1 Introduction
Vascular pulsatile tinnitus (PT) is the aberrant perception of arterial sounds that are time-locked to the cardiac cycle. Etiologically, vascular PT is conventionally trichotomized into arterial, arteriovenous and venous vascular subtypes, with venous mechanisms predominating
[1,
2]. Venous PT originates from turbulent flow within the dural venous sinuses; when the bony plate of the sigmoid sinus or jugular bulb is dehiscent, these hydrodynamic sounds are air-conducted into the mastoid
[3]. Flow acceleration at the transverse–sigmoid junction—further potentiated by ipsilateral transverse-sinus stenosis—amplifies the audible spectrum. Although venous cases are anatomical, a subset is complicated by idiopathic intracranial hypertension (IIH), reported more frequently in Western cohorts
[4]. This venous phenotype must therefore be discriminated from other vascular, non-vascular mimics, and even systemic diseases.
Sigmoid sinus wall anomalies, comprising dehiscence and/or diverticulum, represent surgically remediable substrates. Two operative modalities have evolved: (a) Transmastoid resurfacing restores the defective sinus wall, thereby attenuating air-bone conduction and obliterating diverticular outpouchings that harbor vortices. Additional extraluminal compression can be applied to reduce local flow velocity, but risks iatrogenic intracranial hypertension if the contralateral venous outflow is inadequate
[5]. (b) Endoluminal stenting addresses the hemodynamic disturbance by dilating stenotic transverse-sinus segments; however, radial force exerted on the adjacent parenchyma may precipitate post-procedural headache
[6]. Both strategies are efficacious, yet equipoise persists regarding superiority in effectiveness and safety
[7,
8].
Successful resurfacing mandates complete, high-density occlusion of the dehiscence to extinguish sound transmission into the mastoid
[9,
10]. To this end, microporous bone cement is favored over autologous grafts to preclude material fatigue and late recurrence. Robust, circumferential resurfacing dehiscence with or without reduction of diverticulum constitutes the technical correlate of durable symptom resolution
[3].
Notwithstanding the high success rate when the dehiscence is fully sealed, the kinetics of post-operative symptom extinction remain poorly characterized. Sparse evidence suggests that retained fluid within the mastoid may transiently sustain PT before ultimate silence
[11]. During this interval, subjective tinnitus (ST) and aural fullness have been reported. Moreover, because existing series only declare outcome at a fixed landmark—typically one month—the precise interval required to judge surgical success remains undefined. This raises the question: why does perception persist after anatomical closure and gradually silences eventually? We therefore sought to delineate the time-to-recovery of PT following complete resurfacing and to identify factors that govern this interval, thereby refining our understanding of postoperative PT pathophysiology.
2 Materials and Methods
2.1 Patient data and study design
This prospective, single-centre cohort recruited 26 consecutive patients with SSWA at the PT Specialty and WWQ Specialist Clinics who underwent transmastoid resurfacing with bone cement between July and October 2025. Pre-operative evaluation comprised pure-tone audiometry, complete blood count, high-resolution temporal-bone CT, colour-Doppler ultrasonography of the cervical vascular, and contrast-enhanced magnetic resonance (MR) imaging including MR angiogram and venogram, equivalent to our previous method
[3].
Inclusion criteria were strictly applied:
1) CT demonstration of sigmoid-sinus wall dehiscence and/or diverticulum. Dehiscence was defined as focal absence of the bony plate overlying the sigmoid sinus with an “air-on-sinus” sign; diverticulum was defined as a focal outpouching of the sinus lumen into the adjacent air cells or mastoid cortex.
2) A positive internal-jugular-compression test: ipsilateral compression abolished or substantially reduced tinnitus, whereas contralateral compression augmented it.
3) Pre-operative trans-canal microphone recording confirming a cardiac-synchronous vascular bruit.
Patients were excluded if any complete dataset was missing, if weekly post-operative follow-up was lost, or if concurrent anomalies unrelated to venous PT were identified: arteriovenous malformation, arterial malformation, temporal-bone neoplasm, otosclerosis, Ménière’s disease, or systemic disorders such as anemia or hyperthyroidism.
Preoperatively, all subjects were informed that fluctuations in PT sensation, ST, fluid retention, aural fullness, hearing loss, and ear pain—along with their potential duration—might occur before surgical intervention.
2.2 Surgical procedure: the Inner-Sinodural-Outer Up-Down Air Cell Removal technique (w/wo) diverticulum isolation technique
All surgical procedures were performed by Y-LH under general anesthesia; OSTEOPAL R polymethylmethacrylate bone cement (Jiangsu Lekai Biotechnology CO., LTD, China) was used as the sole resurfacing material in every case.
The “Inner- Sinodural-Outer Up-Down Meticulous Drilling” method starts with the post-auricular incision (See Figure 1 for the surgical philosophy and accompanying illustrative photographs). When a diverticulum bulges through the mastoid cortex toward the subcutaneous plane, a customized thin cuff of periosteum severed from the flap is deliberately preserved over the protrusion to guard against inadvertent rupture (The periosteal island technique, See Figure 2). The mastoidectomy follows the principles previously described, with the following modifications. The dissection is carried forward to no further than the aditus ad antrum and a general dissection depth out-side in to secure a sufficient cement-filling space. The sinodural angle is fully exposed; the tegmen mastoideum and the wall of posterior cranial fossa are thinned to a translucent eggshell layer with a coarse diamond burr from inside-out. If the posterior semicircular canal is shallow and immediately anterior to the sigmoid sinus, its cortical layer is skeletonized to create adequate space for cement placement. The air cells along the outer aspect are consistently removed to prevent sound leakage from the lateral compartments. Once the dehiscence(s) have been delineated, the lateral and medial walls of the sinus—from the sinodural angle to the inferior margin 2–3 mm lower than the defect—are covered with a thick layer of bone cement. Additionally, if present, the emissary vein should be fully skeletonized and encased in bone cement during the filling process. During cement injection, begin with the waterier phase in the lateral compartment; the low-viscosity cement will quickly fill every air cell. Once the mixture thickens, redirect injection to the wall of posterior cranial fossa and the medial sinus wall—less fluid is needed, yet enough fluidity must be retained to ensure complete mural coverage. Finally, employ the more solidified cement on the tegmen and on the deeper medial segment of the posterior wall adjacent to the antrum; the stiffer material minimizes the risk of cement flowing into the antrum. Finally, a hand-sculpted cement cap is then molded to recreate the cortical contour, and the preserved periosteal flap is laid over the reconstruction to re-establish the mastoid surface. The diverticulum is fully skeletonized but left uncompressed to prevent intraoperative hemorrhage; this technique has proven effective in our previous work
[3]. Postoperative CT in 24 hours was taken on all subjects to observe the sinus wall repairing outcome.
To protect the facial nerve from thermal injury, we cover the entire nerve with a saline-soaked gelatin sponge whenever exposing jugular bulb anomalies requires skeletonization and identification of the nerve (Figure 3). In contrast, we do not cover the vascular wall, because the continuous blood flow dissipates heat and maintains a relatively stable temperature. The bone cement is injected to achieve direct contact, eliminating any dead space that could permit sound leakage.
2.3 Qualitative analysis of postoperative subjective outcome and follow-up method
The tinnitus handicap inventory tailored for PT (THI-PT) was completed pre-operatively (baseline, 1 month before surgery) and then post-operatively at 24 hours, weekly thereafter, and continued for at least 2 consecutive weeks after the last documented PT episode to confirm sustained relief. PT is considered eliminated and completely cured when the THI-PT score is 0.
To distinguish PT from ST, the patient-specific prerecorded PT sound was presented to the participant before surgery. ST was considered present if any sound that differed from the presented PT was perceived in the absence of a physical sound source. Whenever ST was reported by the participant, the tinnitus handicap inventory for ST (THI-ST) was administered. Irrespective of spontaneous complaints, the presence of ST was queried at every follow-up visit; once identified, THI-ST was repeated at each subsequent encounter until the ST had resolved. ST is considered recovered when the THI-ST score is 0.
Aural fullness was prospectively tracked with a 0–10 visual analogue scale (VAS). The symptom was explicitly questioned at every follow-up visit; VAS ratings were recorded until the participant reported complete resolution.
2.4 Statistical analysis
All analyses were executed in OriginPro 9.1 (OriginLab, Northampton, MA). Categorical variables were compared with Fisher’s exact test. Continuous data were first screened for normality (Shapiro–Wilk); thereafter, group differences were examined with the paired t-test (parametric) or Mann–Whitney U test (non-parametric).
3 Results
3.1 Demographic Data
The average age of study subjects is 33.8 ± 8.6 years (range: 17–53 years). The female-to-male ratio was 26 : 1. The duration of PT was 26.3 ± 39.5 months (range: 1–168 months). Left-to-right ratio was 7 : 19. The follow-up duration was 8.8 ± 4.2 weeks (range: 2–15.1 weeks). Dehiscence-to-diverticulum ratio was 20 : 7. All diverticulum cases were found to have dehiscence(s) overlying the diverticulum surface and/or sigmoid sinus surface.
3.2 PT Therapeutic Outcome
Among the 26 enrolled subjects, PT was eliminated in 25 cases (96.2%) during the entire follow-up. Immediate resolution was observed in 10 individuals (38.5%). Mean preoperative THI-PT score was 54.6 ± 25.2 (range: 8–100). Scores fell to 19.6 ± 25.9 (range: 0–98) on post-operative at first 24 hours (preoperative THI-PT vs. postoperative THI-PT-24 hours, paired t-test, p < 0.0001) and to 6.9 ± 13.7 (range: 0–50); at week 1 (postoperative THI-PT-24 hours vs. postoperative THI-PT-1 week, paired t-test, p = 0.01537). Among 7 subjects whose PT did not disappear in the first week, PT in 4 subjects later disappeared on the second week. There is a significant difference between week 1 and 2 (postoperative THI-PT-1 week vs. postoperative THI-PT-2 weeks, paired t-test, p = 0.04285). Table 1 concludes the PT therapeutic outcome. One patient experienced recurrence of PT that persisted; early post-operative CT revealed focal inadequate thickness of the sigmoid sinus wall (see Figure 4). Her PT had remitted by week 2, but reappeared 2 weeks later and has remained present through week 10. Except for this subject whose PT recurred, all other subjects (n = 25) experienced PT disappearance by week 4 and thereafter. One subject’s PT reoccurred on week 4 and later subsided on week 6. Throughout the follow-up, four subjects reported episodic, fleeting PT—abrupt in onset, subtle in intensity, lasting only minutes and recurring two to three days per week—yet all rated their THI score as 0, since the residual sensation left their quality of life unaffected.
3.3 Expected Self-Limiting Sequelae
Temporary ear fullness occurred in all subjects at different postoperative time points. Within 2 days, postoperative ear fullness occurred in 24 subjects (92.3%) (mean VAS score: 7.7 ± 3.0; range: 0–10), and all subjects had ear fullness postop after 1 week (mean VAS score: 6.7 ± 2.4; range: 2–10) and 2 weeks (mean VAS score: 4.1 ± 3.3; range: 0–10). There is a significant difference in mean VAS score between week 1 and 2 (postoperative VAS-1 week vs. postoperative VAS-2 weeks, paired t test, p < 0.0001). At week 3 (mean VAS score: 2.5 ± 3.1; range: 0–10), 11 subjects’ ear fullness recovered. There is a significant difference in mean VAS score between week 2 and 3 (postoperative VAS-2 weeks vs. postoperative VAS-3 weeks, paired t-test, p = 0.0015). Week 4 mean VAS score was 0.47 ± 1.8 (range: 0–4). 23 of 26 subjects (88.5%) experienced recovery of ear fullness by week 4. From week 5 onwards, only one subject had ear fullness (VAS score: 2–4) and recovered by week 8. Postoperative CT revealed that all subjects presented various degree of fluid retention in the middle ear cavity.
Transient post-operative ST developed in 11/26 patients (42.3%) and resolved completely within the first post-operative month. Of these, seven (63.6%) first noted the ST after week 1; their mean THI-ST score was 15.3 ± 13.9 (range: 0–44). Three of these seven patients recorded a THI score of 0, describing the tinnitus as barely perceptible. On week 4, ST remained only in one subject and subsided on the fifth week.
Temporary crackling or popping sounds and autophony were observed in 18 (69.2%) subjects after surgery; however, these symptoms did not affect quality of life.
4 Discussion
Although most published series have focused on long-term outcomes after sigmoid sinus wall reconstruction, few have examined short-term fluctuations of PT. In the present cohort, sigmoid-sinus-wall resurfacing produced an anatomically successful seal in 96.2% of cases. Nevertheless, immediate PT abolition was observed in only 38.5%, implying that the majority of patients still perceive PT during the first three post-operative weeks, with the loudness typically diminishing over time.
Several subjects reported a transient increase in PT on post-operative days 2–3, followed by a sharp decline after the first week. Three mechanisms plausibly account for this early variability: 1) Postoperative fluid accumulation: Edema peaks during the first week; because fluid is a more efficient sound conductor than air, venous pulsation can be temporarily augmented. Another possible reason is that the middle ear retention likely enhances the sound or vibrotactile sensation of the carotid canal. This can be seen in subjects with secretory otitis media
[12]. 2) Mucosal inflammation: Mastoid obliteration produces a sterile mucositis; hyperemic mucosae swell and pulsate, generating an additional sound source. This explains why jugular vein compression fails to abolish postoperative PT sensation, indicating that the transmission pathway is likely extravascular. 3) Altered central gain: Acoustic deprivation during healing can unmask tinnitus perceptual circuits, a phenomenon already described after middle-ear surgery.
Nevertheless, fluid accumulation can be observed on 24-hour post-operative CT scans even in patients without post-operative PT. This expatiate on why ear fullness is frequently and rather immediately reported after surgery. In fact, fluid may persist for several months and sometimes years, most often confined to the mastoid cavity when filling material obstructs drainage; in these cases, patients experience neither aural fullness nor PT unless the tympanic cavity is involved. This fluid likely arises from inflammatory interstitial edema and minor post-operative hemorrhage after mastoid obliteration. When the fluid communicates with the middle ear, conductive hearing loss occurs, which is conceivably why 42.3% of our subjects developed ST, typically beginning after the first week and resolving later on as the fluid reabsorbs and aural fullness subsides.
In the broader context, the primary objective of PT via transmastoid surgery is to prevent the transmission of dural venous sinus flow sounds into the middle ear. Our previous study demonstrated that PT could be recorded within the mastoid cavity and was abolished once the sinus wall was completely resurfaced
[3,
13]. This principle applies equally to patients with SSWA, irrespective of co-existing diverticulum or venous sinus stenosis. In the absence of persistent symptoms attributable to IIH or confirmed IIH diagnosis, stenting a stenotic transverse sinus appears premature; moreover, there is no accepted gold standard for defining a “hemodynamically problematic” stenosis
[8]. It should also be recalled that transverse sinus stenosis is present in nearly one-third of the healthy population, and arachnoid granulations within the sinus are common, particularly in older individuals
[14].
We advocate the use of bone cement rather than autologous materials. Bone cement offers distinct advantages due to its phase-specific properties and versatility, allowing it to fill the mastoid cavity completely without leaving pores or voids, thereby forming a tightly sealed, soundproof barrier. It is durable, robust, and biocompatible. In contrast, autologous materials such as bone dust are prone to revascularization and mechanical degradation over time. Resurfacing with bone dust is suboptimal, as it typically results in a less dense construct. Furthermore, because bone dust is biologically active and vascularized, it remains susceptible to disruption by the underlying venous sinus. If the native sinus wall is insufficient to resist pathological forces, it is illogical to expect a degraded, powdered bone dust construct to provide superior resistance. This approach fundamentally contradicts the surgical goal of establishing an “undegradable” soundproof barrier. However, a prospective, long-term study comparing resurfacing materials is required to validate our extrapolation.
Skeletonization extending to the medial wall of the sigmoid sinus, the bony plate of the posterior cranial fossa, and the tegmen is essential, as it enables deep and thicker placement of cement, obliterates all air cells to prevent vascular sound transmission from the medial and anterosuperior directions, and forces the vascular sound-power vector to undergo transmission loss before reaching the antrum and tympanic cavity. This approach effectively seals all residual air spaces (see Figure 5 for bone cement injection depth in details). Notably, we observed that PT may still resolve if the transmission pathway is blocked, even when a few non-aerated air cells remain on the sigmoid sinus surface. This was evident in our study: three subjects retained a few air cells over the sinus surface yet experienced complete resolution of PT. This observation informed the design of our method, which ensures that sound within the cylindrical sinus structure does not propagate directionally into the tympanic cavity.
This study has several limitations. First, because follow-up was restricted to the short term, it is unknown how long the disappearance of PT can be maintained in these subjects. Nevertheless, a recent long-term study that also used bone cement to resurface the dehiscence reported that PT is unlikely to recur once the sinus wall has been completely resurfaced
[9]. Second, the relatively small sample size—especially the low number of patients in the diverticulum group—limits the robustness of the qualitative analysis. However, because the surgical objective was to resurface the dehiscence rather than to reduce the diverticulum, it is postulated that both entities represent the same disease process: a dehiscence may enlarge or evolve into a diverticulum. Third, despite using patient-specific PT samples to anchor perception, some participants could unreliably differentiate ST from persistent or centrally generated PT, risking misclassification on tinnitus severity. For evaluation of postoperative symptoms, a postoperative transcanal recording or the tinnitus psychoacoustic matching could therefore help differentiate PT and ST entities. Fourth, the postoperative development and subsequent regression of fluid retention and the resulting conductive hearing loss were not objectively quantified; serial audiometric testing could provide the evidence needed to support our hypothesis. Nevertheless, postoperative sensations of PT, ear fullness, ST, and autophony can fluctuate dramatically within seconds, hours, days, or even weeks, rendering the establishment of objective, universally applicable measurement criteria exceedingly difficult—particularly when this study captured data only on a weekly basis. Further refinement of the correlation analysis is warranted to elucidate their mutual relationships.
5 Conclusion
Our surgical philosophy has yielded a 96.2% success rate (25/26). Nevertheless, PT does not always disappear immediately after resurfacing; in roughly 70% of cases the residual PT wanes week-by-week and is markedly attenuated by the fourth postoperative week. Early fluctuations are attributed to transient fluid accumulation, mucosal inflammation, and possibly altered central gain. Ear fullness—reported by 42.3% of patients—may likewise result from fluid retention secondary to inflammatory interstitial edema and minor post-operative hemorrhage within the obliterated mastoid, and can precipitate conductive and/or subjective tinnitus. These anticipated, self-limiting sequelae typically resolve within one month and almost invariably subside by six weeks.
The Author(s). This article is published by Higher Education Press at journal.hep.com.cn.
This is an open access article distributed under the Creative Commons Attribution License 4.0 (CCBY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
PDF
(13254KB)
