Función de Optimización

Prim ary culture

A total of 82 childhood brain tum ours were collected and prepared as prim ary cultures. In th e m ajority of cases, the samples were prepared using collagenase digestion and the rem ainder w here tissue fragm ents received were too small for digestion as explant cultures.

P reparation m e th o d N um ber o f cultures G row th success

Digestion 75/82 (91.5%) 74/75 (99%)

C overslip (explcmt) 7 /82 (8.5%) 5 /7 (71%)

In total, 96% (79/82) of the samples grew successfully. Of the tumour types received all 21 meduUoblastomas, 3 PNETs, 7 ependymomas (including an ependym oblastom a), 3 neuroblastomas, 3 choroid plexus tumours, 2 germinomas and one each of craniopharyngiom a, gangliogliom a, histiocytom a, meningioma and a prim itive n eu ro ep ith elial tumour produced prim ary cultures. In addition, 25/26 low (16/16 pilocytic and 9/10 fibrillary) and 7/8 h ig h grade astrocytom as and 2/3 dysem bryoplastic tumours successfully produced a p rim a ry culture. The four cultures w hich resulted in no grow th were derived from a fib r illa r y astrocytom a (IN I882), a dysem bryoplastic neuroectodermal tumour (IN2089), a h ig h grade astrocytom a (IN2041) and a pineoblastom a (IN1861) respectively. Two of these cultures (IN I882 and IN1861) were sm all samples from stereotactic burr-hole biopsies. Even though the other tum our samples received were from larger resections, these cultures also could h a v e contained necrotic or non-tumour m aterial. From the two p rep arativ e techniques, 99% of sam ples prepared using digestion and 71% of those samples prepared as explant cultures were successful in producing a prim ary culture.

N ine MBs & PNETs (9/24, 37.5%), three pilocytic astrocytom as (3/16, 19%), two fibrillary astrocytom as (2/9, 22%) and one ependymoma (14%) culture produced large quantities of floating aggregates cn initial plating. These aggregates were separated from adherent cells and placed into agar-coated flasks. U nfortunately, none of these flo atin g aggregate cultures survived long-term passaging. The aggregates either clumped together or attached (despite the agar) to the surface of the flask and cells grew out as an ex p lan t culture.

All tum our samples were plated out on conventional cell culture plastic-w are and, i f sufficient m aterial w as available, they w ere also plated on different substrates (see below). A difference between the tum our groups in the time taken to reach passage level 1 on p la stic w as observed and summarised in Table 7. Cultures derived from h igh grade and pilocytic

astrocytomas, ependym om as and neuroblastom as tended to reach passage level 1 m ore quickly

th an those derived from MBs, PNETs and low grade fibrillary astrocytom as. A nalysis using the Student's t-test show ed significant differences betw een a small number of groups, nam ely, betw een, neuroblastom as and MBs (p=0.0076) and fibrillary astrocytom as (p=0.043) in th e tim e taken to reach passage level 1. Interestingly, there was nearly a significant difference betw een the neuroblastomas and the pilocytic astrocytom as (p=0.068). The MBs on average took the longest tim e (2.7 weeks), although some individual cultures took much less time to reach confluency. Neuroblastom a cultures took the shortest time, taking less than 2 weeks to reach passage level one. H igh grade and pilocytic astrocytom as, ependymomas and neuroblastom as tended to produce fast, vigorous cultures overall, w hilst MBs and low grade fibrillary astrocytom as grew more slowly. Interestingly, there was a m arked difference betw een the pilocytic and fibrillary astrocytom as in term s of the tim e taken to reach passage

level 1, however, was found not to be significant.

Table 7: Summary of time taken to reach passage level 1

T um our type N u m b er

Tim e ta ken to reach p assage level 1 (in w eeks) M ean (in w eek s) Pilocytic astrocytom as N=16 1 -5 .5 2.31 Fibrillary astrocytom as N=9 1 -5 .0 2.78

H igh grade astrocytom as N =7 1 -5 .0 2.29

M eduUoblastomas N=21 1 -7 .5 2.70

PNETs N=3 1 -4 .0 2.33

Ependymomas* N=7 1 -3 .5 2.07

Neuroblastom as N=3 1.5-2.0 1.66

(* including the ependymoblastoma)

Initial grow th on different substrates

In attem pting to increase the success rate of establishm ent of prim ary cultures, p a rtic u la rly from meduUoblastomas, a number of cultures were grown on different attachm ent factors. In m ost cases there w as "good" (defined as in itial plating, attachm ent of floating cells and th e early signs of grow th of more than half the cell population w ith in 24 hours) i n it ia l attachm ent in prim ary culture in tumour types on most substrates. How ever, there w as

heterogeneity w ithin tumour groups and between the different groups. O verall, it w as observed th at at least 75% of cultures derived from meduUoblastomas attached w ell to plastic, FN, LMN and vitrogen. In contrast, PLL, M atrigel and C ell-tak were found to be less effective and only 50% of cultures on these surfaces attached well. H ow ever, glass w as found to be a very poor substrate for this group of cultures, w here although spheroids attached quite strongly to the glass base, there was little further growth. For cultures derived from both low and high grade astrocytomas, the num ber of substrates th at resulted in good attachm ent w as greater than compared to meduUoblastomas, in th at 75% of these cultures were observed to attach well on plastic, FN, LMN, PLL, vitrogen, Cell-tak and glass. Possibly the electrostatic nature of PLL and glass is im portant in the attachm ent of cells present in this group of tumours. The only difference between the low and high grades was th a t the h igher grade astrocytom a ceUs grew weU on M atrigel, unUke those from low er grade tumours. For cultures derived from ependymomas, more than 75% of cultures attached well on plastic, FN, LMN and occasionally cn glass. However, less th an 30% of cultures derived from these tumours attached weU on vitrogen, M atrigel and Cell-tak. Thus overall, M atrigel, vitrogen and glass w ere the least effective of substrates overall for all the tumours studied. All cultures grew on at least three other surfaces, indicating th at viable cells must have been present in th e original ceU suspension. M ost of the cultures also m aintained grow th to at least passage le v e l

2 on three or more surfaces.

Sustained grow th on different substrates

The results of growing cells derived from 8 meduUoblastomas, 6 low grade astrocytom as (4

pilocytic & 2 fibrillary), 5 high grade astrocytom as, 3 ependymomas and 1 each of a m alignant fibrous histiocytom a, craniopharyngiom a and neuroblastom a on d iffe re n t

substrates is summarised in Fig 6. It is apparent th a t there were m arked differences in th e

ability of different substrates to m aintain the grow th of particular tumour types beyond passage level 3.

Plastic w as consistently able to m aintain grow th of nearly all tumour types. FN w as found to sustain grow th to the same degree for both low pilocytic and fibrillary astrocytom as

but only supported the grow th of 88% (7/8) meduUoblastomas, 67% (2/3) ependymom as, 80%

(4/5) high grade astrocytomas and a single case of neuroblastom a, craniopharyngiom a and a

maUgnant fibrous histiocytoma. LMN supported the grow th of 75% (6/8) m eduU oblastom as,

67% (2/3) ependym om as, 83% (5/6) low (3/4, 75% pilocytic and 2/2 100% fibrillary) and 80% (4/5) high grade astrocytomas and the same tum ours from the misceUaneous group (as for F N above). A lthough LMN was not capable of supporting grow th to the same extent as plastic or FN, there w as little evidence of this being influenced by tum our histology.

O verall, PLL was found not to be an appropriate substrate for sustained grow th of these tum ours in culture, b u t there were a few exceptions. PLL appeared to support grow th of

MeduUoblastomas

Ependymomas

Pilocytic astrocytomas

Fibrillary astrocytomas

High grade astrocytomas

Miscellaneous

Fig 6

P e rc e n ta g e grow th su ccess of p ae d ia tric s h o rt-te rm c u ltu res beyond p as sag e level 3 w h en grow n on different su b strates

—i— 20 —T- 40 — I--- r 60 —T— 80 Ke y □ Plastic □ Fibronectin □ Lamlnin □ Poly-L-lyslne □ Cell-tak □ Matrigel □ Vitrogen ■ Glass 1 0 0 P e r c e n t a g e

both low (83%, all 4 pilocytic tumours, but only one of the two fibrillary) and h ig h (60%) grade astrocytom as b u t to a lesser extent than plastic or FN. The grow th of m eduU oblastom as (50%) and ependymomas (33%) was less successful. It appeared th at low grade pilocytic astrocytom a cultures did well cn this substrate. For example, four such cultures (IN1520, IN I524, IN1533 & IN1591) grew to passage levels 7, 7, 11 and 10 respectively. A culture

derived from each of a MB and a craniopharyngiom a both grew to passage level 8. In contrast,

the neuroblastom a culture (IN1634) did not grow on this substrate.

C ell-tak proved a poor substrate for the grow th of most kinds of childhood b rain tum ours, although a smaU proportion m anaged to grow beyond passage level 3. H ow ever, th e grow th was slow in these cases. On C ell-tak, cultures IN1594 (ependym oma), IN1601

(meduUoblastoma) and IN1634 (neuroblastoma) grew to passage level 7, 9 and 6 resp ectiv ely

giving rise to vigorous grow th and relatively high yields on seeding.

A smaU num ber of cultures grew well on vitrogen including IN1566 (HGA), IN1591 and

IN1533 (LGA, pilocytic) and IN1601 (MB), w here each grew to passage level 8, 10, 4 and 6

respectively. A single ependym om a culture (IN1497) grew to passage level 4 on vitrogen. There were a number of technical difficulties using these grow th substrates, p articu larly M atrigel and vitrogen. W ith M atrigel, prevention of a gel forming on th a w in g was difficult as well as keeping other cell culture disposables at 0°C (such as pipettes and flasks). A nother problem w ith M atrigel was th a t during routine passaging it w as often difficult to retrieve the cells from the gel even using the recommended enzyme dispase. Consequently, this resulted in cell yields w hich w ere 1.3 - 2.9-fold less, w ith a m edian of 1.5- fold cn M atrigel, compared to plastic. Vitrogen was an adequate substrate only if it w as com pletely dry but this was sometimes difficult to achieve. In addition, as w ith a g a r (probably because vitrogen film had not dried efficiently), cells h a d a tendency to grow underneath the film rath er than on it. Again there was often a reduced h arv est of cells cn passaging, so that cell yields were in the order of 3.5x less cn vitrogen than the same culture grow n on plastic. The difference in cell yield ranged between 1.3-18-fold less, w ith a m edian of 1.9-fold cn vitrogen compared w ith plastic. How ever, the three surfaces plastic, FN and LMN increased initial grow th and prom oted sustained grow th w ith hig h er cell yields cn seeding and consequently were used in subsequent culturing of new tumours. For most of th e cultures from the four major groups, there w as little difference in cell yields betw een th ese three surfaces, although there were individual cultures th at h a d h igher cell yields cn one particular substrate. For example, 50% of h igh grade astrocytom a-derived cultures h a d a 2- fold hig h er cell yield cn plastic compared w ith th at cn FN and LMN in the first th re e passage levels, b ut in later passages this difference was no longer apparent. Less than 50% of ependym om a cultures cn FN and LMN produced cell yields between 2-4-fold and 2.5-3-fold greater compeired w ith plastic respectively. For the low grade astrocytom a and m eduUoblastoma cultures less than 20% produced greater numbers of cells cn FN and LMN

com pared w ith plastic. As w ith the high grade astrocytom a and ependymoma cultures th is difference was no longer evident at later passages. In summary, the overall success rate of sustained grow th of cultures cn M atrigel, PLL and vitrogen was poor, but was much better on plastic, FN and LMN. The other substrates, glass and Cell-tak, appeared to be of interm ediate benefit.