, lead toa failure of clot retraction or at least to a prolongedclot retraction time. Normal blood, when allowedto clot in a test tube, shows retraction of the clotwith the expression of a small quantity of clearserum within one hour, and if then the volumes ofthe clot and the serum are determined a quantita-tive estimate of clot retraction may be obtained(Macfarlane, 1939). The error in estimating clotretraction in vitro is considerable, but in idiopathicthrombocytopenic purpura the defect is usually


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

September 95I JACKSON: Haemorrhagic Diseases in Children 455

severe and there is a complete failure of the clot toretract. In haemophilia, when the clot eventuallyforms, it retracts normally.

Bleeding time. This simple procedure measuresthe time taken for the flow of blood from a smallpuncture wound to cease. Using the single punc-ture technique in the lobe of the ear, the normalbleeding time is between two and five minutes andappears to depend both on the ability of thecapillary wall to constrict after injury and on thepresence of a normal number of platelets. Sinceboth capillaries and platelets are unaffected inhaemophilia the bleeding time is essentially normalin this condition, despite a gross defect of thecoagulation mechanism. Even in diseases due to avascular defect, if the platelet count is normal, thebleeding time will also be normal, and it is only inidiopathic thrombocytopenic purpura that it isinvariably prolonged.

Coagulation time. This test determines the timetaken for blood to clot after removal from the bodyand is an assessment of the efficiency of the coagula-tion mechanism. In children the capillary tubemethod of Dale and Laidlaw is convenient, sinceonly a drop or two of blood from a finger prick isneeded. The normal coagulation time by thismethod is not greater than three minutes (fiveminutes in the newborn) and abnormally longtimes are found in haemophilia and the prothrom-bin deficiency diseases. In afibrinogenaemia theblood does not coagulate. The coagulation defectin thrombocytopenic purpura (i.e. platelet factordeficiency) is not demonstrable by this test, a factwhich will be explained in the next paragraph.

Prothrombin consumption test. It has beenthought for some time that a coagulation defectmust exist in thrombocytopenic purpura, despitethe fact that the coagulation time is normal, andthe prothrombin consumption test was designedby Quick (I947) to demonstrate this defect. Theprinciple of the test is simple. In the presence ofan optimum amount of thromboplastin largequantities of prothrombin are converted intothrombin during coagulation. There is, therefore,very much less prothrombin in serum than inplasma. Since only small amounts of thrombin arenecessary to produce enough fibrin for the coagula-tion time to be normal, a relative deficiency ofthromboplastin may exist without prolonging thecoagulation time. The amount of prothrombinconsumed in this case will be less than normal andthe amount left in the serum will be considerable.The prothrombin consumption test measures theamount of prothrombin consumed during the co-agulation of a sample of blood and an estimate ofthe available thromboplastin in that sample is thusobtained. It has already been stated that plasmathromboplastin is derived from the interaction of

platelet factor with thromboplastinogen, so that inthrombocytopenic purpura, where there is a de-ficiency of the former, and in haemophilia, wherethere is a deficiency of the latter, the consumptionof prothrombin during clotting should be low. Inpractice this is found to be the case and the pro-thrombin consumption test provides some con-firmation of the theories of aetiology in these twodiseases. From a practical point of view, the testis not suitable for routine use and, in any case,does not provide sufficient information to make ituseful in diagnosis.

Prothrombin estimation. By ensuring the presenceof an excess of thromboplastin, calcium and fibrino-gen the coagulation time of a sample of bloodbecomes a direct measure of the quantity of pro-thrombin it contains and the various methods ofestimating prothrombin depend on this principle.In practice, however, there are many technicalcomplications which may influence the result, sothat the test gives only a rough indication of pro-thrombin concentration. The normal prothrom-bin time is between 15 and 25 seconds, but theresult is usually compared with that of a sample ofnormal blood (tested under identical conditions)and expressed as a percentage of the normal (pro-thrombin index). The value of the prothrombinestimation lies in detecting diseases of the pro-thrombin deficiency group. In all the otherhaemorrhagic diseases the prothrombin concentra-tion is normal.

Fibrinogen estimation. The fibrinogen contentof normal blood is from 0.2 to 0.4 gm. per cent.,but it is rarely necessary to carry out this estima-tion, as all the cases of afibrinogenaemia so farrecorded have shown a complete absence offibrinogen, resulting in a failure of the blood toclot. It is possible, however, that unusual formsof the disease may occur and the fibrinogen shouldbe estimated where the cause of haemorrhagicsymptoms is obscure.

Aetiology and DiagnosisHereditary haemorrhagic telangiectasia. This

condition is inherited as a Mendelian dominantand affects both sexes. It is due to a develop-mental defect in some of the small vessels of theskin and mucous membranes, the affected vesselsbeing thin walled and dilated. The onset ofsymptoms is usually delayed until late childhoodor early adult life, but may occur as early as twoyears. Bleeding from the mucous membranesarises spontaneously and epistaxis is the com-monest symptom. The skin vessels do not bleedspontaneously and petechial haemorrhages are notseen. The visible telangiectases on the face, in themouth and nose, and under the nails appear onlyin adult life, but isolated 'spider naevi' uncon-


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

456 POSTGRADUATE MEDICAL JOURNAL September I951

TRAUMA

SI4ED LOOD

PLALATELETS-

TH4ROMNL0Sr1COE TEL4T fACTOR

Iz

PROHn4ROMINtiIcNBWLTT

FIBRIJNOqEW TOROMBN

FIBRINFIG. i.-The coagulation mechanism.

nected with this disease are often seen in children.Since all the diagnostic tests are normal and thereis usually a family history, diagnosis from otherhaemorrhagic diseases is not difficult, but when thebleeding is confined to haemoptysis, haematemesisor haematuria detailed investigation may be neces-sary to exclude other local causes of thesesymptoms.

Scurvy. A deficiency of vitamin C results ina grave disturbance of ossification and an increasein the permeability of the capillaries which leads tospontaneous haemorrhages. The age incidence ofthe disease is between six months and two years,the majority of cases occurring at the end of thefirst year. It is rarely seen in breast-fed babies, andin artificially-fed babies it occurs only if there havenot been adequate supplements of vitamin C.

After a period of vague symptoms the mainclinical features appear. There are haemorrhagesin the skin in the form of petechiae or ecchymosesand there is usually some haematuria and sub-periosteal haemorrhage. The gums, where teethhave erupted, become red, swollen and spongy,and the changes in the bones produce beading ofthe ribs and extreme tenderness of the limbs with

pseudo-paralysis. In addition, there are patho-'gnomonic radiological changes, most marked in thelong bones, including generalized rarefaction; thedense metaphyseal ' white line' of calcified carti-lage with a localized zone of rarefaction proximalto it; a similar dense line ringing the epiphysesand, in the later stages, calcification in the sub-periosteal haemorrhages (Fig. 2). The diagnostictests are all normal except for the tourniquet test,which is often positive. The diagnosis of scurvyis thus easily established by a combination of thetypical clinical features, the radiological appear-ances and the normal blood findings. When thereis still any doubt, the amount of ascorbic acidexcreted in the urine after a test dose may beestimated and this will be low in a case of scurvy.There is no convincing evidence that either

vitamin P or the similar compound, rutin, playany part in the causation of scurvy or of any othercondition associated with increased capillary fra-gility (Lancet, 1950).

Non-thrombocytopenic purpura. A number ofdiverse clinical conditions are grouped togetherunder this heading, all of them showing purpuriceruptions and often other forms of haemorrhage,without platelet deficiency. The cause of bleedingin these cases is an increased fragility of thecapillary wall. Two main groups are recognized:the anaphylactoid group, which has a distinctiveclinical picture, and a group in which the purpurais only a symptomatic manifestation of someprimary disease.The term anaphylactoid purpura is used to cover

the various disorders in which the capillary per-meability is considered to be allergic in nature,resulting from sensitivity, mainly to bacterial in-fection, but possibly also to other exciting factors.The separation of these disorders into different~yndromes, such as allergic purpura, Henoch'spurpura and Sch6nlein's purpura, is unnecessary,since they are all fundamentally due to the samecause and differ only in the severity or the site ofthe symptoms. In anaphylactoid purpura, then,there is often a history of a preceding bacterialinfection, such as a streptococcal sore throat, andthe presenting symptoms are a rash, acute ab-dominal symptoms with melaena, and painful non-haemorrhagic swellings in and around the jointsof the limbs. These main symptoms are usuallyall present together, but the severity of one or all ofthem may be extremely variable. Acute nephritisis a common complication (Gairdner, 1948). Therash has a characteristic appearance and distribu-tion. It is primarily maculo-papular with a hae-morrhagic component superadded, but true pete-chiae and ecchymoses are not c3mmon. Thetypical distribution is on the buttocks, the extensorsurfaces of the limbs and around the anHles (Fig. 3).


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

September 1951 JACKSON: Haemorrhagic Diseases in Children 457

.. .................. ..... .... ...

*........ :....................:!:::.:.:..........:..................

.,: '.: .... ..:;: :-.:

*~....:......e'

a.. ........

'~~~~. .. ...': .......... .. .

:~~~ -.. ..:

~. .. ..... ...:-''....~~~~~~~~~~.... .... ...................

ffi;rS;ff;s |ti~i~i~j;(~~

FIG. 2.-Radiological appearances in scurvy. There are dense lines of calcification in the metaphysesand surrounding the rarefied epiphyses. The localized zones of rarefaction adjacent to the meta-physeal lines are well shown in the femora. Calcification of a sub-periosteal haematoma iscommencing at the upper ends of the tibiae.


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

458 POSTGRADUATE MEDICAL JOURNAL September I9ot

There are no abnormal findings in the blood inthis group, although the tourniquet test isoccasionally positive, and the appearance of thecapillaries in the nail bed as seen under a micro-scope is essentially normal. The similarity betweenthe abdominal symptoms of anaphylactoid purpuraand those of intussusception is well known, butthe presence of a rash or joint swellings, or both,will usually help to differentiate the two con-ditions. Other causes of painful, swollen jointsmust also be excluded on clinical grounds, but theskin lesions are usually unmistakable. When thereis a true purpuric eruption the normal plateletcount will exclude the thrombocytopenic purpuras,but not the non-thrombocytopenic group, which issecondary to other diseases. Therefore historytaking and examination must be aimed at elimin-ating such diseases which include the later stagesof the acute specific fevers, subacute bacterialendocarditis, meningococcal infections and acutenephritis. In these diseases, which produce atoxic effect on the capillary wall, bleeding is merelya minor incident and is usually confined to thepetechial haemorrhages in the skin. Where pur-puric haemorrhages occur as a result of mechanicalpressure, as in the paroxysms of whooping cough,the cause will be obvious.

Non-thrombocytopenic purpura in the newbornis a symptom of septicaemia or asphyxia (where thecapillary weakness is due to anoxia), and may alsobe the result of mechanical compression duringlabour. The differential diagnosis of this specialgroup will be discussed in the section on haemor-rhagic disease of the newborn.

Thrombocytopenic purpura. The haemorrhagiccondition associated with thrombocytopenia ismost commonly secondary to some other disease.The thrombocytopenia in this case is due to adepression of the megakaryocytes in the bonemarrow and the bleeding is probably due to acombination of the platelet deficiency and acapillary weakness also brought about by theunderlying disease. This situation occurs, forexample, in leukaemia (Fig. 4), aplastic anaemia,the reticuloses, and drug intoxication with suchdrugs as arsenic, gold and sulphonamides. Inaddition, most of the acute infections, which havealready been described as causing a vascular defectalone, may also cause thrombocytopenia by theirtoxic effect on the bone marrow and, in childrenparticularly, simple upper respiratory infectionsmay produce a similar effect. In most of theseconditions the haemorrhagic manifestations arerelatively unimportant compared with the mainfeatures of the primary disease, but it is obviousthat in any case of thrombocytopenic purpura fullclinical and haematological investigations must becarried out in order to establish, if possible, a

t .:.;...............a

..... ......::

FIG. 3.-The rash in anaphylactoid purpura.

cause for the low platelet count. When no causewhatever can be demonstrated, the condition maybe termed idiopathic thrombocytopenic purpura or

purpura haemorrhagica.This disease is commoner in females than in

males and only about 30 per cent. of cases occurunder the age of ten, the highest incidence being inyoung adults. Occasionally babies born of motherswho are suffering from the disease may developtemporary thrombocytopenic purpura during thefirst few days of life, suggesting the passage ofsome causative factor across the placenta. Theaetiology of the condition remains obscure. Allthat can be said at present is that there is a platelet(and therefore plateletfactor) deficiency, possiblydue to an inhibition of megakaryocyte maturation,combined with a capillary defect. Exactly howthesefactors are related to each other and to thecause of the disease is unknown, but, since


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

September g951 JACKSON: Haemorrhagic Diseases in Children 459

'n*~ .i: .'··

.............

,&Xi

.:?:::::;ii~:,:~:~

ri,'~i.i~ ,;ii~i~'ii'"ii ~.;;; ';

iiii .lI.... ... I;i:iii.·. j~:;::ii:"~iiQliii'

FIG. 4.-Symptomatic purpura. Skin haemorrhagesfrom a rase of chronic myeloid leukaemia with aplatelet count of 65,ooo per c.mm.

removal of the spleen often cures the symptoms, ithas been suggested that some substance is pro-duced by the spleen which acts on both the marrowand the capilla:ics to produce the observed defects.

Idiopathic thrombocytopenic purpura is charac-terized by haemorrhages, either spontaneous orinduced by mild trauma, into the skin and fromthe mucous membranes. Skin petechiae andbruises are particularly common and the bleedingfrom any part of the body may be slight ordangerously severe. The spleen is occasionallypalpable and capillary microscopy shows abnormalbranching and distortion of the capillaries in thenail bed (Macfarlane, I941). The platelet countis always low, usually below 50,000 per cmm., thebleeding time is prolonged, clot retraction is defec-tive and the tourniquet test is usually positive, butmany ofthe cases are recurrent with variable periodsof remission, during which these abnormal findingsmay revert to normal. The coagulation time isnormal and megakaryocytes are present in thebone marrow in normal quantities. By means ofthese tests it is usually a simple matter to excludeother haemorrhagic diseases, but difficulty may beencountered in deciding whether the thrombo-cytopenia is secondary or not when the suspectedprimary cause is of a mild or trivial nature. How-ever, although investigations of the blood may give

identical results in both the idiopathic and thesymptomatic groups, the finding of a decreasednumber of megakaryocytes in the marrow willexclude the former. Any case of thrombocytopenicpurpura with an anaemia more severe than couldbe accounted for by blood loss must be consideredas leukaemia until this diagnosis is disproved byexamination of the bone marrow.

Hereditary haemorrhagic thrombasthenia (Glanz-mann's disease). This title is one of many (e.g.pseudo-haemophilia, von Willebrand's disease)given to a group of rare haemorrhagic disorders ofwhich the aetiology is quite unknown. In someof the cases it is thought that there is an excessivestability of the platelets which leads to a deficiencyof platelet factor, and the findings of Macfarlane(I941) on capillary microscopy in this disease sug-gest that there is a primary capillary defect. Asfar as recorded cases are concerned, the conditionappears to be hereditary and affects both sexes.Haemorrhage may occur, usually as a result ofsome slight injury, at any age and may be of varyingseverity. Glanzmann (I949) himself states thatecchymoses and bleeding from the mucous mem-branes are common, but that he has not knownhaemarthrosis to occur, and he maintains that theconstant finding in the blood is a defective clotretraction. There is no thrombocytopenia; in


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

460 POSTGRADUATE MEDICAL JOURNAL Septenber I951fact the platelet count is often higher than normal,and the bleeding and coagulation times are notprolonged. Very rarely cases have been describedin which the clot retraction is normal but thebleeding time is prolonged, and to this group thename of von Willebrand has been attached.Diagnosis is really a matter of exclusion of thewell-recognized diseases and, for the present atleast, the miscellaneous remainder must be placedin this group. However unsatisfactory this mayseem, it is probably less confusing than attemptingto separate each unusual case or group of cases intosyndromes with eponymous titles.

There. are some individuals, often with similarlyaffected relatives, who appear to bleed more thannormal, although not severely, after dental extrac-tions or tonsillectomy. There is, however, neverany demonstrable defect in the vessels or in thecoagulation mechanism in these cases and thereseems no reason to connect them with thromb-asthenia. Whether this sort of bleeding, theexistence of which is based solely on clinical im-pressions, is within the range of normal or is dueto some defect which present-day in vitro testsare too crude to detect will no doubt be decidedin the future.

Haemophilia. This disease is inherited by meansof a sex-linked recessive gene and affects maleswhose mothers are heterozygous carriers of thedefect. Sporadic cases may appear occasionally,presumably as a result of gene mutation, but noproven case has been recorded in a female. Thenature of the coagulation defect in haemophilia isstill uncertain, but the modern theories whichattempt to explain it must come very near to thetruth. It is reasoned that, since fibrin is notformed adequately and the fibrinogen is normal, athrombin deficiency exists. The prothrombincomplex is also normal, so that the fault lies withthromboplastin, a fact which is confirmed by thelowered consumption of prothrombin during thecoagulation of haemophilic blood. As thrombo-plastin formation depends on the platelet factorand the plasma factor, thromboplastinogen, one orboth of these factors must be ultimately respon-sible for the abnormality of coagulation. Withregard to the platelets, it has been demonstratedby Merskey (1950) that the low prothrombin con-sumption in thrombocytopenia can be remedied bythe addition of platelets from haemophilic blood,and he also showed that the similar defect inhaemophilia can be corrected by normal platelet-free plasma. In other words, in haemophilia theplatelets are normal and the missing factor iscontained in normal plasma. By a process ofelimination, therefore, a deficiency of thrombo-plastinogen appears to be the cause of the disease.Thromboplastinogen is probably identical with

anti-haemophilic globulin, the plasma fractionwhich has been described by Lewis et al. (I946)and which they stated would correct the prolongedcoagulation time in haemophilia. A circulatinganti-coagulant has also been described in somecases of haemophilia, but this is a rare occurrenceand often only appears after repeated trans-fusion.The bleeding in haemophilia is initiated by

trauma, which may only be trivial, and is notcommon under the age of about six months. Itsonset is delayed a little after the injury, but itsduration is considerably prolonged. Any part ofthe body may be affected but the subcutaneoustissues and muscles, the tongue and tooth socketsare the common sites. Petechial haemorrhagesdo not occur, but haemarthrosis, which is rare inother haemorrhagic disorders, is often seen andleads eventually to ankylosis of the affected joints.Episodes of bleeding often recur at regular in-tervals and the coagulation time, although nearlyalways prolonged, may vary with a similarperiodicity. Apart from the low prothrombinconsumption all the other diagnostic tests arenormal. The diagnosis is made on the familyhistory, the clinical features and the laboratoryfindings. In the rare cases in which the coagula-tion time is normal, confirmation of the diagnosismay be obtained from the prothrombin con-sumption test (Merskey, I95i).

Haemorrhagic disease of the newborn. Haemor-rhage is not uncommon in the newborn periodand occurs in some 5 per cent. of babies. Sucheasily recognized causes as trauma or infectionaccount for the majority of cases and only a few(about one in 500 live births) are associated withan abnormally low level of prothrombin in theblood and appear to have no other obvious cause.To the latter group the term haemorrhagic diseaseof the newborn is applied. The prothrombin levelin all newborn babies falls to about 30 per cent. ofnormal by the second day of life; the formation ofprothrombin in the liver depends on the presenceof vitamin K, and during the first few days of lifethere is no intake of this vitamin and there are nointestinal organisms to synthesize it; therefore theprothrombin lost into the placenta immediatelyafter birth is not replaced. By the fourth or fifthday these factors are no longer operating, and bythe seventh day the prothrombin has returned tonormal. This physiological hypoprothrom-binaemia is not usually associated with bleeding,but it may be; and while the majority of infantswith haemorrhagic disease have an unusually lowprothrombin level (below 20 per cent.), theselevels may occasionally exist without symptoms.The concentration of prothrombin is not there-fore entirely responsible for the bleeding in this


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

September 951 JACKSON: Haemorrhagic Diseases in Children 46I

disease and the existence of some additional factor,as yet undetermined, must be postulated.The bleeding in haemorrhagic disease appears to

be spontaneous, although a history of trauma is notdifficult to obtain in the first week of life, and itarises typically from the mucosa of the gastro-intestinal tract. In addition to haematemesis andmelaena there may be bruises in the skin (but nopetechiae) and bleeding from the umbilicus andvagina. The onset of symptoms is between thesecond and seventh days but not after the seventh(except, perhaps, in premature babies) and thediagnosis is supported by a low prothrombin leveland a prolonged coagulation time, the bleedingtime and platelet count being normal. These testsdo not prove the diagnosis, however, and as faras possible the many other causes of haemorrhagein the newborn must be excluded, mainly onclinical grounds. The following conditions mustbe considered:

Birth trauma. tAsphyxia neonatorum.Neonatal sepsis and septicaemia.Neonatal vaginal bleeding.Localized umbilical bleeding.Haematemesis due to oesophagitis.

Congenital thrombocytopenic purpura.Afribrinogenaemia.Of these, birth trauma and asphyxia are obvious,

but since they both may cause intracranial haemor-rhage, it may be impossible to make an exactdiagnosis of the cause of this particular type ofbleeding. Infections of various types are commonin the newborn and, particularly when com-plicated by septicaemia, can cause widespreadhaemorrhage. This sort of bleeding is recognizedby the local signs of the infection and by the bloodculture, which is often positive. Neonatal vaginalbleeding, due in some way to the effect of maternalhormones, is only slight and is not accompanied byhaem6rrhage from other sites. When bleedingfrom the umbilicus is not part of a generalizedhaemorrhagic condition, it is due either to in-efficient ligation of the cord or to a strictly localizedinfection of the cord stump. It is important toexclude haematemesis arising from local con-ditions in the oesophagus before attributing it tohaemorrhagic disease. Even at this early ageulceration of the lower end of the oesophagus canoccur as a result of a congenital 'short oeso-phagus' with a partial intrathoracic stomach, andsimple vomiting due to any cause may produce an

TABLE I

SUMMARY OF THE DIAGNOSTIC FEATURES OF THE HAEMORRHAGIC DISEASES

SkinDisease Pur- Tourni- Platelet Clot Bleeding Coagula- Pro- Other

pura quet Test Count Retraction Time tion Time thrombin Features

Hereditary Hereditaryhaemorrhagic No Negative Normal Normal Normal Normal Normal dominanttelangiectasia Both sexes

Scurvy Yes Positive Normal Normal Normal Normal Normal

Anaphylactoid Sometimespurpura Yes positive Normal Normal Normal Normal Normal

Idiopathicthrombocyto- Yes Positive Low Abnormal Prolonged Normal Normalpenic purpura

Hereditaryhaemorrhagic ? High May be May be Normal Normal Hereditarythrombasthenia normal abnormal prolonged Both sexes

HereditaryHaemophilia No Negative Normal Normal Normal Prolonged Normal recessive

Males onlyHaemorrhagic

disease of the No Negative Normal Normal Normal Prolonged Lownewborn

AbsentAfibrino- No Negative Normal Normal - Normal fibrinogen

gcnaemia No coagulation


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

462 POSTGRADUATE MEDICAL JOURNAL September 95Ig

oesophagitis with bleeding. Congenital thrombo-cytopenic purpura shows the same clinical featuresas the acquired type but occurs only when thedisease has been present in the mother. Haemo-philia, surprisingly, does not cause symptoms inthe first week of life and afibrinogenaemia caneasily be detected by examination of the blood (seeTable i).Of the other types of prothrombin deficiency

little need be said. Congenital idiopathic hypopro-thrombinaemia has been described and, apartfrom an earlier onset of symptoms, produces aclinical picture identical with that of haemophilia.There is no family history but the coagulation timeis prolonged, and if the prothrombin is notestimated these cases will be mistakenly diagnosedas sporadic haemophilia. Failure of absorption ofvitamin K as in obstructive jaundice and chronicdiarrhoea, or failure of formation of prothrombinin the liver in severe disease of that organ both

result in hypoprothrombinaemia, but haemor-rhagic symptoms are not a feature of these con-ditions.

Congenital afibrinogenaemia. About ten cases ofcongenital absence of fibrinogen have been re-corded. The disease is probably hereditary andhas the following features; a tendency to bleedafter minor injuries, from the time of birth; acomplete absence of fibrinogen; failure of theblood to clot even after incubation; a very slowsedimentation rate and a variable bleeding time.Diagnosis is therefore a simple matter once thepossibility of the disease has been considered.

I should like to thank Dr. W. G. Wyllie, Dr.Wilfrid Sheldon and Dr. Bernard Schlesinger forpermission to use photographs of their cases,which were kindly supplied by Mr. D. Martin.

I am greatly indebted to Dr. I. A. B. Cathie forhis helpful criticism.

BIBLIOGRAPHY

GAIRDNER, D. (1948), Quart. J. Med., 17, 95.GLANZMANN, E. (I949), ' Einfihrung in die Kinderheilkunde,'

3rd Ed., Springer: Vienna.LANCET (I95o), li, 690.LEWIS, J. H., DAVIDSON, C. S., MINOT, G. R., SOULIER,

J. P., TAGNON, H. J., and TAYLOR, F. H. L. (1946), J.clin. Invest., 25, 870.

MACFARLANE, R. G. (X939), Lancet, i, 1199.MACFARLANE, R. G. (1941), Quart. J. Med., 10, I.

0

MERSKEY, C. (1950), J. clin. Path., 3, 130.MERSKEY, C. (I95I), Brit. Med. J., i, 906.QUICK, A. J. (1942), 'The Haemorrhagic Diseases,' Sprin,fie d,

Illinois: Charles C. Thomas.QUICK, A. J. (i947), Amer. J. Med. Sci., 214, 272.QUICK, A. J. (1949), Amer. J. Clin. Path., I9, ioI6.STEFANINI, M. (i951), Lancet, i, 606.TOCANTINS, L. M. (1938), Medicine, Baltimore, 17 155.


on Decem


j.bmj.com

/P


j.27.311.453 on 1 Septem

ber 1951. Dow

nloaded from

http://pmj.bmj.com/

the aetiology and diagnosis of haemorrhagic diseases …

Documents