Causalgia: A Meta-analysis of the Literature

Background  Causalgia is not familiar to most physicians whose training and experience are limited to civilian practice.

Hypothesis  Through a thorough review of the literature, we attempted to determine the boundaries of causalgia and separate it from other sympathetically related disorders.

Data Sources  Database search for English-language articles in MEDLINE and Index Medicus up to the year 2000 as both keyword and subject under causalgia.

Study Selection  References that described any new cases referred to as "causalgia" by their authors were included in a meta-analysis.

Data Synthesis  One hundred ten articles contained a total of 1528 cases of causalgia. High-velocity missiles caused at least 77% of the injuries. In 72% and 90% of the cases reported, the time from injury to onset of pain was within 1 week and 1 month, respectively. Median nerve alone or in combination with other nerves (56%) and sciatic trunk injury (60%) were the most common nerves involved. In 92%, the nerve injury was incomplete. The most prominent clinical manifestations included burning pain in 86%, increased sweating in 73%, relief with application of cold in 62%, warmth in 50%, paresthesias in 96%, absence of anesthesia in 81%, and sensitivity to stimuli in 98%. Response to sympathetic blocks was observed in 88%. Finally, a total of 94% of the patients undergoing sympathectomy were cured.

Conclusions  Cases of causalgia are easy to recognize and treat, with excellent results. Causalgia always follows a somatic nerve injury, usually partial, and is associated with near-constant, very severe pain distal to the injury in the extremity, varied in nature but characteristically with a predominantly burning quality. An effective anesthetic block of the appropriate part of the sympathetic chain frequently immediately relieves the pain. Most cases are cured by surgical sympathectomy.

CAUSALGIA COMES from the Greek words meaning heat (καυσος) and pain (αλγος), ie, burning pain. Causalgia, also known as true major causalgia, needs to be differentiated from minor causalgia and many other posttraumatic pain syndromes, such as reflex sympathetic dystrophy, Sudeck atrophy, sympathalgia, and posttraumatic edema. These syndromes are presumed to result from a reflex disturbance initiated by the injury or inflammation in which effective therapy necessitates certain measures not ordinarily required to manage the inciting cause.¹

While reviewing the literature in the process of reporting our cases of causalgia in an earlier report,² we often noted that patients with the same clinical manifestations were labeled with different conditions according to the above terminology. Many of the articles discuss the cause, pathophysiology, clinical presentation, and treatment of these syndromes together as a group.^3-5 Therefore, we performed a meta-analysis of the literature to evaluate all currently reported cases. Our aim was to better define the condition of causalgia as a specific syndrome that could be differentiated from other, less well-defined posttraumatic painful syndromes.⁶

Although patients with different manifestations of sympathetically initiated pain have been treated medically with varying results,^3,7-9 the only successful treatment for major causalgia is complete interruption of all sympathetic pathways to the affected limb, usually achieved by surgical sympathectomy.^2,10,11 Patients with other posttraumatic pain syndromes usually respond less well to this invasive mode of therapy. This emphasizes the importance of a more descriptive definition of the clinical manifestations of causalgia.

Sympathetic-derived pain syndromes present a wide spectrum of manifestations and have been defined by varying nomenclature, producing significant confusion.^1-5 By using a meta-analysis approach, we believe causalgia can be described distinctively enough to be identified as a unique entity among this mixture of syndromes. Causalgia is rarely seen except during wartime (Figure 1) and is consequently unfamiliar to most physicians whose training and experience are limited to civilian practice. As a result, the diagnosis is frequently delayed or missed, causing protracted suffering and increased disability.^2,11 Through this review of the literature, we have attempted to define the true clinical manifestations and correct treatment of this chronic debilitating disease.

All efforts were made to include all reported cases of causalgia in this study. A database search was performed for English-language articles in MEDLINE (1966-2000, as both keyword and subject) under causalgia. This was also done with Index Medicus (up to 1965) under both causalgia and neuralgia, since the latter term was sometimes used for this condition during this earlier period. All references presenting new cases were included in the meta-analysis if they represented "causalgia" according to their authors. Additional references to cases reported in other articles that were not included in MEDLINE or Index Medicus were also included in our study. A total of 378 articles were identified and reviewed in relation to the topic. They were classified as meta-analysis, descriptive, experimental, animal study, or case reports. Articles focusing on animal studies or review of the literature without presentation of new cases were excluded. A group of 109 articles describing cases of causalgia, as well as our recent report (total, 110 articles), form the basis of this analysis. The articles contained case series of causalgia,^2,12-23 patients with causalgia along with other sympathetic-based pain syndromes,¹¹ single case reports, and case series of specific complications of treatment modalities, such as causalgia as a complication of first rib resection for thoracic outlet syndrome²⁴ or thoracoscopic cervical sympathectomy,²⁵ or after venipuncture.⁵

Eight groups of patients were described twice.^19,26-34 Three series of reports, with 4 reports in each group (total of 12 reports), contained more patients in the second of the series (and third in one series) and were then duplicated up to 4 times in the literature.^{11,15,18,35-44} In 2 groups of reports, consisting of 2 articles each, there were patients common in each one from the other group.^29,30,45,46 In one series of reports,^47,48 the group of patients in the first report was included in a larger group of patients in the second group. If individual duplicated patients in consecutive reports from the same institution could be identified, then these duplicated patients and not the report as a whole were excluded. If some data not reported in the selected article were available in the excluded "duplicated" article, then these data were included without duplicating patients. The weighted average of reported means in every article was calculated as the sum (number of patients multiplied by the mean of those data for each individual article) divided by the total number of patients in all articles reporting information on that specific subject.

Thus, we were able to identify 1528 patients labeled as having causalgia by the authors of the articles. The following information was extracted and analyzed for all patients described:

• General

•     Mechanism of injury

•     Nerve and limb involved

•     Time from injury to onset of pain

• Clinical manifestations

•     Type and distribution of pain

•     Sweating

•     Limb warmness

•     Color changes

•     Sensitivity to cold

• Treatment

•     Response to sympathetic block

•     Response to sympathectomy

The first description of the condition, while not referred to as causalgia, is believed to be by Denmark in 1813.⁴⁹ Our analysis includes all articles published from 1867, including the first case of causalgia, which was labeled by Mitchell et al in 1864,¹² to the end of the 20th century. The English-language literature contains a total of 1528 cases of causalgia (as labeled by their authors). The vast majority of these cases are related to major wars (Figure 1), confirming that causalgia is mainly a wartime disease.

The major etiologic factor was high-velocity missile injuries (67% of all cases); the cause was not mentioned in 12% of cases (Table 1). The iatrogenic group (5% of the 1350 cases for which cause was identified) included 5 cases that occurred after laminectomy, 1 after cholecystectomy, 8 after first rib resection,²⁴ and 22 other postoperative cases (in 6 reports). The nonpenetrating causes included high-velocity deceleration trauma in 24 patients, blunt trauma in 2, brachial plexus traction in 4, nerve stretch in 28, joint sprain in 3, entrapment in 9, and fall in 2. Other causes included myocardial infarction in 7, venipuncture in 9,^5,50 chronic pain syndrome in 6,⁵¹ telephone-mediated lightning electrical injury in 2,⁵² and capsaicin treatment in 1⁵³; other medical conditions in 7; and in 19 patients (in 2 reports), no preceding injury was present. If the cases with nonreported causes are excluded, then at least 77% of the injuries were caused by a high-velocity mechanism.

In 72% and 90% of the 926 patients in whom the time from injury to onset of pain was reported, pain occurred within 1 week and 1 month, respectively (Table 2). In 602 cases (39%), time to pain onset was not mentioned. Also, in 92% of the 904 patients in whom nerve injury was reported, the injury was partial transection (Table 3). Complete nerve injury was reported in only 3% and miscellaneous or other kinds of injury in 5% of the remaining patients in whom nerve injury was reported.

Nerves involved in the causalgic symptoms are listed in Table 4. A total of 707 patients did not have the injured nerve specified. Of 462 injuries that had the responsible nerve defined in the upper extremities, in 261 cases (56%) the injured nerve was the median nerve alone or together with other nerves. In the lower extremities (241 patients), 144 injuries (60%) were related to the sciatic trunk alone.

In Table 5, the clinical manifestations are summarized. Among patients for whom symptoms were reported, the most prominent included burning pain in 86%, increased sweating in 73%, relief of pain to some extent with application of cold in 62%, and warmth to the touch in the involved extremity in 50%. There were paresthesias in 96% of the patients, but no anesthesia (81%) in most of them. Sensitivity to foreign stimuli was seen in close to all of the reported cases (98%). Because the color of the involved extremity was not mentioned in 989 (65%) and not reported clearly in another 342 (22%), there is no clear conclusion on this matter. However, cyanotic color (21%) was seen slightly more commonly than red discoloration in the affected limb.

With regard to diagnosis, 822 patients (54%) received sympathetic blocks of the involved extremity. Of those, a beneficial response was observed in 721 (88%) of the patients. In addition, 791 patients (52%) were treated by sympathectomy of the diseased extremity. In 721 patients (91%) the condition responded to this intervention. A second sympathectomy was performed in 21 patients whose condition did not respond to the initial procedure, with a uniform favorable response.¹¹ Therefore, a total of 94% of the patients receiving sympathectomy were cured. The timing of sympathectomy in relation to the time of injury was reported in only 101 patients. Eight patients were operated on within 3 months; 7 were cured by the first procedure, and the last was cured after reoperation. Of the 40 patients operated on more than 3 months after injury (30 between 3 and 6 months and 10 operated on after ≥6 months), all were cured with surgery. Although surgical timing was not explicitly mentioned in the other 743 cases, from the timing of publication of the article to the time the cases were reported, it appears the majority of these patients underwent sympathectomy less than a year after injury.

About half of the patients were described in 13 original case series.^2,12-23 The results in these primary reports correspond favorably with the results obtained on the evaluation of the whole group of 1528 patients, including all review articles.

Our current understanding of this intriguing disease has resulted primarily from wartime observations. The first description by Denmark, approximately 190 years ago, of a case characteristic of causalgia came about as the consequence of his caring for a soldier whose musket ball injury occurred during the war of 1812.⁴⁹ The classic description of the clinical picture by Mitchell and his associates¹² half a century later resulted from their experiences during the American Civil War at a specialized center for neurologic disorders and injuries. World War II provided the opportunity for observation and treatment of hundreds of patients with this disorder and for clarification of its management.^{14-16,18,39,40} Shumacker,¹¹ as the director of the army vascular centers and on the basis of extensive experience extending over 40 years, reported on 110 cases and presented his overview of causalgia. The information derived from these 3 crucial reports matches the results achieved by our meta-analysis. We have included many reports and a large number of patients to summarize cases carrying the same definitions.^{2,12,13,19-23} Although some reports may have been missed, significant efforts were made to include every case from the literature, and we believe we were largely successful.

Although some reports represent cases that only partially match patients with "real causalgia," this study is based on all cases labeled as causalgia by their authors. A subset of 19 patients had no penetrating injury. Circulatory (vascular) compromise, fractures, spondylosis, infection, and other unique pathologic conditions without direct nerve injuries^51-53 were interpreted as being the cause of causalgia. On careful review, most of these were found not to be true causalgia. Generally, most were other sympathetic-derived disorders, such as reflex sympathetic dystrophy, with causalgia in parentheses, implying that these are the same entities.^3-5 By reviewing the clinical manifestations or treatment response, it was evident that the authors were describing a different disease under the topic of causalgia.^7-9,53

There is no doubt that causalgia is mainly a wartime disease. The great majority of causalgia cases are from war casualties; of these, 721 (47% of the total group) are from 13 articles that present cases from 7 major wars, beginning with the American Civil War.^2,12-23 Cases not reported from combat injuries are from similar injuries. Among those with the cause reported, 77% were due to high-velocity missile injuries. The second most common cause was other penetrating nerve trauma (7% of the cases with cause identified), including iatrogenic (5%). The reports coding other causes (12%), such as vascular compromise, fractures, chemical applications, spinal cord abnormalities, and nonpenetrating trauma, all without specific nerve injuries, were carefully examined. They, in large part, do not represent patients with true causalgia.

Ninety percent of patients presented symptoms of causalgia within the healing period of their wounds (<1 month). The closer pain starts to the time of trauma, the more likely that the disorder is causalgia. Also, 92% of the patients had partial rather than complete (3%) nerve interruptions. If only series with patients with typical causalgia are considered (13 reports with 721 patients), up to 99% were reported to have partial nerve injuries.² This primarily distinguishes causalgia from other sympathetically maintained pain syndromes, such as reflex sympathetic dystrophy, Sudeck atrophy, sympathalgia, and posttraumatic edema, in which the inciting cause is not penetrating partial nerve injury.

Some observers state that causalgia is somewhat more common in the upper extremity, with the ratio being about 5:4.¹¹ The ratio of upper to lower extremity involvement in this meta-analysis was 1.7. Median nerve, either alone or combined with other peripheral nerves (56%) in the upper extremity, and sciatic trunk or its major terminal branches (60%) in the lower extremity were the most commonly injured nerves. Burning pain is the hallmark of causalgia; in fact, the Greek root for causalgia has the same meaning. From the first description of the condition by Denmark, in all large series burning pain has been the primary symptom. If reports describing other types of pain or no pain are reviewed, the other clinical symptoms also do not match this disease and the cases should not be classified as true causalgia. The pain is aggravated by many kinds of emotional or physical stimuli in most patients with causalgia (98%). Even the mildest stimuli, such as the sound of traffic, conversational noise, or even water dripping from a water tap, may aggravate the symptoms. If not familiar with the syndrome, many physicians have considered these patients to have an underlying psychotic disorder.⁵⁴ These patients typically keep the involved extremity in a moist towel, guarded from tactile contact, immobile, and supported. Because of the near-constant pain, they excessively focus on their extremity and become detached from their environment.

Paresthesias, but not anesthesia, are also a near-constant feature in the involved extremity. In addition, approximately 73% of the patients had increased sweating of the limb involved. These findings are constant in the reports with a majority of the patients and should be included as manifestations of causalgia. Sixty-two percent of the patients had relief of the pain with application of cold compared with 26% whose pain became worse and 12% who had no change. Fifty percent of the patients had a warmer extremity compared with 26% with a colder limb than the other parts of the body; 24% had no change. Thus, neither these symptoms nor the color of the extremity provides specific diagnostic information. The commonly described presentation of a cool, reddish extremity with symptoms becoming worse with the application of cold is not in agreement with the overall composite of true causalgia.^2,12-23

Reviewing the various operative and nonoperative attempts to manage causalgia before the era of sympathetic interruption is of little importance because none was effective. These included neurolysis, repeated resection of neuroma, proximal resection or chemical block of the affected peripheral nerve trunk, and division of the posterior roots. The individual to first use sympathetic interruption is not known. In 1917, Leriche⁵⁵ referred to periarterial sympathectomy and excision of thrombotic segments of arteries. However, in reviewing his cases, we found that he was not referring to patients with causalgia. According to Shumacker,¹¹ in 1929 Petit-Dutailles and his associates subjected a patient to several operations, the last of which completed, for all practical purposes, sympathetic denervation of the entire cervical chain, which finally caused pain relief. Spurling¹⁰ probably deserves credit for establishing sympathectomy as a curative measure in causalgia from his report of 1930, in which complete relief of pain followed excision of the second dorsal ganglion and division of the rami to the first and stellate ganglia. While occasional cases were subsequently reported, sympathetic interruption was not firmly demonstrated to be the correct method of management in a substantial number of cases until World War II.¹¹

As a rule, in typical causalgia the application of sympathetic block has brought about immediate cessation of pain.^38-40 The response to sympathetic block is diagnostic of true causalgia. During the period of anesthesia, some of these patients remain aware of their paresthesias from the somatic nerve injury. However, the results are immediate and spectacular. Before injection, the patient is in agonizing pain and guards the injured extremity, extremely reluctant to allow any examination. As soon as the injection has been made, adequate examination, difficult or impossible before, is now easily performed. The strength of active movements of the affected hand or foot is vastly improved.² As this study demonstrates, in the 54% of the patients who received blocks, 88% had a favorable response. When relief is not obtained (12%), the 2 following reasons appear to be present: (1) the patient has one of the other diseases related to sympathetically maintained pain syndromes other than causalgia; or (2) in patients with true causalgia, the injection appeared to fail in anesthetizing the appropriate part of the sympathetic chain. If the 40 patients described before the introduction of surgical sympathectomy by Spurling¹⁰ are omitted, 53.5% of the correctly diagnosed patients were treated by sympathectomy and 94% were cured. When sympathetic denervation is complete in a patient with causalgia, the pain and related symptoms disappear. Most authors now agree that immediate sympathectomy should be offered to these patients and that the operation for both upper and lower extremities is simple and safe and the postoperative period of disability is short. No long-term complications are identified. Therefore, one should operate early, before the patient becomes dependent on narcotics, develops irreversible atrophic musculoskeletal changes, or loses motion in affected joints. Minimally invasive thoracoscopic sympathectomy for the upper extremity has also been recently reported, with excellent success.²⁵ Resection of half of the stellate ganglion down to not less than the fourth thoracic ganglion for the upper extremity and the first to third lumbar ganglia for the lower extremity is required to achieve complete sympathetic denervation of the limb.

Sympathetic nerve-based pain syndromes present a wide spectrum of manifestations and have been confused with variable nomenclature. This meta-analysis is an attempt to describe distinctly the discrete syndrome of causalgia. The primary symptoms of causalgia are burning pain that occurs soon after a high-velocity missile or other penetrating injury, causing partial peripheral nerve transection; the pain is aggravated by numerous minor stimuli and displays the effects of hypersympathetic state. Sympathetic denervation following diagnostic sympathetic block dramatically cures the syndrome. In some cases, one or more blocks can achieve a cure. The majority require operative sympathectomy. Spontaneous disappearance of causalgic pain occurs so rarely that treatment should never be deferred. Causalgia should be readily diagnosed, and treatment provides excellent permanent results.

Corresponding author: Ronald V. Maier, MD, Department of Surgery, Campus Box 359796, Harborview Medical Center, 325 Ninth Ave, Seattle, WA 98104 (e-mail: ronmaier@u.washington.edu).

Accepted for publication February 22, 2003.