The Journal of Bone and Joint Surgery 78:594-6 (1996)
© 1996 The Journal of Bone and Joint Surgery, Inc.
Sudden Death of a Child Who Had Pain in the Knee and Varicella. A Case Report*
DYSON L. HAMNER, M.D. ,
ROGER M. LYON, M.D. and
JOHN B. EMANS, M.D. , BOSTON, MASSACHUSETTS
Investigation performed at the Department of Orthopaedic Surgery, Children's Hospital, Boston
 |
Introduction
|
|---|
We report the case of a four-year-old girl who died within hours after she was seen for pain in the knee and varicella. The death was due to fulminant sepsis with group-A Streptococcus. The postmortem examination revealed a bacterial infection of the varicella skin lesions, which resulted in bacteremia leading to generalized sepsis. The pain in the knee was initially thought to be secondary to septic arthritis, varicella arthritis, or osteomyelitis, but it was diagnosed after death as septic popliteal thrombophlebitis associated with fulminant group-A streptococcal septicemia. We report the case of our patient to bring attention to this rare but life-threatening cause of pain in the knee.
 |
Case Report
|
|---|
A four-year-old girl was seen in the emergency room at Children's Hospital because of pain in the left knee of two days' duration. Two weeks before admission, she had an ear infection that was treated with sulfa compounds. Six days before admission, a diffuse vesicular rash, consistent with varicella, developed. Four to five days before admission, the child had a fever while she was at home, but it was not measured. Two days before admission, she began to have pain in the left knee that steadily increased in severity such that, by the time of admission, she was unable to bear weight on the left lower limb. The medical history was unremarkable. She had an allergy to penicillin.
On the morning of admission, the patient was seen at an outside clinic, where laboratory studies revealed a white blood-cell count of 8.7 x 109 per liter (8700 per cubic millimeter), a platelet count of 119 x 109 per liter (119,000 per cubic millimeter) (normal, 150 to 400 x 109 per liter [150,000 to 400,000 per cubic millimeter]), and an erythrocyte sedimentation rate of twenty-five millimeters per hour (normal, zero to twenty millimeters per hour). She was referred to Children's Hospital for additional treatment.
On arrival at the hospital, the child was evaluated in the emergency room for pain in the left knee, which was the primary symptom. She had no history of antecedent trauma. The temperature was 39.2 degrees Celsius, the blood pressure was 103/61 millimeters of mercury (13.73/8.13 kilopascals), the pulse rate was 128 beats per minute, the respiratory rate was thirty-two breaths per minute, and the oxygen saturation on room air was 99 per cent. The patient had diffuse crusted lesions involving the face, neck, trunk, arms, and lower extremities that were consistent with varicella.
A consultation with the orthopaedic service was requested. The patient was irritable but alert and was cooperative with the examination. The left knee was slightly swollen about the medial and lateral joint line, but it had a full range of motion with pain at the terminal degree of flexion and increased warmth compared with the contralateral knee. There was no pain or tenderness elsewhere in the lower extremity, including the calf, and the dorsalis pedis pulses were palpable bilaterally. The osseofascial compartments of the leg were soft to palpation. Plain radiographs of the left knee revealed negative findings. An arthrocentesis was performed, yielding approximately two milliliters of viscous synovial fluid that appeared normal. On the basis of the fever, the tenderness of the left knee, the pain at the terminal degree of flexion, and the normal-appearing synovial fluid, a provisional diagnosis of osteomyelitis of the proximal part of the tibia or the distal part of the femur was made. A bone scan was performed to identify potential sites of osteomyelitis. The flow phase showed reduced tracer activity in the left tibial metaphysis. On the immediate blood-pool study, reduced tracer activity was evident in the proximal part of the left tibia.
While the patient was awaiting the final phase of the bone scan, approximately three hours after the time of presentation, the entire left lower limb became more painful and had swelling, tenderness, and mottling distal to the knee. The femoral pulses were decreased bilaterally, and the popliteal, dorsalis pedis, and posterior tibialis pulses were not palpable on the left. The blood pressure was 90/62 millimeters of mercury (12/8.26 kilopascals), the pulse rate was 132 beats per minute, the respiratory rate was twenty breaths per minute, and the oxygen saturation on room air was 97 per cent. The orthopaedic service was asked to re-evaluate the patient, and a vascular surgery consultation was also requested. Because of the diminished pulses and the swelling of the leg, a 1700-unit bolus of heparin was administered in the emergency room followed by a continuous infusion of heparin at a rate of 340 units per hour. Initial color-flow Doppler studies revealed that the common femoral, superficial femoral, and popliteal arteries were of small caliber but patent bilaterally. Doppler studies also revealed no arterial pulses in the left foot. During the Doppler examination, the child became pale and was immediately returned to the emergency room with a repeat systolic blood pressure of seventy-eight millimeters of mercury (10.4 kilopascals), a pulse rate of 160 beats per minute, and a respiratory rate of sixty breaths per minute. Repeat laboratory studies, performed before infusion with heparin, demonstrated a white blood-cell count of 2.17 x 109 per liter (2170 per cubic millimeter), a hematocrit of 37.2 per cent, a platelet count of 24 x 109 per liter (24,000 per cubic millimeter), a prothrombin time of 17.5 seconds (normal, 10.7 to 12.3 seconds), a partial thromboplastin time of 49.5 seconds (normal, twenty-one to thirty seconds), and a fibrinogen level of 2.09 grams per liter (209 milligrams per deciliter). Ceftriaxone (850 milligrams) and clindamycin (225 milligrams) were administered intravenously in the emergency room approximately nine hours after presentation, and the patient was admitted to the intensive-care unit.
On admission to the intensive-care unit, the patient had a heart rate of 185 to 190 beats per minute and a blood pressure of 78/40 millimeters of mercury (10.4/5.33 kilopascals). Treatment with low-dose dopamine and volume infusion with crystalloid and fresh-frozen plasma was started. Repeat examination revealed tightness of the compartments of the left leg, suggesting the development of a compartment syndrome. The patient suddenly became stiff, apneic, and bradycardic. She was intubated, and cardiopulmonary resuscitation was initiated. The child was pronounced dead after one and one-half hours of unsuccessful cardiopulmonary resuscitation and only twelve hours after the initial presentation to the emergency room.
At autopsy, the results of histological evaluation of specimens of the skin lesions were consistent with a post-infectious stage of varicella zoster. Blood drawn postmortem was positive for group-A Streptococcus on serological assay. Cocci were present in the skin lesions, in the blisters, and within a thrombus in a subcutaneous vessel. Injection of radiopaque contrast medium into the left femoral artery revealed narrowing and attenuation of the popliteal artery and no filling of the tibial or fibular arteries. There was faint filling of the arteries and veins around the ankle. Histologically, the popliteal artery appeared normal. Further evaluation of the left popliteal fossa revealed a septic thrombophlebitis of the popliteal vein. There was no evidence of osteomyelitis of the tibia on examination of histological sections, and there was only a mild reactive synovial hyperplasia on histological examination. No inflammatory cells were present in the synovial tissue. There was no evidence of necrotizing fasciitis. On culture of the knee aspirate, only the broth was positive for group-A Streptococcus serologically. The primary culture plates showed no growth. The fluid was too viscous for a cell count. Multiple septic thromboemboli containing gram-positive cocci were present in the arterioles of all pulmonary lobes. No saddle embolus or thromboembolus was present in the major lobar arteries to explain the sudden decompensation. The brain was edematous, and gram-positive cocci were identified within some vasculature, which was consistent with sepsis. No underlying immunodeficiency was detected on additional studies, including a screening for the human immunodeficiency virus and immunoglobulin analysis.
The cause of death was a fulminant group-A streptococcal septicemia after an infection with varicella zoster, complicated by septic thrombophlebitis and embolization of small septic thrombi to the lungs.
 |
Discussion
|
|---|
Fierman, in a review of the literature, found only twenty reported cases of patients who had arthritis associated with varicella. Cases of patients who had both septic and aseptic arthritis have been reported2,4,5,13,14. The arthritis is primarily monoarticular and most frequently involves the knee4. In our patient, the initial source of the pain in the knee appeared to have been secondary to the septic thrombophlebitis, which was diagnosed at autopsy. More common causes of arthritis associated with varicella include bacterial septic arthritis, viral arthritis, and reactive arthritis. Hematogenous osteomyelitis may also cause pain in the joint and a mild effusion.
Recent reports have suggested an association between varicella and infection with group-A Streptococcus as well as an increase in the prevalence and virulence of group-A Streptococcus bacteremia1,3,6,12-14. It is recognized that a postoperative wound infection with group-A Streptococcus is potentially fatal11. Soft-tissue infections with group-A Streptococcus have been described as presenting with a toxic-shock-like syndrome and are characterized by a high rate of mortality. The production of toxin by the group-A Streptococcus bacteria and the changes produced by the Varicellavirus in the host immune system may explain why otherwise healthy individuals who have varicella are predisposed to severe infection with group-A Streptococcus2,12.
Varicella arthritis of bacterial origin typically occurs after the exanthem phase. Before swelling and erythema of the joint are noted, the only remarkable physical findings may be a limited range of motion of the joint and localized tenderness. Systemic illness is frequently absent. In a study of rabbit knees, Johnson et al. reported three varieties of synovial reactions occurring only three hours after an intra-articular injection of Staphylococcus aureus. All three types of reactions were remarkable for the presence of vacuolar synovial cells containing staphylococci and for the presence of leukocytes in or overlaying the synovial membrane9. It was doubtful that our patient had bacterial septic arthritis because the examination revealed pain only at the terminal degree of flexion; the broth was serologically positive for group-A Streptococcus, but the findings on the primary culture plates remained negative; the synovial fluid appeared normal; and the synovial tissue demonstrated only a very mild reactive hyperplasia on postmortem examination. In addition, no inflammatory cells or organisms were present in the synovial tissue, an unusual finding in a joint that had been infected for two days9.
A viral etiology for varicella-associated arthritis has been proposed by other authors5,10. Priest et al. isolated varicella-zoster virus from a patient who had non-bacterial arthritis. Fink et al. argued against a direct viral invasion. They suggested that the arthritis was due to an activated infiltrate of inflammatory cells rather than to direct viral replication.
Another diagnosis to be considered for patients who have arthritis but negative findings on culture is post-infectious or so-called reactive arthritis. An infectious agent distant from the joint may provoke an inflammatory response in the joint, possibly by immune-complex formation7. Although studies to detect virus or immune-complex formation in the synovial fluid or synovial tissue were not performed in our patient, a reactive form of arthritis is unlikely in the absence of inflammation of the synovial tissue.
Osteomyelitis is another possible source of the pain, and it has rarely been described in association with varicella4,13,14. The presentation and findings on physical examination may be similar to those associated with early bacterial septic arthritis. The metaphyses of immature bones are the common sites for hematogenous osteomyelitis. Howie et al. reviewed the results of bone scans made with use of technetium-99m-labeled phosphate for the diagnosis of osteomyelitis. Although the diagnosis of osteomyelitis is usually made on the basis of areas of increased uptake of the isotope, their study also included three so-called cold scans (photopenic areas). All three cold scans were made at sites of subperiosteal abscesses8. Our patient had no evidence of abscess formation or osteomyelitis on histological examination of the specimens from the proximal part of the tibia.
In our patient, a child with varicella who died of group-A Streptococcus sepsis only twelve hours after being seen in the emergency room, the pain in the left knee was most likely secondary to septic thrombophlebitis of the left popliteal vein, which was discovered at autopsy. Injection of the left femoral artery with radiopaque contrast medium revealed narrowing and attenuation of the popliteal artery and no filling of the tibial or fibular arteries, suggesting compression at the level of the popliteal fossa. The tissues in the popliteal fossa were tense, and histological sections of the popliteal artery were normal. No clot was palpated along the course of the popliteal artery before histological examination. This suggested that septic thrombophlebitis of the popliteal vein produced local inflammation and obstructed venous flow, leading to elevated pressures and subsequent arterial compression. This would explain the photopenic area of the proximal part of the left tibia seen on the flow phase of the bone scan in the absence of subperiosteal abscess formation and the physical findings of tight compartments of the left leg detected immediately before the patient died in the intensive-care unit.
As the physician who is potentially involved in the initial evaluation of a patient who has varicella and pain in the joint, the orthopaedic surgeon should be cognizant of the fulminant and often overwhelming nature of group-A Streptococcus sepsis associated with varicella despite treatment with antibiotics. Before the Doppler studies, the oxygen saturation had decreased in our patient. She became more tachypneic, tachycardic, and hypotensive. In such situations, intensive fluid resuscitation is of paramount importance to maintain the depleted intravascular volume. Establishment of an airway and the administration of oxygen (through a nasal cannula, a mask, or endotracheal intubation) and vasoactive drugs should be initiated immediately. Antibiotics should be started after appropriate blood and synovial specimens have been obtained for culture. Antibiotic therapy alone should not be considered adequate treatment. Had septic thrombophlebitis been considered earlier in the differential diagnosis of pain in the knee associated with varicella in our patient, immediate supportive therapy might have been instituted before the sudden cardiovascular collapse.
 |
Footnotes
|
|---|
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
Department of Orthopaedic Surgery, Children's Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115.
Department of Orthopaedic Surgery, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, Wisconsin 53222.
 |
References
|
|---|
-
Belani, K.; Schlievert, P. M.; Kaplan, E. L.; and |and |Ferrieri, P.: Association of exotoxin-producing group A streptococci and severe disease in children. Pediat. Infect. Dis. J, 10: 351-354, 1991.[Medline]
-
Blank, C. A.; Nagel, J. E.; Adler, W. H.; and |and |Heldrich, F. J.: A transient granulocyte killing defect secondary to a varicella infection. Maryland Med. J, 38: 739-742, 1989.
-
Christie, C. D.; Havens, P. L.; and |and |Shapiro, E. D.: Bacteremia with group A streptococci in childhood. Am. J. Dis. Child, 142: 559-561, 1988.[Abstract/Free Full Text]
-
Fierman, A. H.: Varicella-associated arthritis occurring before the exanthem. Case report and literature review. Clin. Pediat, 29: 188-190, 1990.
-
Fink, C. G.; Read, S. J.; Giddens, G.; and |and |Eglin, R. P.: Chicken pox infection (varicella zoster virus) and acute monoarthritis: evidence against a direct viral mechanism. J. Clin. Pathol, 45: 267-269, 1992.[Abstract/Free Full Text]
-
Givner, L. B.; Abramson, J. S.; and |and |Wasilauskas, B.: Apparent increase in the incidence of invasive group A beta-hemolytic streptococcal disease in children. J. Pediat, 118: 341-346, 1991.[Medline]
-
Goldenberg, D. L.: "Postinfectious" arthritis. New look at an old concept with particular attention to disseminated gonococcal infection. Am. J. Med, 74: 925-928, 1983.[Medline]
-
Howie, D. W.; Savage, J. P.; Wilson, T. G.; and |and |Paterson, D.: The technetium phosphate bone scan in the diagnosis of osteomyelitis in childhood. J. Bone and Joint Surg, 65-A: 431-437, April 1983.[Abstract/Free Full Text]
-
Johnson, A. H.; Campbell, W. G., Jr.; and |and |Callahan B. C.: Infection of rabbit knee joints after intra-articular injection of Staphylococcus aureus. Comparison with joints injected with Staphyloccocus albus. J. Pathol, 60: 165-202, 1970.
-
Priest, J. R.; Urick, J. J.; Groth, K. E.; and |and |Balfour, H. H., Jr.: Varicella arthritis documented by isolation of virus from joint fluid. J. Pediat, 93: 990-992, 1978.[Medline]
-
Quintiliani, R., and |and |Engh, G. A.: Overwhelming sepsis associated with group A beta hemolytic streptococci. J. Bone and Joint Surg, 53-A: 1391-1399, Oct. 1971.[Abstract/Free Full Text]
-
Stevens, D. L.; Tanner, M. H.; Winship, J.; Swarts, R.; Ries, K. M.; Schlievert, P. M.; and |and |Kaplan, E.: Severe group A streptococcal infections associated with a toxic shock-like syndrome and scarlet fever toxin A. New England J. Med, 21: 1-7, 1989.
-
Wheeler, M. C.; Roe, M. H.; Kaplan, E. L.; Schlievert, P. M.; and |and |Todd, J. K.: Outbreak of group A streptococcus septicemia in children. Clinical, epidemiologic, and microbiological correlates. J. Am. Med. Ass, 266: 533-537, 1991.[Abstract/Free Full Text]
-
Wong, V. K., and |and |Wright, H. T., Jr.: Group A beta-hemolytic streptococci as a cause of bacteremia in children. Am. J. Dis. Child, 142: 831-833, 1988.[Abstract/Free Full Text]

CiteULike Connotea Del.icio.us Facebook Technorati Twitter What's this?
|