What is the Natural History of the Triangular Fibrocartilage Complex Tear Without Distal Radioulnar Joint Instability?

Abstract

Background

The triangular fibrocartilage complex (TFCC) tear is a common cause of ulnar-side wrist pain; however, its natural course is not well understood.

Questions/purposes

We sought (1) to determine the natural course of TFCC tears without distal radioulnar joint (DRUJ) instability, and (2) to identify the factors associated with poor prognosis after nonsurgical treatment of TFCC tears.

Methods

Over a 3-year period, we treated 117 patients with TFCC tears who did not have DRUJ instability. The diagnosis was made on the basis of ulnar-sided wrist pain, a positive ulnocarpal stress test or ulnar grinding test, and identification of a tear on MRI or CT arthrography. Of those, 25 were excluded during the initial evaluation period because they met the previously defined indications of surgery on the basis of clinical history. Another 19 patients (20%) were lost to followup before 6 months, and one patient was excluded because of prior wrist surgery, leaving 72 wrists in 72 patients for analysis in this retrospective study, which drew data from a review of electronic medical records of one institution. The group consisted of 42 men and 30 women, with a mean age of 40 years (range, 18-70 years). The study group was followed for a mean of 16 months (range, 6 to 36 months). We evaluated the pain VAS and patient-rated wrist evaluation (PRWE) at the initial visit, at 4, 8, and 12 weeks, and at more than 6 months after the initial visit. A PRWE score ≤ 20 points indicated complete recovery, and a PRWE score more than 20 points was considered an incomplete recovery. We used Kaplan-Meier survival analysis and Cox regression modelling to estimate the time to complete recovery and to identify factors associated with incomplete recovery among the seven possible factors of older age (≥ 45 years), male, obesity (body mass index ≥ 30 kg/m²), dominant-hand involvement, chronic symptoms (≥ 6 months), traumatic tear, and ulnar-plus variance.

Results

The Kaplan-Meier survival analysis showed that estimated cumulative incidence of complete recovery was 30% (95% confidence interval [CI], 20–40) at 6 months and 50% (95% CI, 39–61) at 1 year. We could not find any risk factors among the seven candidate factors, including older age (hazard ratio [HR], 0.608; 95% CI, 0.34–1.087; p = 0.093), male (HR, 1.152; 95% CI, 0.667–1.991; p = 0.612), obesity (HR, 1.433; 95% CI, 0.603–3.402; p = 0.415), dominant hand involvement (HR, 1.808; 95% CI, 0.927–3.527; p = 0.082), chronic symptoms (HR, 0.763; 95% CI, 0.443–1.922; p = 0.133), traumatic tear (HR, 0.756; 95% CI, 0.432–1.32; p = 0.325), and ulnar plus variance (HR, 0.804; 95% CI, 0.461–1.404; p = 0.443).

Conclusions

This study demonstrates that nonsurgical treatment is moderately successful for treating patients with TFCC tears without DRUJ instability. We recommend a minimum of 6 months nonsurgical treatment as the first-line treatment for this injury. Future studies are necessary to clarify predictors of persistent pain with nonsurgical treatment to avoid an unnecessary surgical delay.

Level of Evidence

Level III, prognostic study.

Introduction

The triangular fibrocartilage complex (TFCC) is a ligament-cartilage complex consisting of multiple anatomical subunits that span the sigmoid notch of the distal radius, distal ulna, and ulnar carpal bones. It acts as a primary stabilizer of the distal radioulnar joint and as a stress absorber of loads that transfer through the ulnocarpal joint [9]. A TFCC tear is one of the most-common causes of ulnar-sided wrist pain [28].

However, there is little evidence to support a general consensus about the most-appropriate treatment for TFCC tears. Current recommendations support nonsurgical initial treatment if the distal radioulnar joint (DRUJ) is stable, and surgical treatment is considered if no improvement is observed after a few months of conservative treatment [9, 24, 29, 35]; however, we are not aware of any studies that support this recommendation. Contrary to the large volume of studies regarding surgical treatment of TFCC tears [2, 11-13, 16, 23, 26, 27, 38, 39], to our knowledge, there are no studies on the results of conservative treatment, meaning that the natural course of a TFCC tear has not been described. Consequently, it is not clear when and for whom surgical intervention of a TFCC tear will be beneficial.

Therefore, we sought (1) to determine the natural course of TFCC tears without distal radioulnar joint (DRUJ) instability, and (2) to identify the factors associated with poor prognosis after nonsurgical treatment of TFCC tears.

Patients and Methods

This was a retrospective study on the natural course of TFCC tears performed at a tertiary referral center. We obtained institutional review board approval before the commencement of this study. We searched the electronic medical records database at our institution for patients who presented with ulnar-sided wrist pain from January 2014 to December 2016.

We included patients aged 18 years or older with a diagnosis of a triangular fibrocartilage tear. Criteria for the diagnosis of a TFCC tear were: (1) presenting with ulnar-sided wrist pain, (2) positive ulnocarpal stress test or ulnar grinding test [5, 19, 30, 32], and (3) the presence of a TFCC tear on MRI or CT arthrography. We classified the TFCC tear as traumatic if pain developed after a single traumatic event and there was no evidence of cystic or sclerotic change on the ulnar side of the lunate on wrist x-rays, which were taken for all the patients. Otherwise, we determined that the TFCC tear was degenerative, that is, related to ulnar impaction syndrome. Indications for surgical treatment upon presentation with a TFCC tear during the period of this study were: (1) duration of symptoms more than 3 months; and (2) persistent severe pain despite adequate nonsurgical treatment for more than 3 months prior to presenting to our office, including rest, splinting, physiotherapy, and intermittent NSAID use. Patients who met all these indications were offered surgery at their initial evaluation and were not included in the current study.

Other exclusion criteria for this study were: (1) age younger than 18 years, (2) other causes of ulnar-sided wrist pain, including DRUJ instability, DRUJ osteoarthritis, extensor carpi ulnaris (ECU) tendinitis or subluxation, flexor carpi ulnaris tendinitis, pisotriquetral joint instability/osteoarthritis, lunotriquetral joint instability, or hamatolunate impaction syndrome, (3) wrist fractures, (4) wrist deformity, (5) inflammatory or degenerative wrist arthritis, and (6) neuromuscular disorders affecting wrist function like hemiplegia. Patients with less than 6 months of followup were considered lost to followup and were excluded.

We identified 246 patients who visited our institution with ulnar-sided wrist pain. Of these, 205 patients completed diagnostic workups, and 88 patients were excluded because of a diagnosis other than a TFCC tear (Fig. 1). Of the 117 patients with an TFCC tear, 25 patients who met surgical indications at the initial evaluation period were excluded; 21 of 25 patients underwent surgery (ulnar shortening osteotomy in five patients and arthroscopic repair in 16 patients) and four patients were lost to followup before 6 months, leaving 92 patients eligible for this study. Another 19 patients were lost to followup before 6 months, and one patient was excluded because of previous wrist surgery.

Fig. 1.

A schematic diagram shows patient recruitment. TFCC = triangular fibrocartilage complex; LTIL = lunotriquetral interosseous ligament; DRUJ = distal radioulnar joint.

A total of 72 wrists in 72 patients (42 men and 30 women; mean age, 40 years [range, 18-70 years]) were included in this study (Table 1). Demographic and baseline clinical characteristics were not different between the patients who were included, lost to followup (n = 19), or excluded (n = 1) (Table 2). Twelve patients were current smokers, and the mean BMI was 23 ± 4 kg/m². The dominant hand was involved in 47 patients (65%), and a single traumatic event was associated in 38 patients (53%). The mean duration of ulnar-sided wrist pain was 18 months (range, 1 day–112 months). The study group was followed for a mean of 16 months (range, 6–36 months).

Table 1.

Patient demographic and clinical characteristics

Table 2.

Comparison of demographic and clinical characteristics between the included and excluded patients

Our nonsurgical management consisted of immobilization of the involved wrist with a short-arm, metacarpophalangeal joint-free, removable brace for 4 to 12 weeks depending on the pain level elicited by provocative tests. We chose a short-arm brace rather than a long-arm splint/cast because patients had a very low compliance with the latter, and limiting only ulnar deviation with the short-arm brace reduced pain while allowing the patients to continue workplace activities. When patients perceived their wrist pain to be severe, NSAIDs were administered.

All patients underwent a diagnostic workup, including a physical examination, standard wrist x-rays (posterior-anterior, lateral, both oblique, and both wrist posterior-anterior with power grip views), and 3.0-T MRI (Achieva, Philips Healthcare, Best, The Netherlands) or multidetector CT arthrography (Somatom Definition Flash, Siemens Medical Solutions, Forchheim, Germany). Outcome assessments included the pain VAS and the patient-rated wrist evaluation (PRWE). For the pain VAS measurement, we used a 10-cm line with 0 at the left end and 10 at the right end. A score of 0 indicated no pain and a 10 indicated the worst pain imaginable. The PRWE is a validated 15-item wrist-specific questionnaire that assesses pain (five items for 50 points) and function (10 items for 50 points), scoring from 0 to 100; higher scores indicate a worse condition [18]. The PRWE has been used to assess subjective outcomes in various wrist lesions [4, 6, 7, 31]. The minimal clinically important difference (MCID) for PRWE has been reported to be 11.5, 14, or 17 points, depending on the studies [14, 33, 37]. We conducted the outcome evaluation at the initial visit, at 4 weeks, 8 weeks, and 12 weeks, and at the final followup after 6 months. When the followup period was less than 6 months, we contacted the patients by phone and invited them to participate; four patients received followup only by phone.

We measured ulnar variance using the perpendicular method [34] on the posterior-anterior view x-ray taken in a neutral wrist position. Neutral pronation-supination, which is required for the standardized measurement of the ulnar variance [34], was judged based on the ECU groove criterion [17]. We measured the ulnar variance twice and took an average using semiautomated image analysis software installed in the picture archiving and communication system (PACS; PiViewStar; Infinitt, Seoul, Korea). Ulnar plus variance was defined as an ulnar variance of 1 mm or more, ulnar neutral variance as an ulnar variance less than 1 mm and more than -1 mm, and ulnar minus as an ulnar variance -1 mm or less. Two radiologists (SHK, S-YP) who specialize in the musculoskeletal system interpreted the MRIs (40 patients) or CT arthrography (32 patients) images.

We used an MCID of 14 points to judge clinically meaningful improvements [33]. Patients were considered improved when the final PRWE score decreased by 14 points or more from the initial score, they were considered to show no change when the PRWE score difference ranged from -13 to 13 points, and to be worse when the final score increased by 14 points or more from the initial score. In addition, for survival analysis and predictive model construction, we defined a complete recovery as a final PRWE score of ≤ 20 points; with the PRWE score, which ranges from 0 to 100 points with lower scores indicating less disability of the wrist, 20 points or less is considered a state of no or minimal wrist disability [7]. We defined an incomplete recovery as a final PRWE score of more than 20 points or conversion to surgical treatment [7]. The patients were offered surgery during followup when their pain and disability were unresponsive to adequate nonsurgical interventions for more than 3 months, with symptoms sufficiently severe to interfere with daily and occupational activities.

Statistical Analysis

Statistical analysis was performed using IBM SPSS 23.0 (SPSS Inc, Chicago, IL, USA). The Kolmogorov-Smirnov test revealed that the data sets followed the Gaussian distribution. We used the repeated ANOVA measures with post-hoc Bonferroni correction to examine differences in the pain VAS and PRWE scores between each measurement. We used the Kaplan-Meier product-limit method to estimate the cumulative incidence of complete recovery at 6 months and 1 year. Before we constructed our predictive model, we listed seven potential predictive factors: older age (≥ 45 years), male, obesity (body mass index ≥ 30 kg/m²), dominant hand involvement, chronic symptoms (≥ 6 months), a traumatic tear, and ulnar plus variance, based on the maximum number of variables according to the rule of thumb for Cox regression modelling [25]. Then we constructed a multivariate Cox proportional hazards regression model via backward conditional method with p ≤ 0.05 for entry and p > 0.1 for removal to identify independent factors associated with the incomplete recovery after a TFCC tear and the associated hazard ratio (HR) with 95% confidence interval (CI) estimates. The level of significance was set at p < 0.05.

Results

Natural Course of TFCC Tears

The PRWE score and pain VAS both improved from the initial to final evaluations (Fig. 2). The mean PRWE score decreased from the initial to final evaluations (55 ± 22 versus 29 ± 25, mean difference, 26 [95% CI, 20–31]; p = 0.024) (Fig. 2), although with the numbers available, there were no differences in PRWE scores at the other points. The mean pain VAS score decreased from the initial to final evaluations (5.4 ± 2.3 versus 2.5 ± 2.4, mean difference, 2.9 [95% CI, 2.1–3.6]; p = 0.015) (Fig. 2), although with the numbers available, there were no differences in pain VAS scores at the other points.

Fig. 2 A-B.

The graphs show temporal changes of (A) the patient-rated wrist evaluation (PRWE) and (B) pain visual analogue scale (VAS) scores after nonsurgical treatment of patients with triangular fibrocartilage complex (TFCC) tears. The mean PRWE and pain VAS scores decreased from the initial to final evaluation, but with the numbers available, there were no differences at other time points. Error bars show the SD; ^*p < 0.05 versus 0 week.

Using a PRWE score ≤ 20 as the criterion, 32 of 72 patients (44%) achieved complete recovery at a mean of 34 ± 33 weeks. Forty-five of 72 patients (63%) showed clinically meaningful improvements (defined as a decrease of 14 points or more from the initial evaluation), 24 of 72 patients (33%) showed no change, and three of 72 patients (4%) had a worse PRWE score. According to the Kaplan-Meier analysis, the estimated cumulative incidence of complete recovery was 30% (95% CI, 20–40) at 6 months and 50% (95% CI, 39–61) at 1 year (Fig. 3).

Fig. 3.

The graph shows cumulative incidence of complete recovery in the patients with triangular fibrocartilage complex (TFCC) tears. The number of patients reaching a complete recovery increased rapidly during the first 6 months and then slowed down, eventually reaching a plateau at about 60%.

Risk Factors for Poor Outcomes After Nonsurgical Treatment for a TFCC Tear

With the numbers available, no factors that we explored were independently associated with an incomplete recovery (defined as a PRWE score greater than 20 points) after nonsurgical treatment of TFCC tears (Table 3). Notably, dominant hand involvement (HR, 1.868; 95% CI, 0.807–4.326; p = 0.145) and ulnar-plus variance (HR, 0.568; 95% CI, 0.258–1.253; p = 0.161) were not associated with a poor PRWE score at latest followup.

Table 3.

Hazard ratios of factors for incomplete recovery (defined as PRWE scores > 20 points) after nonsurgical treatment of triangular fibrocartilage complex tear

During the followup period, 11 patients (15%) met the previously defined indications for surgery and all underwent surgical treatment at a mean of 23 weeks (range, 12–49 weeks); three patients underwent diagnostic arthroscopy with ulnar shortening osteotomy and eight patients underwent arthroscopic repair for a TFCC 1b lesion.

Four of 54 patients (7.4%) who were employed at the initial visit changed their jobs due to wrist pain during the followup period. In addition, six of 24 patients (25%) who enjoyed recreational/sports activities involving wrist usage changed their preferred activities due to wrist pain.

Discussion

Although a TFCC tear is a common cause of ulnar-sided wrist pain, its natural course is not fully understood. We found that half of the patients who sustain an TFCC tear will still have pain and disability 1 year after nonsurgical treatment. Many patients received surgical intervention or changed their occupational or recreational actives during the followup period.

There are several limitations to this study. The retrospective study design and absence of a predefined study protocol during the period of time these patients were treated makes it likely that selection bias influenced surgical decisions. However, our chart review confirmed that indeed over the relatively short, 3-year span of the study, reasonably consistent surgical indications were adhered to at the center where these patients were treated. Second, patients with other causes of ulnar-side wrist pain may have been included in the study cohort because a TFCC tear might have been an incidental finding in those patients. However, we do not believe this was a serious limitation because to be included in the study group, patients also needed to have both signs and physical symptoms consistent with the findings of a TFCC tear on advanced imaging. Third, 22% of patients in this study were lost to followup, which might be slightly higher than the acceptable rate of loss to followup in cohort studies [1], introducing a transfer bias. However, the importance of loss to followup depends on the mode rather than the number of the patients lost [15]. As the baseline clinical and demographic variables of the excluded patients were comparable to the ones included in the study, the mode of loss to followup could be considered nonsystematic, suggesting that the estimates of the outcomes were not seriously biased by this level of loss to followup. Fourth, this study might not have had a sufficient number of patients to detect all factors or combinations of factors associated with incomplete recovery.

We found that a nonsurgical method was only moderately successful for the treatment of TFCC tears without DRUJ instability in this study. This finding is consistent with that of one retrospective study [24], in which 43% of 84 patients with traumatic TFCC tears underwent surgery after a minimum of 4 weeks of nonsurgical treatment. In addition, we demonstrated that the slope of the Kaplan-Meier survival curve flattened at a point between 6 to 12 months after nonsurgical treatment was initiated. This finding suggests that the probability of a spontaneous resolution of symptoms is likely to decrease after this point, and, therefore, we recommend a minimum 6 months of nonsurgical treatment before recommending surgical intervention for an unresponsive TFCC tear. Because 4 weeks to 4 months of nonsurgical treatment was recommended by other authors [24, 30, 35], we used 3 months as the cutoff for consideration of surgical intervention during this study period. However, based on this study’s findings, we now try to manage patients without surgery for a full 6 months before considering surgical intervention.

One of the most unexpected findings in this study was that the ulnar-plus variance was not associated with incomplete recovery (defined as persistently high PRWE scores) in patients with TFCC tears. It has been well known that lengthening the ulna relative to the radius dramatically increases the forces applied to the ulnocarpal joint [22]. TFCC perforations are four times more frequent in cadavers with ulnar-plus variance than in those with ulnar-negative variance [21]. Thus, it seems reasonable to presume that ulnar plus variance makes the TFCC more vulnerable to injury, adversely affecting its healing process. However, it has been suggested that ulnar length is inversely correlated with TFCC thickness [8, 20, 36, 40], and that the load transmission across the distal ulna is not different with the numbers available regardless of the length of the native ulna [8], indicating that ulnocarpal joint pressure is not necessarily increased in congenital-type ulnar-plus variance. In this study, we included only congenital-type ulnar plus variance—we excluded patients with a wrist deformity or fracture—which may explain why we did not find ulnar-plus variance to be a factor associated with poorer outcomes scores in this study cohort.

Another interesting finding in this study was that the mode of tear (that is, traumatic versus degenerative) was not associated with prognosis. It seems reasonable to anticipate that degenerative tears, as they are closely related to the ulnar impaction syndrome [29], are more likely to result in recurrent symptoms when the patients resume full activities. However, the frequency of incomplete recovery was not different between the patients with degenerative tears and those with traumatic tears in this study. This finding suggests that even for patients with idiopathic ulnar impaction syndrome, nonsurgical treatment can be successful and, therefore, should be considered as the first-line treatment for these patients. However, caution should be exercised in interpreting this result. It was hard to objectively define a “traumatic” or “degenerative” tear based on advanced imaging. Unlike arthroscopy, advanced imaging could not provide clues of acute trauma, such as well-defined torn edge and intraligamentous hemorrhage, and thus we had to rely solely on the patient’s recall of acute, forceful, traumatic events, which might not be reliable, especially in chronic injuries. We believe future studies with improved methods for differentiating a traumatic tear from a degenerative tear and with large sample sizes will help refine our findings on this point.

Diagnosing a symptomatic TFCC tear is not straightforward because asymptomatic TFCC tears are common, especially in older patients. A systematic review reported that TFCC abnormalities are observed in 27% of patients younger than 30 years, and in 49% of those who were 70 years or older [3]. Another study found a compete TFCC tear in 22% of asymptomatic volunteers as judged by MRI [10]. Both studies showed an increasing prevalence of TFCC tears with increasing age. Therefore, it is critical to establish a causal relationship between the TFCC tear and the patient’s ulnar side wrist pain when diagnosing symptomatic TFCC tears. Although provocative tests such as the ulnocarpal stress test [19] or ulnocarpal grinding test [30] may play an important role in this regard, it is still difficult to establish such a causal relationship because diagnostic properties of those tests are not well studied. The lack of a gold standard makes any clinical and epidemiological studies on the TFCC tear challenging. Although we used three criteria to identify symptomatic TFCC tears, which we believe ruled out patients with incidental TFCC tears, there is still a possibility that some incidental TFCC tears were included in this study cohort. Further studies are necessary to establish a consensus regarding the diagnostic criteria of a symptomatic TFCC tear to make the results more comparable and consistent.

In conclusion, we found that TFCC tears without DRUJ instability, irrespective of whether they are traumatic, can be treated nonoperatively. Patient-perceived pain and disability improved rapidly during the first 6 months after nonoperative treatment was started. However, the success rate of the nonsurgical treatment is only moderate, and surgical treatment may be considered if pain and disability persist after a minimum 6 months of nonoperative treatment. We found no factors that we explored were associated with persistent pain or functional disability after nonsurgical treatment of an TFCC tear. Future studies might seek to identify risk factors for poor prognosis with nonsurgical treatment to avoid unnecessary delay of surgery and to establish better diagnostic criteria for symptomatic TFCC tears.