Tracing the origins of autism: a spectrum of new studies.

The etiology of a medical condition might seem an unlikely subject to arouse intense feelings. Yet few medical disorders have stirred up as much passion and divisiveness among scientists and the general public as autism has in recent years. The heat of the controversy has even attracted attention from periodicals such as The Wall Street Journal, the Columbia Journalism Review, and Wired magazine—seemingly improbable forums for a medical debate. Why all the furor? 
 
At the eye of the storm is the startling climb in the numbers of children who have been diagnosed with one of the autism spectrum disorders (ASDs). The most severe ASD is autistic disorder (which often is called simply “autism”); other forms include Asperger syndrome and the much rarer childhood disintegrative disorder. In the United States, the diagnosis of ASDs increased roughly 10-fold over the course of a decade, from 4–5 children per 10,000 in the 1980s to 30–60 children per 10,000 in the 1990s, according to a report in the August 2003 Journal of Autism and Developmental Disorders. The 5 May 2006 issue of Morbidity and Mortality Weekly Report describes the results of two parent surveys from 2003 and 2004, which suggested that 55–57 children per 10,000 had autism (however, an editorial note points out that, due to the nature of the surveys, parents of children with other ASDs may have reported their children as having autistic disorder). 
 
Some scientists believe that much of the upsurge is the result of increased awareness of ASDs or changes in diagnostic criteria, which would suggest that the true prevalence of the disorders has been stable over time. Others disagree. “It is premature to state that there is no increase in prevalence,” says W. Ian Lipkin, a professor of neurology, anatomy, and neurobiology at Columbia University. “None of the studies to date has been designed to definitively address the issue.” 
 
The prevalence of ASDs plays into the fundamental question of what causes these disorders. If the number of cases is truly on the rise, then it would seem likely that some change in the environment is driving up the total. That’s partly what has divided scientists into opposing camps—they cannot agree on the relative importance of genetic and environmental factors in the disorders’ etiology. 
 
Alas, answering the prevalence question might not end that debate. “Even if the prevalence of autism were stable,” says Lipkin, “you would not be able to rule out the possibility of an environmental trigger.” That’s because very little is known about the mechanisms that cause autism, be they environmental or genetic. 
 
“The study of autism was, until recently, largely dominated by the field of psychology, where characterizing the behaviors and developing reliable instruments for diagnosis have been major areas of research over the past few decades,” says Irva Hertz-Picciotto, an epidemiologist at the University of California, Davis. 
 
Indeed, the core symptoms of ASDs—social disinterest, repetitive and overly focused behavior, and problems in communication, usually appearing before 3 years of age—have been well described. Much less research has focused on the causes of these symptoms. 
 
Several investigations dating back to the 1970s indicate that identical twins have a much higher concordance rate of ASDs than fraternal twins, according to a report in the Spring 1998 issue of Mental Retardation and Developmental Disabilities Research Reviews. Those studies provide some of the best evidence that these disorders have a strong genetic component. But the identity of the genes involved, much less how they produce ASDs, has not been established. Moreover, the concordance rate for identical twins is not 100%, which suggests that at least some cases must be associated with environmental or epigenetic factors. 
 
A few cases of ASDs have been clearly linked to environmental insults. These include prenatal exposure to chemical agents such as thalidomide and valproic acid, as well as to infectious agents such as the rubella and influenza viruses. Here again, the concordance rate is not 100%, which suggests that a genetic predisposition is necessary for chemical and microbial factors to act as triggers. 
 
Tantalizing clues like these are prompting scientists to reconsider the research agenda for ASDs. Martha Herbert, a pediatric neurologist at Harvard Medical School, and her colleagues have been applying the methods of genomics to identify environmentally responsive genes that might be important in these disorders. 
 
“When you realize that the widespread changes we’re seeing in autistic brains may occur in parallel with or even downstream from widespread changes in the body—such as in the immune system—and that these changes may be environmentally triggered, you start looking for ways to think more broadly about genetic vulnerability. It can’t be just about ‘brain genes,’” Herbert says. 
 
Some new epidemiological studies also are looking for gene–environment interactions. According to Diana Schendel, an epidemiologist and project officer for autism research at the CDC, which funds one of the projects, these initiatives will be able to examine many possible causal pathways to ASDs, including both genetic and environmental causes that may lead to the development of the disorders in different subgroups of children. 
 
Some of these projects are already under way, whereas others will begin soon. All of the scientists involved, however, believe their research will finally provide some of the answers that everyone has been looking for.

Interferon has been the subject of scientific interest for over 20 years, first as an antiviral agent and later as an anticancer drug. Early preparations were produced by challenging cells in tissue culture with inducing agents such as viruses or synthetic RNA (Gresser, 1961). Three main cell types were used. Leucocytes derived from buffy coats, or from leukopheresis preparations, produced crude ainterferon of 0.1% purity which could be partially refined by precipitation methods (Cantell et al., 1981;Horowitz, 1981). Human lymphoblastoid cell lines derived from patients with Burkitt's lymphoma resulted in a mixture of a and yinterferon (Strander et al., 1975). Normal fibroblasts grown in tissue culture resulted in /Binterferon of higher specific activity (Giard et al., 1979). It has gradually become clear that there are at least 3 gene families of inferferons: a, mainly derived from leucocytes, P from fibroblasts, and y-interferon produced by mitogen or antigen stimulated lymphocytes.
With the advent of the recombinant DNA technology it is possible to produce vast quantities of interferon of high specific activity by genetic engineering. Human interferon genes have been isolated and cloned in bacterial plasmids (Goeddel et al., 1980). The recombinant plasmids are inserted into E. coli, where they can replicate and produce mRNA, which in turn causes the bacterial protein synthesis machinery to make human interferon. Clones of E. coli have been isolated which produce large amounts of interferon. Using industrial fermentation techniques, unlimited quantities of interferon are now available for clinical use. There may be biochemical differences, such as the absence of glycosylation, between recombinant and natural interferons but the significance of these differences is not known.
Until recently clinical studies have used crude interferons of low purity. Profound side effects were observed which were initially attributed to impurities (Armin et al., 1983). Furthermore, the inadequate supplies and difficulties in purification allowed only limited amounts of interferon to be used. Nevertheless there was a wealth of data showing regression of animal tumours (Gresser, 1977) and cytotoxic activity against tumour cell lines (Gresser, in press). In addition there were encouraging reports of tumour regression in patients with myeloma, non-Hodgkin's lymphoma, breast cancer, melanoma, renal carcinoma and Kaposi's sarcoma (Sikora, 1983). In patients with advanced breast cancer, 30% partial remission rates were reported by two groups using crude buffy coat interferon (Gutterman et al., 1980;Borden et al., 1980). These preliminary results, together with the availability of large amounts of highly purified recombinant interferon, led us to set up a phase II study to look at the use of one preparation, recombinant leucocyte A interferon (rIFN-aA, Hoffman-La Roche) in patients with advanced breast cancer.

Patients and methods
Fifteen patients with readily assessable disease were studied. Those with CNS metastases, malignant pleural effusions, ascites, blastic or mixed lytic/blastic osseous metastases as the only evidence of disease were excluded. Also excluded were patients with hypercalcaemia or impaired haematological, renal or hepatic function. Treatment with non-steroidal anti-inflammatory agents was not permitted during the study, Corticosteroid doses were no greater than physiological replacement. No anti cancer medication (hormones or chemotherapy) was given for the 4 weeks preceding the study. The dose of rIFN-ocA was randomised to 20 x 106 units m-2 daily Correspondence: K. Sikora or 50 x 106m-2 3 times weekly by deep i.m. injection for 12 weeks. It was initially planned to study 60 patients but the study was stopped at 15 in view of the poor response rate. This dose was based on previously reported phase I studies (Gutterman et al., 1982). Dose reductions to 50% and then 10% of the above doses were made in patients with toxic side effects. All patients were aware of their diagnosis and gave written informed consent to the study.
Patients were assessed on entry by full clinical examination, chest X-ray, bone scan, liver function tests, and photography of skin lesions. These investigations were repeated periodically during IFN therapy. Oestrogen receptor status was also determined on a repeat biopsy, performed under local anaesthetic in all patients. Patients were admitted to hospital for initiation of IFN therapy but in most cases were discharged after several days. A community research nurse continued outpatient treatment and provided a crucial link between the patient's home and the clinic. Patients were formally assessed by two of us at regular intervals for signs of tumour regression, toxicity and performance status. A disease marker, such as a lymph node or patch of skin infiltration which could be measured, was chosen as an indicator for response. The following criteria were used: (CR) for disappearance of all tumour; partial response (PR) for 50% reduction in the product of the longest perpendicular diameters; stable disease where no change occurred; or progressive disease when tumour growth continued unabated. We also used the term less than partial response (LPR) for a response not satisfying the criterion for PR. Clinical photographs were taken frequently since the measurement of extensive infiltrative or ulcerative chest wall disease by the above criteria was often difficult.
Toxicity was graded as in Table II and dose reductions made to 50% of the planned dose for two or more minor toxicities; to 10% if moderate toxicity was observed. The appearance of a major toxic effect resulted in the discontinuation of IFN until the effect disappeared.

Results
The ages of the 15 patients ranged between 41 and 69 (mean 60). All were post menopausal and all had received previous treatment with one or more of the following: hormones, radiation, cytotoxic therapy (combination therapy or oral cyclophosphamide). All had progressive disease at the time of entry. Four had lymph node recurrences and the remaining eleven more widespread chest wall disease which was fungating or ulcerated in 4 patients. Three patients had pleural effusions and one pulmonary secondaries. Ten patients had positive bone scans.
In the first 11 patients treated, drastic dose reductions to 10% of the initial dose were necessary in all but one on account of lethargy and CNS toxicity. In the last 4 patients whose starting dose was reduced by 50% (10 x 106unitsm-2 daily or 25 x 106 units m-2 3 times weekly) only one further small dose reduction was necessary in one patient. There were no obvious differences in toxicity between daily or 3 times weekly administration.
Four patients died during IFN therapy. In 2 of these death was possibly related to the treatment, either resulting from direct CNS toxicity or secondary to dehydration and renal failure which followed as a result of impaired cerebration. Two deaths occurred in patients with very advanced carcinomatosis and were considered disease-related.
Side effects CNS All patients experienced varying degrees of lethargy, anorexia, nausea, vomiting, headache and myalgia. Five developed transient paraesthesia. Severe lethargy preceded a dose related state of confusion, disorientation, and dysphasia, if regular IFN administration continued. Subsequent coma developed in the two patients whose deaths were possibly related to IFN therapy. No predisposing metabolic cause could be found for these deaths and there was no evidence of cerebral metastases.
Severe CNS side effects were seen in 6 patients (Smedley et al., 1983). Serum electrolytes, glucose and liver function tests remained unchanged during this syndrome. EEGs showed excess slow wave activity and in subsequent patients serial EEGs showed increasing slow wave activity as interferon therapy progressed despite the absence clinically of major CNS toxicity. The EEG abnormalities improved slowly after stopping treatment in the patients who have been reassessed at this stage. All four patients treated on the reduced dosage developed EEG slow wave changes although none developed major CNS toxicity. CNS toxicity appeared, therefore, to be dose related. There was no major toxicity for doses <10x 106unitsm-2 daily.
Weight loss Nearly all patients lost weight and in some this was 10% or more of body wt. Weight loss resulted from anorexia, nausea, vomiting as well as the total loss of interest in sustenance and self preservation during periods of extreme lethargy and fatigue.
Haematological Nine patients showed a small fall in haemoglobin levels at 4 weeks. All but 2 patients showed a fall in the white cell count which occurred within the first 3 weeks and thereafter returned to near normal levels. The mean count fell from 6.6 x 1091-1 initially to 2.9 x 1091 1. There were no infections resulting from leukopenia and a neutrophilia occurred in two patients in response to infection (one chest infection and one case of cystitis). The platelet count fell in parallel with the white cell count, mean values falling from 315+101 x 1091-1 to 178+100x 1091-1. Two patients who had received extensive chemotherapy developed counts below 100 x 1091 -but there were no purpuric or haemorrhagic complications.
Hepatic The bilirubin and alkaline phosphatase levels did not appreciably alter during IFN therapy except in patients with progressive disease. The SGPT (L-alanine, 2-oxogluterate aminotransferase) rose in the second to third week in all patients and then returned slowly to normal. It appeared unrelated to major CNS toxicity and its significance is unknown (Table I). Fever All patients developed pyrexia shortly after commencing IFN. This was easily controlled by paracetamol and settled within one week. In order to assess the severity of the side effects we have classified them as major, moderate and mild (Table II) and recorded the proportion of time during IFN therapy that each patient suffered these side effects. Finally we have averaged the percentages over all patients (Table III). It is clear that by reducing the starting dose to 10 x 106 units m2 daily major toxicity was avoided, but that one-third of patients still had moderate, and two-thirds mild toxicity. The absence of a "no toxicity" category in this group is due to the fact that at this dose IFN therapy could continue whereas for the first 11 patients the treatment was frequently reduced to 3 x 106 units daily (or equivalent) after about 4 weeks, or stopped temporarily, and not restarted untill all toxic side effects had completely subsided.
We also looked at the time spent in hospital during IFN therapy. Ten patients were able to go home 2 days after starting treatment and only 3 of these needed readmission. The average time spent in hospital was 13 days with a range from 2 to 45 D In the assessment of any new form of cancer treatment the odds are loaded heavily against the new modality in that patients must, for ethical reasons, have very advanced disease. Despite the severe toxicities encountered, we found that many patients were reluctant to stop therapy. Tumour regression was seen in lymph nodes, skin infiltration and ulcerated and fungating lesions. Arm lymphoedema also lessened on objective measurement. Analgesic requirements decreased in 3 patients with painful bone metastases and in one of these the serum alkaline phosphatase level fell. We also noted a phenomenon observed in other studiesregression of disease at one site with progression elsewhere. Of the 12 patients assessed at 4 weeks, 2 were considered to have had true partial responses, and 7 less than partial responses. One patient had stable disease and one had disease progression. The remaining patient had a partial response at the indicator site but developed on going disease in the opposite breast. At 12 weeks, or cessation of therapy, 10 patients were alive. There were only 2 who showed any degree of response at this stage, the others having relapsed. It should be noted, however, that the dose of IFN administered for the last 8 weeks in most patients was much lower than for the first month. We have explored whether there were any obvious predisposing factors governing response (Table IV). Here we tabulate age, oestrogen receptor (ER) status and the number of months from presentation to first relapse as well as the time between this relapse and starting IFN therapy. This indicates the natural history and aggressiveness of the disease process for each patient. We have also summarised response to previous cytotoxic or hormone therapy in an attempt to see whether any patterns of responsiveness emerge.
Other independent variables governing response are the dose of IFN used and the duration of therapy. Many patients had a dose reduction after 4 weeks. We have stated the mean dose of IFN given per week during the first 4 weeks and the subsequent weeks respectively for each patient. The responses at 4 and 12 weeks are presented as well as survival.

Discussion
Response could not be related to the total duration of previous history, previous response to hormones or cytotoxic drugs, oestrogen receptor status or the dose of IFN administered. In the two patients who were still responding at 12 weeks, two factors emerge. Firstly, each patient had very advanced local disease without obvious metastases. The first patient had disease for 46 months from diagnosis and probably for 3 years prior to this. She had shown some improvement with tamoxifen. Clearly she had lived in symbiosis with her disease for a long time. The second patient had presented with a locally fungating carcinoma which she admitted having had for well over a year and we suspect much longer. There had been little response to 5 months tamoxifen therapy but equally little progression. In both cases the relative indolence of the disease could have been a function of the disease itself (i.e. mitotic rate) or of immunological factors in the host, or both. It is interesting, however, that both these patients showed the most consistent and sustained response to IFN. Little is known of the possible antitumour activity of IFN in vivo. Possible mechanisms include the inhibition of DNA-replication in the malignant cell following The second feature common to these two patients is that they completed 12 weeks of IFN treatment with relatively few toxic side effects and consequently with little dose reduction. To explore this further we have calculated the average weekly doses of IFN for the groups of patients 1-11 and 12-15 for the first month of IFN therapy, and subsequently, and compared these with the two responding patients (Table V). There was a marked dose reduction after one month owing to toxicity in the first 11 patients and this was overcome by reducing the starting dose. Patient 4 tolerated a dose which was far higher than the 12 week average for the other patients, and tumour regression may have been dose related. The dose of patient 14 was no different from the other 3 patients in her group who failed to respond, and not very different from the dose used in many of the first 11 patients. It seems unlikely, therefore, that her response was dose related. Patient 4 went on to maintenance IFN for one year, 86 x 106 units twice weekly. She has suffered no appreciable toxicity but has required additional treatment in the form of local irradiation and oral cytotoxic therapy. The response of her tumour to radiation has been excellent. Her EEG has showed persistent slow wave activity but has not significantly altered over the year.
The study would suggest that there is little place for rIFN-axA as a single agent in breast cancer although this study did not investigate its role in combination with other agents or as an adjuvant. There is no doubt, however, that many patients showed signs of tumour destruction at 4 weeks. The objective responses seen however encourage further study to investigate the potential value of IFN and its mechanism of action which are currently unknown. We also noted several responses to irradiation which were unexpectedly good in patients who were receiving, or had recently received, IFN. We are currently looking into this possible synergism, encouraged by the fact that others have made similar observations in relation to chemotherapy and interferon.