Neurofibromatosis type 1 and optic pathway gliomas

Neurofibromatosis Type 1 and Optic Pathway Gliomas Follow-up of 54 Patients Sureka Thiagalingam, MBChB, MPH,1 Maree Flaherty, FRANZCO, FRCOphth,1 Frank Billson, FRANZCO, FRCOphth,1,2 Kathryn North, FRACP, MD3 Objective: To describe the natural history of optic pathway gliomas (OPGs) in patients with neurofibromatosis type 1 (NF1), and to evaluate the current recommended guidelines for monitoring and follow-up of OPGs in this population. Design: Retrospective case series. Participants: Patients with OPGs and NF1 seen in the neurofibromatosis clinic at the Children’s Hospital at Westmead in Sydney, Australia. Methods: Patients with definite NF1 and confirmed OPGs were identified and their medical records searched to obtain data on demographics, details of the OPG diagnosis, and serial ophthalmic examination findings. Patients were stratified into groups according to age and mode of presentation. Main Outcome Measures: Visual acuity was recorded for each eye and grouped into mild (Snellen equivalent ⱖ 6/12), moderate (Snellen equivalent ⫽ 6/15– 6/60), and severe (Snellen equivalent ⬍ 6/60) visual impairment at time of diagnosis, during follow-up, and at the most recent examination. Results: Data were collected on 54 patients, the majority of whom (78%) were seen from 1990 to 2002, with an average follow-up of 8.6 years. The mean age at the time of OPG diagnosis was 5.2 years, with 32 patients having symptoms or signs at the time of diagnosis. Seventeen patients were diagnosed after the age of 6 years (range, 7–15). Twenty-two patients had tumor progression within 1 year of diagnosis, and a further 6 patients showed progression after 1 year. Most patients’ conditions were managed conservatively (68.5%). At follow-up, 17 patients (31.5%) had severe visual impairment (⬍6/60 Snellen equivalent) in their worse eye, and 16.7% had bilateral moderate/severe visual impairment. Conclusions: Contrary to some previous reports, our results show that OPGs in patients with NF1 often present in older children and may progress some time after diagnosis. Given the potential for serious visual consequences, these findings indicate a need for regular ophthalmological monitoring of this population for a long duration. Ophthalmology 2004;111:568 –577 © 2004 by the American Academy of Ophthalmology.

Neurofibromatosis type 1 (NF1) is a common inherited disorder with an approximate incidence of 1:4000. One of the most significant complications of NF1 in childhood is the development of optic pathway gliomas (OPGs), thought to be lowgrade pilocytic astrocytomas, with a reported prevalence of approximately 15% (range, 1.5–24%).1,2 Although these tumors are considered benign, in that they do not greatly increase mortality rates, they can result in significant visual impairment or endocrine disturbance in some patients.3– 6 Originally received: November 18, 2002. Accepted: June 6, 2003.

Manuscript no. 220918.

1

Department of Ophthalmology, The Children’s Hospital at Westmead, Sydney, Australia.

2

Save Sight Institute, University of Sydney, Sydney, Australia.

3

Department of Clinical Genetics, The Children’s Hospital at Westmead, Sydney, Australia. Correspondence to Professor Kathryn North, MD, FRACP, The Children’s Hospital at Westmead, Locked Bag 4001, Westmead, 2145, Sydney, NSW, Australia. E-mail: [email protected].

568

© 2004 by the American Academy of Ophthalmology Published by Elsevier Inc.

Earlier reports suggest that OPGs rarely progress, have a self-limiting growth pattern, and tend to present in younger children,7,8 and those OPGs associated with NF1 are thought to have a more indolent course relative to non-NF1 cases.5,9 –11 The National Institutes of Health (NIH) NF1 Optic Pathway Glioma Task Force has recently published guidelines for the diagnosis, monitoring, and treatment of OPGs, based on this evidence.12 For newly diagnosed NF1 patients without a known OPG, these include a complete ophthalmological examination at diagnosis, followed by annual examinations until the age of 6 years. A longer period between examinations is suggested for children older than 6. The guidelines also recommend neuroimaging only when clinically indicated, rather than as a routine screening procedure. For children with newly diagnosed OPGs, regular ophthalmic examinations accompanied by neuroimaging are suggested, with decreasing frequency from the time of diagnosis. The majority of reports on the natural history of OPGs have been limited by small patient numbers, short follow-up ISSN 0161-6420/04/$–see front matter doi:10.1016/j.ophtha.2003.06.008

Thiagalingam et al 䡠 Optic Gliomas in NF1 Table 1. Demographic Variables by Mode of Presentation at the Time of Optic Glioma Diagnosis Variable Gender Male Female Family history of NF1 Age NF1 diagnosed (mean) Age OPG diagnosed (mean) Age OPG diagnosed (grouped) ⱕ6 yrs ⬎6 yrs Site of glioma Isolated ON involvement Intraorbital only Whole ON Unsure Right ON Left ON Bilateral ON ON and OC involvement Isolated OC involvement Visual impairment at OPG diagnosis (worse eye) Mild/none Moderate Severe No data available Progression of disease At diagnosis 1 year after diagnosis 2 years 3 years 6 years Treatment Surgery Radiotherapy Chemotherapy Combination Conservative Visual impairment at most recent examination (worse eye) Mild/none Moderate Severe

Symptomatic [n (%)*]

Asymptomatic [n (%)*]

14 (43.8) 18 (56.2) 12 (37.5) 2.2 (SE, 0.6) 4.8 (SE, 0.6)

13 (59.1) 9 (40.9) 8 (36.4) 2.3 (SE, 0.5) 5.8 (SE, 0.8)

27 27 20

22 (68.7) 10 (31.3)

15 (68.2) 7 (31.8)

37 17

8 (25.0) 5 3 0 6 5 16 19 (59.4) 5 (15.6)

8 (36.4) 4 2 2 5 2 10 9 (40.9) 5 (22.7)

16 9 5 2 11 7 26 28 10

14 (43.8) 6 (18.8) 11 (34.4) 1 (3.1)

21 (95.5) 0 (0) 0 (0) 1 (4.5)

35 6 11 2

21 (65.6) 2 (6.3) 2 (6.3) 1 (3.1) 0 (0)

0 (0) 0 (0) 1 (4.5) 0 (0) 1 (4.5)

21 2 3 1 1

5 (15.6) 6 (18.8) 1 (3.1) 3 (9.4) 17 (53.1)

0 (0) 1 (4.5) 1 (4.5) 0 (0) 20 (91.0)

5 7 2 3 37

9 (28.1) 6 (18.8) 17 (53.1)

20 (90.9) 2 (9.1) 0 (0)

29 8 17

Total

NF1 ⫽ neurofibromatosis type 1; OC ⫽ optic chiasm; ON ⫽ optic nerve; OPG ⫽ optic pathway glioma; SE ⫽ standard error. *Percentage of total symptomatic or asymptomatic subgroup.

times, and failure to clearly differentiate between NF1 and non-NF1 patients. On this basis, we reviewed our large cohort of patients with NF1 to determine the natural history of patients with NF1 and OPGs. In addition, we evaluated our management protocol for the diagnosis and monitoring of patients with OPGs, which are based on the NIH guidelines, to determine whether this approach was optimal for patient visual outcome.

Materials and Methods This study comprises a retrospective review of patients seen at the neurofibromatosis clinic at the Children’s Hospital at Westmead in Sydney (previously the Royal Alexandra Hospital for Children, Camperdown, Sydney), a tertiary referral center for the diagnosis and management of children with NF1. The majority of patients

were from New South Wales, but occasionally patients from other Australian states, and one patient from Fiji, were referred for an opinion. Ethics approval for the study was obtained from the Children’s Hospital Ethics Committee. The study population was initially identified via the Children’s Hospital neurofibromatosis database. Patients with confirmed NF1, according to the criteria set by the NIH Consensus Development Conference, and OPGs evident on neuroimaging were selected. At the initial clinic visit, all patients underwent a thorough history and examination by a pediatric neurologist or geneticist and were then referred to an ophthalmologist for a routine ophthalmic examination. Details of the clinic procedures have been described elsewhere.13 Patients were routinely reviewed annually, or more frequently if clinically indicated, to monitor disease complications. Frequency of ophthalmology reviews varied according to the patients’ age, symptoms, and signs. Neuroimaging was usually performed only when clinically indicated; however, some asymptomatic patients were referred for imaging by external cli-

569

Ophthalmology Volume 111, Number 3, March 2004

Figure 1. Age at optic glioma diagnosis by mode of presentation.

nicians, either for routine screening or to investigate symptoms unrelated to the optic glioma. Relevant information was obtained from the patients’ medical files, including general demographic data, symptoms and signs of NF1, medical history, family history of NF1, age at diagnosis of OPG, visual symptoms or signs present at the time of diagnosis and at progression of the OPG, treatment modality, and visual outcome. Where necessary, further details of visual findings were obtained by contacting the ophthalmologist involved. Patients with insufficient data were excluded. Data were collected from the initial clinic visit to the most recent records. Relevant details of the patients’ other visits to the hospital before their first neurofibromatosis clinic visit were also included, especially details of eye clinic appointments. Visual field measurements were performed using either a Bjerrum screen or Goldmann or Humphrey automated perimeters. Visual acuity was measured using a variety of methods, depending on the age of the patient. Generally, Teller acuity cards were used up to the age of 2 to 3 years, followed by Kay pictures, Lippman HOTV matching letters, or Sheridan–Gardiner single/linear letters until 5 to 6 years of age, and then Snellen reading charts at 6 m were used. For each eye, visual acuity corrected for refractive errors was recorded and then grouped as follows for analysis: 6/6 to 6/12 Snellen equivalent, mild or no impairment; 6/15 to 6/60, moderate impairment; and ⬍6/60, severe impairment. These Table 2. Symptoms and Signs at Time of Optic Glioma Diagnosis by Age (Grouped) Age Group Mode of Presentation

ⱕ6 yrs

⬎6 yrs

Total

Symptomatic Visual loss Proptosis Nystagmus Strabismus Visual field defect Optic disc atrophy Precocious puberty Asymptomatic Routine scan Scan for other reason

22 17 9 0 6 2 17 0 15 6 9

10 6 1 1 2 3 8 5 7 2 5

32 23 10 1 8 5 25 5 22 8 14

570

groups were chosen because they were more clinically relevant and allowed better comparison of the different visual acuity tests. The optic pathways were imaged using computed tomography (CT) and/or magnetic resonance imaging (MRI). Optic pathway gliomas were diagnosed on neuroimaging according to previously defined criteria.14,15 Clinical progression of the OPG was diagnosed if there was deterioration in symptoms or signs, or a radiological increase in size. Regression of OPG was defined as reduction in size of the tumor and/or improvement in visual impairment. Patients were classified as symptomatic if they had visual symptoms or signs at the time of OPG diagnosis, or asymptomatic if they were diagnosed only on the basis of positive cranial imaging results. Data were analyzed using SPSS 10.0.16 Chi-square tests were used to determine a statistically significant difference between the 2 groups and various variables. Stratified analysis was also performed according to the age at which the OPG was diagnosed.

Results We identified 57 patients with NF1 and optic pathway tumors. The clinic NF1 database contained records on 861 patients, giving a prevalence of 6.6% in our population. Three of these patients were excluded from analysis, as there were insufficient data available. The mean follow-up for the remaining 54 patients was 8.6 years (range, 1–29). The majority of these patients (78%) were first seen between 1990 and 2002. All patients had serial neuroimaging, and MRI was performed on 50 patients. The time interval between neuroimaging ranged between 3 and 24 months, depending on the degree of visual impairment or radiological abnormalities. Our population comprised equal numbers of males and females (27 of each), and 20 participants (37.0%) had a family history of NF1. The average age at NF1 diagnosis was 2.2 years (standard deviation [SD], 3.0). There was only 1 death in our series, a 9-year-old boy who was asymptomatic at diagnosis, and who later developed a malignant sarcoma unrelated to the optic glioma. Sixteen patients (29.6%) had isolated optic nerve involvement of one or both nerves, a further 28 had both optic nerve and optic chiasm involvement, and 10 patients (18.5%) had isolated chiasmal enlargement. Eleven patients had a glioma of the right optic nerve, 7 had a glioma of the left optic nerve, and 26 patients had bilateral optic nerve gliomas. Of the 16 patients

Thiagalingam et al 䡠 Optic Gliomas in NF1

Figure 2. T2-weighted magnetic resonance imaging of a patient with an optic glioma diagnosed at 9 years of age when a routine examination revealed optic disc swelling in the right eye (previous eye examinations had been normal). Her vision remained 6/6 in both eyes for the next year, before rapidly deteriorating to blindness in her right eye over 4 months. A, An axial slice from her initial scan, showing an enlarged, enhancing right optic nerve (arrow) and a normal left optic nerve. B, A slice from the scan performed at onset of progression, demonstrating an obvious increase in size of the right optic nerve.

with isolated optic nerve gliomas, only the intraorbital portion was involved in 9 patients.

Symptomatic versus Asymptomatic Symptoms or signs attributable to the OPG were present in 32 patients (59.3%) at the time of diagnosis, whereas 22 patients were asymptomatic at the initial eye examination. Table 1 summarizes the findings according to these subgroups. Decreased visual acuity was present in 23 (42.6%) patients, proptosis in 10 patients (18.5%), and nystagmus in 1 patient. Twenty-five patients (46.3%)

Figure 3. Gadolinium-enhanced, T1-weighted magnetic resonance imaging (MRI) of a 5-year-old asymptomatic patient who underwent neuroimaging to investigate poor coordination. A, A sagittal slice from the initial scan, showing an enhancing mass in the left hypothalamus (arrow). B, The corresponding slice from a follow-up MRI performed 1 year later, showing complete resolution of these findings.

had evidence of optic disc atrophy or swelling on examination. Visual symptoms were present in 7 patients at the time of diagnosis of NF1. In the asymptomatic group, 8 patients had routine neuroimaging, and 14 patients underwent imaging to investigate other symptoms, most frequently headaches or macrocephaly. Both groups had similar proportions of chiasmal involvement. Figure 1 demonstrates the distribution of age at diagnosis in our population. The average age at OPG diagnosis was 5.2 years (SD, 3.6), with 17 patients (31.5%) diagnosed after 6 years of age. The average age of the symptomatic group was 4.8 years (SD, 3.3), compared with 5.8 years (SD, 3.8) for the asymptomatic group,

571

Ophthalmology Volume 111, Number 3, March 2004 Table 3. Level of Visual Impairment in Each Eye by Mode of Treatment, Location of Optic Glioma, and Age Level of Visual Impairment* Worse Eye Variable Treatment modality Surgery Radiotherapy Chemotherapy Combination therapy† Conservative Site of glioma Isolated optic nerve Chiasmal involvement Age group ⱕ6 yrs ⬎6 yrs

Better Eye

Mild

Moderate

Severe

Mild

Moderate

Severe

Total

0 0 0 0 29

0 4 1 0 3

5 3 1 3 5

4 4 2 1 34

0 2 0 1 1

1 1 0 1 2

5 7 2 3 37

10 19

1 7

5 12

16 29

0 4

0 5

16 38

18 11

6 2

13 4

31 14

4 0

2 3

37 17

*Mild, 6/6 to 6/12 Snellen equivalent; moderate, 6/15 to 6/60; severe, ⬍6/60. Combination of 2 of the modalities surgery, radiotherapy, and chemotherapy.

†

although the difference between the groups was not statistically significant (P ⫽ 0.5). Table 2 depicts mode of presentation at the time of OPG diagnosis according to age (younger and older patients). Proptosis was more common in the younger group, whereas precocious puberty was frequently present in children presenting after 6 years of age. Two patients presenting with precocious puberty had completely normal visual findings at diagnosis, and a third had only temporal pallor in one eye. The older group also had a higher proportion of patients with chiasmal involvement (76.5% vs. 67.6%) and positive family history (45.7% vs. 23.5%); however, the differences between the groups were not statistically significant.

Progression Twenty-eight patients (51.9%) had clinical and/or radiological progression of their OPG during follow-up. Of these, 12 patients (42.9%) had both clinical and radiological features of progression, 4 patients (14.3%) had clinical progression with no change in radiological size of the tumor, and 2 patients (7.1%) had radiological progression without clinical deterioration. The remaining 10 patients all showed clinical progression, with no corresponding follow-up imaging available for comparison, as progression occurred close to the time of OPG diagnosis. Progression occurred within 1 year of diagnosis of OPG in the majority of patients (22 patients, all in the symptomatic group). In this subgroup, 16 patients (72.7%) had chiasmal involvement, similar to the proportion with chiasmal involvement in the whole group (70.4%). The remaining 6 developed symptoms and signs of progression between 1 and 6 years from the time of OPG diagnosis, with an average age of 8.3 years at the time of progression (range, 5–14 years). All 6 patients had optic chiasm involvement. Two of these patients had both visual loss and an increase in size of the tumor on MRI, and Figure 2 shows the progression in one of these patients. Two had marked visual loss with no radiological change, and 1 patient had an increase in radiological size 2 years after diagnosis, with no real change in visual acuity. The sixth patient presented with no light perception in her right eye when she was 8 (there was no record of previous eye examinations or neuroimaging) and was found to have a right optic nerve and chiasmal glioma on MRI. Her vision remained stable for 2 years, and then over a period of 6 months the vision in her left eye deteriorated from 6/5 to 6/36 (Snellen equivalent), with peripheral

572

visual field loss. Repeat MRI showed an increase in size of the chiasmal/hypothalamic tumor. She underwent surgery to debulk this tumor with partial removal of her right optic nerve, and postoperatively her vision improved to 6/12 in the left eye, with an improved visual field measurement. Four years later, at the age of 14, vision in her left eye again deteriorated to 6/120, with marked field loss accompanied by an increase in radiological size, and a second debulking operation was performed at this stage. Histopathology reports from both operations indicated a low-grade astrocytoma. At the most recent follow-up she remained blind in the right eye and had faint perception of light in the left eye. Precocious puberty was present in 13 patients (32.5%): 5 at the time of OPG diagnosis and a further 8 during follow-up. Three patients had evidence of growth hormone deficiency requiring replacement therapy. Two of these patients had prior radiotherapy for chiasmal gliomas, whereas the third patient had a left optic nerve and chiasmal glioma that was managed conservatively. Follow-up MRIs on 6 patients (11.1% of patient total) showed evidence of a decrease in the size of the OPG. The patients’ ages at the time of radiological regression were 4, 5, 6, 7, 9, and 11 years, and the interval between diagnosis of OPG and regression ranged between 1 and 7 years. In 3 of these patients, 1 with bilateral isolated optic nerve involvement and 2 with chiasmal/ hypothalamic involvement, there was complete resolution of the tumor appearance. All 3 of these patients were asymptomatic at diagnosis and throughout their follow-up. Figure 3 illustrates the initial and follow-up MRI findings of one of these patients. The MRIs of the remaining 3 patients showed a radiological reduction in size of their OPGs. Two of these patients had both optic nerve and chiasmal involvement and had visual abnormalities at diagnosis (both had optic disc pallor, and 1 had moderate visual impairment in one eye), whereas the third patient had an isolated chiasmal/hypothalamic tumor and remained asymptomatic. All 6 patients were managed conservatively. None of the patients in our series showed spontaneous improvement in visual acuity throughout follow-up.

Previously Normal Imaging None of our patients had a normal MRI before being diagnosed with an OPG; however, 8 patients had normal CT scans before diagnosis. Of these patients, none had visual symptoms or signs at the time of negative imaging. Five patients (with normal scans at

Thiagalingam et al 䡠 Optic Gliomas in NF1 ages 2, 3, 3, 4, and 4 years, respectively) remained asymptomatic throughout their follow-up. Three had isolated optic nerve gliomas and 2 had isolated chiasmal gliomas. Of the remaining 3 patients, one was a 10-year-old boy who had serial normal CT scans up to the age of 8 and presented 2 years later with a rapid decline in vision due to a chiasmal tumor. The second patient also had no evidence of an OPG on a routine CT scan but a year later, at the age of 2, presented with decreased vision and was found to have a bilateral optic nerve/chiasmal glioma on MRI. Both of these patients subsequently developed severe visual impairment in both eyes. The third patient had a normal routine CT scan at 5 months of age before being diagnosed at 16 months with a bilateral intracranial optic nerve and chiasmal tumor with a repeat routine CT scan. This patient’s details are described more fully below.

Treatment Seventeen patients (31.5%) underwent treatment for OPG. Fourteen were treated with a single modality (surgery, 5; radiotherapy, 7; chemotherapy, 2). Two patients had radiotherapy after chemotherapy, and 1 patient had radiotherapy after surgery. All 3 patients in the treated group with isolated optic nerve tumors underwent surgery to excise the affected nerve. The majority of those treated (88.2%) belonged to the symptomatic group, as shown in Table 1, and treatment was commenced because of a clinical deterioration in visual function. Two patients in the asymptomatic group were given chemotherapy and radiotherapy, respectively, on the basis of an increase in radiological size of their chiasmal tumors. One of these patients, mentioned in the section above, was diagnosed with routine imaging at the age of 16 months. His initial eye examination was normal; however, a repeat MRI a few months later showed an increase in size of the chiasmal glioma. After treatment with chemotherapy, the tumor showed a radiological reduction in size. The patient remained symptom-free until 6 years later when, at the age of 7, he developed visual loss, and MRI indicated a slight growth of both optic nerves and the optic chiasm. A further course of chemotherapy was given at this stage, and his corrected visual acuity stabilized at 6/36 (Snellen equivalent) in the right eye and 6/12 in the left eye.

Visual Outcome Table 3 outlines visual impairment in each eye at the most recent eye examination according to treatment modality, site of OPG, and age at diagnosis. Approximately half (29 patients [53.7%]) had mild or no visual impairment in their worse eye, 8 (14.8%) had moderate impairment, and 17 (31.5%) had severe impairment. The majority had only mild or no impairment in the better eye (83.3%). Patients with isolated optic nerve gliomas had a proportion of severe visual impairment in the worse eye similar to that of patients with chiasmal involvement; however, all 9 patients with moderate to severe impairment in the better eye had chiasmal involvement. Those who were younger than 6 years at diagnosis had a slightly higher proportion with severe visual impairment in the worse eye (35.1%), compared with patients presenting after 6 years of age (23.5%), but this difference was not statistically significant. In one patient with moderate impairment of one eye, amblyopia secondary to ptosis caused by a concurrent plexiform neurofibroma of the upper eyelid may have contributed to the visual loss. Twelve patients who underwent treatment were blind or had severe impairment in their worse eye at their most recent examination, and the remaining 5 patients had moderate impairment. However, the majority of patients in the treatment group (64.7%) had mild or no impairment in their better eye, with 3 patients having moderate impairment and 3 severe im-

pairment. In comparison, only 5 of the 37 patients managed conservatively had severe impairment in the worse eye, and 3 of these patients had moderate or severe impairment in their better eye at the most recent examination.

Discussion Although optic gliomas generally behave as benign tumors, the potential for visual loss has concerned many clinicians and has brought forward questions on how best to monitor patients with NF1 to maximize early detection of OPG, and to optimize monitoring of OPG once diagnosed. A number of studies have attempted to answer these questions; however, the natural history of these tumors is still unclear. Table 4 presents a summary of results of recent studies of OPGs in NF1 patients, with which our results can be compared. The results of our review of 54 patients with NF1 show that the prevalence of OPGs is approximately 6.6% in our population. This figure is considerably lower than in studies in which routine neuroimaging was performed (Table 4), as there is likely to be an under-representation of asymptomatic OPG patients in our series. A number of other studies that also did not utilize routine neuroimaging had prevalence rates ranging from 14% to 16%,6,17,18 but this may be explained partly by selection bias. Our patient cohort is derived from a nonspecific management clinic for all patients with suspected NF1 and therefore is more likely to represent the general NF1 population. The majority of our patients (75%) had optic chiasm involvement, a proportion similar to that of other studies (Table 4). Almost half had moderate to severe visual impairment in the worse eye. Although most patients (83.3%) retained good functional vision with normal or mild impairment in the better eye, 9 patients (16.7%) had moderate/ severe bilateral impairment. The range of visual acuity tests performed on our patients represents a weakness of our study, given that tests such as matching letters can overestimate the true visual acuity and therefore may not be directly comparable to other tests. It highlights the difficulties faced by clinicians in accurately assessing acuity in very young children, and indicates the importance of examining for other signs of an optic glioma, such as optic atrophy (present at diagnosis in almost half of our patients, and in proportions similar to those of visual acuity loss). Our findings raise a number of issues regarding current understanding of the behavior of these tumors. The NIH guidelines on monitoring of OPGs suggest annual ophthalmological examinations only up to the age of 6,12 as earlier investigations indicated that this was the danger period for presentation and progression of OPGs. A study by Listernick et al7 found that, in 33 patients with NF1 who were routinely imaged, the median age at OPG presentation was 4.2 years. Similarly, Packer et al19 reported a median age of 4.0 years in 24 patients, some of whom were routinely imaged. In contrast, almost one third of our patients presented after the age of 6 years, and over half of these children presented with visual symptoms or signs. A number of other authors have also reported older presenting ages, as seen in Table 4.20 It is unlikely that these all

573

Ophthalmology Volume 111, Number 3, March 2004 Table 4. Summary of Previous Publications on Optic Author (Year of Publication)

No. of NF1 Patients

No. with OPG (%)

Age at OPG Diagnosis (yrs)

Asymptomatic at OPG Diagnosis (%)

Routine neuroimaging not performed Present study (2004)

861

54 (6.6)

5.2 (mean)

22/54 (41)

Gayre et al (2001)‡§

—

23

9.8 (mean)

—

Grill et al (2000)¶# Deliganis et al (1996)**

— —

51 16

— 5.25 (mean)

0 —

Kuenzle et al (1994)††

—

21

7.1 (mean)

6/21 (29)

Pascual-Castroviejo et al (1994)‡‡ Lund and Skovby (1991)§§

221 74

31 (14) 16 (22)

4 (mean) 6.4 (mean)

— 31%

Hochstrasser et al (1988)㛳 㛳

133

20 (15)

6.8 (mean)

5 (25)

Routine neuroimaging performed Chateil et al (2001)##

—

14

5.1 (mean)

11/14 (79)

Balcer et al (2001)*** Wolkenstein et al (1996)††† Listernick et al (1994)‡‡‡

— 93 176

43 12 (13) 33 (19)

3.0 (median) 5.8 4.2 (median)

12/20 (60) 100% 25 (76)

Rossi et al (1994)§§§ Packer et al (1988)¶㛳 㛳 㛳

25 —

6 (24) 24

6–21 (range) 4.0 (median)

100% 4/24 (17)

Lewis et al (1984)###

217

33 (15)

16 (mean)

22/33 (67)

OPG ⫽ optic pathway glioma. *Diagnosed with National Institutes of Health criteria, unless indicated with ‡. † Deterioration in symptoms and signs after a clear period of stability. § Gayre GS, Scott IU, Feuer W, et al. Long-term visual outcome in patients with anterior visual pathway gliomas. J Neuroophthalmol 2001;21:1–7. 㛳 Includes results for non-NF1 patients. ¶ Series included patients without NF1, who were excluded from this table. # Grill J, Laithier V, Rodriguez D, et al. When do children with optic pathway tumours need treatment? An oncological perspective in 106 patients treated **Deliganis AV, Geyer JR, Berger MS. Prognostic significance of type 1 neurofibromatosis (von Recklinghausen disease) in childhood optic glioma. †† Kuenzle C, Weissert M, Roulet E, et al. Follow-up of optic pathway gliomas in children with neurofibromatosis type 1. Neuropediatrics 1994;25:295–300. ‡‡ Pascual-Castroviejo I, Martinez Bermejo A, Lo´ pez Martin V, et al. Optic gliomas in neurofibromatosis type 1 (NF-1): presentation of 31 cases. Neurologia §§ Lund AM, Skovby F. Optic gliomas in children with neurofibromatosis type 1. Eur J Pediatr 1991;150:835– 8. 㛳 㛳 Hochstrasser H, Boltshauser E, Valavanis A. Brain tumors in children with von Recklinghausen neurofibromatosis. Neurofibromatosis 1988;1:233–9. ## Chateil JF, Soussotte C, Pe´ despan JM, et al. MRI and clinical differences between optic pathway tumours in children with and without neurofibro***Balcer LJ, Liu GT, Heller G, et al. Visual loss in children with neurofibromatosis type 1 and optic pathway gliomas: relation to tumor location by ††† Wolkenstein P, Fre`che B, Zeller J, Revuz J. Usefulness of screening investigations in neurofibromatosis type 1: a study of 152 patients. Arch Dermatol ‡‡‡ Listernick R, Charrow J, Greenwald M, Mets M. Natural history of optic pathway tumors in children with neurofibromatosis type 1: a longitudinal study. §§§ Rossi LN, Pastorino G, Scotti G, et al. Early diagnosis of optic glioma in children with neurofibromatosis type 1. Childs Nerv Syst 1994;10:426 –9. 㛳 㛳 㛳 Packer RJ, Bilaniuk LT, Cohen BH, et al. Intracranial visual pathway gliomas in children with neurofibromatosis. Neurofibromatosis 1988;1:212–22. ### Lewis RA, Gerson LP, Axelson KA, et al. von Recklinghausen neurofibromatosis. II. Incidence of optic gliomata. Ophthalmology 1984;91:929 –35.

represent late diagnosis of a long present tumor.21 As noted, one of our patients had normal eye findings at the age of 8 years, before presenting with visual loss a year later. Another finding of concern is that even children who were diagnosed at a young age showed clinical deterioration at much later ages than previously reported, often after a period of apparent stability. Grill et al22 recently reported that 47% of their patients with NF1 (n ⫽ 24) had evidence of tumor progression requiring treatment after age 6; however, they did not specify the time taken from diagnosis to progression for this subgroup. Balcer et al23 reported loss of visual

574

acuity in a 17-year-old patient with normal acuity 1 year previously, and Deliganis et al24 have observed the progression of OPGs in patients with NF1 up to 8 years after diagnosis. Although the small number of our patients who had histological confirmation of their gliomas prevents us from making too detailed an analysis, this questions the current understanding of the biologic behavior of these lesions. Hoyt and Baghdassarian8 postulated that optic gliomata are clinically indolent tumors that tend to enlarge, cause symptoms early in life, and remain static thereafter. The sporadic and delayed nature of the progression of some gliomas in our

Thiagalingam et al 䡠 Optic Gliomas in NF1 Gliomas in Neurofibromatosis Type 1 (NF1)* Patients Optic Chiasm Involvement (%)

Progression of OPG after Diagnosis (%)†

38/54 (70)

6/54 (11)

19/23 (83)

11/26㛳

11 (⫾35)/51 12/16 (75)

12/51 (24) 5/16 (31)

14/21 (67)

2/21 (10)

22/31 (71) 75%

— 35%

13/20 (65)

—

7/14 (50)

5 (radiological progression)

39/43 (91) — 18/33 (55)

12/43 (28) — 3/33 (9)

0% 21/24 (88)

— 9/24 (38)

33%

—

Treatment (n) All modalities ⫽ 17 Conservative ⫽ 37 Treatment ⫽ 10/23 Conservative ⫽ 13/23 All treated Surgery ⫽ 1 Radiotherapy ⫽ 10 Surgery/biopsy ⫽ 8 Conservative ⫽ 13 Radiotherapy ⫽ 7 Radiotherapy ⫽ 11 Surgery ⫽ 3 Conservative ⫽ 2 — Radiotherapy ⫽ 2 Chemotherapy ⫽ 1 Conservative ⫽ 11 — — Surgery ⫽ 3 Chemotherapy ⫽ 2 — Radiotherapy ⫽ 10 Chemotherapy ⫽ 2 Conservative ⫽ 12 —

Visual Outcome (Level of Impairment) Mild ⫽ 29 Moderate ⫽ 8 Severe ⫽ 17 (worse eye) ⬎50% still able to drive ⬍25% legally blind㛳 — — Blind in 1 eye ⫽ 5 Blind in both eyes ⫽ 1 Decreased acuity ⫽ 7 Asymptomatic ⫽ 3 — Improved ⫽ 10% Stable ⫽ 55% Worse ⫽ 35% Blind ⫽ 9/20 —

Follow-up (yrs) 8.6 (mean) 9 (mean)㛳 7 (median) 10.2 (mean) 9.0 (mean)

— 5.8 (mean) 6.9 (mean) 3.5 (mean)

Visual acuity ⬍ 20/20 ⫽ 47%¶ — —

3.0 (median) — 3.4 (median)

— —

— 5.0 (median)

—

—

in a single centre. Eur J Pediatr 2000;159:692– 6. Neurosurgery 1996;38:1114 – 8, discussion 1118 –9. 1994;9:173–7. matosis. Br J Radiol 2001;74:24 –31. magnetic resonance imaging. Am J Ophthalmol 2001;131:442–5. 1996;132:1333– 6. J Pediatr 1994;125:63– 6.

series, as well as others, suggests that, though the majority may remain quiescent, clinicians should be vigilant in observing all patients with a diagnosed OPG for longer duration. Our personal experience and that of other authors suggest that follow-up should be at least until the patient is 17 years old. Attempts at determining predictors for increased tumor morbidity or progression have so far been inconclusive. Histological studies indicate that the morphology of OPGs in NF1 patients is similar to that in isolated OPGs, even though survival rates are higher for NF1 patients.5 Tumor

location has also been suggested as a predictor of survival, with some reports indicating that chiasmal gliomas behave more aggressively relative to optic nerve gliomas.25–27 In our series, none of the patients died as a result of the OPG, but all 6 patients with later progression of disease had chiasmal involvement. Although similar levels of visual impairment were present in the worse eye for both locations, a higher proportion of patients with chiasmal involvement had moderate or severe visual impairment in the better eye, indicating that those with isolated optic nerve gliomas had a better overall functional visual outcome.

575

Ophthalmology Volume 111, Number 3, March 2004 There is much controversy regarding the use of screening neuroimaging in patients with NF1 for the early diagnosis of optic gliomas. The NIH guidelines do not recommend routine imaging because it does not alter management of these patients. Others have even postulated that patients who are asymptomatic at diagnosis may have diffuse hyperplastic gliosis rather than a glioma.28 Of the patients in our series who were diagnosed by means of a routine scan, only 2 showed signs of late radiological progression requiring treatment, and neither had visual impairment at the time of treatment. None of the patients in the asymptomatic group had progressed to severe visual impairment at follow-up, compared with over half the patients in the symptomatic group. It is likely that this result is due to patients in the symptomatic group having a more aggressive form of glioma, rather than early diagnosis in the 2 asymptomatic patients resulting in earlier treatment and more favorable outcome. Overall, it seems that routine imaging has not been of great benefit to the majority of asymptomatic patients in our series. Comparison of our results with those of studies that utilized routine neuroimaging (Table 4) was not possible due to insufficient data on the progression and visual outcome of asymptomatic patients, with the exception of Listernick et al,7 who concluded that the usefulness of screening was limited. A number of our patients had normal CT scans before being diagnosed with an OPG, in one case only a year before developing visual loss. Although it is recognized that CT scans are not as sensitive in detecting gliomas as cranial MRI, other reports of similar cases also include previously normal MRI.29,30 A major difference between these studies and our cohort is that our patients were much older at the time of the normal scan, the oldest child being 8. These findings suggest that a negative scan does not rule out later emergence of a glioma, and provide further evidence for possible later progression of OPGs. The majority of our patients were treated conservatively, and patients who underwent treatment were more likely to have severe visual impairment in their worse eye at followup. However, this cannot be attributed simply to treatment effect, as patients who underwent treatment were likely to have severe visual loss before treatment onset, and one of the principal aims of treatment, especially in those who had surgery, was to prevent loss of vision in the better eye, rather than improve it in the worse eye. The majority of patients in our study who had treatment had mild or no visual impairment in their better eye. Gayre et al4 had similar findings, whereby treated patients progressed to have poor vision in the treated eye but retained good vision in the other eye. The descriptive design of this study prevents us from making further conclusions as to the usefulness of treatment. For the clinician, the decision to commence treatment rather than continue conservative management is complicated by reports of spontaneous regression of OPGs in some patients, as demonstrated by serial MRIs.31–35 In the majority of reported cases to date, there was no corresponding improvement in visual function at the time of radiological regression.32–34 Similarly, in our series, of the 6 patients with radiological regression, only 2 had visual impairment,

576

and neither showed clinical improvement at the time of regression. However, other authors have documented spontaneous visual improvement in patients with OPGs,36 and Parsa et al31 recently reported 4 NF1 patients with radiological regression of their OPGs, 3 of whom showed some improvement in visual acuity. The implications of these findings are that there is a need for careful follow-up of patients to establish progression before commencing treatment, and to be aware that when treatment is given, a favorable outcome may be due partly to spontaneous regression. Also of note, the 4 patients described by Parsa were aged 3, 7, 14, and 15 years at commencement of regression, and in our series the age at regression ranged between 4 and 11 years, indicating that regression can occur in older children. In summary, almost half the patients in our cohort with OPGs had moderate or severe impairment of one eye. Although the majority of patients retained reasonable functional vision, it is of some concern that 16.7% had bilateral moderate/severe visual impairment. Particular attention should be paid to monitoring signs in children under 6 years of age, as difficulty in identifying visual loss in this age group is compounded by the young child’s ability to adapt to reduced vision. The behavior of OPGs can be quite variable, with visual loss occurring at a later age than previously reported. Patients with chiasmal involvement seem to be more susceptible to delayed progression and visual morbidity. These findings indicate that patients with NF1 should have regular ophthalmic surveillance until at least 17 years of age, and those with known chiasmal OPGs need continued monitoring into adulthood. Acknowledgments. The authors gratefully acknowledge Dr Adam Steinberg for reviewing the images, and the valuable contribution of the clinicians who provided follow-up data on patients.

References 1. Stern J, Jakobiec FA, Housepian EM. The architecture of optic nerve gliomas with and without neurofibromatosis. Arch Ophthalmol 1980;98:505–11. 2. Lewis RA, Gerson LP, Axelson KA, et al. von Recklinghausen neurofibromatosis. II. Incidence of optic gliomata. Ophthalmology 1984;91:929 –35. 3. Billson FA, Minty CCJ. Tumours of the eye and orbit. In: Jones PG, Campbell PE, eds. Tumours of Infancy and Childhood. Melbourne: Blackwell Scientific Publications; 1976: 411–20. 4. Gayre GS, Scott IU, Feuer W, et al. Long-term visual outcome in patients with anterior visual pathway gliomas. J Neuroophthalmol 2001;21:1–7. 5. Rush JA, Younge BR, Campbell RJ, MacCarty CS. Optic glioma: long-term follow-up of 85 histopathologically verified cases. Ophthalmology 1982;89:1213–9. 6. Lund AM, Skovby F. Optic gliomas in children with neurofibromatosis type 1. Eur J Pediatr 1991;150:835– 8. 7. Listernick R, Charrow J, Greenwald M, Mets M. Natural history of optic pathway tumors in children with neurofibromatosis type 1: a longitudinal study. J Pediatr 1994;125:63– 6. 8. Hoyt WF, Baghdassarian SA. Optic glioma of childhood: natural history and rationale for conservative management. Br J Ophthalmol 1969;53:793– 8.

Thiagalingam et al 䡠 Optic Gliomas in NF1 9. Listernick R, Darling C, Greenwald M, et al. Optic pathway tumors in children: the effect of neurofibromatosis type 1 on clinical manifestations and natural history. J Pediatr 1995;127: 718 –22. 10. Wright JE, McDonald WI, Call NB. Management of optic nerve gliomas. Br J Ophthalmol 1980;64:545–52. 11. Klug GL. Gliomas of the optic nerve and chiasm in children. Aust N Z J Surg 1977;47:596 – 600. 12. Listernick R, Louis DN, Packer RJ, Gutmann DH. Optic pathway gliomas in children with neurofibromatosis 1: consensus statement from the NF1 Optic Pathway Glioma Task Force. Ann Neurol 1997;41:143–9. 13. North K. Neurofibromatosis type 1: review of the first 200 patients in an Australian clinic. J Child Neurol 1993;8:395– 402. 14. Byrd SE, Harwood-Nash DC, Fitz CR, et al. Computed tomography of intraorbital optic nerve gliomas in children. Radiology 1978;129:73– 8. 15. Jacoby CG, Go RT, Beren RA. Cranial CT of neurofibromatosis. AJR Am J Roentgenol 1980;135:553–7. 16. SPSS [computer program]. Version 10.0. Chicago: Statistical Packages for the Social Sciences Inc; 1999. 17. Pascual-Castroviejo I, Martinez Bermejo A, Lo´ pez Martin V, et al. Optic gliomas in neurofibromatosis type 1 (NF-1): presentation of 31 cases. Neurologia 1994;9:173–7. 18. Hochstrasser H, Boltshauser E, Valavanis A. Brain tumors in children with von Recklinghausen neurofibromatosis. Neurofibromatosis 1988;1:233–9. 19. Packer RJ, Bilaniuk LT, Cohen BH, et al. Intracranial visual pathway gliomas in children with neurofibromatosis. Neurofibromatosis 1988;1:212–22. 20. Kuenzle C, Weissert M, Roulet E, et al. Follow-up of optic pathway gliomas in children with neurofibromatosis type 1. Neuropediatrics 1994;25:295–300. 21. Listernick R, Charrow J. Reply: screening for optic gliomas in neurofibromatosis type 1: the role of neuroimaging [letter]. J Pediatr 1995;127:507– 8. 22. Grill J, Laithier V, Rodriguez D, et al. When do children with optic pathway tumours need treatment? An oncological perspective in 106 patients treated in a single centre. Eur J Pediatr 2000;159:692– 6. 23. Balcer LJ, Liu GT, Heller G, et al. Visual loss in children with

24.

25. 26. 27. 28. 29. 30. 31. 32.

33. 34. 35. 36.

neurofibromatosis type 1 and optic pathway gliomas: relation to tumor location by magnetic resonance imaging. Am J Ophthalmol 2001;131:442–5. Deliganis AV, Geyer JR, Berger MS. Prognostic significance of type 1 neurofibromatosis (von Recklinghausen disease) in childhood optic glioma. Neurosurgery 1996;38:1114 – 8, discussion 1118 –9. Stern J, DiGiacinto GV, Housepian EM. Neurofibromatosis and optic glioma: clinical and morphological correlations. Neurosurgery 1979;4:524 – 8. Borit A, Richardson EP Jr. The biological and clinical behaviour of pilocytic astrocytomas of the optic pathways. Brain 1982;105:161– 87. Chutorian AM, Schwartz JF, Evans RA, Carter S. Optic gliomas in children. Neurology 1964;14:83–95. Rubinstein LJ. Pathological features of optic nerve and chiasmatic gliomas. Neurofibromatosis 1988;1:152– 8. Massry GG, Morgan CF, Chung SM. Evidence of optic pathway gliomas after previously negative neuroimaging. Ophthalmology 1997;104:930 –5. Listernick R, Charrow J, Greenwald M. Emergence of optic pathway gliomas in children with neurofibromatosis type 1 after normal neuroimaging results. J Pediatr 1992;121:584 –7. Parsa CF, Hoyt CS, Lesser RL, et al. Spontaneous regression of optic gliomas: thirteen cases documented by serial neuroimaging. Arch Ophthalmol 2001;119:516 –29. Gottschalk S, Tavakolian R, Buske A, et al. Spontaneous remission of chiasmatic/hypothalamic masses in neurofibromatosis type 1: report of two cases. Neuroradiology 1999;41: 199 –201. Schmandt SM, Packer RJ, Vezina LG, Jane J. Spontaneous regression of low-grade astrocytomas in childhood. Pediatr Neurosurg 2000;32:132– 6. Perilongo G, Moras P, Carollo C, et al. Spontaneous partial regression of low-grade glioma in children with neurofibromatosis-1: a real possibility. J Child Neurol 1999;14:352– 6. Brzowski AE, Bazan C, Mumma JV, Ryan SG. Spontaneous regression of optic glioma in a patient with neurofibromatosis. Neurology 1992;42:679 – 81. Liu GT, Lesell S. Spontaneous visual improvement in chiasmal gliomas. Am J Ophthalmol 1992;114:193–201.

577