Lessons and Indications from Three Decades of West-Swedish Cerebral Palsy Data

It is a great honour for me to have been chosen to receive the prestigious “Year 2000 Becker Prize” and, in this connection, to have been asked to present our Gothenburg experience: the lessons learned from the West-Swedish Cerebral Palsy (CP) follow-up project, now in its 3rd decade. I took the initiative in 1971 after I had just been appointed to the chair of Paediatrics II at Gothenburg University. The Medical Director at Bräcke Ostergård (Regional Habilitation Centre), my late friend Ingemar Olow, and my wife Gudrun (epidemiologist) joined me. The project is population-based, which is of considerable importance. It is localized in a part of Sweden with a very uniform neuropaediatric structure. To date, it covers the birth years 1954 - 1994 (1649 CP cases). The data has been clinically evaluated and epidemiologically analysed in separate time sectors: the first covered the birth years 1954 - 1970 in one cohort [[7]], followed by five successive four-year birth cohorts (ages at inclusion and evaluation in each at 4 - 8 years) [[9]]. The birth period of 1991 - 1994 has just been analysed [[6]].

Lesson 1 is a basic one, generally known, but not often adhered to. The message, which should be always remembered, is that CP is not a disease but a sociomedical symptom complex, like mental retardation and epilepsy. Biologically, CP is therefore an artificial concept, inappropriate to be applied as though it were some kind of uniform and unspecified diagnostic label for early brain damage. Nevertheless, clinically sorted into well-defined syndromatic subgroups, the CP sector can provide important medical perspectives and information, which might also be useful for future preventive work [[2], [3], [7], [9], [20]].

Lesson 2. This point relates to the importance of unlimited availability and good reliability of the data that are presented. This data should then be uniformly statistically analysed. The presentation should be based on identical clinical/diagnostic definitions. The same epidemiological tools should be systematically applied throughout. The errors are then at least systematic! In our Swedish project, we feel we have maintained this kind of such strictness and continuity. Moreover, in field studies like ours, every neuropaediatric examiner should use identical terminology and classifications that are agreed on by the leading team. We think that having the same clinical and epidemiological project leaders since 1971, with collaborators over longer or shorter periods who trained at the same centre, has been essential for our research design and its validity.

Lesson 3. This message refers to CP epidemiology in relation to population statistics. One such particularly important “lesson” from cohorts born in the 1970s was that, in spite of considerable advances in perinatal preventive care of preterm births, the final outcome nevertheless involved a significant increase in the numbers of preterm children with spastic CP [[8], [9]]. This was shown to be related to a simultaneously marked increase in the survival of very preterm infants in general. Analyses of “gains and hazards” in intensive neonatal care [[8]] therefore demonstrated that there were about ten times as many non-spastic, very preterm survivors as spastic ones, i.e. this, in fact, indicated a considerable gain in terms of healthy survivors which was, however, not visible in the health statistics. Our recent data from the late 1980 and the early 1990 birth cohorts confirm that the prevalence of CP among preterms is no longer increasing, in spite of additional improvements in survival [[9], [20]].

Lesson 4. This message relates to developmental brain biology. The brain of full term babies is biologically different from that of preterms. In fact, these two types of brains represent developmentally totally different biological “species”, which is particularly relevant when it comes to the brain of the extremely preterm baby [[12]]. Takashima and his team were recently elegantly able to demonstrate the close correlation between characteristic developmental changes in human cerebral vessels, those perforating the medulla, and the occurrence of periventricular leucomalacia (PVL) [[12]]. MRI studies [[13], [18], [19]] confirm that the brain morphology of lesions is an indicator of the developmental stage of the brain when the damage occurred. Abnormal clinical patterns can therefore be first understood when findings of brain pathology at MRI and CT are related to gestational age (GA) for the likely timing of brain impairment. The GA-dependent occurrence of PVL is obvious. At present, we believe that PVL occurs in utero among full term “unexplained” spastic diplegics whereas in the preterm, PVL mainly occurs during the neonatal period. In this connection the recently described brain dysmaturity index for automatic detection of high-risk infants can be of future importance [[11]].

Lesson 5. I would like to refer to the recurrent experience as a neuropaediatrician of the considerably changing clinical presentation of preterm spastic CP over time. This is apparent when retrospectively considering the first, classical description of Little disease given in 1862. Up to 1970, in our project we still saw cohorts of diplegic children presenting with classic Little disease, then mainly moderately preterm CP children. The extreme preterms we see today with differently presenting spastic CP were non-existent at that time. Atypical diplegics now constitute an ever-increasing group which differs clinically from the classic Little disease [[6], [9]].

Lesson 6. This is a warning referring to the earliest age of the child at which the definite clinical classification of CP should be finally settled for research purposes. According to our experience: never definitely classify too early! This refers particularly to preterm spastic diplegias of the more severe type. During the first years of life, quite a few of these children present with a marked dystonic neurology which can still persist, and even be dominant, up to 3 - 4 years of age, and sometimes even longer. These children not infrequently present with a dominant yet transient dystonic pattern; they show tonic neck reactions and are not seldom mistakenly registered as dystonic/dyskinetic CP. Figs. [1] a and b show a classic severely diplegic girl in her upper teens in contrast to her marked dystonic presentation when she was 2¿ years of age. In this connection, it should also be remembered that not a few preterm children early demonstrate a transient diplegic dystonia that resolves completely [[5]], sometimes, however, not until 4 years of age.

Fig. 1 Classic preterm spastic diplegia, Little disease: a the early pronounced but transient dystonic presentation, still dominating at age 2¿ years; b the final pure spastic lower limb dominated presentation in the late teenager.

Lesson 7. This message refers to collaborative CP research in which more than a few centres participate. It is vital to stick to non-controversial subgroups which can be sorted clearly and defined uniformly. Different countries and even different teams in the same country not infrequently apply different terms in their classifications. We learned many valuable lessons in our closely collaborative CP project with the Tübingen team. This project started in the 1980s [[14]]. In spite of periods in which visiting collaborators joined one another's centres, to obtain the correct transferable data, we nevertheless finally found it necessary to concentrate on “bilateral spastic CP” with the definitions and the classifications so well drawn up by Michaelis et al [[14]]. This classification, with the strict limitation of CP types included, was then successfully applied in the joint studies between Tübingen and Gothenburg in a series of papers throughout the 1990s [[14], [15], [16], [17]]. Our Swedish classification, which is perhaps more complete, but also more complicated, could be used in a project like our West-Swedish one with the same project leaders throughout and with an identical subgroup identification of complex/transitional cases.

Lesson 8. We learned that there was one particularly complex CP group which was difficult to “transfer in research” between centres. This comprised the 10 % of dyskinetic CP cases. In addition to being complex, it also changes considerably in terms of clinical pattern over time. Our first dyskinetic cohorts therefore mainly contained classical choreoathetosis with Kernicterus as the pathogenic background. Today, this kind of basal ganglia damage (with diffuse tissue changes) is very rare. In fact, it is virtually non-existent in Sweden. However, the 10 % of dyskinetic cases within our CP panorama are completely different and are dominated by severe asphyxia damage at birth in severely compromised full terms. The pathology differs from that of posticteric choreoathetosis, nowadays with the putamina more selectively affected, in the past with far more diffuse basal ganglia lesions [[5]]. Parallel to this, the clinical presentation at the 4- to 8-year evaluation has changed. There are therefore many more very severely motor-disabled patients with posthypoxic impairments, but practically none with milder and clinically different posticteric CP presentations with their relatively good practical motor performance. This latter group from earlier periods, on the other hand, was not infrequently complicated by considerable neurogenic hearing defects, which are not seen today.

Lesson 9. This refers to simple non-progressive ataxia syndromes - without complicating simultaneous bilateral leg spasticity - which have been shown to have a special, completely different background and presentation compared with other CP types. The brain pathology is most heterogeneous. Neurodevelopmentally, many of them are quite complex. The group comprises a variety of inborn ataxic syndromes, many of them with a well-established genetic background. As CP cases we have included children with defined syndromes like Angelman, Joubert and so on [[1]] which might be disputed. We have noted that quite a few small Rett-syndrome girls may still present with ataxic symptomatology at age 4 to 5 years, however transient and developmentally dependent, the so-called jerky truncal ataxia [[4]]. This changes in pattern with increasing age and finally disappears more or less completely. As I see it, this “ataxia” is probably related to the very special Rett brain pathology, the dendritosynaptic developmental errors. These age-dependent transient “ataxias” do not qualify as part of the ataxic CP group. Even after the age of 6 - 7 years, individual children with non-progressive ataxia can be extremely difficult to pinpoint as being ataxic (which they finally are). These children not infrequently simulate a pronounced isolated hypotonia and demonstrate their classic ataxic deviations much later, sometimes first at school age. The so-called Foerster's atonic-astatic CP appears also to be developmentally dependent. In my experience, this CP type does not exist in its pure form in older teenagers [[10]].

References

• 1 Escher E, Flodmark O, Hagberg G, Hagberg B. Non-progressive ataxia: origins, brain pathology and impairments in 78 Swedish children.  Dev Med Child Neurol. 1996;  38 285-296
• 2 Hagberg B. Epidemiological and preventive aspects of cerebral palsy and severe mental retardation in Sweden.  Eur J Pediatr. 1979;  130 71-78
• 3 Hagberg B. Epidemiologie der Cerebralparesen.  Neuropediatrics. 1981;  2 (Suppl) 141-154
• 4 Hagberg B. Rett syndrome: clinical pecularities and biological mysteries.  Acta Paediatr. 1995;  84 971-976
• 5 Hagberg B, Hagberg G. The origins of cerebral palsy. David TJ Recent Advances in Paediatrics. XI. Edinburgh, London; Churchill Livingstone 1993: 67-83
• 6 Hagberg B, Hagberg G, Beckung E, Uvebrant P. The changing panorama of cerebral palsy in Sweden. VIII. Prevalence and origin in the birth year period 1991 - 1994.  Acta Paediatr. 2000; 
• 7 Hagberg B, Hagberg G, Olow I. The panorama of cerebral palsy in Swedish children born 1954 - 1974.  Neuropädiatrie. 1977;  8 (Suppl) 516-521
• 8 Hagberg B, Hagberg G, Olow I. Gains and hazards of intensive neonatal care: an analysis from Swedish cerebral palsy epidemiology.  Dev Med Child Neurol. 1982;  24 13-19
• 9 Hagberg B, Hagberg G, Olow I. The changing panorama of cerebral palsy in Sweden. VII. Prevalence and origin in the birth year period 1987 - 1990.  Acta Paediatr. 1996;  85 954-960
• 10 Hagberg B, Sanner G, Steen M. The dysequilibrium syndrome in cerebral palsy. Clinical aspects and treatment.  Acta Paediatr Scand. 1972;  Suppl 226 1-63
• 11 Holthausen K, Breidbach O, Scheidt B, Frenzel J. Brain dysmaturity index for automatic detection of high-risk infants.  Pediatr Neurol. 2000;  22 187-191
• 12 Inage Y W, Itoh M, Takashima S. Correlation between cerebrovascular maturity and periventricular leucomalacia.  Pediatr Neurol. 2000;  23 204-208
• 13 Krägeloh-Mann I, Hagberg B, Petersen D, Riethmüller J, Gut E, Michaelis R. Bilateral spastic cerebral palsy - pathogenic aspects from MRI.  Neuropediatrics. 1992;  23 46-48
• 14 Krägeloh-Mann I, Hagberg G, Meisner C, Schelp B, Haas G, Edebol Eeg-Olofsson K. et al . Bilateral spastic cerebral palsy - A comparative study between southwest Germany and western Sweden. I. Clinical patterns and disabilities.  Dev Med Child Neurol. 1993;  35 1037-1047
• 15 Krägeloh-Mann I, Hagberg G, Meisner C, Schelp B, Haas G, Edebol Eeg-Olofsson K. et al . Bilateral spastic cerebral palsy - A comparative study between southwest Germany and western Sweden. II. Epidemiology.  Dev Med Child Neurol. 1995;  36 473-483
• 16 Krägeloh-Mann I, Hagberg G, Meisner C, Haas G, Edebol Eeg-Olofsson K, Selbman H K. et al . Bilateral spastic cerebral palsy - A collaborative study between southwest Germany and western Sweden. III. Aetiology.  Dev Med Child Neurol. 1995;  37 191-203
• 17 Krägeloh-Mann I, Petersen D, Hagberg G, Vollmer B, Hagberg B, Michaelis R. Bilateral spastic cerebral palsy - MRI pathology and origin. Analysis from a representative series of 56 cases.  Dev Med Child Neurol. 1995;  37 379-397
• 18 Krägeloh-Mann I, Toft P, Lunding J, Andresen J, Pryds O, Lou H C. Brain lesions in preterms: origin, consequences and compensation.  Acta Paediatr. 1999;  88 897-908
• 19 Sie L TL, van der Knaap M S, Oosting J, de Vries L S, Lafeber H N, Valk J. MR patterns of hypoxic-ischemic brain damage after prenatal, perinatal or postnatal asphyxia.  Neuropediatrics. 2000;  31 128-136
• 20 Stanley F, Blair E, Alberman E. Cerebral Palsies: Epidemiology and Casual Pathways. Cambridge; Mac Keith Press 2000