Recommended Guidelines for Screening Programs for Hyperphenylalaninemia in Newborn Infants

(1)

396 SCREENING FOR CONGENITAL METABOLIC DISORDERS

Guidelines

for Screening

Programs

for

Hyperphenylalaninem

ia in Newborn

Infants

In 1966 the U.S. Department of Health, Educa-tion, and Welfare issued Recommended Guidelinesfor PKU Programsfor the Newborn’ to assist health departments and others in estab-lishing screening services. Programs now function in most states, and it would be pointless to reiterate the information conveyed ten years ago. However, there are problems that must be addressed, particularly in light of new possibili-ties:

1. PKU screening is often of less than optimal (and attainable) effectiveness.

2. Parents are poorly informed about screen-ing, the use of specimens, and results.

3. Neonatal screening tests for other conditions are available, but the question arises of when they should be used on a population-wide basis.

New developments related to these areas are reviewed in the first part of this report. Recom-mendations are presented in the second part.

This document is intended primarily for those involved in the organization and regulation of screening as a service (Fig. 1). As the objective “is to find affected subjects at a time when interven-tion may prevent the ill effects of the disease,” responsibility includes assuring that efficacious therapy will be effectively provided. Procedures involved in the diagnosis and management of PKU that will be of interest to those providing care to infants with presumptive positive screen-ing tests are dealt with in other publications.2 :

NEW DEVELOPMENTS PKU

PrognosLi. There can no longer be any doubt that the early institution of a diet low in pheny-lalanine is efficacious in preventing retardation from PKU. The IQ (mean ± SD) at 4 years of age in 111 children with PKU identified as a result of neonatal screening was 93 ± 16.4. This is much higher than that in untreated or late-treated

children with PKU, including siblings of those in the study group. The best outcome was observed in those in whom the diet was started by 3 or 4 weeks of age.45 Because screening has become widespread, the admission of children with PKU to mental institutions has virtually ceased.6

Transient Elevations and Variants. Not all infants who have an elevation of serum phenylal-anine level identified by screening have classical PKU. Eighty-five percent of infants with presumptive positive screening test results have normal concentrations on the follow-up test. Only 5% will prove to have confirmed PKU. Most of the remaining infants with initial elevations will eventually have phenylalanine concentrations in the normal range, but 1% or 2% will have persistent, moderate elevations of blood phenylal-anine concentration up to 20 mg/ 100 ml on regular diets.7 These variants probably represent several different defects, but most of these patients are not at risk for mental retardation.89 Unmonitored restriction of dietary phenylalanine in a variant form of hyperphenylalaninemia may result in a severe deficiency state that could be harmful in itself. Thus, careful attention must be paid to the intensity of hyperphenylalaninemia

and its response when therapy is started2 in any infant with a positive screening test result.

A new form of “PKU” causing severe retarda-tion has been reported.’#{176} The pathology of the central nervous system is unresponsive to dietary restriction of phenylalanine in this variant. In some of these infants, the phenylalanmne concentration may not rise excessively when they are challenged with a regular diet.” This form of

Adapted from a document prepared by Neil A. Holtzman, M.D., for the Health Services Administration, U.S. Depart-ment of Health, Education, and Welfare, Washington, D.C.

at Viet Nam:AAP Sponsored on September 8, 2020

www.aappublications.org/news

(2)

INPUT

INFANTSIN INFANTSAT INFANTSIN

NEWBORN HOME DOCTORS OFFICES

NURSERIES AND CLINICS

PARENTS HEALTH

I

LABORATORY

POSITIVE TESTS NEGATIVE TESTS REPORT RESULT REPORT RESULT ANDREQUEST TO:

RECORD REPEAT FROM:

RESULTS; 1) HOSPITAL 1) HOSPITAL

APPROPRIATE 2) BABY’S 2) BABY’S RECORD ACTION PHYSICIAN PHYSICIAN

TAKEN 3) HEALTH 31 HEALTH DEPARTMENT DEPARTMENT

4) PARENTS

3

REPEAT SCREENING TEST

POSITIVE NEGATIVE

+

DEFINITIVE TEST

POSITIVE: VARIANT NEGATIVE DISEASE

DIAGNOSIS

3

MANAGEMENT. COUNSELING

3

OUTCOME

API tps he

qvaeaapa!of CONDITION PARENTS ad shoId bethe ASYMPTOMATIC RECOGNIZE

SpO(Shhty ofa DUE TO RISKS OF Coi,aI Aafhoay TREATMENT RECURRENCE

FIG, 1. Screening: process and outcome. cultured skin fibroblasts; because these infants

may benefit from early treatment with other modes of therapy, skin biopsy for culture should be considered in any infant with persistent eleva-tions of phenylalanine. The abnormal enzyme in these subjects is dehydropteridine reductase, and an abnormal excretory pattern of pterin metabo-lites may facilitate diagnosis; this possibility is under investigation.

False-Negatives. Between 5% and 10% of infants ultimately proved to have PKU are not detected by newborn infant screening of blood samples (false-negatives).’2 Most of these infants are missed because of the early age at which they are screened; the blood phenylalanine concen-tration in infants with PKU may not rise above the usual cutoff concentration until 4 days of age or later.” Although poor feeding contributes to the possibility of normal values in the first few days of life, it is not the only factor.” Most infants in the United States are screened on or before the fourth day of life.’2 When infants are screened later, as in the United Kingdom, there are virtually no false-negatives.’5

It is abundantly clear that the urine ferric chloride test for PKU (or the Phenistix dip stick) will fail to detect a significant number of infants with PKU,’ regardless of their age. This test should not be used as a screening test.

Laboratory Error. As used, the screening test does not always measure the true phenylalanine concentration in the specimen.’7 Error at concentrations close to the demarcation value assigned to distinguish the normal test from the abnormal explains why laboratories differ one from another in the reported frequency of presumptive positive tests. Some programs report an incidence of positive tests as low as 5 in 100,000; others are as high as 275 in 100,000 in infants tested on the third day of life.’2 Because the blood phenylalanine concentration in infants with PKU may be only minimally elevated in the first four days of life, laboratory error contributes to the failure to detect some affected infants.

Delays. The time consumed to retrieve an infant with abnormal test results is sometimes excessive. As a result of the delay, some harm to the central nervous system may be sustained by affected infants. The average time between the first test and the follow-up is 25 days in North America, and in more than 22% of infants it took longer than one month to obtain a follow-up.’2 Centralization, consolidation of testing programs, and the establishment of lines of communication between those collecting specimens, laboratories analyzing them, and those providing follow-up

care could alleviate some of these problems.’5

Unresolved Problems. The successful manage-ment of PKU in early childhood has raised other problems:

1. The question of when and whether the diet should be terminated. An answer may be forth-coming from the PKU Collaborative Project, which is terminating the low-phenylalanine diet in one half of its subjects, selected at random, at 6 years of age. The feasibility of such a study was recently reported.’

(3)

398 SCREENING FOR CONGENITAL METABOLIC DISORDERS Ethical and Legal Issues

Failure to define objectives, benefits, and risks; to obtain consent; and to assure confidentiality

(particularly

in chromosome screening and sickle cell carrier detection) has prompted concern for safeguarding individual rights in all genetic screening programs, including those for PKU. Recently, a committee of the National Academy of Sciences reviewed biomedical, legal, and ethical aspects of genetic screening and offered procedural guidance.2

Informed Consent. Most laboratory tests are performed because the patient seeks help. However, it is the patient who is sought in screening, and many healthy subjects are conscripted in the testing program. Consent for screening, therefore, must be viewed differently from consent for tests that result from patient-initiated contact. Informed consent need not be a barrier between the screener and the screenee but could, in fact, eliminate misunderstandings. The provision of information to parents before their infant is screened for harmful disorders for which treatment is available could allay anxiety at the time of the test and result in greater willingness to bring the infant in for a second test, if that is necessary. However, if the requirements for informed consent resulted in significantly fewer infants being screened, a mandatory program might be legally justified under the parens patriae

doctrine that the state acts to protect those who cannot protect themselves.

Any use of a specimen for purposes other than the demonstrated benefit of the infant being screened constitutes research and must conform to established guidelines for research in chil-dren.2’ The doctrine of parens patriae would not apply when proof is lacking that the screening test affords protection to the infant. Thus, when a specimen (or part of it) will be used to validate a new test or identify infants in whom a new treatment will be tried, failure to inform the parents about all aspects of the procedure and to obtain consent cannot be justified.

Right to Information. In newborn infant screening programs it may not be feasible to inform parents of normal test results. This is legally permissible if the parents are advised of it in advance and agree. Investigators may also wish to withhold results in situations when the spec-imen (or part of it) will be used to identify infants with conditions whose natural history is uncer-tain. Infants may be recruited into such studies if their parents are informed in advance and given

an opportunity to refuse to allow them to partici-pate.

Except for statistical data compiled without reference to the identity of the subjects, informa-tion regarding test results should not be released (for example, to school authorities) without the consent of the infant’s parents. There is no threat to the public health-as in the case of communic-able diseases-to justify release, and it may lead to unwarranted stigmatization of the infant or his family.

Regulation of Screening. As it becomes possible to screen for additional disorders, legislators may be influenced to pass laws dealing with each of them. There already are state laws dealing with sickle cell screening, histidinemia, galactosemia, and other conditions. Recognizing that separate laws for each new disease may be cumbersome and may not reflect the best interests of its citizens, the Maryland State Legislature created a Commission on Hereditary Disorders, which has the authority to promulgate regulations and stan-dards for the detection and management of genetic conditions in the state. As a result of having staggered terms of office, the Commission always has experienced members on it. Conse-quently, it is in a better position than the legis-lature, to which it is responsible, to evaluate the pros and cons of screening for specific disorders. The inclusion of consumer members provides greater assurance that programs recommended by the Commission serve the best interest of the public, and that rights of informed consent and confidentiality are safeguarded.

Regionalization. In some states the number of births per year may be too small to allow efficient screening, and several states may contract with one center to provide laboratory services and clinical consultation. A regional network of screening covers Alaska, Montana, and Oregon, and a similar network includes Maine, Massachu-setts, and Rhode Island. Regional networks also exIst in the Province of Quebec in Canada22 and in several other countries.

Recently the Secretary of Health, Education, and Welfare suggested the establishment of regional networks for laboratory work related to prenatal diagnosis.23 If this were to occur, newborn infant screening tests might eventually be performed in regional laboratories. This would improve efficiency if procedures were established concomitantly for the rapid transmission of spec-imens to the regional laboratory and of results from the laboratory to minimize delays in follow-up.

(4)

Screening for Other Conditions

PKU screening are given at the end of this

Newborn infant screening tests have now been reported for more than 20 disorders.21 Most of them require specimens from infants who are a few days old because they measure metabolites whose accumulation in affected infants is, at least in part, dependent on dietary intake. Cord blood or capillary blood from the newborn infant can be used for assays to detect hypothyroidism (see the first part of this supplement for T4 screening) and galactosemia.2l Screening has also been used to identify women with PKU because the infants of mothers with this disorder have high phenylal-anine concentrations in cord blood but not in specimens obtained 24Amino acid chromato-graphy of blood, particularly when obtained in the second week of life, enables detection with one test of PKU as well as of other abnormalities. Urine chromatography on specimens collected at 2 to 4 weeks of age has also proved helpful in detecting conditions that may be missed by blood screening at an earlier age.224

The availability of a test is insufficient reason for instituting a public screening program. Many tests require further validation. Some of the chemical disorders that are detected are not associated with any disease state. For others, the relation is uncertain and additional study is needed. As a minimum condition, newborn infant

screening should meet at least one of the

following criteria before it is offered as a routine service:

1. Affected subjects who are discovered will benefit, as in the case of PKU, from the early institution of therapy.

2. Affected subjects and their families will be apprised of hazardous situations that should be avoided (e.g., genetically determined drug and food intolerances).

3. The families of affected subjects discovered to have genetic disorders will be counseled about risks of recurrence.

RECOMMENDATIONS General Recommendations

A mechanism for supervising screening oper-ations for any disorder-from innovation, through implementation, to evaluation-is presented here. Laboratory responsibilities are described later in the report. These recommendations are intended for laboratory personnel as well as for those in the screening authority to whom the laboratory is responsille. Specific recommendations regarding

report.

Regional Advisory Committee (or Commission)

Screening is more than a laboratory service (Fig. 1). To improve operation of the entire program, there should be one authority (usually under the state health department) or a regional grouping, with an advisory board or commission consisting of (1) a representative of hospitals who send specimens, (2) a specialist in metabolic disease, (3) a nurse and/or nutritionist involved in management of patients, (4) a representative of practicing physicians, (5) the director of the laboratory performing the tests, and (6) consum-ers. Each of these persons can contribute to better operation and acceptance of the program. In some areas the committee may have regulatory authority.

The advisory commission could have the following responsibilities.

Determine Which Conditions Should Be Screened For. This decision should be based primarily on the following considerations: (1) frequency and severity of the condition; (2) avail-ability of a therapy of proven efficacy; (3) extent to which detection by screening improves the outcome; (4) validity, reliability, and safety of the screening test; (5) adequacy of resources to assure effective screening and follow-up; (6) costs; and (7) acceptance of the screening test by the community, including both consumers and prac-ticing physicians. It must be emphasized that the availability of a test is an insufficient condition for undertaking routine screening.

(5)

Determine Procedures for Transmission of Spec-imens and Results and for Recording of Data. All steps up to and including referral of infants with presumptive positive tests should be performed with sufficient rapidity so those in whom the diagnosis is established receive maximum benefit. To help assure that the test is performed on every

infant for whoni the parents consent, a notation

on the hospital medical record or birth certificate should be required.

Determine Adequacy of Laboratory (or Labora-tories) in the State. The advisory committee can request that the laboratory evaluate its tests periodically (quality control) and define the procedures for reporting of results.

Referral Center

The center to which an infant with a presump-tive positive test is referred should be capable of confirming the diagnosis, initiating and moni-toring therapy, and counseling the family. To accomplish these objectives, the center should have experience with that disorder or related conditions; it should also have effective communi-cation with the advisory committee, the screening laboratory, and the practicing physician who provides primary care to the patient. The advi-sory committee should assure that no infant discovered by screening to have a genetic disorder be denied treatment for financial reasons.

Education and Information. Physicians, col-leges and schools, and the mass media should be used to educate the community about the avail-ability of screening and the reasons for it. Infor-mation should be provided to pregnant women and their spouses concerning the objectives of the test, when and how it will be done, and which results will be transmitted and how. The advisory committee might make form letters available that can be used to request follow-up tests for presumptive positive and unsatisfactory screening results. Practicing physicians must be kept informed about which conditions are being screened and the limitations of the tests. False-negatives are possible with any test; in older

infants with suspicious findings, the diagnosis

should not be ruled out on the basis of a history of a negative screening test.

Deteriii inc lufornied Consen t Procedu res. Infor-ination can be presented to the parents either prenatally or after birth, but not when the mother is in labor or immediately postpartum. A state-ment can be included that, unless the parents specifically object, the test will be performed. A more positive consent can be obtained orally and

noted on the infant’s record, or it can be obtained by a written statement. If part of the specimen will be used for investigational purposes, the consent should note this fact and describe the type of investigation. A brochure providing infor-mation about the test(s) should be distributed. A mechanism to obtain the appropriate informed consent should be implemented and evaluated.

Evaluation. The advisory committee should periodically evaluate the individual components of the program as well as the entire program itself

(see

“Reporting and Evaluation” under the section on “Screening for PKU”).

Laboratory Responsibilities

Quality Control. Because of problems in the performance and interpretation of screening tests, specimens should be analyzed in a large, central facility. This may be one laboratory within a state or, when the number of births in a state is too small to permit efficient utilization, one labora-tory serving several states. If more than one laboratory must be used within a state, the health department must rigorously check the quality of each laboratory. In addition, to provide equal coverage of all infants born in the state, all of the tests deemed routine by the advisory committee should be performed by all laboratories autho-rized to perform tests.

The Center for Disease Control in Atlanta will provide standard specimens that can be used to check the reliability of methods in each partici-pating laboratory; this assures consistency of results between laboratories. Bacterial assays and chromatography require visual comparisons of an unknown and a standard. To minimize error, standards should be included in the daily runs and, from time to time, specimens with elevated concentrations should be inserted as unknowns. Intraobserver and interobserver variability should also be checked.

The possibilities of error are too great to base a diagnosis on a single test result; this is true of any laboratory test, but particularly of screening tests in which the procedure itself, as well as the number of specimens being handled simulta-neously, increases the chance of error. Thus, a confirmatory test on a second specimen is always part of the routine management of positive screening tests.

Changes in supplies, personnel, instruments, or climatic conditions can alter the results of tests routinely performed. Consequently, there is a need to review periodically results obtained and to alter procedures or cutoff points accordingly. Methods reported in the literature should be

(6)

NUMBER OF INDIVIDUALS

Mean lsd 2sd 3sd

CONCENTRATION followed closely, but cutoff points should not be

accepted without validation on a sample of the local population, which may differ in age or environmental (e.g., nutritional) and genetical factors from the population on which the litera-ture report was based. Handling of the specimens may also differ.

Determination of the Cutoff Point and Optimal Age for Testing. A cutoff point that distinguishes all of those with overt manifestations of a condi-tion from those who do not have the condition may not distinguish all, or even most, presympto-matic infants from those without the condition. In affected subjects in the immediate postnatal period in particular, the concentration of substances destined to reach much higher levels may be only minimally elevated.

Ideally, to determine how frequently the cutoff level selected will miss affected infants, another test should be performed in every infant. Either a more definitive method should be used simulta-neously or the test should be repeated when the infants are older. If the second test reveals infants with the condition who had levels below the cutoff point on the screening test, either the cutoff point should be changed or the test should be performed when infants are older.

As the cutoff point approaches the mean of all test results, more and more subjects who do not have the condition will exceed it (Fig. 2). A large number of such false-positives inflates the cost of screening because they must all be followed up. Thus, rather than alter the cutoff point, it may be more efficient to screen infants when they are somewhat older, if therapy will still be effec-tive.

A less satisfactory method of determining an optimal cutoff point than performing two tests on all infants involves performing second tests only on those whose screening test results fall outside of some multiple of the standard deviation of the mean on the abnormal side (Fig. 2). For

condi-tions of low incidence, a higher multiple of the

standard deviation can be selected to increase the ratio of true- to false-positives among those receiving a second test. This approach assumes that all subjects with the condition fall within the population on which another test is performed.

For tests that are already in routine use, extreme changes in the mean or standard devia-tion in the most recent period compared to the past suggest a need for revalidation. If the mci-dence of the condition is lower than that reported on comparable populations, revalidation is also indicated.

Storage of Specimens. Specimens should be

FIG. 2. Distribution of test results in a population screened during first week of life (solid lines). Small curves represent affected infants; solid line, their results in first week of life; dashed line, their results a few weeks later. sd = standard deviation for the mean of all test results during first week of life. To detect all affected infants during first week of life would require repeated testing of all infants who exceeded mean by 1 SD. At later age, repeated testing would be needed only on those who exceeded the mean by 3 SD. Test results in unaffected subjects are much less affected by age. Diagram also applies when affected infants have reduced, rather than elevated, concentrations of compound being

screened.

saved, under conditions to maximize stability, for as long as possible. Phenylalanmne and some other amino acids in dried blood spots on filter paper are stable for several years. When an infant whose screening test was reported as negative is discov-ered to have the disease for which screening was performed, a retest can indicate whether the laboratory was in error. Saved specimens may be used for validating tests for other conditions, if parental consent was given. Should a serious disease be shown to have a biochemical aberra-tion, analysis of the specimen saved from affected infants may indicate that the condition is detect-able in the newborn period.

Recording of Results. The laboratory should keep permanent records of all results and period-ically tabulate them to determine changes in the frequency of both true- and false-positives. The interval it takes for each hospital to send spec-imens should be checked periodically and delays corrected. A system should be developed to flag incomplete records on which a follow-up is due.

The records of all results are confidential and should be released only to the hospital responsible for the collection of the test so they may be placed on the infant’s medical record or given to the infant’s physician or other person specifically

agreed to by the parents.

(7)

Conditions and New Tests. The consent of the parents is required, unless the specimen is used anonymously, to establish the reliability of new methods. Validation of tests requires follow-up of the patient; therefore, consent must be obtained. When a new test is being considered for a condition for which screening is already routine, it should be run simultaneously with the old on the same specimens.

Screening for PKU

Benefits of PKU screening are sufficiently well established to recommend testing of all infants. The exception might be when no other screening test is being performed in parallel and when the population is at extremely low risk for PKU, for example, a black population or Ashkenazi Jews. The advisory committee should be the judge of the capability of its jurisdiction to screen for other conditions under such circumstances.

Information and Consent. Every pregnant woman should receive information concerning PKU screening. The advisory committee should distribute to obstetricians, prenatal clinics, and others responsible for the care of pregnant women a brochure containing the following infor-mation: (1) definition of PKU and reasons for the test; (2) indication that the test will not give information about other causes of mental retarda-tion; (3) description of the test procedure, including when and where it will be performed and how the specimen will be obtained; (4) description of the risks (the approximate number of tests already performed and the number of adverse reactions, or the lack of them, should be included); (5) the reason why follow-up tests will be requested (a positive screening test does not, by itself, indicate the presence of PKU); and (6) description of how the results will be communi-cated. If negative results will not be communi-cated, the parents should be so informed; they should also be told how they can obtain the results if they want them. Parents should be informed of the time it will take for notification of positive and unsatisfactory test results.

It may be appropriate to include information on PKU screening with information concerning other recommended procedures (e.g., thyroid hormone screening in newborn infants, Tay-Sachs testing in the adult Jewish population, and sickle cell screening in the adult black population). Information on PKU can also be communicated, perhaps for a second time, after the woman is delivered of her infant. If parents do not consent to the test, their decision should be recorded in the infant’s medical record.

Timing of the Test. A blood test to screen for elevated concentrations of phenylalanine should be performed on all newborn infants no sooner than 24 hours after onset of milk feeding and as close to discharge as possible. Premature infants should be tested between the fifth and seventh day after birth, or as soon thereafter as protein intake is adequate (whether receiving antibiotics or not). If the test results are positive, further investigation to distinguish between neonatal hyperphenylalaninemia and other forms of hyper-phenylalaninemia is indicated.

Even when an infant is discharged before having ingested milk feedings for 24 hours, a blood specimen should always be obtained on nursery discharge. In addition, arrangements should be made with the parents of these infants to bring the infant back to the hospital for an additional blood specimen during the second week of life. If this is not feasible, mothers should be given a filter paper card on which their physician or clinic can collect a drop of the infant’s blopd and send it to a specified laboratory as soon as possible after the infant’s discharge.

There is current controversy about whether or not a second blood specimen for PKU testing should always be obtained. A second blood spec-imen does not seem necessary for the detection of PKU if milk intake is adequate, the first test is obtained no earlier than the fourth day of life, and all screening tests are performed in a well-organized laboratory whose personnel have expertise in PKU testing procedures.

The second blood specimen can be collected any time after the first week of life. Postponement can only delay the onset of therapy if the later test is positive.

Infants born into families with a history of PKU should be tested daily during the neonatal hospital stay; if the results remain negative, the test should be repeated again at 2 weeks of age.

Responsibility for Testing. The responsibility for collection of the specimen rests with the person in charge of the institution in which the infant is born (or his designated representative) or, if an infant is born outside the institution, with the person delivering the infant or required by statute to file the certificate of birth. If an infant is discharged from the institution earlier than 24 hours after the onset of milk feeding or before a blood test is collected, the physician and/or institution should be responsible for ensuring that a satisfactory test is obtained by 2 weeks of age. If tests are to be performed routinely after infants are discharged from the hospital, the responsi-bility can be assigned to the infant’s physician or

(8)

well-baby clinic, to a health visitor in the home, or to the hospital in which the infant is born by requesting the parents to return at a specified time.

Collection of Specimens. The following infor-mation should be affixed to each specimen: name, sex, race, date of birth, identifying number of the infant, time and date of the first feeding, and time and date at which the specimen was obtained. Birth weight, type of feeding (breast or bottle), volume of milk ingested in last 24 hours, condi-tion of the infant at discharge, and the name of the infant’s physician are other useful items. Supplies for the performance of the test sould be provided by the advisory committee.

All specimens should be forwarded by first-class mail to the laboratory within 24 hours of collection. The envelopes in which they are mailed should be marked or colored to indicate that they require prompt handling.

Report of Test Results. The laboratory should report results to the institution or individual responsible for the collection of the specimen within five days after its receipt. When test results are positive, telephone communications may be used as well.

The report of each test should be entered on the infant’s medical record and, if the test results are negative, be available to parents on request. No test result should be released by the labora-tory, the institution, or any other individual, except as already described, without the written authorization of the parents.

Repeated Tests for Presumptive Positive Test Results. The institution or individual, or his designate, responsible for collection of the screening specimen should be responsible for obtaining a repeated test when necessary and should do so promptly. If the infant cannot be retrieved for a second test, the advisory committee should be notified and a further follow-up attempted by the local health depart-ment. Care should be taken not to alarm the parents unduly, particularly if the second test is requested simply because the first screening test was technically unsatisfactory. The family’s physician should be notified at the same time. Parents should be notified of the results of follow-up tests, even if they are negative, so they do not remain concerned.

Confirmed positive test results should be reported by the laboratory to a central office, whose responsibility it is to notify the institution or the individual who collected the specimens and the family’s physician. The aims are to assure referral to a physician experienced in the field of

genetic metabolic disorders for confirmation of the diagnosis and management, and to develop a cooperative relationship between the family physician and the consultant which will benefit the infant and the family.

Criteria for Diagnosis of PK U. The publication,

Diagnosis and %lanagemen t of Plic’ii ylketon 12

offers a detailed description of diagnostic procedures and management. To establish a diag-nosis of PKU after a positive test is obtained, at least two separate determinations, by a quantita-tive method, should yield a blood phenylalanine concentration of 20 mg/100 ml or higher and a tyrosine concentration of 5 mgI 100 ml or less. Both of these specimens should be collected when the infant is receiving adequate amounts of protein in the diet. If these criteria are satisfied, dietary phenylalanine should be restricted even if the ferric chloride test is negative. (Alternative procedures may indicate the presence of pheny-lalanine metabolites in the urine.) After the infant is started on a low-phenylalanine diet, frequent monitoring of serum or blood phenylalanine concentration is needed. Lower than expected levels suggest the possibility of a variant form of hyperphenylalaninemia. The diagnosis should be further confirmed, either by challenging the infant with a formula containing normal amounts of phenylalanine or by the demonstration from serial monitoring that the dietary phenylalanine tolerance cannot exceed 500 mg/day without provoking hyperphenylalaninemia in excess of 16 to 20 mg/100 ml.

Acceptable Screening Procedures. Laboratory methods have been reviewed elsewhere.#{176}2 The Guthrie bacterial inhibition assay, the McCaman and Robins fluorometric assay, and chromato-graphic procedures on either paper or thin-layer plates are the preferred screening methods currently in use. The fluorometric method is the most sensitive and precise, and paper chromatog-raphy is the least sensitive. The determination of the cutoff value by any of these methods should be made in each laboratory according to the preceding section on laboratory responsibilities. Neither the Guthrie bacterial inhibition assay of urine or the ferric chloride or Phenistix tests are suitable screening tests at any age.

(9)

receipt by the laboratory for each hospital. It should also report the time required to process specimens (the time between receipt in the

labo-ratory and the mailing of results) and the interval

between the first positive test and follow-up. The disposition of all infants with positive tests should be determined by the advisory commission from the appropriate records and the incidence of PKU, variants, and presumptive positive test results calculated and compared to similar data from earlier years and other regions.

On the basis of this information, the advisory committee can suggest appropriate modifications of procedures by the hospitals, laboratories, or others involved in the program.

REFERENCES

1. Maternal and Child Health Service, HSMHA: Recom-mended Guidelines for PKU Programs for the Newborn. US Department of Health, Education, and Welfare, 1971.

2. Diagnosis and Management of Phenylketonuria. US Department of Health, Education, and Welfare, to be published.

3. Guidelines to Diet Management of PKU Infants and Preschool Children. US Department of Health, Education, and Welfare, to be published.

4. Dobson J, Williamson M, Azen C: Intellectual assess-ment of 97 four year old children with phenylketo-nuria. Unpublished manuscript.

5. Kang ES, Sollee ND, Gerald PS: Results of treatment and termination of the diet in phenylketonuria (PKU). Pediatrics 46:881, 1970.

6. MacCready RA: Admissions of phenylketonuric patients to residential institutions before and after screening programs of the newborn infant. J Pediatr 85:383, 1974.

7. Cunningham GC: Phenylketonuria testing-its role in pediatrics and public health. CRC Crit Rev Clin Lab Sci 2:45, 1971.

8. Berman JL, Ford R: Intelligence quotients and intelli-gence loss in patients with phenylketonuria and some variant states. J Pediatr 77:764, 1970.

9. Levy HL, Shih yE, Karolkewicz V. et at: Persistent mild hyperphenylalaninemia in the untreated state. N EngI J Med 285:424, 1971.

10. Kaufman 5, Holtzman NA, Milstien 5, et al: Phenylke-tonuria due to a deficiency of dthydroptendine reductase. N EngI J Med 293:785, 1975. 11. Milstien 5, O’Flynn M, Holtzman NA, et al:

Hyperphe-nylalaninemia due to a deficiency of dihydropter-idine reductase. Unpublished manuscript.

12. Holtzman NA, Meek AG, Mellits ED: Neonatal screening for phenylketonuria: I. Effectiveness. JAMA 229:667, 1974.

13. Holtzman NA, Meek AG, Mellits ED, Kallman CH: Neonatal screening for phenylketonuria: III. Altered sex ratio; extent and possible causes. J Pediatr 85:175, 1974.

14. Dontanville VK, Cunningham GC: Effect of feeding on screening for PKU in infants. Pediatrics 51:531,

1973.

15. Starfield B, Holtzman NA: A comparison of effective-ness of screening for phenylketonuria in the United States, United Kingdom and Ireland. N EngI J Med

293:118, 1975.

16. Medical Research Council Working Party on Phenylke-tonuria: Present status of different mass screening procedures for phenylketonuria. Br Med J 4:7, 1968.

17. Ambrose JA: Report on a cooperative study of various fluorometnc procedures and the Guthrie bacterial inhibition assay in the determination of hyperphe-nylalaninemia. Health Lab Sci 10: 180, 1973. 18. Holtzman NA, Welcher DW, Mellits ED: Termination

of restricted diet in children with phenylketonuria: A randomized controlled study. N Engl J Med

293:1121, 1975.

19. MacCready BA, Levy HL: The problem of maternal phenylketonuria. Am J Obstet Gynecol 113:121,

1972.

20. Genetic Screening-Programs, Principles and Research. National Academy of Sciences, 1975.

21. Federal Register, part III, August 8, 1975.

22. Clow CL, Fraser FC, Laberge C, et at: On the applica-tion of knowledge to the patient with genetic

disease. Prog Med Genet 19:159, 1973.

23. Culliton BJ: Aminocentesis: HEW backs test for pren-atal diagnosis of disease. Science 190:537, 1975.

24. Levy HL: Genetic screening. Adv Hum Genet 4:1, 1973.

(10)

1977;60;396

Pediatrics

Newborn Infants