Myasthenia gravis

MGFA Clinical Classification.

This classification (table 1) is designed to identify subgroups of patients with MG who share distinct clinical features or severity of disease that may indicate different prognoses or responses to therapy. It should not be used to measure outcome. It defers quantitative assessment of muscle weakness to the more precise Quantitative MG Score for Disease Severity, defers response to therapy to the MGFA Postintervention Status and the Quantitative MG Score, and defers the status of medication to the Therapy Status classification.

The fluctuating extent and severity of MG, and the variable predominance of the muscle groups involved, makes it extremely difficult to classify these patients. Most existing classifications are modifications of Osserman’s,1 separating patients with purely ocular involvement from those with generalized weakness, and further separating those with mild, moderate, or severe generalized weakness. Osserman classifications have included categories based on the course of the disease, such as “acute fulminating” and “late severe,” and at times also included categories for muscle atrophy and childhood onset. Experienced clinicians have devised other classifications to monitor response to treatment, some based on the degree of disability or age at onset, and others that include quantitative measurements of specific muscle function, such as arm abduction time and vital capacity.

In general, these classifications use subjective assessments and lack quantification. What one physician may regard as “mild,” another might regard as “moderate” or “severe.” Furthermore, some experienced clinicians believe that oropharyngeal involvement is more dangerous, and perhaps different than limb weakness, and thus should be identified by the classification system. Some feel that “crisis” (respiratory insufficiency necessitating intubation and assisted ventilation) is the result of coincidental infection or other stress in susceptible patients, whereas others regard crisis as defining the ultimate in disease severity.

The Task Force discussed these issues at length, and the members think that a uniform classification is necessary if meaningful comparison of data is to be achieved. The Task Force sees no alternative but to accept the inherent imprecision of a clinical classification, and it recommends that the MGFA Clinical Classification be used to supplement or to replace the classifications now in use. It also recommends that the most severely affected muscles be employed to define the patient’s Class and that the “maximum severity” designation be used to identify the most severe pretreatment clinical classification status. The “maximum severity” designation may be made historically and is employed as a point of reference. The maximum severity remains the point of reference thereafter, with any worsening of the MG being reflected in the postintervention status determination.

MGFA Therapy Status.

The MGFA Therapy Status (table 3) defines the treatment regimen of the patient at a given time and is most useful when used with the MGFA Postintervention Status. The Therapy Status, at any point in time, is defined by using a single designation or a combination of the designations. In addition, the duration of this status, the current doses of all pertinent medications, and the schedule of plasma exchanges and IV immunoglobulin (Ig) should be recorded.

MGFA Postintervention Status.

The Postintervention Status (table 4) is designed to assess the clinical state of MG patients at any time after institution of treatment for MG. Use of this classification requires that specific forms of therapy be recorded separately in the Therapy Status section.

The criteria defining the Postintervention Status require that the patient be examined by someone skilled in the evaluation of neuromuscular disease. Criteria for change in the patient’s status should be defined in each study protocol based on quantitative assessment of strength in pertinent or sentinel muscles. What constitutes a sustained substantial change in medication should also be defined specifically in each study protocol. Composite analysis (“bundling”) of postintervention categories is discouraged because it circumvents the goal of response-specific analysis. However, if composite analysis is employed, the individual categories should also be analyzed. Pending further study, it is recommended that treatment and strength be stable for at least 3 months as a baseline before beginning prospective studies. If the patient has attained CSR, PR, or MM status, the change in status should be indicated as well.

Isolated weakness of eyelid closure was thought not to be a sign of active disease and therefore was not considered an exclusionary criteria from CSR or PR status. In contrast, patients receiving cholinesterase inhibitors are excluded from PR and MM-1 status because these medications mask myasthenic symptoms.

Thymic pathology.

There are no comprehensive guidelines for the evaluation of the nonthymomatous thymus removed from patients with MG. Incomplete and variable sampling techniques14 and the lack of comprehensive diagnostic guidelines may account for the conflicting reports regarding the relationship between the histologic changes in the thymus and the response to thymectomy.

To determine the prognostic and therapeutic importance of pathologic changes in the thymus removed during thymectomy, uniform criteria for specimen management, sampling techniques, criteria for the diagnosis of hyperplasia and involution, characterization of immunocytochemical changes, and other determinants need to be defined and standardized.12,15,16

The histologic classification and grading of thymic neoplasms, including the identification of noninvasive and invasive thymomas, other neoplasms of the thymus, and their relation to MG results, also need to be standardized.15,17-24 Patients with thymoma undergoing thymectomy for MG should be analyzed separately from those without thymoma, because combining these patient cohorts has made it difficult to determine whether the presence of a thymoma, even when noninvasive, alters the prognosis.

The Task Force recommends that the multiple issues involved be addressed with the development of guidelines for the evaluation of the nonthymomatous thymus in MG and a universally acceptable single classification of thymic neoplasms.

MGFA Thymectomy Classification.

Multiple techniques are described for removal of the thymus in MG. The debate regarding which technique is preferable is not resolved. Although, classically, “total thymectomy” is considered the goal of surgery, it has not been demonstrated unequivocally that this is necessary, nor is it clear that all the resectional techniques do achieve this goal.

To resolve the issues regarding the choice of thymectomy technique and whether there is a relationship between the resectional technique employed and the rate of remission and improvement, the type of thymic resection used needs to be defined in as objective terms as possible. In addition, multiple resectional techniques should not be reported as a single cohort.

The Thymectomy Classification (table 5) is based on published reports. The techniques are grouped according to the primary approach (transcervical, videoscopic, transsternal, or combinations) and are described briefly. Referenced reports are recommended for details. Because, within each category, there may be variations in the extent of the resection from surgeon to surgeon, the extent of the resection for each patient cohort must be recorded. In all prospective studies it is recommended that detailed descriptions of the operative technique be supplied, accompanied by drawings and photographs of typical specimens. Ideally, a video of the technique should also be available.

At this time, two types of transcervical thymectomy are performed. The “Basic” resection employs an intracapsular extraction of the mediastinal thymus via a cervical incision and is limited to the removal of the central cervical–mediastinal lobes (figure, A and B). No other tissue is removed in either the neck or the mediastinum.25,26 The original “Extended” resection employs a special manubrial retractor for improved exposure of the mediastinum. The mediastinal dissection is extracapsular and includes resection of the visible mediastinal thymus and fat. Sharp dissection may or may not be performed on the pericardium. The neck exploration and dissection varies in extent and may or may not be limited to exploration and removal of the cervical–mediastinal extensions.27,28 Variations include the addition of a partial median sternotomy29 and the associated use of mediastinoscopy.30 FIGURE

• Download figure
• Open in new tab
• Download powerpoint

Figure. Anatomy of the thymus. This illustration represents what is now generally accepted as the surgical anatomy of the thymus.41 The frequencies (percent occurrence) of the variations are noted. Black = thymus; gray = fat that may contain islands of thymus and microscopic thymus. A-P window = aorto-pulmonary window. Source: Neurology 1997;48(suppl 5):S52–S63.

A number of variations in videoscopic technique are being developed to assist in the performance of a thymectomy. The “Classic” video-assisted thoracic surgery (VATS) technique employs unilateral videoscopic exposure of the mediastinum (right or left) with removal of the grossly identifiable thymus and variable amounts of anterior mediastinal fat. The cervical extensions of the thymus are usually removed from below.31,32 The Video-assisted Thoracoscopic Extended Thymectomy (VATET) employs bilateral thoracoscopic exposure of the mediastinum for improved visualization of both sides of the mediastinum. Extensive removal of the mediastinal thymus and perithymic fat is described, the thymus and fat being removed separately. A cervical incision is performed with removal of the cervical thymic lobes and pretracheal fat.33

There are several variations to the Transsternal Thymectomy approach. The “Standard” technique was originally designed to remove the well-defined central cervical–mediastinal lobes. At this time, although a complete34 or partial35,36 sternotomy may be performed, the resection is more extensive than originally described, with removal of all visible mediastinal thymus. Mediastinal fat, varying in extent, may or may not be removed. The cervical extensions of the thymus are removed from below, with or without some adjacent cervical fat. Variations of this technique include a video-assisted technique using a complete median sternotomy via a limited lower sternal transverse skin incision.37 The “Extended”38 procedure is also known as Aggressive Transsternal Thymectomy39 and Transsternal Radical Thymectomy.40 These resections remove the entire mediastinal thymus and most of the mediastinal perithymic fat. They vary somewhat in extent in the mediastinum and may or may not include all tissue removed by the T-4 techniques. The cervical extensions are removed from below, with or without additional tissue, but without a formal neck dissection.

The combined transcervical and transsternal thymectomy procedures are known as Transcervical–Transsternal Maximum Thymectomy41 and Extended Cervical–Mediastinal Thymectomy.42 These resections routinely use wide exposure in the neck and a complete median sternotomy with en bloc removal of all tissue in the neck and mediastinum that may contain gross or microscopic thymus anatomically. The resections include removal of both sheets of mediastinal pleura and sharp dissection of the pericardium. A similar procedure, although a less extensive resection in the neck and mediastinum, has been described by Lennquist et al.43

It is recommended that the thymectomy classification, with modifiers as necessary, be employed when reporting the results of thymectomy for MG.

Formal Clinical Trial and Outcomes Research.

A prospective, randomized clinical trial remains the preferred method to evaluate therapy (Class I evidence in the American Academy of Neurology [AAN]) nomenclature).45 When a randomized trial does not appear feasible, a prospective risk-adjusted outcome analysis of nonrandomly assigned treatment (Class II evidence in the AAN nomenclature) is recommended.46 In prospective studies, in addition to the use of the classification, definitions, grading systems, and methods of analysis recommended herein, the CONSORT guidelines47,48 are recommended.

“Survival” instruments, which are used in the analysis of remissions, are fundamental in the comparative analysis of therapeutic programs for MG. Although different levels of clinical improvement should be evaluated in the analysis of all forms of therapy, Complete Stable Remission remains the primary focus of the analysis, at least in the assessment of thymectomy.

Qualify-of-life instruments should also be employed because therapy for MG is usually not innocuous and frequently does not produce a completely stable remission. Quality-of-life measures evaluate the impact of intermediate levels of clinical improvement and morbidity of therapy, and complement information provided by remission and clinical improvement analysis. Although a functional status instrument assessing activities of daily living has been developed for MG,49 there are no disease-specific quality-of-life instruments for MG at this time. The Task Force recommends that these be developed. The steps necessary to accomplish this have been defined.50

An Outcomes Analysis guideline and an accompanying diagram defining their interrelationships were developed to provide background information on the available analytic techniques.44 Experts in the field of biostatistics and outcomes analysis should be consulted in the design of all studies, and in the collection and evaluation of the data.

Data bank.

Multi-institutional studies utilizing the data bank concept and fulfilling the requirements discussed next are recommended. This method of study should be particularly useful and practical for multiple institutions to compare the relative value of the many therapies, including various thymectomy techniques.

For a data bank program to be successful, it must be developed appropriately and monitored rigorously. Computer-based patient records, including the “human language” component,51 are required. Definitions, classifications, and standardized forms must be agreed on and used. Standardization of numeric grading for all important variables is required. Mechanisms must be in place to review all the clinical records and to monitor the quality of the database, including validation for completeness and accuracy through a rigid auditing process. The monitoring requirement involves a major commitment by the sponsoring institutions, professionals, and staffs. Mechanisms to defray costs would need to be developed.

Amendments.

The MSAB of the MGFA has established a Standing Committee for review of the Clinical Research Standards. The goal of any proposed amendment is to improve the guidelines based on demonstrable errors in the existing guidelines, development of new data, and common sense. The Review Committee will review recommendations from national and international neurologic organizations and centers, journal editorial boards, neurologists, immunologists, surgeons, biostatisticians, nurses, respiratory therapists, and others working in the field of MG. The Review Committee will also serve as a clearinghouse for questions, as they arise, concerning the application of the Clinical Research Standards.

Please submit questions and proposals in writing to Chairperson, Clinical Research Standards Review Committee, Medical Scientific Advisory Board, Myasthenia Gravis Foundation of America, Inc., 123 West Madison, Suite 800, Chicago, IL 60602; e-mail: myastheniagravis@msn.com