February 2019 Edition. Volume XIX

The following material was kindly provided to the Burton ReportÒ by Carl Eden who contacted the Swedish National Board of Welfare to inquire as to the reasoning behind their having rejected Pantopaque/Myodil and their views on water-based contrast agents. This communication (reproduced as written) serves as another significant contribution regarding the global history of oil myelography.

Dear Mr Eden,

I am sorry it took so long to answer your questions but we have consulted a number of various specialists. The following answers come from Melker Lindqvist, associate professor of neuroradiology, Department of Neuroradiology, Karolinska Hospital, S-171 76 Stockholm. We would appreciate very much if your company could pay us for the costs of consulting professor Lindqvist. We have paid him 15.000 SEK for the many hours he has spent investigating these complicated matters, which have resulted in the very valuable information below. The amount 15.000 SEK should be put inte bankaccount number 1 56 16-6, belonging to Postgirot Bank Sweden. Please also note “The national board of health and welfare, unit 335, project 72 11 13.”

From a paper in the American Journal of Neuroradiology (AJNR) in 1995 (ref. 18) about  the Stockholm School of Neuroradiology  by the two famous Swedish neuro-radiologists, Professor Erik Lindgren and Professor Torgny Greitz, the following paragraph is quoted: “The Stockholm School took a stand against the use of these (oil based) contrast media at an early stage, and they were hardly ever used in Sweden. Considering the number of postmyelographic arachnoiditis that are now reported, this may seem to have been a wise decision.”

Instead of oil based contrast agents gas (air or oxygen) was used in Sweden for ventriculography and encephalography for investigation of intracranial lesions. Gas myelography was used for investigation of the spinal cord, and from the late 1940´s most centres in Sweden used the ionic water soluble contrast medium methiodal sodium (Abrodil, Kontrast U) for lumbar myelography. The same techniques were used in most centres in the other Scandinavian countries and to some extent in Germany and France. The rest of the world used almost exclusively the oil based contrast agent iophenylate (Pantopaque, Myodil) for both intracranial and spinal examinations.

What were the reasons for these fundamental differences in neuroradiologic technique ? Why did the Swedish radiologists avoid oil based contrast media? Or, why were not gas myelography, pneumoencephalography and lumbar myelography with water soluble contrast agent accepted as alternatives to examinations with oil based media in the rest of the world ? The interesting question seems to be: Could the large number of postmyelographic arachnoiditis, that have been reported over the years, have been avoided ?

The general view of the Stockholm School of Neuroradiology on these issues are presented in a book chapter by Lindgren (ref. 13) and in the above mentioned paper by Lindgren and Greitz (ref. 18).

Questions 1 – 2:

1. Why did Sweden never use oil based contrast agents ?

2. If Sweden believed that there was a danger with oil based contrast agents what research did it perform to verify this?

The risk of arachnoiditis with oil based contrast media. In the 1930´s and 1940´s the department of radiology at the Serafimer Hospital in Stockholm was headed by the famous neuroradiologists Erik Lysholm and Erik Lindgren working together with an equally famous neurosurgeon, professor Herbert Olivecrona. Patients from all over the world were referred to professor Olivecrona, and many of these patients had been examined elsewhere with oil based contrast media. During the radiological examinations of these patients at the Serafimer Hospital the neuroradiologists noted remaining contrast medium in the subarachnoid space, both in the head and in the spinal canal. The oil deposits were often fixed (did not move with change of the patients position), and they were also associated with local thickening of the meninges. This was interpreted as arachnoiditis caused by the contrast medium, and in a few cases this arachnoiditis caused difficulties in the removal of a tumor at surgery (ref. 13).

In 1928 Odin, Rundström and Lindblom published a supplement of Acta Radiologica entitled “Iodized oils as an aid to the diagnosis of lesions of the spinal cord and a contribution to the knowledge of adhesive arachnoiditis” (ref. 25). These authors, working in Sweden, used their own home-made oily media out of soya and sesame oils in addition to Lipiodol. Their preparations made out of sesame oil had less adverse effect than Lipiodol. They observed clinical signs of acute meningeal reaction similar to acute meningitis. However, they did not perform a long term follow up of their patients.

Except for this supplement (ref. 25) no scientific studies seem to have been performed in Sweden to test the early or late adverse effects of the oil based contrast media. Some papers on late and persisting sequelae from the use of Myodil and Pantopaque started to appear in the international literature in the 1940´s and 1950´s: Tarlov 1945 (ref. 29), Luce 1951 (ref. 19), Hurteau 1954 (ref. 7), Davies 1956 (ref. 4). A case of death following Pantopaque myelography was reported by Erickson et coll. in 1953 (ref.6).

Thus the decision already around 1935 by the leading Swedish neuroradiologists not to use oil based contrast media was mainly based on clinical observations among their own patients rather than on published or unpublished scientific examinations. As an example of this type of clinical observations the following  case was reported by Lindgren in a lecture in 1984. The manuscript of this lecture was published in Swedish by Nycomed, Stockholm in 1994 (ref. 17). (The quoted part is translated by Professor Lindqvist.)

“It may seem strange that oily based contrast media, especially Pantopaque, were used to such a large extent (especially in America), since these contrast media doubtless caused arachnoiditis, although the clinical importance of these changes were and are debated. We had a nurse with some vague symptoms and performed a pneumoencephalography on her with negative result. She was convinced that she had an intracranial lesion, so she went to America, where she was examined with positive contrast medium, also with negative result. She came back to Sweden, and after two years she developed serious symptoms. We performed a new pneumoencephalogram. No air entered the intracranial space in a normal way. An air ventriculography was performed and showed a dilated ventricular system, including the fourth ventricle. At surgery extensive adhesions were found in the posterior fossa. In my opinion it is thus clear that positive contrast media, at least in some cases, can cause clinically significant arachnoiditis.”

What other reasons, except for the risk of arachnoiditis, contributed to the Swedish decision ?

The physical properties of the oil based media were not ideal. They did not mix with the cerebrospinal fluid (CSF), had high viscosity and had a tendency to be split up in droplets. For these reasons they did not fill out narrow spaces, for instance the subarachnoid pockets surrounding the spinal roots and the narrow parts of the intracranial subarachnoid cisterns, and thus did not allow visualisation of fine anatomical details.

The Swedish neuroradiologists were masters of air studies and claimed that air encephalography and ventriculography in the hands of skilled neuroradiologists had proved capable of giving equally good results as examinations performed with oil based media (ref. 17). Several publications on neuroradiologic investigations with gas (air or oxygen) emanated from “The Swedish School of Neuroradiology” at the Serafimer Hospital. Examples of important papers are: Lysholm et coll. published “Das Ventrikulogramm” part I -III 1931 – 1937 (ref. 20-22). Lindgren published “On the diagnosis of tumors of the spinal canal by aid of gas myelography” in 1939 (ref. 15), “A pneumographic study of the temporal horn with special reference to tumors in the temporal region” in1948 ( ref. 16) and “Some aspects on the technique of encephalography” in 1949 (ref. 14).

The neuroradiologists at the Serafimer Hospital became famous for their clinical and scientific achievements and attracted many colleges from abroad who came to the Serafimer Hospital for training. One of them was James Bull from London, and he made the following statement at the Symposium Neuroradiologicum in Rotterdam in 1949: “There is no doubt that air or oxygen is the contrast material of choice. The degree of success one obtains with air is an index of ones ability as a neuro-radiologist. In my view the more often one falls back on positive contrast material, the less capable one proves oneself as a neuroradiologist.” (ref. 13).

The organisation of Swedish radiology. The first and main prerequisite for the international success of Swedish radiology was the fact that independent departments of radiology were created already when diagnostic radiology was introduced in Sweden. The man who had the main responsibility for this was Gösta Forssell, the first director of the Roentgen Institute at the Serafimer Hospital in 1908. He managed to convince the authorities that each hospital should have only one x-ray department, and the heads of roentgen departments in Sweden should have the same position as other clinical heads. (In most European countries, as well as in the US, each clinical speciality had its own small roentgen department, usually run by a clinician.) The Swedish organisation meant that specially trained radiologists took the full responsibility not only for the “interpretation” of the images, but also for the choice of technique, including the risk of complications involved. (ref. 18). For this reasons Swedish radiologists were in charge of all parts of the diagnostic procedures, for instance lumbar puncture and the injection of air for the air studies and all parts of the angiographic procedures.

Technical development. The independent status of the radiology department at the Serafimer Hospital gave the radiologists the opportunity not only to develop examination technique, but also to work together with the local industry in developing new technical equipment for radiology. To give just one example: Lysholm, in co-operation with the Elema Schoenander company, in 1931 constructed the “Lysholm skull table” that for many years was used for encephalography and angiography.

Question 3:

Has Sweden found any problems with water based contrast agents ?

Gas myelography was not a good technique for the diagnosis of lumbar disk disease because the air did not give enough contrast. Oil based media had the physical disadvantages mentioned above and were thus not ideal. The Swedish radiologist Arnell had been experimenting with the water soluble contrast medium methiodal sodium (Abrodil, Kontrast U) and finally reached the conclusion that Abrodil could be used for lumbar myelography (ref.2). However it was so irritating that it had to be combined with spinal anaesthesia, and even so the medium was so toxic that it was not allowed to reach the spinal cord. Thus the examination had to be restricted to the lumbar region, but in that region it proved to give excellent visualisation of the lumbosacral nerve roots and their sleeves. Improved technique for lumbar myelo-graphy with Abrodil was published by Lindblom in 1946 (ref.11). For Swedish radiologists the need for spinal anaesthesia, before the injection of the contrast medium, was not regarded as a big problem since they were in charge of the whole procedure. The contrast medium was excreted via the kidneys, and no residual contrast medium could be seen in the spinal canal after one day. In 1937 Lindblom published an investigation of complications of myelography by Abrodil (ref. 12). The investigation was based on 721 cases collected from several Swedish hospitals. Several cases with complications were noted, for instance drop in blood pressure with shock, lumbar pain, spasm in the legs and even 3 cases with “secondary shock, and paralysis of the legs and sphincters for weeks.” However, no complications with persistent neurological symptoms were found among these 721 patients.

In spite of the limitations Abrodil had such great advantages for the diagnosis of lumbar disk disease that it soon became the medium of choice for lumbar myelography. For many years permanent complications to myelography with methiodal sodium were unknown. In 1956 Munroe (ref. 24) and in 1959 Söderberg et coll. (ref. 28) published a few cases with cauda equina syndrome after myelography with methiodal sodium. However, their papers were not published in the radiological literature, and for a long time their observations remained unknown to the radiologists. In 1972 came the first publication (ref. 3) on adhesive arachnoiditis after myelography with ionic water soluble contrast media, and since then several papers have been published with reports on adhesive arachnoiditis after water soluble contrast media (ref. 1,5, 8, 9, 10, 26, 27 ). By that time the new nonionic contrast medium metrizamide was already on its way, and methiodal sodium was withdrawn from the market in 1974.

Since the introduction of the water soluble nonionic contrast media metrizamide (Amipaque) and later iohexol (Omnipaque) complications of clinical importance are practically unknown.

Swedish neuroradiologists were strongly opposed to oil based contrast media because of their tendency to give arachnoiditis, and still they were so late to discover the same type of complications caused by methiodal sodium (Abrodil, Kontrast U) which was used extensively in Sweden for many years. The explanation for this seems to be that the water soluble media, as opposed to the oil based, disappear completely from the subarachnoid space. If a patient was examined a second time there was thus nothing in the pictures that indicated that the arachnoiditis like changes that had developed since the previous examination were caused by the contrast medium. The incidence of postmyelographic arachnoiditis was also low with methiodal sodium as compared both to oil based media and to later introduced water soluble ionic contrast media like Conray and Dimer-X.

Question 4:

What is Sweden´s position on epidural steroid injections? Are they considered safe? Conclusions from Professor Melker Lindqvist.

“The question about the safety of epidural steriod injections is difficult for me, as a neuroradiologist, to answer. I have consulted Björn Meyerson who is professor of neurosurgery with treatment of chronic pain as his special interest. He in turn has discussed the problem with associate professor Staffan Arnér, head of the pain treatment clinic at the Karolinska Hospital. They agree that the beneficial effect of epidural steroid injections in patients with arachnoiditis and similar chronic pain syndromes can be questioned, and this type of treatment is not much used in Sweden. However, except in the case of inadvertent injection into the subarachnoid space, they do not think this treatment is harmful. In an extensive report in 2000 on “Back and neck pain” by the Swedish Council on Technology Assessment in Health care (SBU Report 145) the conclusions seem to be the same.”

With kind regardsMarianne Thorén Principal Administrative Officer The Division of Medical Quality Development The Department of Health and Medical Services The National Board of Health and Welfare S-106 30 Stockholm SWEDEN


1. Ahlgren P: Long term side effects after myelography with water soluble contrast media: Conturex, Conray meglumin 282 and Dimer-x. Neuroradiology (1973), 206

2. Arnell T et coll.: Myelography with skiodan (Abrodil). Acta Radiologica 12 (1931), 287

3. Autio E et coll: Adhesive arachnoiditis after lumbar myelography with meglumine iothalamate (Conray). Acta Radiological Diagnosis 12 (1972), 17

4. Davies FL: Effect of unabsorbed radiographic contrast media on the central nervous system. Lancet 2 (1956), 747

5. Dullerud R et coll.: Adhesive arachnoiditis after lumbar radiculography with Dimer-X and Depo-Medrol. Radiology 119 (1976), 153

6. Erickson TC et coll.: Late meningeal reaction to ethyl idophenylundecylate used in myelography. Report of a case that terminated fatally. JAMA 153 (1953), 636

7. Hurteau EF et coll.: Arachnoiditis following the use of iodized oil. The Journal of Bone and Joint Surgery 36-A (1954), 393

8. Irstam L et coll.: Water-soluble contrast media and adhesive arachnoiditis. I. Reinvestigation of nonoperated cases. Acta Radiologica Diagnosis 14 (1973), 497

9. Irstam L et coll.: Water-soluble contrast media and adhesive arachnoiditis. II. Reinvestigation of operated cases. Acta Radiologica Diagnosis 15 (1974), 1

10. Irstam L et coll.: Water-soluble contrast media and adhesive arachnoiditis Acta Radiologica Diagnosis 15 (1974), 356

11. Lindblom  K: Lumbar myelography with Abrodil. Acta Radiologica 27 (1946), 1

12. Lindblom K: Complications of myelography by Abrodil. Acta Radiologica 28 (1947), 69

13. Lindgren E: A history of neuroradiology. In: Newton, Potts, eds. Radiology of the skull and brain. Skull. Vol 1, Book 1. St Louis: C.V.Mosby, 1971:1-25

14. Lindgren E: Some aspects on the technique of encephalography. Acta Radiologica 31 (1949), 161

15. Lindgren E: On the diagnosis of tumors of the spinal cord by the aid of gas myelography. Acta Chirurgica Scandinavica 82 (1939), 303

16. Lindgren E: A pneumographic study of the temporal horn with special reference to tumors in the temporal region. Acta Radiologica, suppl. 69 (1948)

17. Lindgren E: En tillbakablick på neuroradiologiens tidigaste utveckling. (In Swedish) Nycomed, Stockholm 1994

18. Lindgren E and Greitz T: The Stockholm school of neuroradiology. AJNR 16 (1995), 351

19. Luce JC et coll.: Pantopaque meningitis due to hypersensitivity. Radiology 57 (1951), 878

20. Lysholm E et coll.: Das Ventrikulogramm, part 1. Acta Radiologica, suppl. 24 (1931) 1

21. Lysholm E et coll.: Das Ventrikulogramm, part 3. Acta Radiologica, suppl. 25 (1935) 1

22. Lysholm E et coll.: Das Ventrikulogramm, part 2. Acta Radiologica, suppl. 26 (1937) 1

23. Mason MS et coll.: Complications of Pantopaque myelography. Case report and review. Journal of Neurosurgery 19 (1962), 302

24. Munro D: Lumbar and sacral compression radiculitis (herniated lumbar disk syndrome). The New England Journal of Medicine 254 (1956), 243

25. Odin M et coll.: Iodized oils as an aid to the diagnosis of lesions in the spinal cord and a contribution to the knowledge of adhesive arachnoiditis. Acta Radiologica, suppl. VII, 1928

26. Rådberg C et coll.: Late sequelae following lumbar myelography with water-soluble contrast media. Acta Radiologica Diagnosis 14 (1973), 507

27. Skalpe IO: Adhesive arachnoiditis following lumbar radiculography with water-soluble contrast agents. A clinical report with special reference to metrizamide. Radiology 121 (1976), 647

28. Söderberg L et coll.: Neurological complications following myelography with water soluble contrast medium. Acta Ortopaedica Scandinavica 28 (1959), 220

29. Tarlov IM: Pantopaque meningitis disclosed at operation. JAMA 129 (1945), 1014

30. The Swedish Council on Technology Assessment in Health Care. SBU Report 145, May 2000. Nachemson A ed: Back and neck pain.

Burton Report is an independent and non-commercial internet journal which was first published on January 1, 2000 and is dedicated to the principle that health care and the health care process MUST reflect truth and integrity as well as the best interests of the patient.

The information presented in Burton Report is intended for dissemination without alteration.

© Burton Report® 2000-2019, All Rights Reserved