Asherman's syndrome

Dextran solutions are favored distending media for many hysteroscopic procedures because they are easy to administer, distribute uniformly within the uterine cavity, and are relatively nontoxic. We present the case of a 26-year-old woman who developed hypotension, noncardiogenic pulmonary edema, and hemorrhagic diathesis following hysteroscopic surgery with 32% dextran 70. A medical literature review indicates that following hysteroscopic surgery in which dextran solution has been used, "dextran syndrome" has been diagnosed in some patients. This syndrome is characterized by acute hypotension, hypoxia, coagulopathy, and anemia. We speculate on the pathogenesis of this condition and offer recommendations on how to evaluate and treat this rare dextran-related complication.

(CHEST 1997; 111:513-18)

Key words: bronchoalveolar ravage; dextran 70; disseminated intravascular coagulation; hypotension; hypoxia; hysteroscopy; pulmonary edema

Abbreviations: PRBC=packed RBC; SG=Swan-Ganz

Hysteroscopy is a major diagnostic and therapeutic tool allowing the surgeon a panoramic view of the intrauterine cavity. Thirty-two percent dextran 70 (Hyskon; Hyskon Division, Pharmacia; Piscataway, NJ) and 10% dextran 40 (Rheomacrodex; Baxter-Travenol; Tieste, Italy) are distending media commonly used for this procedure.[1,2] Complications rarely linked to dextran solutions include anaphylactic and anaphylactoid reactions, ascites, oliguria, disseminated intravascular coagulation, transient prolongation of bleeding times, and hemodilution.[1-8] Presumably, most of these complications occur after dextran enters the intravascular space.[9-12]

Acute pulmonary edema is also a dramatic and potentially life-threatening complication of dextran administration.[9,10] The osmotic effect of intravascular dextran may cause volume overload and cardiogenic pulmonary edema.[13-16] Alternatively, dextran may have direct pulmonary toxicity resulting in noncardiogenic pulmonary edema.[9-11]

We review the clinical course of a patient, in whom "dextran syndrome" was diagnosed after hysteroscopic surgery with 32% dextran 70. This syndrome is characterized by acute hypotension, noncardiogenic pulmonary edema, anemia, and coagulopathy. Based on information obtained from her BAL fluid specimen, Swan-Ganz (SG) catheterization, and a medical literature review, we conclude that factors other than acute volume overload due to intravascular absorption of dextran contribute to the clinical manifestations of this syndrome.

CASE REPORT

A 26-year-old woman (gravida 2, pare 2) with Asherman's syndrome (intrauterine adhesions) experienced postpartum bleeding which led to a dilatation and curettage. Despite 18 months of hormonal therapy, she failed to resume normal menses, and a hysterosalpingogram revealed extensive intrauterine synechiae. Hysteroscopic lysis of intrauterine adhesions using 32% dextran 70 was performed without complications, and a second procedure was scheduled. Her past medical history was unremarkable; she did not smoke, take medications, or have a tendency to bleed.

Preoperatively, her physical examination and vital signs were normal. Coagulation studies were not performed, but an [O.sub.2] saturation level was 99% and a hematocrit value was 38%. She received epidural anesthesia with 2% lidocaine hydrochloride and fentanyl citrate, followed by an intrauterine injection of 300 mL of 32% dextran 70 diluted equally with normal saline solution. Her laparoscopic survey disclosed no abnormalities, and multiple thick intrauterine adhesions were lysed. During the 97-min procedure, she received 900 mL of lactated Ringer's solution, 50 [Micro]g of fentanyl, and 4 mg of midazolam; estimated blood loss was 200 mL. At case completion, her vital signs were normal although she continued to have modest uterine bleeding.

In the recovery room, she became cyanotic. Her BP was 88/52 mm Hg with a pulse of 122 beats per minute. Arterial blood gas levels with the subject breathing 12 L [O.sub.2] by face mask were as follows: pH, 7.32; [PaCO.sub.2], 41 mm Hg; and [PaO.sub.2], 43 mm Hg. After receiving a 3-L bolus of lactated Ringer's solution and a unit of packed RBCs (PRBCs) over a period of 2 h, her systolic BP remained greater than 90 mm Hg, but urine output was less than 30 mL/h. Ancillary studies included a chest roentgenogram demonstrating diffuse airspace disease involving the upper two thirds of both lungs (Fig 1); an ECG showing sinus tachycardia with ST depression anterolaterally; and an echocardiogram revealing mild pulmonary hypertension, normal biventricular contractions, and no evidence of air embolism. She was intubated, and red frothy sputum was suctioned from her endotracheal tube. Hematologic studies included a WBC count of 13,600/mm3; hemoglobin, 7.3 gm/dL; prothrombin time, 14.9 s (normal, 11.0 to 13.0 s); partial thromboplastin time, 45.9 s (normal, 23.2 to 31.5 s); platelet count, 140,000/[mm.sup.3]; fibrinogen, 82 mg/dL (normal, 150 to 430 mg/dL); fibrin-split products were greater than 250 (normally negative). Shistocytes were seen on a peripheral blood smear. Values in an electrolyte battery were within normal limits. Urine osmolarity was 394 mOsm/Kg and was reportedly "extremely viscous" and difficult to measure.

[Figure 1 ILLUSTRATION OMITTED]

Twelve hours after surgery, she remained hypotensive, hypoxemic, and had a total urine output of 1,000 mL. SG catheter measurements included central venous pressure, 8 mm Hg; pulmonary artery systolic pressure, 32 mm Hg; pulmonary artery diastolic pressure, 17 mm Hg; pulmonary capillary wedge pressure, 9 mm Hg; cardiac index, 3.8 L/min/[m.sup.2]; and systemic vascular resistance, 1,102 mm Hg. A BAL performed 48 h later yielded 28 mL of hemorrhagic fluid with 12.4 x [10.sup.6] leukocytes, 95% neutrophils, and many erythrocytes. Protein was not measured, and cultures returned negative. Over the next 3 days, she received 6 units of fresh frozen plasma, 3 units of whole blood, and 2 units of PRBCs. Without additional blood products, her bleeding and coagulopathy resolved 24 h later.

Despite the fact that numerous cultures were negative for bacterial organisms and despite administration of empiric antibiotics, she remained febrile and hypoxemic, and x-ray films continued to show pulmonary infiltrates. Gradually, her roentgenographically evidenced abnormalities improved, and 17 days after surgery, she was extubated. She was discharged from the hospital while receiving 2 L of [O.sub.2]. Flow-volume curves 7 weeks later included an [FEV.sub.1] of 70% predicted, an FVC of 73% predicted, and an [FEV.sub.1]/FVC ratio of 96% predicted. At the present, 3 years later, the patient remains dyspneic with exertion but has declined a formal pulmonary evaluation.

DISCUSSION

Hysteroscopy using dextran solutions provides a panoramic view of the uterine cavity. Diagnostic hysteroscopy, lasting only a few minutes, is associated with minimal side effects, while therapeutic hysteroscopy, requiring greater time and more insufflating medium, is associated with greater risks.[2]

Our patient's course was complicated by hypotension, pulmonary edema, anemia, and coagulopathy following hysteroscopic surgery utilizing a dextran solution. A review of medical literature indicates that this quadrate of medical problems has been reported in other patients (Table 1).[1-6,9,10,13] All patients with dextran syndrome had hematocrit reductions, and several required blood products, presumably to maintain adequate blood volumes or to reverse a hemorrhagic diathesis.[2-4,10] Patients were at greatest risk for these complications if surgery lasted more than 45 min (range, 45 to 120 min), if more than 500 mL of distending media was used, or if large areas of endometrium were traumatized (Table 1). Injection pressures greater than 150 mm Hg may be another risk factor leading to increased dextran intravasation.[10]

[TABULAR DATA 1 NOT REPRODUCIBLE IN ASCII]

Although dextran syndrome has not been associated with fatal complications, significant hypoxia may ensue and several patients have required short-term mechanical ventilation (Table 1).[1-6,9,10,13] With supportive treatment, however, most improved rapidly and were discharged from the hospital within 1 week of surgery.

The cause of dextran-associated pulmonary edema is controversial. Several authors suggest that rapid intravascular absorption of dextran solution via the uterine cavity could result in acute volume overload with cardiogenic pulmonary edema.[1-5,14] They cite the calculations of Lukacsko[16] who estimated that every 100 mL of intravascular 32% dextran 70 would increase intravascular volume by 860 mL.

While 32% dextran 70 can act as a volume expander, a significant intravascular infusion would be required in order for cardiogenic pulmonary edema to develop in young, generally healthy patients. Although large increases in intravascular volume might cause blood pressure elevation, this was not described in the cases we reviewed. Our patient had pulmonary compromise, but at no time were filling pressures elevated. Perhaps additional factors contribute to the development of pulmonary edema in this setting.

In 1975, Kaplan and Sabinil postulated that dextran can damage alveolar endothelial cells directly. Arguing that the lack of abnormal auscultatory findings, electrocardiographic changes, or elevation in cardiac enzymes favored a noncardiogenic cause, they believed dextran 40 caused pulmonary edema in their patient. Others adopted this original hypothesis to explain how pulmonary edema occurred in their patients who were exposed to 32% dextran 70, although SG catheter measurements were not used to support their contentions.[2,6,10]

Based on canine experiments, Mangar[12] also suggested that intravascular dextran could directly cause pulmonary endothelial damage. He measured protein in BAL fluid before and after infusions of whole blood or 32% dextran 70. Protein was detected only in those animals which received 32% dextran 70. He concluded that 32% dextran 70 "altered pulmonary microvascular membrane permeability, causing alveolar flooding with plasma proteins and possibly accounting for the deterioration of oxygenation and pulmonary compliance seen in patients." Unfortunately, BAL fluid protein was not assayed in our patient.

The cause of dextran syndrome-associated coagulopathy is also unsettled. Dextran infused at a dose greater than 1.5 g/kg will prolong bleeding times, will impair platelet aggregation and platelet procoagulant activity, and can cause a modest reduction in plasma von Willebrand concentration.[1-3,7,14] Tissue factors released after dextran intravasation could also promote fibrinolysis and a consumptive coagulopathy. Hemodilution of clotting factors could further account for coagulopathic changes, although this could occur only under the most extreme circumstances of dilution.[17] Specific coagulation factors have not, however, been assayed during episodes of acute cardiovascular compromise. Relevant hematologic findings noted in our patient included the manifestation of shistocytes on a peripheral blood smear, elevations in fibrin-split products, reduction in fibrinogen, and prolonged prothrombin and partial thromboplastin times, suggesting a consumptive coagulopathy.

We speculate that intravascular absorption of dextran solution, through a combination of factors, leads to "dextran syndrome" (Fig 2). In our patient, SG catheter measurements 12 h after the onset of symptoms indicated that volume overload was not the major factor contributing to her pulmonary edema at that fume. Additionally, the findings on BAL, done on postoperative day 2, were more consistent with acute alveolar hemorrhage and not cardiogenic pulmonary edema. Despite normal firing pressures, she required [O.sub.2] throughout her hospital stay and at discharge 21 days later. The slow resolution of her radiographic abnormalities and the presence of persistent pulmonary symptoms 3 years later are more indicative of a chronic process, perhaps owing to dextran-induced endothelial damage.

[Figure 2 ILLUSTRATION OMITTED]

In patients who become hypotensive after receiving dextran, a fluid challenge is indicated. SG catheter measurements may be required to better clarify the nature of pulmonary edema in those who do not rapidly improve. Diuretics do not speed dextran clearance and when fluid overload is not present should not be used. Dextran is filtered rapidly at the glomerulus and concentrated in the tubules and collecting ducts as saltwater is reabsorbed. Because it exerts an oncotic pressure, renal filtration may diminish transiently. This, coupled with an increased intratubular concentration of dextran, may explain why several patients with dextran-induced toxicities have decreased urine volumes.

Our patient was previously exposed to dextran during her first hysteroscopic procedure. We wonder if this represents an additional risk. Dextran may induce immune complexes contributing to pulmonary injury.[18] Steroids possess anti-inflammatory and immune-modulating effects. Although steroids were not given to our patient, they could be considered in those who fail to improve rapidly. In summary, we have presented a patient who after a second exposure to a dextran solution during hysteroscopic surgery, rapidly developed life-threatening hypotension, hypoxia, coagulopathy, and anemia. Although her clinical course is the most severe yet published, a review of the medical literature indicates that these complications are occasionally seen with dextran use and may be linked to several diverse pathophysiologic processes. With supportive care, most patients do well and very few will be left with long-term pulmonary defects.

ACKNOWLEDGMENT: The authors wish to thank Jane Lopez for the preparation of this manuscript.

REFERENCES

[1] Vercellini P, Rossi R, Pagnoni B, et al. Hypervolemic pulmonary edema and severe coagulopathy after intrauterine dextran instillation. Obstet Gynecol 1992; 79:838-89

[2] McLucas B. Hyskon complications in hysteroscopic surgery. Obstet Gynecol Surv 1991; 46:196-200

[3] Mangar D, Gerson Jl, Constantine Rm, et al. Pulmonary edema and coagulopathy due to Hyskon (32% dextran 70) administration. Anesth Anal 1989; 68:686-87

[4] Choban JM, Kalhan SB, Anderson RJ, et al. Pulmonary edema and coagulopathy following intrauterine instillation of 32% dextran 70 (Hyskon). J Clin Anesth 1991; 3:317-19

[5] Jedeikin R, Olsfanger D, Kessler I. Disseminated intravascular coagulopathy and adult respiratory distress syndrome: life-threatening complications of hysteroscopy. Am J Obstet Gynecol 1990; 162:44-5

[6] Brandt RR, Dunn WF, Ory SJ. Dextran 70 embolization: another cause of pulmonary hemorrhage, coagulopathy, and rhabdomyolysis. Chest 1993; 104:631-33

[7] Witz CA, Silverberg KM, Burns WN, et al. Complications associated with the absorption of hysteroscopic fluid media. Fertil Steril 1993; 60:745-56

[8] Mangar D. Anaesthetic implications of 32% dextran-70 (Hyskon) during hysteroscopy: hysteroscopy syndrome. Can J Anaesth 1992; 39:975-79

[9] Leake JF, Murphy AA, Zacur HA. Noncardiogenic pulmonary edema: a complication of operative hysteroscopy. Fertil Steril 1987; 48:497-99

[10] Zbella EA, Moise J, Carson SA. Noncardiogenic pulmonary edema secondary to intrauterine installation of 32% dextran 70. Fertil Steril 1985; 43:479-80

[11] Kaplan AI, Sabin S. Dextran 40: another cause of drug-induced noncardiogenic pulmonary edema. Chest 1975; 68: 376-77

[12] Mangar D. Hyskon-induced pulmonary edema. Surg Gynecol Obstet 1993; 177:561-64

[13] Golan A, Seidner M, Bahar M, et al. High-output left ventricular failure after dextran use in operative hysteroscopy. Fertil Steril 1990; 54:939-41

[14] Ljungstrom KG. Safety of 32% dextran 70 for hysteroscopy [letter]. Am J Obstet Gynecol 1990; 163:2029

[15] Schinagl EF. Hyskon (32% dextran 70), hysteroscopic surgery and pulmonary edema [letter; comment]. Anesth Analg 1990; 70:223-24

[16] Lukacsko P. Noncardiogenic pulmonary edema secondary to intrauterine instillation of 32% dextran 70 [letter]. Fertil Steril 1985; 44:560-61

[17] Clavarella D, Reed R, Counts R, et al. Clotting factor levels and the risk of diffuse microvascular or bleeding in the massively transfused patient. Br J Haematol 1987; 67:165

[18] Hedin H, Richter W, Messmer K, et al Incidence, pathomechanism and prevention of dextran-induced anaphylactoid/anaphylactic reactions in man. Joint WHO/IABS Symposium on the Standardization of Albumin, Plasma Substitutes and Plasmapheresis, Geneva 1980, Develop Biol Standards 1980; 48:179-89

(*) From the Division of Internal Medicine, the Everett Clinic, Everett, Wash (Dr. Ellingson), and the Division of Hematology/ Oncology, Virginia Mason Medical Center, Seattle (Dr. Aboulafia).

Manuscript received November 30, 1995; revision accepted June 19, 1996.

Reprint requests: Dr Aboulafia, Section of Hematology/Oncology, Virginia Mason Medical Center, 1100 Ninth Avenue (H14-HEM), PO Box 900, Seattle, WA 98111

COPYRIGHT 1997 American College of Chest Physicians
COPYRIGHT 2004 Gale Group