Primary CNS Lymphoma

Background

The majority of primary CNS lymphomas are large B-cell immunoblastic and EBV positive in nearly all cases. An increase in overall incidence of primary CNS lymphoma is most likely due to the increasing numbers of immunsuppressed patients in the setting of HIV and post-transplant immunosuppression. Virtually all AIDS-associated primary CNS lymphomas have EBV co-infection. After EBV infection normal host responsees mediated by T-lymphocytes suppress EBV induced proliferation. In depressed T-cell immunity patients, EBV transformed clones proliferate leading to the development of lymphoma. primary CNS lymphoma is considered an AIDS defining condition. This disease behaves aggressively and is far less radioresponsive than other lymphomas.

About 1,000 cases of primary CNS lymphoma are diagnosed per year. This amounts to 2% of all CNS tumors. The median age at diagnosis is 55 years in immunocompetent patients and 35 years in immunocompromised patients. Among immunocompetent patients males predominate by 2:1. Among immunocompromised patients males are 95% of the population. These diseases are nearly exclusively diffuse large B-cell lymphomas. About 15% have ocular involvement, typically bilateral involving: vitreous, retina, choroid more than the optic nerve. 75% of patients who present with ocular lymphoma will develop CNS involvement. Systemic NHL is often associated with orbital lymphoma.

Isolated spinal cord involvement is rare. Less than 5% will present with isolated spinal cord or menigeal involvement. Despite this, more than 1/3 will have CSF involvement.

Primary CNS lymphoma is considered an extranodal single site disease, making it a Stage IE, by definition. Primary CNS lymphomas are mostly supratentorial (75%) and intracranial disease is considered more radioresistant than other lymphoma sites. 88% of 8,315 patients in an RTOG review had persistent intracranial disease after 60 Gy. B-cells are not normally found in the CNS, and the mere presence of B-cells in the CNS suggests a pathologic process.

Immunocompetent patients tend to be less often seen with multifocal disease although 50% do have multifocal disease than immunocompromised patients. AIDS patients nearly always (100%) have multifocal or diffuse disease. More than 90% of those with macroscopically unifocal disease are microscopically multifocal. Around 2% - 13% of AIDS patients will develop CNS lymphoma. All of these have EBV positive disease.

Nearly all patients will have a negative systemic workup for lymphoma outside of the CNS. Therefore, if extra-CNS lymphoma is uncovered, the disease is considered NHL with CNS involvement.

Workup and Staging

Primary CNS lymphoma affects all ages with a pkea incidence in non-immunocompromised patients in the sixth and seventh decades, and fourth decade in immunocompromised patients. Most present as an intracranial mass with symptoms of an intracranial mass lesion. Signs are specific to the site of the tumor with focal cerebral deficits occuring in half. 75% are supratentorial. Personality changes and changes in level of alertness are common presenting systems. Other signs include increased intracranial pressure resulting in headaches.

Primary CNS lymphoma affects the deep brain structures rather than the cerebral cortex, so seizures are less common than with other primary brain tumors. Primary CNS lymphomas tend to grow more rapdily than gliomas and symptoms present with a history of weeks to a few months prior to diagnosis.

Primary CNS lymphoma is usally disseminated within the CNS at diagnosis. Over 40% will present with CSF positive cytology as many of these lesions are peri-ventricular, allowing disease access to the CSF. 42% will have demonstrable leptomeningeal seeding based on positive CSF,unequivocal imaging studies or pathologic examination. Despite this, patients rarely have symptoms of leptomenigeal lymphoma.

Brain lesions are multifocal in 40% of immunocompetent and 100% of AIDS patients. Multiple lesions may cause diagnostic confusion with brain metastases particularly because 13% of primary CNS lymphoma patients have a history of prior systemic malignancy. Even absent imaging confirmation of diffuse disease, the autopsy data demonstrated diffuse microscopic involvement even in areas of completely normal MRI imagery. Primary CNS Lymphoma is restricted to the CNS, even in autopsy examinations.

The number 1 location is the deep frontal lobe white matter and frequently the periventricular region. These structures include: Corpus callosum, basal ganglia, brain stem and cerebellum. according to the International Extranodal Lymphoma Study Group.

Occular lymphomas involve the eye and are distinct from orbital lymphomas. They most commonly affect the vitrious, retina or choroid. The eye is a direct extension of the brain. About 20% of primary CNS lymphomas present with occular involvement. 80% - 90% of those with initial ocular lymphoma eventually will develop primary brain lymphoma, usually after a several year latency. Most patients will develop bilateral disease, even if the initial presentation is unilateral. Occular lymphoma usually presents with blurred vision or floaters. Cellular infiltrates can only be visualized by slit lamp examination making this exam essential to the workup of primary CNS lymphoma. The diagnosis of ocular lymphoma is made by vitrectomy.

Within the differential diagnosis, toxoplasmosis is often confused with primary CNS lymphoma. Suspected toxoplasmosis is treated with antibiotics or if necessary biopsied. Other possible diagnoses include secondary metastatic lymphoma, metastatic carcinoma, abscess, hemorrhage, multiple sclerosis, sarcoidosis and in AIDS, toxoplasmosis.

Workup

Suspected CNS lymphoma workup inlcudes and H&P with slit lamp opthalmoscopy to exclude ocular involvement, LDH, EBV titer and HIV status, CSF cytology, MRI imaging of the CNS axis and tissue diagnosis.

The MRI should cover the brain and the spine if there are spinal symptoms. CT of the Chest/Abdomen and Pelvis to rule out other primary lymphoma and a SPECT if there is immunocompromise. MRI positive CNS lymphoma appear as fluffy borders on T1/gad imaging with periventricular sites common and ring enhancement due to central necrosis.

Biopsy in AIDS patients may be avoided in the case of EBV positive, SPECT positive findings. The sensitivity and specificity is 90% - 100%. If both are negative, treat empirically for toxoplasmosis. If either one is positive, a confirming biospy is necessary. In AIDS patients a Toxoplasmosis titer in addition to bone marrow biopsy and CT CAP is necessary.

CSF fluid abnormalities are usually elevated protein (85%), hypoglycemic (33%) increased LDH and increased &beta-2 microglobulin.

Biopsy of brain, globe or CSF most definitively establish the primary CNS lymphoma diagnosis.

Staging

Primary CNS staging is simple. It is an extranodal single site. Therefore the Ann Arbor Staging is IE.

Treatment and Prognosis

The general paradigm for primary CNS lymphoma is: high dose methotrexate or multiagent chemotherapy. If there is a complete response, observe especially in those > 60 years. Radiation therapy is used for recurrence.

Steroids should be held prior to biopsy to avoid skewing biopsy results. Once diagnosis is confirmed steroids should be initiated. If KPS ≥ 40 initiate high dose methotrexate ± radiation therapy after completion of chemotherapy. If ocular disease, consider intra-ocular chemotherapy or radiation to the globe. WBRT may increase toxicity especially in age > 60 and may be with held at least in the primary treatment. Radiation therapy may be reserved for recurrence.

Prognosis

The International Extranodal Lymphoma Study Group IELSG devised a series of prognostic factors that are harbingers of poor outcomes based on Ferrari's report in 2003 JCO. 378 patients were examined from 1980 - 1999. All were HIV negative and all were treated with various regimins (± chemotherapy; ± radiotherapy):

Other factors identified that signal poor prognosis include poor response to chemotherapy, AIDS, multifocal disease.

The median survival with chemo-radiotherapy is 40 months compared with 12 months with radiation alone. OS-5 is 30% with chemo-RT or 5% with RT alone. The outcome with ocular lymphoma is uniformly fatal with median survival times of olny 6 - 18 months.

Surgery

Surgery has a highly limited role, primarily that of biopsy which should be performed prior to administration of steroids, if possible.

Chemotherapy

No large prospective trials have compared chemotherapy + RT to RT alone. Accumulated data from multiple phase II trials have demonstrated chemosensitivity of primary CNS lymphoma to systemic chemotherapy. Most studies have looked at pre-RT chemotherpay to permit an assessment of response to treatment. Almost all patient have a complete but short lived response to radiation therapy. therefore no mearurable disease is present to assess the adjuvant chemotherapy. Chemotherapy in advance of radiation (especially MTX) may reduce the risk of late neurotoxicity. Radiation Therapy opens the blood brain barrier which may persist for weeks to months after RT completion. Chemotherapy should be completed prior to cranial RT to minimize the normal brain exposure to potentially neurotoxic chemotherapeutic agents. This enhanced neurotoxic potential likely applies to other agents besides MTX.

Steroids should be delayed until after biopsy, then commenced. Steroids prior to biopsy may occlude the diagnosis because of an initial 90% tumor regression yielding non-diagnostic results. Only in the rare situation where the patient is unstable should steroids be administered prior to brain biopsy.

Various chemotherapy regimens have been used. What is clear is that radiotherapy alone is insufficient treatment. Median survival with radiation alone is less than 12 months. Adding chemotherapy increases this to 40 months. Conventional lymphoma chemotherapies have not been effective in CNS primary lymphomas. These agents do not cross the blood-brain-barrier in sufficient quantities to produce therapeutic results.

RTOG 8806 and two other randomized controlled studies (Schultz 1996 JCO) demonstrated CHOP or CHOD was ineffective. Schultz showed no benefit for CHOP/CHOD (cyclophosphamide/doxorubicin/vincristine/dexamethasone) due to inability to penetrate the blood brain barrier.

High Dose Methotrexate

HD-MTX is the single most effective agent for the treatment of primary CNS lymphoma. MTX was chosen for its known activity against lymphoma and its ability ot penetrate the blood-brain barrier. It is now the cornerstone of PCNSL. Several studies ahve shown (Phase II) high dose MTX based chemotherapy in combination with radiation (WBRT) have all shown improved survival with a medain of 33 to 51 months over WBRT alone. The best results have been seen in combination with vincristine and procarbazine prior to WBRT, giving a median survival of 51 months. All MTX based regimens when combined with WBRT carry a significant risk of irreversible neurotoxicity manifest by ataxia, dementia and incontinence. Patients aged 60 and older are most vulnerable to this toxicity.

Due to the toxicity of WBRT after MTX, several sole chemotherapy regimens have been designed. The addition of vincristine and procarbazine to MTX achieved a median survival of of 29 months in age ≥ 60 years whether or not WBRT was included. Patients who did receive WBRT died of neurotoxicity. Those who did not died of progressive PCNSL.

Radiation Therapy

PCNSL is a veyr radioresistant disease. Nearly all have an initial complete response, but that response is extremely short lived. OS-5 is 4%, which is dismal compared with NHL where the extranodal disease response is 90%.

RTOG 8315 Phase II trial compared WBRT to 40 Gy → cone down to 60 Gy. Median survival was unchanged at 11.5 months. 80% failed within the boost field.

MSKCC (Abrey, JCO 2000) examined age and MTX+WBRT and found substantial increased risk of neurotoxicity in age > 60. This study is a Phase II study examining 52 patients. Median survival was 60 months. HD-MTX x 5 cycles at 3.5 g/m² was alternated with intrathecal MTX (12 mg) → procarbazine/vincristine + WBRT 45 Gy → high dose cytosine arabinoside. There was no difference in survival between RT v. no RT in those over 60, but DFS was worse if there was no RT. Those who had RT over 60 had an 83% risk of neurotoxicity. For those under 60, the risk dropped to 6%.

This study demonstrated a response rate to pre-RT chemo of 56% complete response and partial response rate of 33%, giving an overall response rate of 89%.

RTOG 9310 compared 36 Gy at 1.2 Gy BID to 45 Gy daily (Fisher 2005) and demonstrated no difference in control but increased toxicity. This study was a prospective study of the Abrey regimen followed by 45 Gy compared with 36 Gy BID if CR to chemotherapy. 63 patients received 45 Gy and 16 recieved 36 Gy. Median survival was 37 months. This study showed no difference in control or survival but worse neurotoxicity (23% compared with 4% in standard fractionation arm).

NABTT 96-07 examined observation after CR to HD-MTX (Batchlor 2003 JCO) . This study examined IV HD-MTX given at 8 g/m² every 2 weeks until CR or 8 cycles. Once there was a CR the patient received an additional two cycles of HD-MTX for every two weeks followed by HD-MTX every 28 days for 11 additional cycles. Median survival has not been reached at 22.8 months, there was no neurotoxicity and CR was 52% with PR 22%.

MGH studied salvage RT (Nguyen, 2005 JCO) which suggests 36 Gy WBRT was a useful salvage dose. Median survival after radiation therapy was 10.9 months and overall median survival was 30 months. There was neurotoxicity in 3 patients aged > 60 and in those who received > 36 Gy at 31% compared with 0%.

Typical response rates to salvage radiation therapy were CR 37% and partial response of 37% for those failing initial chemotherapy. .

RTOG 9310 (Deangelis 2002 JCO) examined IV/IThecal MTX/vincristine/procarbazine &rarr RT → ARA-C. This study found a median survival of 3 years. This study was updated by Fisher who examined toxicity and found no difference in control or survival but worse neurotoxicity.

Treatment Volumes

Areas of high risk in primary CNS lymphoma are the posterior retina and the brain down to C2. These areas need to be covered in the field and are the critical volumes. If the patient presents with primary ocular lymphoma, WBRT to the level of C2 and the entire bilateral orbits to 36 Gy. Cone down to WBRT with posterior retinal coverage and continue the dose to 45 Gy. An alternative to this is to treat with intra-ocular chemotherapy.

Due to the known risk of neurotoxicity in the elderly, (Abrey 2000 JCO/MSKCC) study, it is reasonable to omit WBRT in patients > 60. Unfortunately, WBRT will cause radionecrosis in 83% of such patients and omitting RT will cause death from progression of the PCNSL.

German data reported by Pels 2003 in JCO supported the omission or deferral of radiation after HD-MX/Ara-C. Consolidation Ara-C is an additional treatment option after radiation therapy for partial responders to initial chemotherapy.

Rituximab can be used with MTX/procarbazine/vincristine as induction regimen if followed by reduced dose WBRT. WBRT dose is reduced to 23.4 Gy for complete responders and continued to 45 Gy for partial responders, then followed by Ara-C consolidation chemotherapy.

A MSKCC study by Shah (2007 JCO) reduced WBRT dose with Rituximab found OS-2 at 67% with 2/3 having a complete response. These patients received the reduced dose radiation therapy.

Doses

WBRT radiation dose after a complete remission to chemotherapy is 24 - 36 Gy and consideration of ommission of radiation altogether in people > 60.

For partial response to chemotherapy, increased doses are used 36 Gy - 45 Gy are used. The initial 36 Gy to the whole brain followed by focal radiation to the gross disease to a cumulative dose of 45 Gy.

RTOG 8315 Dose escalation study of RT alone demonstrated significant neurotoxicity at higher doses. There was a high local recurrence in the brain at 61% at doses of 40 Gy + 20 Gy boost.

IELSG randomized MTX x 4 cycles compared with MTX/cytarabine → WBRT(Ferreri, 2009, Lancet) and found complete response rates improved from 18% on MTX alone to 46% on MTX/cytarabine. The ORR was 40% for MTX alone and improved to 69% with the addition of Ara-C.

Special Cases: AIDS/Immunocompromized Patients

Good risk immunocompromized patients with PCNSL include non-HIV immunosuppression and HIV positive with CD4 counts > 200.

AIDS positive PCNSLymphoma should initially be treated with toxoplasmosis antibiotics. If there is no response then proceed to biopsy. Chemotherapy is not well tolerated. Intrathecal MTX should be considred. In the alternative, palliative radiation therapy may be useful alone to 30 - 45 Gy. If the patient is severely immunocompromised, HAART drugs should be considered as first line.

The general treatment paradigm in severely immunocompromised HIV patients is: WBRT to 36 Gy → 45 Gy with concurrent HAART.

Special Cases: Leptomeningeal PCNSL

Leptomeningeal disease is rare. It should be treated with intrathecal MTX or with CSI to 36 Gy → boost to 45 - 50 Gy.

Lacrimal gland dose tolerance is 36 Gy.

RTOG 9310 DeAngeles reported severe delayed neurotoxicity especially in those > 60 years old.

In the German Trials 9% mortality was associated with chemotherapy alone.

Radiation Oncology Synopsis