G A S T R O C Y S T O M A O F T H E S P I N A L C O R D 635

granulation tissue (Fig. 799), while the deeper part of the The formation of peptic ulcers in jejunal and ulcer was composed of fibrous tissue, which had colonic transplants is not surprising, as it is well completely replaced the muscularis (Fig. 800). The known that the mucosa of these two organs is less arteries showed marked endarteritis obliterans (Fig. ~oI), resistant to hydroch~oric acid than that of the while signs of healing were seen at the edge of the ulcer in the form of epithelium growing over the granulation stomach. tissue (Fig. 802). The changes were those of a chronic benign peptic ulcer, typical in every respect except that it affected the colon.

1 am indebted to Professor J. c. Goligher for permission to publish this case ; to Professor W. D.

FIG. 8or.-Artery showing endarteritis obliterans. (H. and E.) FIG. 8oz.-Epitheliurn growing over granulation tissue. ( x 60.) (H. and E.) ( x 250.)

DISCUSSION Jejunal and colonic replacements are designed to

overcome post-gastrectomy ' dumping ' symptoms. Henley (1953) and Moroney (1953) report that both these procedures reduce the incidence and severity of this condition, although the former author records the formation of an ulcer in the jejunal transplant in 4.1 per cent of cases, while the latter records an ulcer in the transplant in z out of 150 cases of colonic replacement.

Newcomb for helpful criticism j and to Mr. A. Beasley and the Photographic Department of St. Mary's Hospital Medical School for technical assistance.

REFERENCES HENLEY, F. A. (1953)~ Ann. R. Coll. Surg., Engl., 13,

MORONEY, J. (1953), Zbid., 12, 328. 141.

GASTROCYSTOMA OF THE SPINAL CORD

BY GEOFFREY KNIGHT, TREVOR GRIFFITHS, AND IOLO WILLIAMS, LONDON

CONGENITAL intradural turnours of the spinal cord other than dermoids and epidermoids are rare. In 1931 Hosoi collected from the literature 10 cases of teratoma occurring in patients of all ages. In 1951 Furtado and Marques discussed 21 recorded examples of intraspinal teratoma, of which 10 occurred in patients over the age of 10 years, the remainder arising in younger subjects. The majority of these recorded examples have contained tissue elements derived from only z of the primary germinal layers, ectodermal and mesodermal. In only I case,

recorded in 1934 by Puussepp, has intestinal epithelium been recognized within the tumour. Hence it would appear that an example of intra- medullary gastrocystoma arising in a child of z is unique and worthy of report.

It is curious that intraspinal teratomas should be so few in number, for the location of teratomas is almost always median in relation to the axis of the body, a fact to be explained by their origin from the embryonic stem and its invaginated derivatives, and yet, in relation to the spinal axis they are usually

636 T H E B R I T I S H J O U R N A L O F S U R G E R Y

seen only as extraspinal and extradural tumours in the sacral and sacrococcygeal regions.

In reviewing the recorded examples of the intraspinal group the following facts emerge : Intraspinal teratomata may occur in association with other congenital abnormalities such as spina bifida and polydactylism or with abnormalities of the spinal cord such as syringomyelia. They are distributed at

FIG. Ro3.-Anteroposterior radiograph and line drawing and dorsal spine.

almost all levels in relation to the spinal axis, the terminal portion of the column being the site of preference. They are only rarely intramedullary in relation to the cord, but may be either intradural and adherent to the cord or extradural with an intradural extension. Teratomatous tumours in general are cystic, the cyst being variable in size and containing either whitish fluid, turbid from the presence of degenerating cells and m u c h or inspissated caseous material. The histological structure is characterized firstly by the disordered grouping of the various tissue elements, which are distributed without the slightest attempt at organization into layers. The majority of the tumours contain tissues derived from two germinal layers only, being bidermoid or teratoid rather than true teratomas ; as Hosoi (1931) and Bucy and Buchanan (1935) point out, the endoderm is the germinal layer which is not represented among the tissues present, all the structures commonly met with being of ectodermal or mesodermal origin. Skin fragments, stratified epithelium, sebaceous and sudoriferous glands, portions of nervous tissue, bundles of neuroglial tissue, choroid epithelium, and ganglion cells have been observed together with muscle-fibres, cartilaginous and osseous tissue, and

structures which are hard to identify, including cysts lined with columnze ciliated and columnae and cuboidal cells, and frankly embryonal tissues. Many varieties of mature and embryonal tissue are found in varying stages of differentiation arranged in a disordered array.

From this review of previously recorded teratomas it would appear that the present case of gastrocystoma

is unique and is, we believe, the only recorded example of such a lesion. I t differs from the recorded examples of teratoid tumour in many respects j in the remarkable organization of the cyst into orderly inner and outer layers, in the high degree of maturity of the glandular tissue which gives the appearance of full differentiation, and capacity of function. I t would appear to be not a teratoma but a congenital ectopic cyst which has been extruded into the spinal canal through an area of anterior spina bifida in the vertebral bodies, and may, or may not, be related to a corresponding tumour in the mediastinum.

CASE REPORT G. H., aged I year 3 months, was

admitted to the South-East Metropolitan Regional Neurosurgical Centre at the Brook General Hospital on May 21, 1953. At the age of 10 months the child had been investigated at a Children’s Hospital on account of fretfulness, cervical rigidity, and tension in the anterior fontanelle. Tuberculous meningitis was suspected, but lumbar puncture had yielded a clear colourless fluid containing I polymorph and 10 lymphocytes per c.mm. Protein 400 mg. per cent, sugar 40 mg. per cent, and chlorides 680 mg. per cent.

of cervical No organisms or tubercle bacilli had been isolated and a tentative diagnosis of meningo-encephalitis having been made,

the child was discharged on May 16. However, the child remained irritable and fretful and failed to sit up, and on account of these persistent symptoms was admitted to the Regional Neurosurgical Centre for investigation.

ON EXAMINATION.-The child screamed when disturbed, but exhibited no meningeal signs. It could not be brought into the sitting position without support. The arm reflexes were normal, but sustained ankle clonus and bilateral extensor plantar responses were demonstrated. Lumbar puncture showed a faintly xanthochromatic fluid containing I cell per c.mm. and protein 2 mg. per cent. Examination of radiographs taken at the time of admission to the Children’s Hospital showed extensive abnormalities in the lower cervical and upper dorsal spine. Further films, including lateral tomographs, were requested, and Dr. Trevor Grifliths reported as follows :

X-ray Report.-The greater part of the dorsal spine showed gross abnormalities ; these are of such a degree that identification of the individual vertebrae is some- what uncertain. Thc abnormalities include hemi- vertebrae and partially fused vertebrae. The bodies of the first to the sixth dorsal vertebrae all exhibit varying degrees of fusion to each other. In the fourth and fifth dorsal vertebrae this fusion is complete. In addition, the 7th cervical and Ist, znd, 4th, and 5th dorsal vertebrae show incomplete fusion of the lateral halves of the vertebral bodies. In the case of D.I and D.2 the

G A S T R O C Y S T O M A O F T H E S P I N A L C O R D 637

development of the lateral halves of the bodies is not of vertebra. An aspirating needle was introduced into the equal degree on each side. These abnormal vertebral substance of the cord and a cyst containing 3 C.C. of bodies are broader than normally developed vertebrx, milky fluid was aspirated, producing a sufficient reduction so that the distance between their pedicles is necessarily of tension to enable the dura to be opened with safety. greater than normal. Several of the pedicles arc, On opening the dura the cord was seen to be widely

expanded, with cystic cavities lying superficially within the dorsal portion of its substance.

Careful examination of the lateral aspects of the cord failed to reveal any plane of demarcation between the cord and the tumour-indeed, the space between the line of emergence of the anterior and posterior roots was increased, as it is with intramedullary tumours. Accord- ingly, a vertical incision was made between the posterior columns strictly in the midline of the cord at the site of the cystic areas. At a depth of a few millimetres a strawberry-coloured tumour was revealed occupying the midline of the expanded cord (Fig. 805). The two halves of the cord which passed down on either side of the mass were lightly adherent to its surface at one or two zones and were readily separated from these points of light attachment by soft dissection. Around the greater part of the circumference of the tumour the cord tissue was separated from the growth by cyst formation which had occurred between the tumour surface and the cord substance, and was easily separated, revealing a tumour arising in connexion with the bodies of the vertebra and projecting backwards to become enwrapt within the cord. The vessels entering the anterior aspect of the tumour from the vertebral bodies were secured with silver clips and the tumour was rotated upwards from its bed and delivered.

After a period of pyrexia and urinary retention the child regained movement in the left leg on July 8 and both legs on July 13 and passed urine spontaneously on July 23. He was moving both legs freely on Aug. 25. At the age of I year and g months the child was sitting up easily in

The post-operative recovery was uneventful.

bed, moving both legs and attempting to pull himself up

spine. plantar response was flexor, the left was extensor. In FIG. goq.-Lateral tomograph and line drawing of upper dorsal by the Ankle clonus had disappeared, the right

FIG. Eos.--Photoarauh of oval tumour which was strawberrv coloured, ?& cm. in longest diameter and 1 3 cm. in cross-sectioA at its widest part.

however, thinned, some of them to such an extent that they cannot be identified. There is an anterior spina bifida involving the 5th dorsal vertebra. The lateral view shows that the posterior surfaces of the vertebral bodies in the affected region are scooped out. This finding was confirmed by tomography. The vertebral arches appeared to be intact at all levels. (Figs. 803,804.) The radiographs demonstrate :-

I . Developmental abnormalities of the vertebrae between C.7 and D.7.

2. Gross enlargement of the neural canal in much of this region.

AT OPERATION (June 23, 1953).-A cervicodorsal laminectomy was performed under general aniesthesia. The dura was found to be widely expanded and stony hard from the level of the 7th cervical to the 7th dorsal

FIG. 806.-Section through tumour showing cystic cavity surrounded by smooth muscle. At site A there is gastric mucosa and site B central nervous tissue, part of the central nervous tissue and not part of the tumour itself.

view of the anterior spina bifida noted in the 5th dorsal vertebra, special investigations were performed to establish whether the intraspinal mass arose in connexion with a corresponding mediastinal tumour. RadioIogicaI investigations with special reference to the mediastinum failed to reveal any tumour within the thorax.

PATHOLOGICAL REPoRT.-Dr. 1010 Williams reports as follows : The tumour on removal from the spinal cord was strawberry coloured and approximately oval in shape being 34 cm. in its longest diameter and 13 cm. in cross- section at its widest part (Fig. 805). On cut section it was cystic, with a white opaque irregular lining and a surrounding tissue of darker colour (Fig. 806). The fluid contained in the cyst and removed by aspiration during the operation was 20 ml. in volume and of an opaque milky appearance. Spun deposit of the fluid

638 T H E B R I T I S H J O U R N A L O F S U R G E R Y

showed the milkiness to be due to amorphous cellular debris and occasional leucocytes. There was no evidence of an inflammatory exudate and no tissue cells were demonstrated.

Histological examination showed the cystic space to be lined by upper alimentary epithelia which is thrown into folds in places and to be surrounded by smooth muscle with a mass of central nervous tissue at its margins but

if this tumour is a teratoma ; the diagnosis could only be sustained if sections of other parts showed other kinds of tissue. These tissues could well be due to a simple developmental heterotopia since they are indigenous to this particular level of the spinal cord. If I am right in supposing the lesion to be the result of a developmental dislocation of the alimentary canal it is strictly comparable with the entrocystomas and the rarer gastrocystomas of

FIG. Eo7.--Section of tissue from site A showing gastric mucosa with gastric glands containing oxyntic and chief cells. ( X 72.) epithelium. ( x 72.)

FIG. SoB.--Section of tissue showing stratified epithelium resembling that of c~ejophagus continuous with which is ciliated

mainly at one pole, the latter being probably part of the spinal cord and not a part of the tumour itself. The epithelia includes gastric mucosa with gastric glands containing parietal and chief cells (Fig. 807), stratified epithelium resembling that of the oesophagus, continuous with which there is some ciliated epithelium which is probably also Esophageal, the esophagus during develop- ment, being lined by ciliated epithelium (Fig. 808). The gastric mucosa contains superficial mucus-secreting columnar cells as shown by P.A.S. staining, with chief and parietal cells in the body of the glands. Haematoxylin and eosin stains show parietal cells with centrally disposed rounded nuclei and faintly eosinophilic cytoplasm of triangular appearance, one of the sides being related to the basement membrane. The chief cells are low columnar in type, with oval nuclei and a blue reticulated cytoplasm which tends to border the oxyntic cells, preventing their access to the lumen (see Fig. 807). There is no area of rapid growth or invasion in any part of the tissues studied. Serial sections through one half of the tumour show no evidence of any other heterotopic tissues. The tumour was originally considered to be a cystic intraspinal teratoma but the sections were submitted for consideration to Professor R. A. Willis. Professor Willis in a personal communication writes. “ I doubt

the abdominal cavity and mediastinum. I do not know of any previously reported case of such a malformation in the spinal cord.”

COMMENT It is possible on theoretical grounds that the

endodermal developing gut became nipped off before the fusion of the scleroderms. T h e ectopic gut growing within the neural canal would thus interfere with the development of the vertebrae at the same level, accounting for the abnormalities of formation of the vertebrz demonstrated in this case.

BIBLIOGRAPHY Bucu, P. C., and BUCHANAN, D. N. (r935), Surg. Gynec.

FURTADO, D., and MARQUES, V. (1951)~ J . Neuropath., 10,

Hosor, K. (193I), Arch. Path. (Lab. Med.), I I, 875. INGRAHAM, E., and BAILEY, 0. (1946), J . Neurosurg., 3,

PWSSEPP, L. (1g34), Rev. neural., 41, 879.

Obstet., 60, 1137.

384.

511.