Fracture : a study of 13 cases and review of the literafure S R Arulenthiranl, G M P Sirimanna2, G L Punchihewa3

Sri Lanka Journal of Dermatology, 201.1., 15, 35-39

Introduction road traffic accidents and struck by a motor vehicle. Four were due to fall from height and one sustained Fracture blisters are a relatively rare entity manifested when a heavy metal bar had fallen from the by vesicles or bullae that arise on markedly swollen top of a heavy vehicle. skin overlying a fracturel. They commonly form following fractures caused by high energy trauma or Fracture pattems associated with fracture severe twisting type of injuries and occur in charac- formation in this series included two tibial-plateau fracture blisters teristic locationsl. Reports conceming fractures, two tibial-plateau fractures with proximal here the are lirnited in the literature and we report fibular fractures, one tibial bi-condylar fracture, four clinical characteristics of fracture blisters seen in 13 tibial shaft fractures with fracture fibula, two tibial patients. malleolar fracfure, one calcaneus fracfure associated with metatarsal and one supracondylar Method fracture of humerus.

2010 june 2011, 13 During the period from July to Fracture blisters were found to be located around patients who developed blisters after acute injuries the knee, over the tibia and on dorsum of foot and were referred to the Dermatology Unit of National in patients with lower extremity fractures and -near Hospital of Sri Lanka. These patients were pros- the elbow in the patient with supracondylar pectively followed up. Demographic data and data fracture. Number of blisters, their size and position concerning the mechanism of , {racture typte, about each fracture were variable. One patient had time course in the development of lesions, clinical single blister and the others had multiple blisters with charac-teristics, their management and impact of the number ranging frornZ to more than 10. Blister blisters on the fracture management and outcome size ranged from few millimeters to as larger as 10 cm were studied. in diameter. Blisters were discrete and unilocular in eight patients but in patients who had more than 10 Results blisters, several tiny vesicles localized to one area were also noted, some of them coalesced to form large There were thirteen patients who developed fracture bullae. blisters during this period; ten males and three females with the age ranging from L4 to 72 years. Blister fluid was either clear or haemorrhagic. Twelve patients had lower extremity fractures and Some had combination of both clear and haemorr- one had upper extremity fracture. Two patients had hagic blisters. In addition, some of the blisters which multiple fractures involving the same extremity and contained clear fluid initially became haemorrhagic one had fractures in both lower limbs. No one had after 3-4 few days as they became older; and in one other organ involvement. instance a blister with clear fluid became blood- stained when it refilled after its content had been Past medical history was significant in eight aspirated for analysis. patients. This included insulin dependent diabetes mellitus in one, non-insulin dependent diabetes Blisters were tense when they appear and mellitus in five (one was newly diagnosed), became flaccid as they become older. hypertension in two and affective disorder in one whose medications included sodium valproate. Two Marked skin and soft tissue swelling were seen of them were smokers and alcohol abusers. in all patients. Seven patients had associated ery- thema, five had ecchymoses and four had superficial The mechanism of fracture injury was a high skin abrasions. Compartment syndrome occurred in energy trauma in most cases. Eight were involved in one patient. The blisters were aq/mptomatic in all.

lsenior Registrar,2Consultant Dermatologist,3Consultant Orthopaedic Surgeon, National Hospital of Sri Lanka, Colombo. 36 S R Arulenthiran, G M P Sirimanna, G L punchihewa

In most patients blisters developed within 24 - trauma and may occur postoperatively as well. In 48 hours of acute injury; the interval ranged from 4 this series the shortest time interval observed was 4 hours to 7 days. Blisters occurred within 24 hours in hours and in majority cases (69.2oh) blisters occurred nine, between} to 48 hours in two and48-72in one. within 48 hours of injwy. Fracture blisters rarely occur In two patient blisters developed postoperatively. with open fractures. One patient with compound Blisters were discovered as an incidental clinical fracture and compartment syndrome developed finding at initial presentation or dressing change or blisters after ORIF and fasciotomy. in the surgical theatre. Hence the exact time interval between the injury and the appearance of fracture blister was difficult to be determined in every case. In They occur in characteristic locations, most two patients new blisters occurred daily over the commonly over the tibia, , elbow and knee - period of2-3 days. where there is little soft tissue between bone and skin or in areas of restricted skin motility. All the blisters Blister fluids aspirated from intact blisters in nine in our patients were confined to sites distal to mid- patients were found to be sterile on microbial studies. shaft of humerus in upper exkemity *d distal to knee Skin biopsy samples in one displayed intra epidermal in lower extremity as observed by varela et all in their and sub epidermal blisters without inflammation and series. Patients with existing co-morbidities like another in patient showed only sub corneal diabetes mellitus, hypertension, peripheral vascular separation. disease, smoking history, alcohol abuse and lymphatic obstruction may be at increased risk of Blisters were left intact and allowed to heal spon- developing fracture blistersl. Ten out taneously in ten patients. When blisters ruptured of thirteen patients in this series had one spontaneously they were covered with non adherent or more comorbid sterile dressing with/without topical antibiotic conditions and it is interesting to note that six of them had diabetes as a cream. Blister bed re-epithelialization occurred in L3 comorbid condition. The exact role to 20 days (mean of 15.3 days). In three patients played by these conditions in the formation of blisters were ruptured after 1,5, 7 and 2 days (but not fracture blisters is not known. Undue joint or limb deroofed) under sterile conditions and covered with manipulation, dependent positioning, heat povidone iodine paint and dry dressing, and re- application in patients at risk may produce fracture epithelialization noted after 6 days, 8 days and 11 blister in an otherwise relatively minor injuryr. days respectively. Fracture blisters may occur singly or in multiples When blisters occurred preoperatively, they had and the size can vary from few mm to several cm. an impact on the patient mau.ragement, causing delay Clinically the blister may contain either clear/serous in surgery, change in operative plan, and prolonging or haemorrhagic fluid. They may be associated with the hospital stay. Four patients underwent open skin and soft tissue swelling, erythema and reduction and internal fixation and seven patients ecchymoses. Generally they are tense blisters but were managed non-operatively when soft tissue older blisters tend to swelling had subsided and blister bed re- become more flaccid and are more likely epithelialized. The overall mean delay in definitive to contain haemorrhagic fluid. Blister fluid within the intact management from the time of presentation was 17.9 blister has been shown to be sterile days, range 12-28 days. Patient management was transudatel. unaffected when blisters developed postoperatively. Histologically, location of the blister may be intra All these patients were empirically treated with epidermal or at dermo epidermal junction. Skin antibiotics. There were no major complications biopsies were done only in few cases in out study directly related to the presence of blisters. One patient and findings were similar to the observations made with NIDDM and tibial plateau fracture who under- by other authors. Biopsy examination of 15 blisters went ORIF developed pseudomonas infection of the by Varela et al1 showed sub corneal blister located surgical which responded to appropriate superficial to the granular layer in 13 and sub antibiotics without any sequelae. A11 blisters healed epidermal blister in 2. h a dinical and histological without any scarring. None had any complications study performed related to fracfure union. by Giordano et al3, the authors showed cleavage injuries at dermoepidermal junction. There was complete separation of dermis from Discussion epidermis in haemorrhagic blisters whereas there The incidence of fracture blisters has been reported were scattered areas of retained epidermal cells on as 2.97o of all acute fractures requiring hospitali- the dermis in clear blisters which authors believed zationl. They usually appear within 24-48 hours of contributed to rapid re-epithelialization and less acute injury but may occur as late as 3 weeks after the morbidity seen with clear blisters.

Sri Lanka lournal of Dermatology Fracture blisters 3/

Table 1. Salient clinical features

Serial Age Type of Blister Soft Time Associated Blister Delay in num- and fracture charac- tissue interoal illnesses managemefit sutgery ber Sex teristics injury before no, size, blister type, site

48 /M Tibial Multiple (4), Marked 18 h Diabetes None. 18 days plateau large (3-4cm), Oedema. (<24h) mellitus D.y (oRrF) and fibula clear and ecchymoses dressing blood, on knee only. and shin

60 /M Tibia and Multiple (1.4), Oedema, 20 h None Ruptured after 22 days fibula varying size ecchymoses (<24h) 15 days (ORIF) segmental (few mm-4cm), and then povidone clear and blood, abrasions. iodine dressing. on knee, shin

42/M Tibial Multiple Erthema, (<24 h) Diabetes Ruptured after 15 days plateau veicled and Oedema, mellitus 15 days then (POP and proximal bullae, varying ecchymoses povidone iodine cast) fibula size (few mm and abrasions dressing -7cm) clear and blood, on leg

37 /M Tibial Multiple (4), Erythema, 20 h Depression None. 24 days plateau lense, 1-3 cm, oedema, (<24 h) on sodium Dry dressing (oRIF) fracture clear and blood, ecchymoses valproate only orrknee and upper leg.

1,4/M Supracondylar Single,2 cm, Oedema 24 - 48 h None Dry dressing 18 days. fracture of initially clear MUA and left humerus -r haemorrhagic POP later. on elbow back slap

30/M Calcaneus and Multiple (5), Oedema 24 -48 h None None 18 days metatarsal bone 1-3 cm, clear -+ (MUA) Lis Franc fracture blood on dorsum dislocations of foot.

53/M Upper tibia Multiple (>10), Marked <24 h Diabetes Left intact. Blisters and fibula few mm - 7 cm oedema, (Blisters mellitus povidone appeared (compound) clear and blood compartment appeared iodine after ORIF knee, leg (medial) syndrome. after ORIF dressing and fascio- and fascio- tomy tomy)

32 /M Tibial Multiple (2), Oedema 07 days Diabetes Dry Blisters bimalleolar large 3-4 cm after mellitus dressing appeared fracture tense haemor- ORIF 07 days rhagic ankle after ORIF

9 45/M Tibial and Multiple (6), Oedema <24 h Smoker, Left intact. 14 days fibiual shaft large L4 cm alcohol povidone MUA and tense clear -+ abuse iodine dressing POP haemorrhagic (Continued)

Vol. 15, 2011 G L Punchihewa 38 S R Arulenthiran, G M P Sirimanna,

Smoker, Left intact. 15 days 10 54 Tibial Plateau Multip1e (6), Oedema, 48 -72h /M and fracture large 1 - cm abrasions alcohol Povidone MUA tense clear abuse iodine dressing POP -+haemorrhagic

2 days POP LL 72/F Tibial and Multiple > 30, Oedema, 24h Diabetes Povidone cast fibiual shaft feuumm - 10 cm erYthema mellitus, iodine tense clear hYPertension dressing -+ haemorrhagic

24h Hypertension Ruptured. 13 daYs L2 34/F Tibial Plateau Multiple (10), Oedema, POP fracture tense clear erYthema Povidone -+ haemotrhagic iodine dressing

<24h Diabetes Ruptured. 28 daYs 13 47 Medial Multiple (4) Oedema, /F ORIFL mellitus tense clear erYthema, mellitus Povidone -+ haemorrhagic abrasions, iodine ecchYmoses dressing

Under Anaesthesia, POP - Plaster of M - MaIe, F - Female, OI{IF - Open Reduction and Internal Fixation, MUA - Manipulation Paris ffi ffi 3 t{i l*':find

demonstrating Medial view right leg one day after injury show in Lateral view left ankle one day after marked oedema and injury clear vesicles and marked edema, ecchymoses, and haemorrhagic bullae. bullae.

Sri Lanka lournal of DermatologY Eracture blisters 39

The pathophysiology of fracture blisters is multi- Giordano and Kovals compared these three methods factorial. Fracture blisters are thought to be caused of managing fracture blisters in 53 patients and found by dermal-epidermal separation secondary to strains no significant difference in the outcome. The authors created in the skin during the initial fracture defor- recommended leaving blisters intact and to deroof mation. When a critical strain is reached during the and cover only those blisters that have spontaneously fracture mechanism, the differing elasticity aird ruptured. Strauss Eric et al6, :using a prospective vesico-elastic properties of the dermis and epidermis protocol, deroofed all blisters at presentation and cause the two layers to separate and then the treated with silvasulfadiazine dressings. They found combination of the inflammatory cascade and this was successful in minimizing soft tissue Starling forces causes the fluid to pass into the complications by promoting re-epitheliazation in all potential spaces between the dermis and epidermis. non diabetic patients; two diabetic patients developed The results of the biome-chanical study involving full thickness skin breakdoum at blister bed. There is uniaxial strain test at several levels on cadaver ankle no compelling evidence to support any method over skin specimens performed by Giordano et al support anotherT. If feasible the blisters may be allowed to hypothesisa. this resolve spon-taneously which may take 10 to 14 days and, surgical treatment can be delayed; or Early stabilization of fractures and elevation of alternatively they can be treated aggressively which limbs to reduce oedema and vascular congestion may may produce faster resolution, within 5 to 10 daysaJ. decrease the incidence fracture blisters. Ear1y of It appears reasonable to avoid incisions tfuough a surgical intervention before blister develops may non-epithelialized blister bed, particularly a reduce the incidence of blister formation. In one study haemorrhagic blister, if possibleT. Varela et al1 have shown that that patients who under- went open reduction and internal fixation within 24 The clinical characteristics of fracture blisters observed hours of injury had a significantly lower incidence of in our patients are similar to those described in fracture blister formation (2'/o) cornpared to those previous sfudies. Six patients had diabetes mellitus patients whose surgery delayed for >24 hours (8%). as a co-morbidity. No one developed any blister bed complications. But postoperative wound infettion Presence of fracture blisters may delay the occurred in one fiabetic patient who underr,rrent open surgery / alter the operative plan and/may be reduction and intemal fixation. Further studies associated with increased risk of complications like with infection, skin breakdown, and delayed wound large number of patients and prospective treatment protocol are needed to compare the outcome of various healing, scarring/ and problems in fracture unionl. treatment methods. Haemorrhagic blisters appear to be associated with increased risk of complications. Lr one study, Varela et all found major postoperative wound infections References when blisters were present at the time of surgery and recommends avoiding surgical incisions through 1. Varela CD, VaughanTK, CarrJB, Slemmons BK. Fracture blisters: clinical and pathological aspects. fracture blisters and associated damaged soft tissues. lournal of Orthopaeilic Trauma 1993; 7(5); 417-27. However Giordano et als obsewed no skin or wound complications when incisions were made through 2. McCann S, Gruen G. Fractwe blisters: a review of literature. D ermatolo gy N ur sin g 1997 . clear-fiIled blisters or adjacent to either type of blisters. But wound healing complications developed in two 3. Giordano CP,KovalKJ,ZuckermanJD, Desai P. Fracture patients in whom incisions were made through blisters. Clin Orthop 1994; 307: 2l+21,. blood-filled blisters. 4. Giordano CP, Scott D, Koval KJ, Kummer F, Atik T, Desai P. Fracture blister formation: alaboratory study. The loumal of Trauma: Injury, lnfection, and Critiul Care 1995;38(6): \A/hen fracture blisters have occurred variety of 907-9. treatment options are available, including: 5. Giordano CP, Koval KJ. Treatment of fracture blisters: a L. Sterile deroofing and application of silver prospective study of 53 cases. I Orthop Trauma 1995;9: sulfadiazine and / or non-adherent dressings, 171-6. 2. Sterile aspiration alone (with maintenance of 6 . Strauss ![, Petrucelli GB, Matthew K, Kenneth IE Kenneth overlying roof), and AB. Associated with lower-extremity fracfure: results of a prospective treatment protocol. lournal of Orthopaedic 3. Leaving the blister intact. Trauma 2006; 2O: 618-22. 7. Bu.cholz RW, Court-Brown C, Heckman ]D. Rockwood In ten of our patients blisters were left intact and in and Green's Fractures in Adult s; 7th editon,2009; Volume three blisters were ruptured (but not deroofed), at the 1, chapter 56, Pilon fractures; p1932. discretion of treating surgeon, after varying intervals. 8. Chapter 50, General Principles in Fracture Treatment. Number of patients was small and no standard p3075. protocol was followed to make any stastical 9. Frances B, Michele M, John M. Fracture blisters. I Am Acad comparisons. Dermatol L994; 3O 1033-4.

Vol. 15, 2011