SCIENTIFIC ARTICLE

Radiographic classification for fractures of the fifthmetatarsal base

Alexander T. Mehlhorn & Jörn Zwingmann &

Anja Hirschmüller & Norbert P. Südkamp &

Hagen Schmal

Received: 11 August 2013 /Revised: 14 November 2013 /Accepted: 25 December 2013 /Published online: 18 January 2014# ISS 2014

AbstractObjective Avulsion fractures of the fifth metatarsal base(MTB5) are common fore foot injuries. Based on aradiomorphometric analysis reflecting the risk for a secondarydisplacement, a new classification was developed.Materials and methods A cohort of 95 healthy, sportive, andyoung patients (age≤50 years) with avulsion fractures of theMTB5 was included in the study and divided into groups withnon-displaced, primary-displaced, and secondary-displacedfractures. Radiomorphometric data obtained using standardoblique and dorso-plantar views were analyzed in associationwith secondary displacement. Based on this, a classificationwas developed and checked for reproducibility.Results Fractures with a longer distance between the lateraledge of the styloid process and the lateral fracture step-off andfractures with a more medial joint entry of the fracture line atthe MTB5 are at higher risk to displace secondarily. Based onthese findings, all fractures were divided into three types: typeI with a fracture entry in the lateral third; type II in the middlethird; and type III in the medial third of the MTB5. Addition-ally, the three types were subdivided into an A-type with afracture displacement <2 mm and a B-type with a fracturedisplacement≥2 mm. A substantial level of interobserveragreement was found in the assignment of all 95 fractures tothe six fracture types (κ=0.72). The secondary displacementof fractures was confirmed by all examiners in 100 %.Conclusions Radiomorphometric data may identify fracturesat risk for secondary displacement of the MTB5. Based onthis, a reliable classification was developed.

Keywords Radiomorphometry . Fracture . Fifthmetatarsalbase . Classification

Introduction

Avulsion fractures of the base of the fifth metatarsal are acommon foot injury that causes dorso-lateral forefoot pain[1–3]. These injuries have to be clearly distinguished fromthe “true” Jones fracture, which is located at the junctionbetween the proximal diaphysis and the metaphysis of thefifth metatarsal involving the 4th to 5th intermetatarsal artic-ulation [4] and from the proximal diaphyseal fracture, which islocated distally to the 4th to 5th intermetatarsal articulation[5].

Avulsion fractures of the fifth metatarsal base occur byindirect force when the foot is plantar-flexed and inverted.The exact pathomechanism is controversially discussed, butrecently some cadaver studies revealed that the causing forceis traction of the lateral cord of plantar aponeurosis and theshort peroneal muscle tendon [6, 7].

Therapy recommendation of non-displaced fractures of thefifth metatarsal base includes conservative treatment withprotected weight bearing in a walking cast, hard-sole-shoe,or a compression wrap [4, 8, 9]. Extended non-weight bearingwas identified as a predictor of a poor functional outcome[10]. The therapy of displaced fractures is controversiallydiscussed, both conservative and operative treatment withopen reduction, screw fixation, or tension band wiring[11–13] is possible [14]. Decision for treatment depends onthe patients’ age, activity level, and co-morbidities; for healthyand young patients with athletic ambitions, operative treat-ment is more favorable.

In the present paper, the hypothesis was tested whetherradiomorphometric parameters of fractures of the fifth meta-tarsal base allow the reliable prediction of a possible

A. T. Mehlhorn : J. Zwingmann :A. Hirschmüller :N. P. Südkamp :H. Schmal (*)Department of Orthopaedic Surgery, University of Freiburg MedicalCenter, Hugstetter Str. 55, 79106 Freiburg, Germanye-mail: hagen.schmal@freenet.de

Skeletal Radiol (2014) 43:467–474DOI 10.1007/s00256-013-1810-5

secondary displacement. Along with this, a classification wasdeveloped considering these observations.

Materials and methods

Patients

From 2001 to 2009, a cohort of 95 patients treated for avulsionfractures of fifth metatarsal base was included in the study.Only healthy (patients graded 1 or 2 according to the ASAclassification [15]), young (age≤50 year) and sportive patients(at least one sportive activity was performed at time of injury)were included in the present study. All patients consideredtheir sports activities as an essential component for theirquality of life. In the same time period, a total of 183 patientswere operatively treated for metatarsal fractures. Multipleinjured patients were excluded and four patients had addition-al fractures of the foot skeleton. Considering the overall rate ofoperative treatment following presentation in the emergencydepartment with 70 %, the inclusion rate was 40 % of allpatients with fractures of the fifth metatarsal base treated in theexamined time period (95 of 238 patients).

After clinical examination, the diagnosis was confirmed byX-rays of the affected foot in a dorso-plantar and an obliqueplane. All included observed fractures were proximal to the4th to 5th intermetatarsal articulation in both radiographicplanes and were related to an indirect forefoot trauma. Thestudy was approved by the Ethical Board.

The patients were distributed into three groups: group 1included patients with non-displaced fractures (step-off<2 mm), who were treated conservatively for 6 weeks. Patientswith fractures showing a step-off exceeding 2 mm in at leastone radiographic plane in initial X-rays were considered pri-marily displaced and were treated operatively (group 2).Group 3 included patients with initially non-displaced frac-tures, who were treated conservatively and secondarilydisplaced (fracture step-off>2 mm). Following this, they werereferred to operative treatment. Conservative treatmentconsisted of 15 kg partial weight bearing for 6 weeks in alower leg cast (VACOpedes, OPED, Germany). Operativetreatment was performed with open reduction and internalfixation using a bicortical screw or tension band wiringfollowed by 15-kg partial weight bearing for 6 weeks in alower leg cast. Secondary displacement was defined when theinitial step-off increased from less to higher than 2 mm. Dur-ing conservative treatment, control radiographs were per-formed 0.5, 1, 2, 4, and 6 weeks after trauma.

Radiographic assessment

All radiographs were performed in two standard planes in-cluding an oblique and a dorso-plantar view by one

investigator (A.T.M., 7 years of experience as a dedicated footand ankle surgeon). Radiographic measurements were per-formed in the dorso-plantar view using a radiomorphometricanalysis tool (Tiani J-Vision, Version 3.3.16, Tiani MedgraphAG, 1150 Wien, Austria/Europe).

The lateral fracture step off at the lateral border of the fifthmetatarsal base, the tip of the styloid process, the medialfracture step at the metatarso-cuboid joint, and the tip of themedial edge of the fifth metatarsal base were defined as theanatomical landmarks. In order to standardize for bone sizeand to exclude radiographic magnification errors, the follow-ing proportions and angles were calculated:

& joint fracture angle (JFA) α (Fig. 1, left)& joint ratio (JR) C/A (Fig. 1, right)& lateral edge length (LEL) B (Fig. 1, right)& the lateral edge ratio (LER) B/A (Fig. 1, right)

The joint-fracture angle (JFA)αwas measured between theproximal articular surface of the fifth metatarsal base and theline connecting the lateral and the proximal fracture step off(Fig. 1, left). The joint ratio (JR = C/A) was calculateddividing the line tangential to the metatarso-cuboid jointreaching from the tip of the styloid process to the fracturestep-off (C) through the total length of the fifth metatarsal base(A) (Fig. 1, right). The lateral edge length (LEL) was calcu-lated measuring the distance from the tip of the styloid processto the lateral fracture step-off of the lateral surface of the fifthmetatarsal (B) (Fig. 1, right). The lateral edge ratio (LER =B/A) was calculated dividing the LEL (B) through the totallength of the fifth metatarsal base (A) (Fig. 1, right). Regard-ing functional anatomy, the LEL and C correspond with theinsertion area of the peroneus brevis tendon and the lateralaspect of the plantar aponeurosis [16].

Interobserver reliability and clinical practicabilityof classification system

Based on the results of the radiomorphometric evaluation, aclassification system was developed, which considers param-eters of an increased fracture displacement risk. All X-rayswere collected and numerated; patient identification data wasrendered anonymously. Then, the X-rays were evaluated bytwo different observers. Observer 1 was an experienced or-thopedic surgeon with special interest in foot surgery. Observ-er 2 was a fellow for orthopedic surgery also with specialinterest in foot surgery. None of the observers had any expe-rience with the classification before study onset to exclude theinfluence of training on reliability. At the beginning of thestudy, the classification was provided to the observers in awritten form. In addition, the first author gave a 15-minintroductory explanation of the classification. The evaluationof the X-rays by the two observers took place independently.

468 Skeletal Radiol (2014) 43:467–474

Statistical assessment

Numerical data were analyzed by computer software packagefor statistical analysis (SPSS statistical program, version 11.5,SPSS Inc. Chicago, IL, USA). All values are reported as mean± standard deviation. Statistical significance was determinedusing the Mann–WhitneyU test and the Kruskal–Wallis H testfor comparing the medians of two and multiple populations,respectively. Categorical data are presented as absolute fre-quency and percent distribution. Data (incidences) were ar-ranged in R by C tables, and statistical significance of differ-ences calculated using a χ2 test. A significance level of 0.05was used throughout the study.

Percentage agreement and interobserver reliability wereassessed with the JMP statistical package, Vers. 6 (JMP Soft-ware, Cary, NC, USA). For intraobserver and interobserverreliability, the kappa statistic function of the JMP statisticalpackage was used measuring kappa values (ĸ) to describe theagreement between observers [17]. Kappa values wereinterpreted using the guidelines proposed by Landis and Koch[18].

Results

Epidemiological data

The mean age of all patients was 39±18 years (41 females and54 males). Fifty-six fractures, representing group 2, were

treated with open reduction and internal fixation (ORIF) witha bicortical screw (n=43) or a tension band wiring (n=13).The remaining fractures (n=39) were initially treated non-operatively. A total of ten patients, representing group 3, weretreated by ORIF following initial non-operative therapy be-cause of secondary displacement. The remaining patients withinitial conservative treatment (n=29) represent group 1. Allpatients were immobilized in a cast as described and per-formed partial weight bearing for 6 weeks. Secondary dis-placement was diagnosed within the first 2 weeks of conser-vative treatment, all fractures consolidated after 3 months.

Radiographic assessment

The Spearman rank correlation revealed a significant correla-tion between the joint ratio (JR) and the lateral edge length(LEL) (ρ=0.63), and the JR and the lateral edge ratio (LER)(ρ=0.65) for all fractures (p<0.0001).

The joint-fracture angle (JFA) α was 51±18° in group 1,55±16° in group 2, and 57±15° in group 3 (Fig. 2a). Therewas no significant difference between any group.

In group 1, the JR was 0.48±0.25, in group 2, 0.63±0.26,and in group 3 0.84±0.16 (Fig. 2b). There was a significantdifference between group 1 and 3 (p=0.033) as well as group2 and 3 (p=0.0004). There was also a significant differencebetween group 1 and 2 (p=0.018).

The LEL was 9.8±4.2 mm in group 1, 12.5±4.2 mm ingroup 2, and 15.6±2.7 mm in group 3 (Fig. 2c). There was asignificant difference between group 1 and 3 (p=0.012) as

Fig. 1 Overview about theevaluated radiographicparameters used for analysis.Radiographic measurements of allfifth metatarsal base fractureswere performed in the dorso-plantar view. The joint-fractureangle (JFA) α, the joint ratio(JR = C/A), the lateral edge length(LEL = B), and the lateral edgeratio (LER = B/A) werecalculated based on thesemeasurements

Skeletal Radiol (2014) 43:467–474 469

well as group 2 and 3 (p=0.0006). There was also a significantdifference between group 1 and 2 (p=0.009).

In group 1, the LER was 0.44±0.18, in group 2 0.56±0.20,and in group 3 0.66±0.11 (Fig. 2d). There was a significantdifference between group 1 and group 3 (p=0.014) and group2 and group 3 (p=0.0007). There was also a significantdifference between group 1 and 2 (p=0.030).

Development of classification system

As shown above, a secondary displacement was associatedwith a higher JR. An association of secondary displacementwith the LER could also be demonstrated, but both parametershighly statistically significantly correlated. Therefore, we de-cided to build up the classification system only on the JRbecause the additional parameter LER would not add furtherinformation, but complicate the arrangement. The AO classi-fication is based on a 3 class system for all anatomical regionsand has been shown to be able to predict outcome [19]. Byfollowing this strategy, the total length of the joint surface ofthe fifth metatarsal base was divided into three equal parts andassigned to three fracture types as follows: type I fractures

included fractures of the lateral third, type II fractures of themiddle third, and type III fractures of the medial third. Type Ifractures correlate with a JR≤0.33, type II fractures with aJR>0.33 and≤0.66, and type III fractures with a JR >0.66.Fractures with a step-off exceeding 2 mm were generallyconsidered to influence the clinical outcome [11–13] anddefined as displaced. Therefore, we introduced an A-type (I-IIIA) without a relevant displacement and a B-type (I-IIIB)with a fracture-step-off ≥2 mm (Fig. 3).

Validation and interobserver reliability of the classificationsystem

In our study, we classified 26 % of fractures as type I (n=24)with a JR of 0.33 or less. Of type I fractures, 50 % (n=12)were non-displaced (IA) and 50 % (n=12) were displaced(IB). We classified 37 % of the fractures as type II (n=35)with a JR of 0.34 to 0.66. Forty percent (n=14) were non-displaced fractures of type IIA and 60 % (n=21) weredisplaced type IIB fractures. We classified 37 % (n=35) offractures type III with a JR of greater than 0.66. Type IIIA25 % (n=9) were non-displaced and type IIIB 75 % (n=26)

Fig. 2 a–d The patients were distributed into three groups: group 1included patients with non-displaced fractures (step-off<2 mm) whowere treated conservatively. Patients with fractures showing a step-offexceeding 2 mm in at least one radiographic plane in initial X-rays wereconsidered primarily displaced and were treated operatively (group 2).Group 3 included patients with initially non-displaced fractures who weretreated conservatively and secondarily displaced (fracture step-off>

2 mm). The joint fracture angle α (shown in part a), the ratio of baselength/basal fracture length (JR, shown in part b), the lateral edge length(LEL, shown in part c), and the ratio of the lateral edge length and baselength (LER, shown in part d) were measured, calculated and comparedbetween the different groups. All values are reported as mean ± SD.Results for statistical comparisons are indicated in the figures

470 Skeletal Radiol (2014) 43:467–474

were displaced. This classification was a consensus decisionof three surgeons including the author.

The positive predictive value for a secondary displacementwas 0 in type I fractures (0 of 12 cases), 0.0625 in type IIfractures (one case of 16), and 0.45 in type III fractures (nineof 20 cases). Correspondingly, the risk for secondary displace-ment of initially non displaced fractures is 0 % in type Ifractures, 6.25 % in type II fractures, and 45 % in type IIIfractures. The distribution of risk for secondary displacementis statistically significantly different comparing all groups (p=0.019). This is confirmed calculating the individual risk dif-ferences between type I and III fractures (p=0.01) and type IIand III fractures (p=0.006). If all type III fractures would haveundergone primary operative treatment, 55 % would havebeen stabilized unnecessarily.

The JFAwas higher in type I than type II and III fractures(60±21° vs. 52.5±15° vs. 51.7±14° mm, p=0.033, Fig. 4a).In order to confirm the classification, results for JR are alsoindicated, although this was the basis for developing theclassification types. The JR was significantly higher in typeIII fractures compared to type II and type I fractures (0.29±0.11 vs. 0.53±0.14 mm vs. 0.90±0.09 mm, p<0.0001,Fig. 4b). The LELwas significantly higher in type III fracturescompared to type II and type I fractures (15±4 mm vs. 12±3 mm vs. 8±3 mm, p<0.0001, Fig. 4c). The lateral edge ratio(LER) was also compared between all three fracture types

(Fig. 4d). It was significantly higher in type III than in typeII and type I fractures (0.67±0.18 vs. 0.51±0.11 vs. 0.36±0.13, p<0.0001). pvalues represent the tests for multiple com-parisons; pvalues for individual comparisons may be seen inFig. 4.

The visually analyzed fractures using the introduced clas-sification system revealed in 74 % an identical result. Asubstantial level of agreement (κ=0.72) according to Landisand Koch [18] was found in the evaluation of the differentfracture types by observer 1 and 2. The secondary displace-ment of fractures was confirmed by all examiners in 100 %.

Discussion

Fractures of the fifth metatarsal base are common fore footinjuries and are found following indirect trauma. The fracturemechanism is considered to be a fore foot supination withplantar flexion, which leads to an extendedmuscle force of theperoneus brevis tendon, the metatarso-cuboid ligaments, andthe lateral cord of the plantar fascia at the fifth metatarsal base.Thus, the fracture morphology is strongly dependant on theanatomic predisposition and the types of forces affecting thefifth metatarsal base. We analyzed a total of 95 fractures of thefifth metatarsal base for epidemiologic and radiomorphometriccharacteristics as well as for fracture stability during follow-up.Based on these findings, we developed a new classification forfractures of the fifth metatarsal base considering the risk forsecondary displacement. All observations taken together, thelargest fragments had the highest risk for secondarydisplacement.

Fractures of the fifth metatarsal base are generally found atall ages and gender. In young patients, they have to be stronglyseparated from the apophysitis at fifth metatarsal base in thegrowing skeleton [20]. Recent epidemiologic studies on oste-oporotic women have shown a high prevalence of fractures ofthe fifth metatarsal in women older than 65 years and a strongassociation with a low bone mineral density [21, 22]. Incontrast, the introduced analysis focuses more on a populationwith a high degree of sports activities, which is reflected bythe mean age of 39 years.

The treatment of fifth metatarsal base fractures ranges fromnon-operative treatment with a lower leg cast and non-weightor partial bearing for 6 weeks to open reduction and internalfixation with screws or tension band wiring [9, 23–25]. Ingeneral, a good healing potential of the fifth metatarsal base isassumed due to its sufficient circulation. Therefore, someauthors concluded that metaphyseal fractures not extendingbeyond the distal end of the 4th to 5th intermetatarsal articu-lation should be treated functionally, regardless of the numberof fragments, displacement, and intraarticular involvement[26]. Based on the samemeta-analysis, it was further conclud-ed that only meta-diaphyseal fractures located at the distal end

Fig. 3 A classification system was developed based on an increased riskfor secondary displacement of fractures with a more medial joint entry ofthe fracture line at the fifth metatarsal base. Dependant on the joint entryof the fracture line at the fifth metatarsal base (lateral one-third, middleone-third, and medial one-third) type I, II, or III were defined. Fractureswithout displacement were summarized as A-type (I-IIIA) and with afracture-step-off >2 mm as B-type (I-IIIB)

Skeletal Radiol (2014) 43:467–474 471

of the 4th to 5th intermetatarsal articulation or just distallyrequire early ORIF. This is mainly based on the results of thestudy by Mologne et al. [27] showing shorter return to sportperiods in the intervention group with screw fixation. Theconclusion for the proximal fractures of the fifth metatarsalbase not extending beyond the distal end of the 4th to 5thintermetatarsal articulation is only based on trials comparingdifferent conservative treatment methods [28, 29], but notoperative versus non-operative intervention. Especially in apopulation with a high sports level, early return to physicalactivities may play a decisive role. Since continuous tractionof the peroneus brevis tendon may result in fracture instabilitywith possible development of a non-union, which is certainlya problem [27] in these fractures, the risk for secondarydisplacement should be reflected by a fracture classification.The highest risk (45 %) was found to be associated with typeIII fractures with the step-off in the medial part of the fifthmetatarsal base. Regarding the risk for development of aposttraumatic osteoarthritis (OA) of the metatarso-cuboidjoint, some authors strongly recommend operation fordisplaced, intraarticular fractures [12, 13]. Hereby, a fracturestep-off exceeding 2 mm is widely recognized as a risk factorfor OA [30] and has been introduced in the presented classi-fication system. Although classifications inaugurated for thisanatomical region by Stewart [31], Lawrence [4], or Holzach[32] sufficiently discriminate between meta-diaphyseal Jonesfractures and proximal fractures of the fifth metatarsal notextending beyond the distal end of the 4th to 5thintermetatarsal articulation, they all do not reflect the risk fora secondary displacement within the second fracture entity.

Therefore, our radiomorphometric analysis of the 95 fracturesdistinguished between non-displaced, primary displaced, andsecondary displaced fractures in order to deduce a risk-adaptedclassification. It could be shown that secondary displacedfractures had a longer distance between the lateral edge andthe lateral fracture step-off (=LEL). Anatomical studies haveshown that the peroneus brevis tendon has a lateral insertion upto 15 mm at the fifth metatarsal base [6]. Since the muscle pullof the peroneus brevis tendon is considered as one of the majorcauses for avulsion fracture mechanisms at the fifth metatarsalbase, it is obvious that LELs larger than 15 mm increase therisk for displacement. In this case, all muscle fibers are attachedto the small proximal fracture piece and pull it away moreeasily, a fact that could be confirmed in the presented analysis.However, it remains unclear why some fractures primarilydisplaced and other after a few days. We assume that insecondary displaced fractures the initially impacting energywas lower compared to primary displaced fractures, but wewere not able to clearly confirm this assumption by the histo-ries of our patients. Actually, there are only a few studiesexactly describing the acting forces as by Orendurff et al.[33] describing the largest bending moments to the fifth meta-tarsal during acceleration maneuvers (20±13.1 N/cm2) follow-ed by running straight (11.6±8 N/cm2). Unfortunately, such anexact measurement will not be possible within the setup of aclinical trial. Further, it can be supposed that for initially notdisplaced fractures, a lower leg cast is not able to completelyreduce the pulling force of the peroneus brevis tendon,allowing the secondary displacement despite cast treatment.Casting the hind-foot in valgus position is maybe an option to

p<0.0001

p<0.0001p<0.0001p<0.0001

p<0.0001p<0.0001

p<0.0001

p<0.0001p<0.0001

EFig. 4 a–e Theradiomorphometric parametersjoint-fracture angle α (JFA,shown in part a), ratio of baselength/basal fracture length (JR,only shown in part b to confirmaccuracy of classification), lateraledge length (LEL, shown in partc), and the lateral edge ratio (LER,shown in part d) were comparedbetween all three fracture types.All values are reported as mean ±SD. Results for statisticalcomparisons are indicated in thefigures. Part e demonstrates anexample for secondarydisplacement of a type III fracture

472 Skeletal Radiol (2014) 43:467–474

reduce the tension of the peroneus brevis tendon. Since avul-sion fractures are caused by exceeding muscle force, it can besuggested that non-displaced fractures of the fifth metatarsalbase can displace secondarily during conservative treatment.Heineck and colleagues [34] found that reduced activity of theperoneus muscle is critical for adequate treatment and can beachieved by partial weight bearing but not by immobilizationof the talo-crural joint. This is in concordance with the fact thatthe risk of fracture displacement depends on fracture morphol-ogy related to the anatomical onset of the lateral cord of plantaraponeurosis and of the short peroneal muscle tendon.

A second risk factor for a secondary displacement was ahigh joint ratio (JR); a high JR stands for a medial fractureentry into the metatarso-cuboid joint. As recent studies re-vealed, another avulsion force for fractures of the fifth meta-tarsal base is mediated by the lateral aspect of the plantaraponeurosis and of the plantar metatarso-cuboidal ligaments[6, 7]. From this point of view, it could be assumed that afracture portion with a more medial proximal entry within themetatarso-cuboid joint is more unstable, because the fracturedoes not run through the middle of the plantar ligamentousstructures but medial of its main cord. We additionally couldshow that fractures with moremedial joint entry are associatedwith a longer lateral edge length (LEL), which destabilizes thedistal fracture piece additionally as discussed above.

According to these findings, we developed a classificationsystem based on the JR and the fractures step-off. The lastcriterion has already been considered as a general prognosticcriterion for fifth metatarsal fractures [30]. The lateral edgelength (LEL) was not directly included in the classification;however, since the LEL and the lateral edge ratio (LER)statistically significantly correlated with the joint ratio (JR),both were indirectly taken into account. Further, it was shownthat LEL and LER were significantly higher in type III than intype II and I fractures.

The clinical practicability and interobserver reliability of theclassification systemwere tested by two surgeons with differentexperience levels and no experience with the classificationbefore study onset. A substantial level of agreement was foundin the assignment of all fractures to the different fracture typesof our classification system. Especially, a definitive distinctionbetween fractures of type II and III seems to be crucial sincelater therapy decision might depend on this correct classifica-tion. Therefore, we recommend an exact radiomorphometricanalysis (digital or manual, e.g., using a ruler or an electronicmeasurement device) in case of doubt. Based on the averagelength of the fifth metatarsal base of 22 mm, the measurementaccuracy should be between 1 and 2 mm considering a confi-dence interval of less than 10 %. Overall, the suggested classi-fication was only examined for reproducibility and needs to befurther validated in a separate data set.

The limitations of this study are the small number of casesin group 3, which included the secondarily displaced

fractures. Despite this fact, evaluation resulted in statisticalsignificance for the made observations with concordantdata of two independent observers. The informative valueof the study is further restricted by the limited number ofobservers.

Summarizing the results of this study, at first,radiomorphometric characteristics of fifth metatarsal base frac-tures were evaluated identifying risk factors for a secondarydisplacement following initial conservative treatment. Based onthis, a new classification system dividing the articular surfaceinto thirds was introduced, demonstrating good interobserverreliability. Considering a fracture step-off larger than 2 mm as arisk factor for a posttraumatic OA, open reduction and internalfixation (ORIF) of all type B fractures may be recommended.However, a fracture treatment remains individually decided andis dependent on general health conditions and functional de-mands. Furthermore, all initially non-displaced fractures with amore medial joint fracture step-off (type IIIA) may, consideringthe same circumstances, be recommended for ORIF because ofits risk for secondary displacement. At least the risk of 45 %should be part of the discussion with the patient about treatmentrisk when deciding for conservative treatment. If non-operativetreatment is preferred, fractures at risk for secondary displace-ment should undergo frequent X-ray controls within the first2 weeks, in case a secondary displacement would trigger thedecision for ORIF.

Conflict of interest Each author certifies that he has no commercialassociations (e.g., consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest inconnection with the submitted article.

References

1. Anderson LD. Injuries of the forefoot. Clin Orthop Relat Res.1977;122:18–27.

2. Kavanaugh JH, Brower TD, Mann RV. The Jones fracture revisited. JBone Joint Surg Am. 1978;60(6):776–82.

3. Quill Jr GE. Fractures of the proximal fifth metatarsal. Orthop Clin NAm. 1995;26(2):353–61.

4. Lawrence SJ, Botte MJ. Jones’ fractures and related fractures of theproximal fifth metatarsal. Foot Ankle. 1993;14(6):358–65.

5. Dameron Jr TB. Fractures of the proximal fifth metatarsal: selectingthe best treatment option. J AmAcad Orthop Surg. 1995;3(2):110–4.

6. Richli WR, Rosenthal DI. Avulsion fracture of the fifth metatarsal:experimental study of pathomechanics. AJR Am J Roentgenol.1984;143(4):889–91.

7. Theodorou DJ, Theodorou SJ, Kakitsubata Y, Botte MJ, Resnick D.Fractures of proximal portion of fifth metatarsal bone: anatomic andimaging evidence of a pathogenesis of avulsion of the plantar apo-neurosis and the short peroneal muscle tendon. Radiology.2003;226(3):857–65.

8. Clapper MF, O’Brien TJ, Lyons PM. Fractures of the fifth metatarsal.Analysis of a fracture registry. Clin Orthop Relat Res. 1995;315:238–41.

Skeletal Radiol (2014) 43:467–474 473

9. Strayer SM, Reece SG, Petrizzi MJ. Fractures of the proximal fifthmetatarsal. Am Fam Physician. 1999;59(9):2516–22.

10. Vorlat P, Achtergael W, Haentjens P. Predictors of outcome of non-displaced fractures of the base of the fifth metatarsal. Int Orthop.2007;31(1):5–10.

11. Hansen ST. Foot Injuries in Skeletal Trauma. In: Browner BD, editor.Foot Injuries in Skeletal Trauma. Philadelphia: Saunders, W.B; 1992.p. 1984–6.

12. Leumann A, Pagenstert G, Fuhr P, Hintermann B, Valderrabano V.Intramedullary screw fixation in proximal fifth-metatarsal fractures insports: clinical and biomechanical analysis. Arch Orthop TraumaSurg. 2008;128(12):1425–30.

13. Rettig AC, Shelbourne KD, Wilckens J. The surgical treatment ofsymptomatic nonunions of the proximal (metaphyseal) fifth metatar-sal in athletes. Am J Sports Med. 1992;20(1):50–4.

14. Ackermann C, Lam Q, Linder P, Kull C, Regazzoni P. Problems inclassification of fractures of the proximal humerus. Z UnfallchirVersicherungsmed Berufskr. 1986;79(4):209–15.

15. Saklad M. Grading of patients for surgical procedures.Anesthesiology. 1941;2:281–4.

16. Kelikian AS, Sarrafian SK. Sarrafian’s anatomy of the foot and ankle:descriptive, topographical, functional. 3rd ed. Philadelphia: WoltersKluwer Health/Lippincott Williams & Wilkins; 2011.

17. Bahrs C, Schmal H, Lingenfelter E, Rolauffs B, Weise K, Dietz K,et al. Inter- and intraobserver reliability of theMTM-classification forproximal humeral fractures: a prospective study. BMCMusculoskelet Disord. 2008;9:21.

18. Landis JR, Koch GG. The measurement of observer agreement forcategorical data. Biometrics. 1977;33(1):159–74.

19. Swiontkowski MF, Agel J, McAndrewMP, Burgess AR, MacKenzieEJ. Outcome validation of the AO/OTA fracture classification sys-tem. J Orthop Trauma. 2000;14(8):534–41.

20. Dameron Jr TB. Fractures and anatomical variations of the proximalportion of the fifthmetatarsal. J Bone Joint SurgAm. 1975;57(6):788–92.

21. Hasselman CT, Vogt MT, Stone KL, Cauley JA, Conti SF. Foot andankle fractures in elderly white women. Incidence and risk factors. JBone Joint Surg Am. 2003;85-A(5):820–4.

22. Varenna M, Binelli L, Zucchi F, Beltrametti P, Gallazzi M, SinigagliaL. Is the metatarsal fracture in postmenopausal women an osteopo-rotic fracture? A cross-sectional study on 113 cases. Osteoporos Int.1997;7(6):558–63.

23. Den Hartog BD. Fracture of the proximal fifth metatarsal. J AmAcadOrthop Surg. 2009;17(7):458–64.

24. Fine WW. Comminuted fracture of fifth metatarsal; open reductionwith wire fixation. J Natl Assoc Chirop. 1957;47(10):509–11.

25. Portland G, Kelikian A, Kodros S. Acute surgical management ofJones’ fractures. Foot Ankle Int. 2003;24(11):829–33.

26. Polzer H, Polzer S, Mutschler W, Prall WC. Acute fractures to theproximal fifth metatarsal bone: development of classification andtreatment recommendations based on the current evidence. Injury.2012;43(10):1626–32.

27. Mologne TS, Lundeen JM, Clapper MF, O’Brien TJ. Early screwfixation versus casting in the treatment of acute Jones fractures. Am JSports Med. 2005;33(7):970–5.

28. Gray AC, Rooney BP, Ingram R. A prospective comparison of twotreatment options for tuberosity fractures of the proximal fifth meta-tarsal. Foot (Edinb). 2008;18(3):156–8.

29. Wiener BD, Linder JF, Giattini JF. Treatment of fractures of the fifthmetatarsal: a prospective study. Foot Ankle Int. 1997;18(5):267–9.

30. Egol K, Walsh M, Rosenblatt K, Capla E, Koval KJ. Avulsionfractures of the fifth metatarsal base: a prospective outcome study.Foot Ankle Int. 2007;28(5):581–3.

31. Stewart IM. Jones's fracture: fracture of base of fifth metatarsal. ClinOrthop. 1960;16:190–8.

32. Linder P, Heim D, Braun R, Holzach P. Base fracture of the 5thmetatarsal. Classification and therapy using a strapping bandage. ZUnfallchir Versicherungsmed Berufskr. 1986;79(4):257–60.

33. Orendurff MS, Rohr ES, Segal AD,Medley JW, Green 3rd JR, KadelNJ. Biomechanical analysis of stresses to the fifth metatarsal boneduring sports maneuvers: implications for fifth metatarsal fractures.Phys Sportsmed. 2009;37(2):87–92.

34. Heineck J, Wolz M, Haupt C, Rammelt S, Schneiders W. Fifthmetatarsal avulsion fracture: a rational basis for postoperative treat-ment. Arch Orthop Trauma Surg. 2009;129(8):1089–92.

474 Skeletal Radiol (2014) 43:467–474