Pattern of Intracranial Space-Occupying Lesions: The Experience of the King Khalid University Hospital

Introduction

Over the past three decades, many reports have suggested that both incidence and pattern of intracranial neoplasms are subject to considerable geographic and racial variation.^1–11 Knowledge of the regional peculiarities of these lesions may, therefore, help in identification of possible risk factors and also in establishing measures for improved diagnosis, treatment, and outcome.

Up-to-date data on intracranial space-occupying lesions in Saudi Arabia are lacking almost entirely. This is largely because neurosurgical practice in this country was limited mostly to emergency treatment and care of head-injured patients, and it was not until after 1975 when the King Faisal Specialist Hospital in Riyadh was established that modern neurosurgery became available in Saudi Arabia. However, because of the referral mode to that hospital and other considerations, only highly selected cases of intracranial tumors received treatment there. This situation has changed greatly over the past few years, as modern neurosurgical units have evolved in most of the major cities of the country. The Neurosurgical Division of the King Khalid University Hospital (KKUH) in Riyadh started in 1982. Besides being an academic and referral center, it provides primary care to all Saudi citizens and government employees of other nationalities. Therefore, the cases admitted there are unselected and should be representative for the whole population.

This paper attempts to summarize and analyze statistically all intracranial space-occupying lesions diagnosed and treated at the KKUH over the past 5 years so as to provide, for the first time, preliminary data on the relative frequencies and locations, as well as age and sex distributions of intracranial space-occupying lesions in Saudi Arabia.

PATIENTS AND METHODS

Between October 1982 and September 1987, 212 patients with intracranial space-occupying lesions were diagnosed and/or treated at the Neurosurgical Division of KKUH in Riyadh. The patients were referred from all parts of the Kingdom. In this study, the term “intracranial space-occupying lesion” is defined as any neoplasm, benign or malignant, primary or secondary, as well as any inflammatory or parasitic mass lying mostly within the cranial cavity. Thus, a lesion originating in a contiguous structure may be included under this definition if its intracranial part prevails. On the other hand, intracranial hematoma of all kinds, aneurysms, and arachnoid cysts have been excluded.

All cerebral neoplasms in this study were grouped basically according to the classification of the World Health Organization.¹² Abscesses and granulomas were considered separately. For the purpose of simplification and uniformity of the differing terminology used by the various pathologists, the term “malignant astrocytoma” in the subgroups of the neuroepithelial tumors has been applied to cover anaplastic astrocytomas (grade III and IV of Kernohan¹³), as well as glioblastoma multiforme.

Each patient had at least one cranial computed tomographic (CT) scan. Of the 212 patients, the lesion was confirmed histopathologically in 150 cases (71%). In 33 of the 62 cases without histopathologic verification (15% of the total patient group), it was possible to ascertain the type of lesion on the basis of the clinical background, such as pathologically elevated pituitary hormone levels, response to a specific treatment (e.g., antituberculous), and characteristic appearance on combined radiographic tests (CT scan and angiography). Thus, on the whole, possible diagnostic inaccuracy was relatively small and limited to only the remaining 29 histopathologically unconfirmed cases (14% of the total) in which the clinical findings were less specific, and the diagnosis was based mainly on the cranial CT scan and the course of the disease.

RESULTS

Among the total of 212 patients with intracranial space-occupying lesions, there were 184 (87%) cerebral neoplasms and 28 (13%) nonneoplastic masses (mostly abscesses and granulomas). The 150 cases with histopathologic verification revealed nearly the same percentages of these two components (Table 1).

Table 1. Types of all and histopathologically confirmed cases of intracranial space-occupying lesions in both sexes.

Intracranial Neoplasms

Of all intracranial neoplasms 51.8% presented between the ages of 20 and 50 years with maximum of occurrence during the fourth decade of life (Figure 1). In 12% of the cases, the patients were less than 15 years of age. The overall male-to-female ratio was 1.45:1. The pattern of age-specific relative frequencies for each sex alone was rather irregular, probably due to the small number of patients in each group.

Of the 184 tumors 139 (75.5%) were located above the tentorium while 39 (21.2%) were below the tentorium. The remaining six patients (3.3%) had multiple lesions that were both supra- and infratentorial. The ratio of supratentorial to infratentorial lesions was lowest (1.1:1) in children and young adults below the age of 20 years. Midline lesions constituted 44.6%, whereas bilateral processes were found in 3.8% of all cases.

Table 1 gives the numbers of relative frequencies of the various types of brain tumors for all 184 cases. In comparison, the 129 cases with histopathologic verification showed basically a similar pattern of the different neoplasms except for the comparatively lower percentages for pituitary adenomas and metastatic lesions. This difference may be attributed to a reduced rate of surgery in these two types of lesions. Figure 2 shows the percentage distribution of the different brain tumors in both sexes.

Neuroepithelial tumors: These were the most common intracranial neoplasms and comprised 39.7% of the total. The numbers and percentages of the different histologic types of this group are shown in Table 2. A total of 67.1% of these tumors occurred during the first four decades of life. Furthermore, they constituted 73% of all brain tumors seen below the age of 15 years. The male-to-female ratio was 2:1. Whereas malignant astrocytomas, with 41.1% the most common type of neuroepithelial tumors, occurred preferentially during the second half of life, low-grade astrocytomas, ependymomas, and medulloblastomas were predominantly tumors of children and young adults. There was only one case of oligodendroglioma (1.4% of all gliomas). The 18 unspecified gliomas affected mainly brain stem, basal ganglia, and the pineal region, areas considered unsuitable for conventional neurosurgical intervention for which no biopsy was obtained. Some 35.6% of all glial tumors and 57.9% of those occurring before the age of 15 years were located infratentorially.

Table 2. Relative frequency of the different types of neuroepithelial tumors.

Metastatic tumors: Secondary neoplasms constituted 8.2% of the 184 intracranial tumors. Males were far more commonly affected than females (4:1). In five cases metastases into the cranial cavity occurred by continuous growth from adjacent structures, mainly nasopharynx and paranasal sinuses (three carcinomas) and soft tissue (two sarcomas). In the remaining 10 cases, cerebral metastases originated from the following distant primaries: lung cancer (four cases), breast carcinoma (one case), hepatoma (one case), rectal carcinoma (one case), and unknown (two cases).

Pituitary adenomas: These accounted for 16.8% of the 184 cerebral tumors. They occurred most frequently during the third and fourth decades of life and, below the age of 20, were quite rare. The male-to-female ratio was 1.65:1. Sixty-one percent of all adenomas were hormonally active, predominantly prolactinomas (12 of 19).

Meningiomas: These tumors constituted 22.8% and were the second most common type of all intracranial neoplasms. Females were affected almost twice as often as males. As a result, meningiomas were the commonest cranial neoplasm of the female. They occurred exclusively in the middle and higher age groups with maximum incidence during the fourth, fifth, and sixth decades of life. Only one case (2.4%) had multiple supra- and infratentorial meningiomas; 9.5% of all meningiomas were located infratentorially.

Neurinomas and malformative tumors: The incidence of neurinomas and malformative tumors (craniopharyngeoma, epidermoid, colloid cyst, and lipoma) was 3.8% and 4.3%, respectively. Vascular malformations comprised 1.6%, whereas tumors of vascular origin (hemangioblastoma and hemangiopericytoma) constituted a further 1.1%. Germinoma, chordoma, and osteoma each occurred once in this series and totaled 1.6% of the intracranial neoplasms.

Comparison of the pattern of intracranial neoplasms in Saudi and non-Saudi patients: Saudi citizens accounted for 120 of the 184 patients with intracranial neoplasms. The remaining 64 patients represented a rather nonhomogenous group of different other nationalities with 56 non-Saudi Arabs, five Southeast Asian nationals, and three Europeans and Americans. The relative frequency of meningiomas was significantly higher among Saudi patients (28.3% versus 12.5%), whereas for pituitary adenomas, the situation was reversed (13.3% versus 23.4%). The remaining intracranial neoplasms revealed a similar incidence in both groups (Table 3).

Table 3. Comparison of the pattern of intracranial neoplasms in Saudi and non-Saudi patients.

DISCUSSION

Although limiting this study only to histopathologically confirmed cases would have helped eliminate statistical errors potentially arising from erroneous diagnosis of unconfirmed lesions, such a measure may introduce other types of selection bias. In a country like Saudi Arabia where autopsies are extremely rare and limited to medicolegal cases, histopathologic studies are restricted to premortem operative specimens. Such studies will not include or only partially include lesions that can be diagnosed efficiently by means other than histopathologic examination, such as hormonally active pituitary adenomas, as well as lesions that have, for one reason or another, considered inoperable, such as brain stem gliomas or multiple metastatic cerebral lesions. Consequently, the data obtained would not coincide with the real distribution of the various lesions in the examined population. The statistical errors introduced herewith are hardly predictable. In contrast, the possible error arising from pathologically unconfirmed cases is estimated to be small and limited to less than 14% of all cases. For this reason, the inclusion criteria in this series have not been based only on histopathologic confirmation. Yet, in studies based on the experience of a single referral hospital, selection biases cannot be avoided completely, since the referral procedure itself is a selective process which affects the composition of the admitted cases.

Despite these limiting factors, this study shows that the present 184 cases of intracranial neoplasms share several features with other published series. Both age and sex distributions lie within the estimated ranges in other reports. In this study as in most series from Asian countries,^11,14–16 brain tumors occurred mostly during the fourth decade of life while in Western countries^2,7,18 during the fifth and sixth decades of life. This could be due to different age characteristics of the populations as well as different case ascertainment in the two country groups, with a higher rate of autopsies in the latter. The percentage of pediatric brain tumors (occurring below the age of 16 years) in the present series was 13% compared with 8% in Germany,² 13.7% in the United States,¹⁹ 16.8% in India,²⁰ 18.6% in China,¹¹ and 28.4% in Thailand.⁸ This figure seems to be related to the size of the pediatric population in each country. The male-to-female ratio of 1.45:1 in the present 184 patients was slightly higher than that reported from Japan¹⁶ and Western countries,^2,18 but less than that found in China,¹¹ India,¹⁴ and Ceylon.¹⁵ However, when Saudi patients in this series were considered alone, the male-to-female ratio was 1.2:1, suggesting that the higher preponderance of males in the whole group of patients was probably due to the higher proportion of males among expatriates working in Saudi Arabia.

Unlike other series of brain neoplasms which showed a higher incidence of infratentorial than supratentorial tumors in children,^19,21 this study shows continuous predominance of the supratentorial location throughout all age groups. Because of the small number of pediatric brain tumors in this series, this finding may not be definite. Nevertheless, a similar result was reported from Japan,¹ and it seems to correspond well with results from recent surveys²² which, compared with earlier studies, also indicate a diminishing predominance of infratentorial brain tumors in childhood, probably due to improved diagnostic aids.²³

As in all other series, tumors of neuroepithelial origin were also in the present study the most frequent type of intracranial neoplasm. They seem to be less common in Saudi Arabia than in Western countries,^2,24 China,¹¹ and Ceylon,¹⁵ but as frequent as in Japan^1,25 and Thailand.⁸ Comparison of the various subgroups of neuroepithelial tumors in the different series is more difficult because of the lack of uniformity in their classification. Nevertheless, astrocytomas, together with glioblastoma multiforme, constitute between 45.7% and 68% of all neuroepithelial tumors in most series.^1,26 The corresponding percentage in this study was 57.5%. The relative frequencies of medulloblastomas and ependymoma were within the reported limits. In contrast, oligodendroglioma was a comparatively rare tumor in this series. Two tumors were localized in the pineal region, but due to the absence of histopathologic tissue diagnosis, they were included under unspecified gliomas. All neuroepithelial tumors showed a varying degree of increased incidence in males over females; the estimated sex ratio in this study was in agreement with most of the other series.^2,8,11,14,24

The relative frequency of meningiomas in this report of 184 intracranial neoplasms was 22.8%, which is higher than the rates reported from all Western and other Asian countries.^{1–3,8,11,14,15,24} Only Percy et al²⁷ found that 35% of 170 primary cerebral neoplasms from Rochester were meningiomas. This unexpectedly high percentage is most likely due to the autopsy rates in Rochester since 57% of the reported meningiomas in that study were first discovered at necropsy. Another interesting aspect in this series is that meningiomas constituted 28% of the 120 brain tumors in Saudi patients compared to only 12.5% of the 64 cerebral neoplasms in the non-Saudis, whereas neuroepithelial tumors showed almost identical relative frequencies in both groups (39% and 40%). Considering that, in incidence, meningiomas have a clear preponderance of females over males, their apparently low number among the non-Saudis could have resulted from a disproportionately small female fraction among the expatriate population. In contrast, the high incidence of meningiomas among the Saudi patients cannot be simply attributed to a certain pattern of admission as there are no restrictions regarding patient's admission to KKUH. In addition, meningiomas are probably the most gratifying intracranial tumors on which to operate, and no neurosurgeon would like to refer them away. Therefore, it is concluded that meningiomas are probably more frequent in Saudi Arabia than in other countries. This view is supported by the findings of Chowdhary et al²⁸ who reported a ratio of 26% for meningiomas in a small series of 54 intracranial tumors from the Eastern Province. Also, the recent statistics about neurosurgical diseases at King Faisal Specialist Hospital show such a high proportion of meningiomas.²⁹

An explanation for this increased incidence of meningiomas in Saudis is not readily available. None of the patients in this series showed any of the signs of the neurocutaneous syndromes, and furthermore, there was no involvement of several members of the same family. Thus, it seems less likely that genetic predisposition alone would account for this finding. In addition, there are no differences in either age or sex distribution of the present series of meningiomas compared with those reported from other countries which might favor a different etiology. Currently, the high figures for meningiomas in this report need to be confirmed in a well-controlled and population-based study before other potential etiologic factors could be considered. A comparably high percentage of meningiomas has been published from several African countries.^4,6,7 However, it is debatable whether the available data truly indicate a higher absolute incidence of meningiomas on that continent. Indeed, the high figures for meningiomas might reflect the low incidence of gliomas among Africans, as demonstrated by Froman and Lipschitz⁴ among the Transvaal Bantus.

The relative incidence of the malformative tumors in this series was markedly below the rates reported from Japan,¹ Thailand,⁸ and China,¹¹ but within the ranges given by most Western series.^2,26 Nerve-sheath cell tumors and vascular malformations were far less frequent than the average of other reports.^1,2,8,11,24

The ratio of secondary brain tumors in the present study was near the upper limits estimated in other series^{1,2,8,11,14,15,24} which could be due to the relatively wide definition of metastatic lesions adopted here. As in reports from other countries,^11,18 lung cancer was also in this series the commonest distant source of cerebral metastases. However, the limited number of metastatic brain tumors in the present study does not allow further conclusions. This is especially true for the three females in this group, of whom one (a Saudi) had breast carcinoma while the other two (a Saudi and an American) suffered from lung cancer. In general, non-Saudis showed again a higher incidence of cerebral secondaries than Saudis (10.9% versus 6.7%). In a recent survey from the United States, secondary intracranial neoplasms were found to be as frequent as all primary brain tumors combined.¹⁸ The study disclosed that only about 20% of the cases diagnosed as secondary intracranial neoplasms were verified histopathologically. Therefore, it is to be expected that in a surgical series of brain tumor, the incidence of brain metastasis will be underestimated.

The total absence of cerebral lymphomas (either primary or secondary) in the present series is noteworthy, especially considering that lymphomas are the commonest primary cancer in Saudi Arabia.³⁰ Several factors could have contributed to this situation. The referral practice of lymphoma patients with the possibility of hospitals of preference might have helped eliminate these patients from this series. Since generalized spread, the lymphomatous meningitis, is the commonest intracranial manifestation of lymphoma, it is expected that most of these patients will escape neurosurgical attention as they do not require surgical intervention. In addition, the tendency of lymphomas to occur deep in the brain³¹ makes them less accessible to surgery and, thus, more likely to be misdiagnosed, especially the primary forms.

As in most of the southern Asian countries, tuberculosis is endemic in Saudi Arabia. Expectedly, the incidence of cerebral tuberculoma in the present series was significantly higher than that estimated in Western countries and Japan^1,2,24,26 but less than rates reported from the Indian subcontinent.^14,32 Non-Saudis, all of whom were citizens of developing countries, were affected at the same rate as Saudis (5.4% and 5.7% of all intracranial space-occupying lesions, respectively). Since there is no investigation, including the CT scan, which is absolutely diagnostic of intracranial tuberculomas, these lesions should always be a constant part of the differential diagnosis of intracranial space-occupying lesion in Saudi Arabia. Furthermore, considering the generally good prognosis following adequate treatment, a sufficiently long trial of antituberculous therapy should be implemented in all cases where tuberculoma is suspected.

The relative frequency of brain abscesses in the present study is small compared with other series¹⁵; however, the unusually high proportion of fungal abscesses (21%) is remarkable. Further observations are still required to substantiate this finding.

CONCLUSION

Despite the different factors that may limit the validity of this study, the available data show that the pattern, as well as age and sex distribution, of intracranial space-occupying lesions in Saudi Arabia share several features with those reported from other countries. The remarkably high frequency of meningiomas among the patients of this series may reflect a higher incidence of this type of tumor in Saudi Arabia. However, a broader, probably population-based study is necessary to confirm this finding. As in most of the southern Asian countries, intracranial tuberculomas are a common lesion in this country and, therefore, should always be considered in the differential diagnosis.