A Comparison Between the Clinical Features of Pseudotumor Cerebri Secondary to Tetracyclines and Idiopathic Intracranial Hypertension

Abstract

BACKGROUND-

A comparison between the clinical features of tetracycline-induced pseudotumor cerebri (PTC-T) and those of idiopathic intracranial hypertension (IIH) is absent in the literature. We hypothesized that significant clinical differences between these two etiologies do exist and could be better understood by thoroughly analyzing signs, symptoms, and patient demographics.

METHODS-

We retrospectively reviewed charts of patients who were evaluated for pseudotumor cerebri syndrome (PTCS) at a single neuro-ophthalmic referral center. We identified patients who developed PTC-T after therapy with minocycline, doxycycline, and tetracycline, as well as patients with IIH. All patients met the recently revised diagnostic criteria for PTCS. Patients with PTC-T and IIH were compared by demographics, body mass index (BMI), visual exam at presentation, treatment, course of illness, and visual outcomes.

RESULTS-

We identified 52 cases of PTC-T and 302 cases of IIH. Minocycline was identified as the causative tetracycline antibiotic in 41 of 52 cases of PTC-T (78.8%). Obesity rates among the two groups (PTC-T vs. IIH) were 43.8% vs. 79.2% (P < 0.001). Mean age at diagnosis was 19.8 vs. 28.1 years (P < 0.001). Diplopia occurred at a rate of 40.4% vs. 20.1% (P = 0.001). Mean duration of illness was found to be 18.3 vs. 62.9 weeks (P < 0.0001). Recurrence rates among the two groups were 4.0% vs. 16.5% (P < 0.001).

CONCLUSION-

We identified several significant clinical differences between PTC-T and IIH. PTC-T patients tend to be younger, often non-obese, and more likely to experience diplopia. Disease recurrence is less frequently seen with PTC-T than IIH. The frequency of aggressive treatments between the two groups (e.g. shunting procedures and optic nerve sheath fenestrations) was found to be similar. Vision loss, while uncommon, occurs with similar frequency, despite a significantly shorter duration of illness in PTC-T. Recognition of these important clinical differences may help guide further investigation into the poorly-understood pathogenesis of IIH and PTC-T.

INTRODUCTION

Pseudotumor cerebri syndrome (PTCS) is characterized by increased intracranial pressure with normal brain parenchyma, including absence of hydrocephalus, mass lesion, and underlying infection or malignancy¹. PTCS can be classified as primary, i.e. idiopathic intracranial hypertension (IIH), or secondary^{2, 3}. Secondary causes of PTCS include, but are not limited to, venous sinus stenosis, CNS infections, Graves’ disease, and medications such as vitamin A and the tetracycline class of antibiotics³. Tetracyclines are commonly prescribed in the treatment of acne vulgaris and have long been associated with PTCS, though a causative relationship has yet to be firmly established⁴. More than 100 cases of tetracycline-induced PTCS (PTC-T) have been described in the literature, with some patients showing improvement of symptoms following discontinuation of the antibiotic ^4,5.

The association between these antibiotics and elevated intracranial pressure was first noticed in infants in the 1960s. It was fairly well known among pediatricians that tetracyclines could lead to bulging of the anterior fontanelle, typically 12 hours to 4 days after initiation of therapy⁶. Of note, this response was not associated with CSF abnormalities or an abnormal neurologic examination.

As seen in patients with IIH, PTC-T patients commonly present with headache, papilledema, and diplopia^{2, 4}. These two etiologies appear to result in similar signs and symptoms; however, a detailed comparison of the clinical features of PTC-T and IIH is absent in the literature. It is well understood, for instance, that patients with PTCS tend to be obese young females¹. What remains less clear is whether or not obesity rates differ among patients with primary disease and patients with tetracycline-induced disease. The same can be said for age of patients, duration of illness, and rate of recurrence. A better understanding of the clinical differences between these etiologies could potentially help guide research relating to the pathogenesis of tetracycline-induced disease. This data might also benefit physicians who prescribe tetracycline antibiotics by providing them with a better understanding of the natural history and management of PTC-T.

In 2014, Paley et al published a retrospective observational case series that described differences between primary and secondary PTCS among pediatric patients⁷. Secondary causes in their cohort included tetracycline use, chronic kidney disease, glucocorticoid withdrawal, and lithium use. They found that secondary PTCS patients tended to be older, but that clinical features were otherwise similar between the two groups. They noted that obesity rates were similar between the two groups but were far greater than pediatric obesity rates nationally. Our study differs in that it includes both pediatric and adult patients and focused on comparing features of IIH and PTC-T rather than multiple secondary causes of PTCS.

We hypothesized that there are important clinical differences between PTC-T and IIH. We focused our efforts on comparing differences in patient demographics, BMI, presenting symptoms, family history, duration of illness, treatment modalities, and outcomes to help identify features that set these disease etiologies apart from one another.

METHODS

We retrospectively reviewed charts of patients who were evaluated for PTCS at the John A. Moran Eye Center at the University of Utah between June 1993 and June 2013. We began by identifying all patients who were assigned ICD-9 codes 348.2 (benign intracranial hypertension/pseudotumor cerebri) and/or 377.01 (papilledema associated with increased intracranial pressure). We then compiled this patient data into our REDCap PTCS database. We identified all patients who were suspected of having tetracycline-induced disease as well as those who were diagnosed with IIH. We verified that patients met the recently revised diagnostic criteria for PTCS, which includes normal neurologic examination except for cranial nerve abnormalities, normal neuroimaging, elevated lumbar puncture (LP) opening pressure (≥ 250 mm of water), and normal cerebrospinal fluid composition⁸. We compared multiple variables, including demographics, body mass index (BMI), visual examination at presentation, treatment, course of illness, and visual outcome.

RESULTS

Unless otherwise noted, comparisons are written as PTC-T value vs. IIH value.

We identified 460 cases of all-cause PTCS. Among these, we identified 52 cases of PTC-T and 302 cases of IIH. We found that minocycline was the antibiotic most commonly associated with the PTC-T diagnosis, accounting for 41 of 52 cases (78.8%). Tetracycline (the specific agent, as opposed to the antibiotic class) accounted for 7 cases (13.5%) and doxycycline accounted for the remaining 4 (7.7%). In 37 of these 52 PTC-T cases, the indication for tetracycline antibiotic therapy was documented in the patient’s chart. Acne vulgaris was the indication in 35 cases (94.6%), with the other two being rosacea and hidradenitis suppurativa. The mean duration of antibiotic therapy prior to the onset of the first symptom was 14.4 weeks, with a median duration of 6.4 weeks. Additional secondary causes of PTCS among the patients in our study are shown in Figure 1.

Figure 1.

Comparing Tetracycline induced Pseudotumor Cerebri and IIH

PATIENT DEMOGRAPHICS AND BODY MASS INDEX (Table 1)

Table 1 –

Patient Demographics & Body Mass Index

		PTC-T (N =52)	IIH (N = 302)	P-value
		N (%)	N (%)
Sex	Male	4 (7.7)	22 (7.5)	1.000
	Female	48 (92.3)	271 (92.5)	1.000
Age (years)	Mean (median)	19.8 (18.0)	28.1 (27.0)	<0.001
	<10	0 (0)	3 (1.1)	<0.001
	10–19	34 (65.4)	52 (17.7)	<0.001
	20–29	13 (25)	119 (40.6)	<0.001
	30–39	4 (7.7)	79 (27)	<0.001
	40–49	1 (1.9)	34 (11.6)	<0.001
	≥50	0 (0)	6 (2.0)	<0.001
BMI	Mean (median)	29.4 (27.9)	35.9 (35.0)	<0.001
	<18.5	1 (2.1)	1 (0.4)	<0.001
	18.5–24.9	8 (16.7)	12 (4.5)	<0.001
	25–29.9	18 (37.5)	42 (15.8)	<0.001
	30–34.9	15 (31.3)	76 (28.7)	<0.001
	35–39.9	3 (6.3)	61 (23)	<0.001
	≥40	3 (6.3)	73 (27.5)	<0.001

A statistically significant difference in age of presentation existed between the two groups, with PTC-T patients initially presenting at a mean age of 19.8 years compared to 28.1 years in the IIH group (P < 0.001). Obesity (BMI ≥ 30 kg/m²) was less common among patients with PTC-T than those with IIH (43.7% vs. 79.3%, P < 0.001). No significant difference in gender distribution was noted between the PTC-T and IIH group—the vast majority of patients in both groups were female (92.3% vs. 92.5%). A family history of PTCS was not present in any of the PTC-T cases, and only present in 6 cases of IIH (2.0%).

CLINICAL PRESENTATOIN AND INITIAL EXAMINATION

We found that the mean duration of symptoms prior to presentation was 7.5 vs. 20.0 weeks, with a range of 1.0 to 37.0 weeks vs. 0.4 weeks to 5.0 years (P = 0.343). At initial examination, a relative afferent pupillary defect (RAPD) was present in 13.5% vs. 9.9% of patients (P = 0.002). A CN 6 palsy was documented in 11.6% vs. 6.8% of cases (P = 0.057). The mean LP opening pressure at presentation was 414 vs. 369 mm H₂O (P = 0.07). We also noted non-significant differences in best corrected visual acuity (BCVA), papilledema grade, and Humphrey Visual Field Mean Deviation (HVF MD) at presentation (Table 2).

Table 2 –

Presentation & Initial Examination

		PTC-T (N = 52)	IIH (N = 302)	P-Value
		Mean ± SD (median)	Mean ± SD (median)
Sx Onset to Presentation (Weeks)		7.5 ± 7 (5)	20 ± 38 (7)	0.343
LP Opening Pressure (mm of water)		414 ± 130 (430)	369 ± 96 (350)	0.070
		N (%)	N (%)
BCVA	Moderate (20/60–20/125)	2 (3.9)	3 (1)	0.176
	Severe (20/200 or worse)	2 (3.9)	5 (1.7)	0.176
Papilledema	Moderate (Stage 3)	11 (21.2)	55 (18.8)	0.826
	Severe (Stage 4–5)	13 (25)	67 (22.9)	0.826
HVF MD	−4.99 to −2.00	16 (30.8)	103 (35.2)	−
	−14.99 to −5.00	13 (25.0)	67 (22.9)	−
	≤ −15.00	4 (7.7%)	19 (6.5)	−

Some of the most commonly reported presenting symptoms in both groups were headache (90.4% vs. 85.7%), pulsatile tinnitus (48.1% vs. 46.4%), and transient visual obscurations (32.7% vs. 43.3%). Diplopia was reported twice as frequently among PTC-T patients when compared with IIH patients (40.4% vs. 20.1%, P = 0.001). No other presenting symptoms occurred at significantly different rates between the two groups (Table 3).

Table 3-.

Presenting Symptoms

	PTC-T (N=52)	IIH (N=302)	P Value
Symptoms at Presentation	N (%)	N (%)
Headache	47 (90.4)	251 (85.7)	0.361
Pulsatile tinnitus	25 (48.1)	136 (46.4)	0.825
Transient visual obscurations	17 (32.7)	127 (43.3)	0.151
Blurred vision	20 (38.5)	89 (30.4)	0.248
Nausea/vomiting	12 (23.1)	74 (25.3)	0.738
Diplopia	21 (40.4)	59 (20.1)	0.001
Photophobia	7 (13.5)	58 (19.8)	0.282
Visual loss	10 (19.2)	30 (10.2)	0.062
Other	8 (15.4)	28 (9.6)	0.205
Neck Pain	2 (3.8)	21 (7.2)	0.550
Dizziness	2 (3.8)	18 (6.1)	0.750
Photopsia	2 (3.8)	17 (5.8)	0.750
Retrobulbar pain	2 (3.8)	10 (3.4)	0.699
Paresthesias	2 (3.8)	9 (3.1)	0.674
Positive visual phenomena	0 (0)	7 (2.4)	0.600
Asymptomatic	0 (0)	6 (2)	0.597
Confusion	1 (1.9)	1 (0.3)	0.279

CLINICAL COURSE AND FINAL EXAMINATION (Table 4)

Table 4 –

Clinical Course & Final Examination

		PTC-T (N=52)	IIH (N=302)	P-Value
		Mean ± SD (median)	Mean ± SD (median)
Duration of Illness (Weeks)		18.3±16.1 (13.0)	62.9 ±73.0 (37.0)	<0.0001
		N (%)	N (%)
BCVA	Moderate (20/60 – 20/125)	1 (1.9)	7 (2.3)	0.966
	Severe (20/200 or worse)	2 (3.8)	5 (1.7)	0.966
HVF MD	−4.99 to −2.00	10 (19.2)	69 (23.5)
	−14.99 to −5.00	1 (1.9)	18 (6.1)
	≤ −15.00	3 (5.8)	11 (3.8)
Residual Visual Field Defect		8 (21.1)	54 (25.1)	0.591
Recurrence of PTC		2 (4.0)	47 (16.5)	< 0.001

Illness duration was documented in 29 of 52 PTC-T cases (55.8%) and 99 of 302 IIH cases (32.8%). Among these, we found the mean duration of illness to be 18.3 vs. 62.9 weeks, with a median of 13.0 vs. 37.0 weeks (P < 0.0001). The frequency of aggressive treatments was similar between PTC-T and IIH cases. Shunting procedures were performed in 3.8% of patients in both groups and optic nerve sheath fenestrations were performed in 7.7% vs. 9.2% of cases. Residual optic nerve pallor was seen in 11.5% vs. 6.5% of patients (P = 0.240). RAPDs were present at final examination in 12.0% vs. 8.3% of cases (P = 0.509). Recurrence of disease was documented in 4.0% vs. 16.5% of patients (P < 0.001). No significant differences were noted in persistent decrease in visual acuity, HVF MD, or residual field defects, as shown in Table 4.

COMPARING RESULTS WITH PREVIOUSLY-PUBLSIHED DATA

We conducted a literature search for reports of tetracycline-induced PTCS that were easily accessible electronically and identified 55 articles reporting on 107 individual cases (See Supplemental Table). The mean age of presentation in these studies was 18.6 years old (compared with 19.8 years old in our study). Females made up 81.9% of patients, compared with our finding of 92.3%. Minocycline was determined to be the triggering agent in 67 cases (62.6%, compared with our 78.8%), tetracycline in 28 (26.2%, compared with 13.5%), and doxycycline in 12 (11.2%, compared with 7.7%). The mean duration of antibiotic use prior to symptom onset from 37 cases was 14.1 weeks (14.4 weeks in our study). 26 articles commented on presenting symptoms, with headache, nausea, and diplopia being the most common. These findings are similar to our own. LP opening pressure was documented in 51 cases, with a mean of 405 mmH20 (414 mmH20 in our study). The most common disease treatments from the literature search were the same as in our own findings, i.e., antibiotic withdrawal and acetazolamide. 35 articles commented on visual outcomes, with 14 of these reporting visual field defects and 6 reporting deficits in visual acuity—suboptimal visual outcomes are recognized in our own PTC-T group as well.

DISCUSSION

The results of our study support the hypothesis that significant clinical differences exist between patients diagnosed with PTC-T and those diagnosed with IIH. Most notably, our findings suggest that PTC-T patients tend to be younger, are less likely to be obese, and more likely to experience diplopia. PTC-T patients also tend to have shorter duration of illness and are less likely to experience disease recurrence than their IIH counterparts. It is also important to note that many similarities exist between the two etiologies, including gender distribution, presenting symptoms, and frequency of aggressive treatments.

Tetracycline antibiotics, particularly minocycline, are commonly used in the treatment of acne. It is, therefore, more common to see PTC-T in adolescents and young adults than in older patients^{4, 9}. The BMI of PTC-T patients in our study varied widely, with fewer than half being classified as obese. One study reported an obesity rate of 71.4% among PTC-T patients, while other studies have reported less^{4, 9}. Despite wide variation in these rates, most are lower than reported obesity rates among IIH patients, with 79.3% in our study and rates as high as 87.8% in others⁴.

To date, there have been no case-control studies proving a causative link between tetracycline antibiotics and PTCS. Furthermore, while it is hypothesized that these agents may lead to intracranial hypertension by inhibiting formation of cyclic adenosine monophosphate at the arachnoid villi, the mechanism by which tetracyclines contribute to PTCS is not fully understood^{3, 4}. One study suggests that minocycline directly interferes with metabolism of endogenous retinoic acid—another substance commonly linked to development of PTCS ¹⁰. While our study cannot prove causation, it further substantiates the widely-recognized association between tetracycline antibiotics and PTCS. As with previous reported cases, symptoms generally improved among the patients in our study following discontinuation of tetracycline antibiotics; however, disease recurrence was documented in two instances of antibiotic withdrawal. Kesler et al also reported recurrence of disease following antibiotic withdrawal, suggesting that tetracyclines, at least in some instances, may serve only as an aggravating factor in patients already predisposed to PTCS⁵.

The use of tetracycline antibiotics is generally considered low risk; however, there exists potential for serious long-term consequences. One PTC-T patient in our study was left with hand motion vision in one eye and light perception in the other, which persist after 10 years of follow-up. Patients receiving these medications should, therefore, be routinely monitored for common presenting symptoms of PTCS, including headache, pulsatile tinnitus, and diplopia, and should be examined for papilledema.

CONCLUSION

The identification of clinical differences between patients with idiopathic intracranial hypertension and those with pseudotumor cerebri secondary to tetracycline use may be beneficial in guiding future research and for educating clinicians who treat patients with these antibiotics. Clinicians might erroneously view pseudotumor cerebri as homogenous in presentation, regardless of the underlying etiology. This study demonstrates that such an assumption is inaccurate, and that a secondary etiology can present quite differently from primary disease. It is reasonable to suspect that significant clinical differences may also exist among the various secondary etiologies of pseudotumor cerebri syndrome that were not investigated in this study. Further related research may help answer important questions regarding the poorly-understood mechanisms of this syndrome. The findings of this study also maintain relevance at the level of direct patient care. For clinicians who regularly prescribe tetracycline antibiotics, a basic understanding of these clinical features will help them detect early signs of disease and initiate appropriate therapies. These findings can also better prepare clinicians to educate patients on the characteristic symptoms and disease course of tetracycline-induced pseudotumor cerebri syndrome.